The present disclosure relates generally to monitoring electrical devices, and more specifically, to a system and method for (of) monitoring an operating condition of an electrical device when in operation. Moreover, the present disclosure also relates to a computer program product comprising a non-transitory computer-readable storage medium having computer-readable instructions stored thereon, the computer-readable instructions being executable by a computerized device comprising processing hardware to execute the aforesaid method. Embodiments of the present disclosure concern a type of technology that is capable of reducing Carbon Dioxide (CO2) emissions, as a result of improving operation of electrical devices, for example electrical machines, at higher efficiency and avoiding unnecessary maintenance.
An example electrical device transforms electrical energy, defined by electrical current and voltage, into another form of energy, for example light energy, heat energy, or motional (kinetic) energy (e.g. an electrical motor) or vice versa (e.g. an electrical generator). Alternatively, an electrical device (e.g. a transformer) may transfer or couple electrical energy from one circuit to another. In industrial applications, large numbers of such electrical devices are employed. These electrical devices have to be periodically monitored manually or by using monitoring systems to detect unexpected faults that can haft or hamper operation, for example production; such monitoring is performed, for example, proactively. A cumulative annual expenditure of $700 billion can be attributed either to faults in electrical equipment or their maintenance, wherein the expenditure includes additional damage, corrective work, and other consequential impacts.
When a fault occurs in a given electrical device, an electrical power consumption of the given electrical device in a faulty condition potentially differs from an electrical power consumption of the given electric device under corresponding conditions prior to occurrence of the fault condition. For example, in a case of an assembly having an electric motor, failure or deterioration in a condition of the electrical motor may bring about either an increase or a decrease in the power consumption of the electrical motor, depending upon a nature of the failure. Changes in conditions other than failures or deterioration of the electrical motor may also potentially affect the power consumption of the electrical motor. For example, a change in a load experienced by the electrical motor may affect its power consumption.
Existing known monitoring systems and methods that are employed for monitoring electrical devices are cumbersome, expensive, and/or are limited to identifying only a few failure modes of the electrical devices. Furthermore, the existing monitoring systems and methods are inaccurate in prognosing and/or diagnosing, potential problems or faults. One such existing known monitoring system employs a vibration analysis, which is used to detect faults in rotating equipment (e.g., fans, motors, pumps, gearboxes etc.) such as imbalance, misalignment, rolling element bearing faults, and resonance conditions. Another existing known monitoring system employs an electrical motor current signature analysis (MCSA) that measures current on a single electrical phase or three electrical phases of an electrical motor to monitor an operating condition of the electrical motor. However, this existing known system is not capable of analyzing distorted voltages and has a limited fault-type application.
Known model-based voltage and current (MBVI) systems are inflexible, have limited diagnostic capability and provide a high level of false alerts, resulting in a need for expert interpretation of their outputs. Furthermore, known model-based voltage and current (MBVI) systems are too difficult to deploy due to their inflexible connection and configuration abilities. Such a combination of high costs for typical model-based voltage and current (MBVI) systems, installation challenges, and the cost of expert interpretation render them too expensive for widespread adoption.
Therefore, in light of the foregoing drawbacks and problems encountered with known monitoring systems, there exists a need to overcome the aforementioned drawbacks in existing known systems and methods used for monitoring an operating condition of an electrical device.
In a published European patent application EP2725370A1; “A method for the diagnostics of electromechanical system based on impedance analysis”; inventors Ottewill (GB), Orman (PL) and Pinto (IN); applicant ABB Technology AG (DE); there is described a method for diagnosing a state of electromechanical systems in which electrical rotating machinery is used, wherein the method employs an analysis of impedance estimated from at least two currents and two voltages, measured during an operation of the electromechanical system. The method is potentially useful when performing condition monitoring of electric motors and generators. The method includes combining such information from both the voltage and current signals measurable at the motor terminals. Specifically, the measurements of voltage and current from two or more phases of a polyphase electrical machine are combined to estimate an impedance of the machine, wherein the impedance is a resistance to the flow of current that a circuit exhibits when a voltage is applied thereto.
In a published granted U.S. Pat. No. 6,308,140B1; “Motor condition and performance analyzer”; inventors Dowling (US), Lang (US), Desai (US) and Ricciuti (US); applicant Crane Nuclear Inc.; there is described a method of detecting broken rotor bars, wherein the method includes sensing currents applied to a motor, converting the sensed currents to digital data signals, and storing the digital data signals. A processor evaluates the digital data to provide a current signal spectrum at various selected sidebands. Sideband peaks having an amplitude within a predetermined value are identified, indicating potential motor problems or developing problems, including winding faults and broken rotor bars. Advanced signal processing techniques are used to evaluate further stored data and to provide trending information.
In a published PCT application WO2017/075254A1; “Interior magnet machine design with low core losses”; Campbell (US) and Dang (US); applicant Faraday & Future Inc. (US); there are described methods and apparatus for estimating and minimizing core losses in interior magnet machines. Such methods, for example, include creating, modifying, or receiving a finite element analysis (FEA) model to represent at least one portion of a motor in a computer system, placing at least one coil at a first location within a rotor iron or stator iron of the motor in the FEA model, calculating a time-domain flux density B of the at least one coil, converting the calculated flux density function to a frequency-domain spectrum, receiving material core loss parameters for at least some frequencies indicated by the frequency-domain spectrum, and determining a core loss of the at least one portion of an electric motor by a weighted combination of the material core loss parameters. Coils are placed, for example, manually by a user through a user interface, or are placed automatically.
Embodiments of the present disclosure seek to provide an improved system and method that can be used for monitoring an operating condition of an electrical device.
The present disclosure provides an electrical condition monitoring system that is connectable to an electrical device, and that is configured to monitor the electrical device when the electrical device is in operation, wherein:
The invention is of advantage in that removal of effects of signal distortion enables an operating state of the electrical device to be determined more accurately and reliably.
The present disclosure also provides a method of (for) using an electrical condition monitoring system that is connectable to an electrical device, and that is configured to monitor the electrical device when the electrical device is in operation, wherein the method includes:
The present disclosure also provides a computer program product comprising a non-transitory computer-readable storage medium having computer-readable instructions stored thereon, the computer-readable instructions being executable by a computerized device comprising processing hardware to execute the above method.
Embodiments of the present disclosure substantially eliminate, or at least partially address, the aforementioned drawbacks in existing known systems and methods used for monitoring an operating condition of an electrical device.
Additional aspects, advantages, features and objects of the present disclosure are made apparent from the drawings and the detailed description of the illustrative embodiments construed in conjunction with the appended claims that follow.
It will be appreciated that features of the present disclosure are susceptible to being combined in various combinations without departing from the scope of the present disclosure as defined by the appended claims.
The summary above, as well as the following detailed description of illustrative embodiments, is better understood when read in conjunction with the appended drawings. For the purpose of illustrating the present disclosure, exemplary constructions of the disclosure are shown in the drawings. However, the present disclosure is not limited to specific methods and instrumentalities disclosed herein. Moreover, those in the art will understand that the drawings are not to scale. Wherever possible, like elements have been indicated by identical numbers.
Embodiments of the present disclosure will now be described, by way of example only, with reference to the following diagrams wherein:
In the accompanying drawings, an underlined number is employed to represent an item over which the underlined number is positioned or an item to which the underlined number is adjacent. A non-underlined number relates to an item identified by a line linking the non-underlined number to the item. When a number is non-underlined and accompanied by an associated arrow, the non-underlined number is used to identify a general item at which the arrow is pointing.
The following detailed description illustrates embodiments of the present disclosure and ways in which they can be implemented. Although some modes of carrying out the present disclosure have been disclosed, those skilled in the art would recognize that other embodiments for carrying out or practising the present disclosure are also possible.
The present disclosure provides an electrical condition monitoring system that is connectable to an electrical device, and that is configured to monitor the electrical device when the electrical device is in operation, wherein:
The present system thus monitors the operating condition of the electrical device (e.g. a “health” of the electrical device) more accurately than existing known monitoring systems, over a period of time. The present system helps to avoid performing unnecessary maintenance works on the electrical device. The present system can analyze problems or faults that are developing in the electrical device before they become critical and provide suggestions to solve the problems or faults at a suitable time to avoid unexpected breakdowns of the electrical device.
Optionally, the electrical device is an electrical power-consuming device. Alternatively, the electrical device is an electrical power-generating device. Yet alternatively, the electrical device is capable of functioning as an electrical power-consuming device and an electrical power-generating device; for example. The electrical device, when in operation, is capable of functioning temporally alternately as an electrical power-consuming device and an electrical power-generating device, for example a turbine generator of a water reservoir used for energy storage, wherein power is generated by the turbine generator when water is release from the reservoir via the turbine generator, and wherein power is stored by the turbine generator when water if pumped upwards against gravity into the reservoir. Similar considerations pertain, mutatis mutandis where the reservoir is substituted for a compressed-air energy storage reservoir, a tidal water reservoir, and an inertial flywheel energy storage and generating device, a vehicle electrical motor and regenerative braking system, and so forth. By “vehicle” includes one or more of road vehicles, railway vehicles, ships, ferries, submarines, aircraft employing electric-motor propellers and so forth.
The present system thus manages to process at least one of current or voltage measurements, or both, having distortion and infer the sensed data from measured voltage or current signals, or both, despite the voltage and current signals are being distorted, for effectively monitoring the operating condition of the electrical device. The present system is more reliable than the aforesaid existing known monitoring systems in detection of problems or faults in the electrical device, thereby avoiding false alarms which can be very disruptive and costly to an industrial facility using the electrical device (e.g. an electrical motor, an electrical generator and so forth). Furthermore, the present system can be easily connected with the electrical system including such an electrical device, and may not require any configurational changes in the electrical device to be made when deploying the present system. The present system helps to save a maintenance cost of the electrical device to a larger extent than hitherto feasible by adopting a condition-based maintenance strategy for the maintenance of the electrical device.
The present system can reduce, for example minimize, energy wastages by promptly analyzing the problems in the electrical device in real-time and by quantifying benefits based on information associated with energy wastages occurred in the electrical device. The present system can generate a baseline profile (e.g. a capability of acceptance of the sensed data, a lifetime of the electrical device etc.) of the electrical device at a Factory Acceptance Test (FAT). The present system may perform a subsequent test (e.g. a Post-Installation Test (PIT) or a commissioning test) on the electrical device on commissioning thereof to confirm that no damage has occurred during delivery and installation processes of the electrical device. The present system may generate automatic written suggestions and graphical reports (e.g. analysis data) that can be edited by a user through the user interface (for example, a graphical user interface (GUI)) to add any specific information. The present system may provide the analysis data to a concerned person (e.g. to a maintenance team or a maintenance manager), whoever the user chooses.
The present system helps, for example, to monitor operating conditions of inaccessible electrical devices such as a submerged pump, a borehole pump, an in-tank pump, a cryogenic pump and a roof mounted fan, a remote unmanned hydroelectric generator, a remote unmanned wind-turbine, a remote unmanned tidal turbine, and remote unmanned wave-energy turbine, a heliostat turbine electrical generator, etc. The present system may help to monitor or test an operating condition of a given electrical device without requiring shutting down or requiring access to the inside of a cabinet of the electrical device in some circumstances. The present system can automatically analyze faults present in the electrical device and provide suggestions on a nature of the faults, a corrective action required and a timescale in which the corrective action should be taken based on measurements of:
The present system can also monitor the electrical devices while the electrical devices are operating at different speeds and different loads.
The electrical condition monitoring system can identify a wide range of failure modes and assess a degree to which the electrical device is suffering from the failures by generating the sensed data based on the voltage or current, or both, drawn by and/or supplied from the electrical device when in operation. In an embodiment, the electrical condition monitoring system is portably connected to the electrical device (e.g. an electrical motor and/or an electrical generator) to measure the voltage or the current, or both, that are being drawn to the electrical motor when in operation and/or generated from the electrical generator when in operation. The electrical condition monitoring system may periodically measure the voltage or the current, or both, supplied to and/or supplied from the electrical device for monitoring the operating condition of the electrical device.
The sensing arrangement may efficiently measure the voltage or the current, or both, supplied to and/or supplied from the electrical device that is operating at different speeds and different loads. The sensing arrangement may include current clamps to measure the current and dolphin connectors or magnetic connectors to measure the voltage to the electrical device. In an embodiment, the electrical condition monitoring system is electrically connected to the electrical device for temporally sensing the electrical supply to the electrical device using the sensing arrangement. The sensing arrangement may filter and process the electrical supply (e.g. the supply of voltage and/or the supply of current) to the electrical device and generate the corresponding sensed data.
The electrical condition monitoring system may generate the temporally-changing Fourier spectrum for at least one of the current or the voltage (optionally, or a calculated residual current or residual voltage), by performing a Fast Fourier Transform (FFT), or a similar type of signal spectral transformation, on at least one of the current and the voltage that is supplied to and/or supplied from the electrical device (optionally, or a calculated residual current or a calculated residual voltage). The analysis data may include a trend graph showing the changes in magnitudes and frequencies of the signal harmonic components based on the sensed data; optionally, the trend graph may also show non-harmonic signal components of a stochastic or random nature. The analysis data may include a report that indicates an operating condition of the electrical device. The analysis data may be generated in .ppt format, .doc format, XML® format etc. In an embodiment, the electrical condition monitoring system generates analysis data in a convenient format such as Microsoft Word® that provides a range of information including raw data (e.g. voltage and current waveform of the sensed signals, a diagnosis, a prognosis, information on the fault and recommended action). In an embodiment, the outputs may include a trend graph, information related to at least one of generic or specific faults detected in the electrical device, suggestions for faults detected in the electrical device, an operating condition of the electrical device in different time periods, energy wastage occurred due to the faults in the electrical device.
In an embodiment, the electrical condition monitoring system retrieves information pertaining to parameters of the electrical device (e.g. an electrical motor, an electrical power inverter, an electrical linear actuator, an electrical generator) from a database of the electrical condition monitoring system when the electrical device is previously tested by the electrical condition monitoring system. The parameters may include nominal voltage, nominal current and nominal speed that is taken from the electrical device. The electrical condition monitoring system may generate an automated report for the electrical device that includes at least one of: imbalance or misalignment, bearing problems, foundation looseness, transmission looseness or rubbing, motor rotor bars, motor stator problems, generator rotor bars, generator stator problems, electrical odd harmonics or other spectrum peak beyond normal expected values, stochastically varying information (e.g. arising from sporadic electrical insulation breakdown). The electrical condition monitoring system may generate a report that includes information on electrical parameters associated with the electrical device which include at least one of active power, reactive power and power factor, voltage balance, current balance, total harmonic distortion (THD) in the voltage or THD in the current.
In an embodiment, the electrical distortion may include for example distortion caused due to switching noise, non-linearity noise (e.g. magnetic saturation noise), spark discharge noise, present in the temporally-changing Fourier spectrum, distortion caused due to an operation of the electrical device, distortion caused due to variations in motor current or generator current caused by variations in the electrical device behaviours or distortion caused due to induced noise coming from other electrical devices/equipments.
In an embodiment, the trend graph is displayed either as the magnitude of a specific line frequency (e.g. the magnitude of 40 Hertz (Hz) peak) or the magnitude of a highest peak or local maximum in the family of locations corresponding to the same underlying phenomenon (e.g. a 10 Hz phenomenon occurring in the electrical device that is fed by, and/or generates, a supply with a frequency of 50 Hz may show up on the temporally-changing Fourier spectrum at either 40 Hz (50−10) or 60 Hz (50+10)). Sometimes it may also appear as sidebands on higher harmonics of line frequency, such as 90 Hz or 110 Hz (i.e. 2*50 Hz +/−10 Hz), at 140 Hz or 160 Hz (i.e. 3*50 Hz +/−10 Hz) or at 190 Hz or 210 Hz (i.e. 4*50 Hz +/−10 Hz)) etc. The trend graph may be plotted based on a highest peak in such a family of peaks. In an embodiment, when identifying the highest peak or local maximum in the family of possible peaks or local maxima, the magnitude of the highest peak may be determined based on either the absolute magnitude of the highest peak, or the magnitude relative to a height of that peak during the base-lining process or the magnitude relative to an average from a number of other electrical devices that is being used for comparison purposes. The magnitude of the highest peak may be calculated either on absolute numbers or a number of standard deviations.
In an embodiment, the data processing arrangement identifies the at least one local maximum corresponding to a phenomenon on the temporally-changing Fourier spectrum by applying well-established rules (e.g. 1× rotational speed is an indicator of imbalance or eccentricity and 2× supply frequency is an indicator of stator winding problems). In an embodiment, a location on the temporally-changing Fourier spectrum at which the at least one local maximum appears is mfl+/−nfp, where fl is a line frequency and fp is frequency of the phenomenon of interest and m and n are integers. In an embodiment, any particular phenomenon provides local maxima on either side of the line frequency and/or on harmonics of the line frequency. It will be appreciated that the data processing arrangement is operable to identify potential faults depending upon whether m and n are indicative of odd or even harmonics.
In an embodiment, the electrical condition monitoring system employs a change-point technique (e.g. a Cumulative Sum (CUSUM) algorithm) to identify changes in the parameters of the electrical device (e.g. a nominal voltage, a nominal current and a nominal speed). The voltage and current signals can be noisy from many reasons, for example due to the electrical switching noise, namely a form of signal distortion. However, it will be appreciated that unexpected magnetic saturation can also potentially give rise to signal distortion, that the electrical condition monitoring system is capable of detecting. The change-point technique is a technique that is typically used for monitoring change detection. The change-point technique may help the electrical condition monitoring system to detect 1% change in a mean of the voltage and current signals which is varying by +/−50% or so. Optionally, for example, the change-point technique may help the electrical condition monitoring system to detect up to a 5% change in a mean of the voltage and current signals which is varying in a range of +/−10% by +/−50%.
The electrical condition monitoring system may generate the trend graph (i.e. the analysis data) by employing a curve fitting process on the trend graph which is derived from the aforementioned temporally-changing Fourier spectrum. The curve fitting process (e.g. a spline curve fit, but not limited thereto) refers to a process of constructing a curve, or mathematical function, that has the best fit to a series of data points, optionally subject to constraints. In an embodiment, the trend graph is extrapolated using a curve fitting technique (e.g. linear, parabolic or exponential) for future purpose.
The electrical condition monitoring system may determine the change-point technique and the curve fitting process to be employed based on normalized sets of data associated with different loads and different speeds. The process of determining the change-point technique and the curve fitting process is very complex; moreover, it is also very sensitive to the load. For example, if the load of the electrical device is changed, the current supplied to the electrical device can be changed without changing the supply of voltage. Thus, parameter values of the linear models are sensitive to changes in the load and speed, as aforementioned. As the parameter values of the linear models are used to identify the problems inside the electrical device (e.g. electrical motor and driven equipment or the generator and driver equipment) or in the electrical supply, the electrical condition monitoring system has less tendency to generate false alerts (for example, avoids generating false alerts), if the change is due to the load or speed of the electrical device. The electrical condition monitoring system may compare the parameter values of the linear model or stored model with parameter values of a real-time model to determine the expected parameter values at a particular load and speed. The electrical condition monitoring system compares the parameter values by creating a set of individual cells and each cell corresponds to a particular range of load and speed (e.g. frequency is 40 Hz to 45 Hz and the supply of current is in a range of 100 Amperes to 110 Amperes). During a baseline process which is generally carried out after initial installation and commissioning of the electrical device, each time a set of sensed data is taken, their results are updated in their corresponding cell. In an embodiment, the electrical condition monitoring system determines characteristics of a relationship between the parameter values and the load and speed of the electrical device. The electrical condition monitoring system uses the characteristics of the relationship to assess an expected parameter value for a measured load and speed of the electrical device, without having to have previously populated data in this particular cell.
According to an embodiment, the data processing arrangement computes parameter differences between the real-time model and the one or more stored models, and to generate a predictive model describing the parameter difference to generate the information indicative of an operating condition of the electrical device. According to another embodiment, the data processing arrangement generates from the predictive model one or more alerts relating to: one or more energy utilization trends of the electrical device; Carbon Dioxide (equivalent) generation being caused by the electrical device; Carbon Dioxide (equivalent) generation being saved by the electrical device; and maintenance, replacement or repair of the electrical device.
According to yet another embodiment, the user interface (e.g. a graphical user interface (GUI)) is configured for user entry of user defined parameters for generating a range of multivariate cells in a database of the electrical condition monitoring system and the data processing arrangement is configured to update the sensed data of the electrical device in their respective multivariate cells, wherein the multivariate cells store the temporally-changing Fourier spectrum associated with the sensed data and the one or more stored models associated with the electrical device.
The sensed data may include both the parameter values and the magnitudes of the local maxima and local minima on the temporally-changing Fourier spectrum. A large set of sensed data are taken and each sensed data is updated to the values in their respective multivariate cells to generate an updated average and a standard deviation for the sensed data, which is further updated with each new sensed data; by “large” is meant in a range of 2 to 100000 separate readings, more optionally in a range of 50 to 500 readings. The electrical condition monitoring system may compare the new sensed data with the mean values in the appropriate multivariate cell once the baseline process is completed. The electrical condition monitoring system may quantify the parameter difference in terms of the number of standard deviations in the multivariate cells. In order to update the new sensed data occurring at a combination of the load and the speed and that has not previously had a multivariate cell, the electrical condition monitoring system may use a process of interpolation and extrapolation from the multivariate cells to identify where the sensed data is stored in the multivariate cells and where there is no previous data. In an embodiment, the electrical condition monitoring system characterizes a way that the parameters vary across the multivariate cells (e.g. imagine a 3-dimensional surface (“surface” in claims below) drawn over a matrix of voltage and current) and estimates the expected values in the multivariate cell on the basis of obtaining actual sensed data in only a small number of the multivariate cells.
According to yet another embodiment, the data processing arrangement is configured to, namely “operates to” when in operation, update pre-existing analysis data in their respective multivariate cells with the analysis data to generate an updated average and a standard deviation for the sensed data. According to yet another embodiment, the data processing arrangement is configured to, namely “operates to” when in operation, decimate data points in the sensed data based on calculated frequency to determine a number of relevant data points for generating the temporally-changing Fourier spectrum. In an embodiment, the temporally-changing Fourier spectrum are plotted to line up perfectly when plotted in terms of orders and when different samples occurred at widely different supplied frequencies; “perfectly” in such a context means that the points on each spectrum will line up with one another, when individual spectra associated with variable speed machines, created at different supply frequencies and therefore different equipment speeds, have to be either stretched or compressed along the frequency axis in order for corresponding features to line up.
According to yet another embodiment, the electrical device includes at least one of: a synchronous 1-phase electrical motor, a synchronous multi-phase electrical motor, a synchronous 3-phase electrical motor, a switched reluctance motor, a switched stepper motor, a D.C. electrical motor, a compressor, a pump, a fan, a conveyor, an agitator, a blender, a blower, a mixer, a Permanent Magnet (PM) motor, a PM generator, a power transformer, an air conditioner, a ventilator, an oven, a solar panel, an electromagnetic linear actuator, an electromagnetic linear generator, a light source for illumination, a synchronous 1-phase electrical generator, a synchronous multi-phase electrical generator, a synchronous 3-phase electrical generator, a D.C. electrical generator, a Permanent Magnet (PM) generator. However, it will be appreciated that embodiments of the present disclosure can be used to monitor other types of electrical devices, for example industrial microwave ovens (e.g. to detect electrode erosion or vacuum degradation), resistive heating element ovens (e.g. to detect heating element progressive oxidation or heating element micro-cracking), solar cell generating performance (e.g. condensation, element ageing, solar cell cracking). According to yet another embodiment, the user interface is configured for user entry of machine parameters describing the electrical device, wherein the machine parameters include at least one of: generic nominal operating voltage, generic nominal operating current, generic nominal supply frequency, generic nominal rotational speed, number of vanes on an impeller of the electrical device, bearing type codes of the electrical device or belt drive dimensions of the electrical device.
According to yet another embodiment, the user interface is configured for user entry of machine parameters describing analysis type to be executed by the data processing arrangement, wherein the analysis type allows user-setting of temporal resolution and a sensing data sample length. According to yet another embodiment, the user interface allows user setting of voltage and/or current readings from the electric device at a sample rate of up to at least 2 MegaHertz (2 MHz) to 10 MegaHertz (10 MHz).
Optionally, sample rates of 100 KHz are employed, decimated into a selectable frequency range of 2.5 kHz to 25 kHz. However, it will be appreciated that the sample rate accommodated is optionally up to 1 MHz. Yet more optionally, the sampling rate is in a range of 2.5 kHz to 125 MHz, for example selectable within this range.
According to yet another embodiment, the user interface allows user-selection for implementing D-Q phase and A-B phase transformations, computed in the data processing arrangement, of the sensed data, wherein the D-Q phase transformation transfers three-phase stator and rotor quantities into a single rotating reference frame to eliminate an effect of time-varying inductances, wherein the A-B phase transformation refers to a mathematical transformation that is implemented to simplify the analysis of three-phase stator and rotor. According to yet another embodiment, the data processing arrangement is configured to create a linear model of one or more relationships between voltage and current signals obtained from the electrical device; for example, various spline fits to sample data can be employed for the one or more relationships.
According to yet another embodiment, the data processing arrangement is configured to compute an analysis of at least one of a numerical algorithm for Sub-Space State-Space System Identification (N4SID), a Multivariable Output Error State Space (MOESP) algorithm, a Past Outputs Multivariable Output Error State-Space (PO-MOESP) algorithm or Canonical Variate Analysis (CVA) to generate the real-time model based on the sensed data. The subspace identification methods (SIMs), such as past outputs multivariable output-error state space (PO-MOESP), numerical algorithms for subspace state-space system identification (N4SID), and Canonical Variate Analysis, generate a real-time model (e.g. a state-space model) based on the sensed data. Multivariable Output Error State Space (MOESP) and Numerical algorithms for Subspace State Space System Identification (N4SID) algorithms are two known subspace identification techniques. The numerical algorithm for Subspace and State-Space identification is typically faster, when implemented correctly, than prediction error methods, because the numerical algorithm for Subspace and State-Space identification are basically non-iterative, exploiting basic algorithms from numerical linear algebra, such as LQ-decomposition (transpose of the QR decomposition) and the singular value decomposition to generate the real-time model based on the sensed data. In an embodiment, the data processing arrangement is configured to compute an analysis of at least one of the numerical algorithm for Sub-Space State-Space System Identification (N4SID), the Multivariable Output Error State Space (MOESP) algorithm, the Past Outputs Multivariable Output Error State-Space (PO-MOESP) algorithm or the Canonical Variate Analysis (CVA) to generate the stored model or linear model based on the voltage and the current signals obtained from the electrical device. However, use of aforesaid known methods of performing data processing is not known in the context of the present disclosure.
According to yet another embodiment, the data processing arrangement is operable to:
In an embodiment, the electrical condition monitoring system identifies each point on the temporally-changing Fourier spectrum and travels right and left on the temporally-changing Fourier spectrum, for example as shown in a graphical form (with reference to an abscissa frequency axis) to a given user via the aforementioned user interface. The electrical condition monitoring system determines a value that increases towards the point or a value that decreases from the point. In an embodiment, if the values on each side of the point are lower, then the point is identified as a local maximum. In another embodiment, if the values on each side of the point is higher, then the point is referred as a local minimum. The electrical condition monitoring system may identify all the local maxima and local minima on the temporally-changing Fourier spectrum and divide the entire temporally-changing Fourier spectrum into local maxima, which abut one another at the local minima.
The present disclosure also provides a method for (of) using an electrical condition monitoring system that is connectable to an electrical device, and that is configured to monitor the electrical device when the electrical device is in operation, wherein the method includes:
According to an embodiment, the method includes using the data processing arrangement to compute parameter differences between the real-time model and the one or more stored models, and to generate a predictive model describing the parameter difference to generate the information indicative of an operating condition of the electrical device.
According to another embodiment, the method includes using the data processing arrangement to generate from the predictive model one or more alerts relating to:
According to yet another embodiment, the method includes configuring the user interface for user entry of user defined parameters for generating a range of multivariate cells in a database of the electrical condition monitoring system and using the data processing arrangement to update the sensed data of the electrical device in their respective multivariate cells, wherein the multivariate cells store the temporally-changing Fourier spectrum associated with the sensed data and the one or more stored models associated with the electrical device.
According to yet another embodiment, the method includes using the data processing arrangement to update pre-existing analysis data in their respective multivariate cells with the analysis data to generate an updated average and a standard deviation for the sensed data.
According to yet another embodiment, the method includes using the data processing arrangement to decimate data points in the sensed data based on calculated frequency to determine a number of relevant data points for generating the temporally-changing Fourier spectrum.
According to yet another embodiment, the electrical device includes at least one of: a synchronous 1-phase electrical motor, a synchronous mufti-phase electrical motor, a synchronous 3-phase electrical motor, a switched reluctance motor, a switched stepper motor, a D.C. electrical motor, a compressor, a pump, a fan, a conveyor, an agitator, a blender, a blower, a mixer, a Permanent Magnet (PM) motor, a PM generator, a synchronous 1-phase electrical generator, a synchronous mufti-phase electrical generator, a synchronous 3-phase electrical generator, a D.C. electrical generator, a power transformer, an air conditioner, a ventilator, an oven or a solar panel. However, the electrical device optionally includes static devices, for example ovens, solar panels, and similar.
According to yet another embodiment, the method includes configuring the user interface for user entry of machine parameters describing the electrical device, wherein the machine parameters include at least one of: generic nominal operating voltage, generic nominal operating current, generic nominal supply frequency, generic nominal rotational speed, number of vanes on an impeller of the electrical device, bearing type codes of the electrical device or belt drive dimensions of the electrical device.
According to yet another embodiment, the method includes configuring the user interface for user entry of machine parameters describing analysis type to be executed by the data processing arrangement, wherein the analysis type allows user-setting of temporal resolution and a sensing data sample length.
According to yet another embodiment, the method includes using the user interface for user setting of voltage and/or current readings from the electric device at a sample rate of at least 2 MegaHertz (2 MHz) to 10 MegaHertz (10 MHz). Optionally, sample rates employed are in excess of 10 MHz.
Optionally, sample rates of 100 KHz are employed, decimated into a selectable frequency range of 2.5 kHz to 25 kHz. However, it will be appreciated that the sample rate accommodated is up to 1 MHz. Yet more optionally, the sampling rate is in a range of 2.5 kHz to 125 MHz, for example selectable within this range.
According to yet another embodiment, the method includes using the user interface to allow user-selection for implementing D-Q phase and A-B phase transformation, computed in the data processing arrangement, of the sensed data, wherein the D-Q phase transformation transfers three-phase stator and rotor quantities into a single rotating reference frame to eliminate an effect of time-varying inductances, wherein the A-B phase transformation refers to a mathematical transformation that is implemented to simplify the analysis of three-phase stator and rotor.
According to yet another embodiment, the method includes configuring the data processing arrangement to create a linear model of one or more relationships between voltage and current signals obtained from the electrical device
According to yet another embodiment, the method includes configuring the data processing arrangement to compute an analysis of at least one of a numerical algorithm for Sub-Space State-Space System Identification (N4SID), a Multivariable Output Error State Space (MOESP) algorithm, a Past Outputs Multivariable Output Error State-Space (PO-MOESP) algorithm or Canonical Variate Analysis (CVA) to generate the real-time model based on the sensed data.
According to yet another embodiment, the method includes operating the data processing arrangement to:
The advantages of the present method are thus identical to those disclosed above in connection with the present system and the embodiments listed above in connection with the present system apply mutatis mutandis to the present method.
The present disclosure also provides a computer program product comprising a non-transitory computer-readable storage medium having computer-readable instructions stored thereon, the computer-readable instructions being executable by a computerized device comprising processing hardware to execute the aforesaid method.
The advantages of the present computer program product are thus identical to those disclosed above in connection with the present system and the embodiments listed above in connection with the present system apply mutatis mutandis to the computer program product.
Embodiments of the present disclosure thus monitor the operating condition of the electrical device (e.g. a health of the electrical device) more accurately than the existing monitoring systems, over a period of time. Embodiments of the present disclosure may help to avoid performing unnecessary maintenance works on the electrical device; such avoidance is especially valuable when the electrical device is at a remote location that is costly and/or difficult to access, for example remote pumps, remote motor-actuated barriers, remote hydroelectric generators, remote wind turbines, remote tidal turbines, remote geothermal turbines, remote ocean wave turbines and so forth. Embodiments of the present disclosure thus manage processing of current and voltage measurements having distortion and infer the sensed data from measured voltage and current signals, despite the voltage and current signals are being distorted, for effectively monitoring the operating condition of the electrical device. Embodiments of the present disclosure may be easily connected with the electrical system and may not require any configuration changes in the electrical device to deploy it. Embodiments of the present disclosure help to save the maintenance cost of the electrical device to a larger extent by adopting a condition-based maintenance strategy for the maintenance of the electrical device.
Embodiments of the present disclosure concern a type of technology that is capable of reducing Carbon Dioxide (CO2) emissions, as a result of improving operation of electrical devices, for example electrical machines, at higher efficiency and avoiding unnecessary maintenance.
Although sensing and monitoring of a single electrical device is described in the foregoing, it will be appreciated that the electrical condition monitoring system 100 is capable of being arranged, namely configured, so that the sensing arrangement 106 senses currents and/or voltages of connection points that are coupled to a plurality of electrical devices; for example, the sensing arrangement (for example, implemented as voltage probes and/or clamp-on current sensors) 106 is coupled at a main electrical junction box that feeds, or receives power from, an array of electrical devices. In such an embodiment, the filtering module 204, the spectrum analysis module 206, the real-time model generation module 208 and the model comparison module 210 are arranged to employ one or more models corresponding to an aggregate of a plurality of electrical devices. However, the electrical devices are potentially operating in mutually different states (for example, some motors or generators being stationary whereas other of the motors or generators being operable at mutually different power magnitudes or rotation rates) that makes the system 100 more complex in its operation. However, such a manner of operation is potentially cost-effective, because installing the sensor arrangement 106 may be a costly personnel-intensive (namely, labor-intensive) activity, for example where remote telemetry is also required.
In other words, the system 100 is able to connect to multiple motors, multiple generators, or a combination of one or more motors and one or more generators, and diagnose faults or other phenomena occurring in any one of the motors and/or generators. The sensing arrangement 106, for example, is conveniently coupled at a switchboard associated with the multiple motors, multiple generators, or combination thereof.
Employing the system 100 to monitor and diagnose concurrently multiple electrical devices is more complex, because signals generated from the sensor arrangement 106 in respect of the multiple electrical devices is effectively aggregated, namely mutually merged into a single signal. The system 100 is able to identify individual parameters in a meaningful way from such an aggregate signal. Where the different motors, alternatively different generators, or both, are driving different types of equipment, the profile of each type of equipment is different, and at least some of the fault frequencies pertaining to the equipment will be different. By building up a profile of each motor in turn, mutatis mutandis each generator in turn, the system 100 is able to build an expected profile in total of all the motors running at that time, mutatis mutandis generators running at that time, and then compare it with the actual overall profile (namely spectrum). Where the motors, mutatis mutandis the generators, are driving, or being driven by, identical types of equipment, the fault frequencies will be identical for each item of equipment. The system 100 is able to identify a nature of the fault, but not which motor, mutatis mutandis generator, is responsible for the fault. Optionally, narrowing down a given motor, mutatis mutandis a given generator, with the fault is achieved in the system 100 by correlating a presence or absence at a given time of the fault signal with which motors, mutatis mutandis which generators, are or are not running at that given time.
The system 100 is capable of being operated to determine and record a set of baselines corresponding to a range of combinations of operational parameters that pertain to one or more motors, and/or one or more generators. Such operation parameters concern, for example, a relationship between load and rotation speed of a given electrical device, for example manifest as a relationship between current and rotational frequency of the electrical device; in the system 100, subsequent monitoring and diagnostics are implemented by comparing subsequently monitored measurements or derived data against such baselines.
As aforementioned, the real-time model generation module 208 generates a real-time model of the electrical device based on the analysis data. This real-time model corresponding to baseline information is conveniently stored in multivariate cells within computing hardware, that are filled by actual measurements, in other words sensor data, during a baselining process when initially characterizing a given electrical device, for example an electrical motor or an electrical generator. However, in a case where not every relevant combination of operational parameters has been encountered during such a baselining process, data appropriate to unencountered combinations of operational parameters are optionally created by a suitable mathematical method; this mathematical method is optionally implemented as follows, for example:
It will be appreciated that by interpretation of “model parameters”, these may be parameters of the mathematical models used in the system 100 or may be parameters that describe physical quantities of a given motor or generator, such as rotor resistance, rotor reactance, stator resistance, stator reactance.
It will further be appreciated that the electrical condition monitoring system 100, namely operating as a form of diagnostic system, identifies distortions in a sensed current signal that has a current waveform that have not been caused by distortions of a corresponding voltage signal that has a voltage waveform, to create a residual current signal, which is then used for executing further analysis. Temporal changes in a given electrical device, for example a load applied to a motor or output from a generator, causes a change in the overall current drawn or supplied by the electrical device, without causing significant change in the voltage signal, so such load or output changes result in apparent changes in the residual current. These apparent changes, in earlier known diagnostic system, potentially lead to spurious identification of faults. Embodiments of the present disclosure by capturing a range of “typical” readings and their associated spectra and model parameters, and storing associated data representative of the spectra and model parameters in individual cells of a matrix defining a relationship between load and rotation speed (i.e. rotation frequency), or output and rotation speed (i.e. rotation frequency). Optionally, embodiments of the present disclosure capture multiple sets of data for each combination of load and speed for a motor, mutatis mutandis output and speed for a generator, and average them or aggregate them as a part of a baselining process employed within the electrical condition monitoring system 100. The diagnosis of problems is then based in the system 100 on a comparison of a reading at any particular time against the baseline established for that combination of load and speed for the given motor, mutatis mutandis output and speed for the given generator. The greater the number of cells employed in the aforesaid matrix, the closer the comparison between the monitoring and baseline readings, and therefore the more valid the diagnosis provided by the system 100 when in operation. However, the larger the number of cells in the matrix, the lower the likelihood of all cells being encountered during the baselining process, and hence the need to fill in these empty cells in some way, for example by way of using extrapolation or interpolation as aforementioned.
As aforementioned, the mathematical methods employed within the system 100 for identifying distortions on a current waveform that have not been caused by distortions on a corresponding voltage waveform involves creating mathematical models of the relationship between voltage and current measurements. Moreover, as aforementioned, these models utilize a plurality of model parameters, whose values potentially temporally change each time a measurement is taken using the system 100. By relating such model parameters to a physical model of an electric motor, mutatis mutandis an electrical generator, the values of the physical parameters, such as rotor resistance, stator resistance and so forth, may be assessed, in particular by identifying changes in the model parameters that may be related to underlying physical changes.
Detection of such changes, for example, allows detection to be made of early stages of electrical insulation breakdown, for example rotor and/or stator electrical winding breakdown; such an early detection is achieved by measuring the capacitance and resistance of the insulation at a range of frequencies, by analyzing a relationship between voltage and current at a range of different frequency components by using harmonics already present in the voltage and current signals, for example isolated by performing aforesaid Fourier analysis of measured current and voltage signal captured by the system 100. In contradistinction, traditional “simple” methods of assessing integrity of insulation involves measuring a resistance from a conductor abutting the insulation to Earth. This resistance is normally very high, namely in an order of MegaOhms. If the insulation has already failed, the insulation will have a resistance that will be very much lower, possibly in an order of a few Ohms only; often, the insulation is a polymer plastics material that can be decomposed to Carbon that is an electrical conductor. Such a traditional technique is not good at identifying gradual deterioration, and so is not very helpful at giving predictive warnings of insulation breakdown; moreover, these traditional techniques are generally only possible when the given motor, mutatis mutandis the given generator, is disconnected from the electricity supply powering the motor, mutatis mutandis is disconnected from the electricity network to which the given generator is providing power. Thus, it will be appreciated that the system 100 is capable of providing real-time continuous monitoring of insulation integrity of electrical motors, mutatis mutandis electrical generators.
A more sensitive technique employable in the system 100 for detecting earlier a deterioration of a given insulation, and hence giving an earlier warning of impending problems, relies on a characteristic that a given insulation gradually changes chemically over a period of time, for example due to chemical oxidation processing and/or metal migration processes. These chemical changes result in a change in its electrical behavior of the given insulation; principally, there is a reduction in a capacitance exhibited by the given insulation and a change, for example an increase or decrease in its resistive behavior, which results in a change in the phase angle (or “tan delta”) between the voltage applied to the insulation and a (very small, for example nano-Amperes or micro-Amperes) leakage current flowing through it, when in use. The system 100 is configurable to operate to measure the phase angle (or “tan delta”).
Computations implemented in the system 100 are based a behavior of a perfect insulator being conceptually thought of as an array of little capacitive elements connected in matrix manner in series and in parallel. On account of there being all capacitive elements, there is no DC current if a DC voltage is applied to the perfect insulator; however, when an alternating AC voltage is applied to the perfect insulator, there will be a capacitive current flowing that will increase as an alternating frequency increases (basis for computation: I=C.dv/dt, wherein I is a current, C is capacitance, and dv/dt is a temporal rate of voltage change applied to the capacitance C). As a pure capacitive current, this will flow at an orthogonal angle of 90 degrees (90°) to the applied voltage signal, with current leading voltage (engineering hint for understanding: “CIVIL”—in a capacitor, C, I before V; in an inductor, L, V before I; wherein V is an alternating sinusoidal voltage and I is an alternating sinusoidal current flowing). Over a period of time, when the perfect insulation is in use, the chemical change in the perfect insulation results in some of these little capacitive elements changing in nature to small resistive elements. While only a few of the capacitive elements have changed to being resistive elements, there is still no path for DC through to Earth through the aged perfect insulation, but the AC current will now have a different phase angle, as some of the AC current is now flowing through the resistive elements. By taking a plurality of readings at different frequencies, the values of the capacitive elements and resistive elements and phase angle can be calculated. Suh a technique for assessing early stage insulation problems is already used by specialist bench testers, for testing motors in a workshop, or when still in situ but de-energized (namely, disconnected from an electrical supply). The specialist bench testers connect across between conductors of the motor and Earth. Conventionally, a series of test signals is applied at a range of frequencies from which the “health” (namely, a degree of ageing) of the insulation can be assessed. In contradistinction to such conventional approaches, the system 100 is capable of assessing the health (namely, “ageing”) of the insulation during normal operation by utilizing measured harmonic components in an alternating voltage applied to the insulation, to assess the capacitive elements and resistive elements of the insulation, namely determinable from the aforesaid “tan delta” angle between current and voltage signals associated with the insulation.
Optionally, the system 100 measure transients when a given motor or generator coupled to the system 100 is started up into operation, for example rotationally accelerated from standstill to a rotating operational state. The aforesaid model used to describe the given motor or given generator (namely, as provided by the real-time model generation module 208 and utilized by the model comparison module 210 of the system 100) beneficially includes parameters in the baseline that describe transient start-up, optionally shut-down, characteristics of the given motor or given generator. Such a manner of operation of the system 100 enables potential faults in switchgear to be identified that deliver or receive power to or from the given motor or generator, respectively.
The aforementioned multivariate cells are not limited to merely current and voltage measurements, whether obtained directly from measurements or generated by interpolation or extrapolation of data from neighboring multivariate cells, but optionally are additionally based on at least one of: temperature measurements, humidity measurements, vibration measurements. For example, temperature measurements allow coefficients of resistivity of winding elements to be taken into account, and humidity measurements enable insulation to be modelled more accurately (because polarized water molecules are able to modulate conduction characteristics of aged and compromised insulation materials).
Modifications to embodiments of the present disclosure described in the foregoing are possible without departing from the scope of the present disclosure as defined by the accompanying claims. Expressions such as “including”, “comprising”, “incorporating”, “have”, “is” used to describe and claim the present disclosure are intended to be construed in a non-exclusive manner, namely allowing for items, components or elements not explicitly described also to be present. Expressions such as “may” are to be construed as meaning “optionally”, “optional”, “is capable of” or “are capable of”, as appropriate. Reference to the singular is also to be construed to relate to the plural.
Number | Date | Country | Kind |
---|---|---|---|
1802619.5 | Feb 2018 | GB | national |
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/EP2019/053964 | 2/18/2019 | WO | 00 |