Example embodiments of the present disclosure generally relate to the field of switchgears, and more specifically, to a method and an apparatus for training a model for predicting a temperature rise of a heating element in a switchgear, a method and an apparatus for determining a temperature rise of a heating element in a switchgear, a computer readable storage medium, and a computer program product.
A switchgear is an electrical device in a power system that can achieve on/off, control or protection function during the process of power generation, transmission, distribution, and power conversion. As the use time of the switchgear increases, the conductive connection (such as contacts) in a main circuit of the switchgear may increase its resistance due to mechanical vibration, wear and manufacturing process defects, thereby causing the temperature of the contacts to rise or even burning the contacts. It may cause a safety incident. Moreover, inadequate switchgear on-site installation and component damages are also the main factors that contribute to an anomaly temperature rise τn the switchgear.
A conventional way of detecting such an anomaly temperature rise τs to use a portable infrared device to detect the temperature of the contacts in the switchgear via a viewing window. Then, the detected temperature is compared with a temperature limit value of the contacts under a rating current to judge whether there is the anomaly temperature rise τn the contacts of the switchgear. However, such a detecting method is not always accurate and is inefficient. Recently, an online temperature monitoring system is applied in the switchgear. However, such an online temperature monitoring system mainly focuses on temperature collecting without effective anomaly temperature rise detection.
Thus, there is a need for an improved solution for temperature rise monitoring of the switchgear.
In view of the foregoing problems, example embodiments of the present disclosure propose a real time regression model for predicting temperature rise of a heating element in the switchgear and a method for detecting anomaly temperature rise of the switchgear. Embodiments of the present disclosure are adaptive to different operating conditions and setups of the switchgear. Embodiments of the present disclosure can reduce the outage time of the switchgear and improve the reliability of the switchgear.
In a first aspect, example embodiments of the present disclosure provide a method for training a model for predicting a temperature rise of a heating element in a switchgear, comprising: obtaining a model for predicting the temperature rise, the model comprising a plurality of inputs, an output, and several parameters to be determined; obtaining n+1 sets of physical quantities in relation to the heating element, each set of physical quantities being collected at a corresponding one of n+1 time points spaced apart from each other by a time step in a normal operation state of the heating element, each set of physical quantities comprising a current and an actual temperature of the heating element and an ambient temperature; converting the actual temperature in each set of physical quantities into an actual temperature rise based on the corresponding ambient temperature; and training the model with the current, the actual temperature rise and the ambient temperature to determine the several parameters.
The proposed model is a modified transient heat balance equation with undetermined thermal and energy parameters. Time constant as a function of temperature rise and dynamic current loads are both taken into consideration when building the model. The model trained by the method according to the first aspect may be used to predict transient or steady state temperature rise under any current load. This model is designed with strong physical constraints and few parameters which could prevent over fitting on most switchgear workloads. With these embodiments, the outage time of the switchgear can be reduced and the reliability of the switchgear can be improved.
In some embodiments, training the model with the current, the actual temperature rise and the ambient temperature comprises: for each time step, creating an equation by using the current, the actual temperature rise and the ambient temperature corresponding to a starting time point of the time step as the inputs of the model and using the actual temperature rise corresponding to an ending time point of the time step as the output of the model, to obtain n equations; and solving the n equations to determine the several parameters. With these embodiments, through plugging the n+1 sets of physical quantities into the model, the n equations can be obtained and solved to determine the several parameters.
In some embodiments, each of the n equations is created by using a discrete approximate scheme or Runge-Kutta iterative method. With these embodiments, the n equations can be obtained reliably with the discrete approximate scheme or Runge-Kutta iterative method.
In some embodiments, the n equations are solved by a least square method. With these embodiments, the n equations can be solved by the least square method to determine the several parameters.
In some embodiments, the heating element comprises at least one of a busbar contact, a circuit breaker upper contact finger, a circuit breaker lower contact finger, and a cable contact. With these embodiments, the temperature rise of these contacts may be predicted reliably with the trained model.
In a second aspect, example embodiments of the present disclosure provide a method for determining a temperature rise of a heating element in a switchgear, comprising: predicting the temperature rise of the heating element in the switchgear with a model trained by using the method according to the first aspect. With these embodiments, the temperature rise of the heating element in the switchgear may be predicted reliably with the trained model.
In some embodiments, predicting the temperature rise of the heating element with the model comprises: obtaining a real-time current of the heating element and an ambient temperature collected in a first time step; and predicting the temperature rise of the heating element in the first time step through inputting the real-time current, the ambient temperature, and the predicted temperature rise τn a second time step prior to the first time step into the model. With these embodiments, the temperature rise τn the present time step can be predicted reliably by using the current, the ambient temperature, and the predicted temperature rise τn the previous time step.
In some embodiments, the method further comprises: obtaining a real-time temperature of the heating element collected in the first time step; converting the real-time temperature of the heating element into a real-time temperature rise based on the ambient temperature; and triggering an alarm in response to a difference between the real-time temperature rise and the predicted temperature rise exceeding a predetermined threshold. With these embodiments, the anomaly temperature rise τn the switchgear can be detected timely and thus corresponding maintenance may be performed on the switchgear.
In some embodiments, predicting the temperature rise of the heating element comprises predicting at least one of a transient temperature rise or a steady state temperature rise of the heating element. With these embodiments, the transient or steady state temperature rise of the heating element may be predicted under any current load.
In a third aspect, example embodiments of the present disclosure provide an apparatus for training a model for predicting a temperature rise of a heating element in a switchgear, comprising: at least one processor; and at least one memory comprising instructions stored thereon which, when executed by the at least one processor, cause the at least one processor to perform acts comprising: obtaining a model for predicting the temperature rise, the model comprising a plurality of inputs, an output, and several parameters to be determined; obtaining n+1 sets of physical quantities in relation to the heating element, each set of physical quantities being collected at a corresponding one of n+1 time points spaced apart from each other by a time step in a normal operation state of the heating element, each set of physical quantities comprising a current and an actual temperature of the heating element and an ambient temperature; converting the actual temperature in each set of physical quantities into an actual temperature rise based on the corresponding ambient temperature; and training the model with the current, the actual temperature rise and the ambient temperature to determine the several parameters.
In some embodiments, training the model with the current, the actual temperature rise and the ambient temperature comprises: for each time step, creating an equation by using the current, the actual temperature rise and the ambient temperature corresponding to a starting time point of the time step as the inputs of the model and using the actual temperature rise corresponding to an ending time point of the time step as the output of the model, to obtain n equations; and solving the n equations to determine the several parameters.
In some embodiments, each of the n equations is created by using a discrete approximate scheme or Runge-Kutta iterative method.
In some embodiments, the n equations are solved by a least square method.
In some embodiments, the heating element comprises at least one of a busbar contact, a circuit breaker upper contact finger, a circuit breaker lower contact finger, and a cable contact.
In a fourth aspect, example embodiments of the present disclosure provide an apparatus for determining a temperature rise of a heating element in a switchgear, comprising: at least one processor; and at least one memory comprising instructions stored thereon which, when executed by the at least one processor, cause the at least one processor to perform acts comprising: predicting the temperature rise of the heating element in the switchgear with a model trained by using the apparatus according to the third aspect.
In some embodiments, predicting the temperature rise of the heating element with the model comprises: obtaining a real-time current of the heating element and an ambient temperature collected in a first time step; and predicting the temperature rise of the heating element in the first time step through inputting the real-time current, the ambient temperature, and the predicted temperature rise τn a second time step prior to the first time step into the model.
In some embodiments, the acts further comprises: obtaining a real-time temperature of the heating element collected in the first time step; converting the real-time temperature of the heating element into a real-time temperature rise based on the ambient temperature; and triggering an alarm in response to a difference between the real-time temperature rise and the predicted temperature rise exceeding a predetermined threshold.
In some embodiments, predicting the temperature rise of the heating element comprises predicting at least one of a transient temperature rise or a steady state temperature rise of the heating element.
In some embodiments, the at least one processor comprises at least one of a local processor or a remote processor.
In some embodiments, wherein the remote processor comprises a cloud computing node.
In a fifth aspect, example embodiments of the present disclosure provide a computer readable storage medium having instructions stored thereon, the instructions, when executed by at least one processor, cause the at least one processor to perform the method according to the first and second aspect.
In a sixth aspect, example embodiments of the present disclosure provide a computer program product comprising instructions when executed by at least one processor, causing the at least one processor to perform the method according to the first and second aspect.
Drawings described herein are provided to further explain the present disclosure and constitute a part of the present disclosure. The example embodiments of the disclosure and the explanation thereof are used to explain the present disclosure, rather than to limit the present disclosure improperly.
Throughout the drawings, the same or similar reference symbols are used to indicate the same or similar elements.
Principles of the present disclosure will now be described with reference to several example embodiments shown in the drawings. Though example embodiments of the present disclosure are illustrated in the drawings, it is to be understood that the embodiments are described only to facilitate those skilled in the art in better understanding and thereby achieving the present disclosure, rather than to limit the scope of the disclosure in any manner.
The term “comprises” or “includes” and its variants are to be read as open terms that mean “includes, but is not limited to.” The term “or” is to be read as “and/or” unless the context clearly indicates otherwise. The term “based on” is to be read as “based at least in part on.” The term “being operable to” is to mean a function, an action, a motion or a state can be achieved by an operation induced by a user or an external mechanism. The term “one embodiment” and “an embodiment” are to be read as “at least one embodiment.” The term “another embodiment” is to be read as “at least one other embodiment.” The terms “first,” “second,” and the like may refer to different or same objects. Other definitions, explicit and implicit, may be included below. A definition of a term is consistent throughout the description unless the context clearly indicates otherwise.
As discussed above, the conventional way of detecting the anomaly temperature rise τs not always accurate and is inefficient, and the conventional online temperature monitoring system mainly focuses on temperature collecting without effective anomaly temperature rise detection. According to embodiments of the present disclosure, a real time regression model for predicting temperature rise of a heating element in the switchgear and a method for detecting anomaly temperature rise of the switchgear are proposed; and the time regression model and the anomaly temperature rise detecting method are adaptive to different operating conditions and setups of the switchgear. The above idea may be implemented in various manners, as will be described in detail in the following paragraphs.
Hereinafter, the principles of the present disclosure will be described in detail with reference to
Referring to
In order to predict the temperature rise of the heating element in the switchgear 100, embodiments of the present disclosure propose a model hypothesis. Since a part of the heat generated by the heating element during operation of the switchgear 100 is dissipated and the other part of the generated heat is absorbed by the heating element, it is supposed that a heat balance equation is satisfied for the heating element in the switchgear 100 during the time period dt, represented by an equation (1):
pdt=K
τ
Aτdt+cmdτ (1),
In engineering, the forms of the heat dissipation include heat conduction, heat convection and heat radiation. For the sake of simplicity, the above three forms of the heat dissipation are combined, that is, KτAT is used to represent the overall heat dissipation process. The selection of KT and A typically depends on the specific application scenarios. KT is simplified on the basis of the KT expression obtained by a similarity theory, thereby obtaining an equation (2):
K
τ=λτγ (2),
In order to solve the equation (1), two example approaches are provided in the present disclosure.
It is assumed that an integral calculation of the equation (1) is performed in a short time period from t0 to t1, with an initial condition of τ=τ0 when t=t0. Since the time period from t0 to t1 is short, the temperature rise of the heating element is small. Thus, the heat dissipating coefficient KT may be regarded as a constant during this period and calculated by an equation (3):
K
T=λτoγ (3),
Then, an equation (4) is obtained by integration of the equation (1) with the initial condition of τ=τ0 when t=t0:
A time constant T is defined by
Since the parameters c, m, A are undetermined, it is assumed that the time constant T is represented by an equation (5):
The total heating power p of the heating element is related to the current, electrical resistance and other loss of the heating element. Thus, in some embodiments, it is assumed that total heating power p is represented by an equation (6):
p=η*I
2*[1+0.00381*(τ+Tamb)]*ρ0 (6),
It is to be understood that in other embodiments, the total heating power p of the heating element may be represented by other equations. The scope of the present disclosure is not intended to be limited in this respect.
Then, through plugging the equations (3), (5), and (6) into the equation (4), an equation (7) is obtained:
The equation (7) may be rewritten as an equation (8) through letting
and b=β:
The equation (8) is the result of the integral calculation of the equation (1) in the time period from t0 to t1 with the initial condition of τ=τ0 when t=t0.
In some embodiments, the constraints of the equation (8) may be set as α ∈ (0,1000], b ∈ [−1000,1000], γ∈ [0,0.45], d ∈ [0,1]. The constraints are approximate parameter ranges derived from multiple times of experiments. In other embodiments, the constraints of the equation (8) may be set to be other ranges. The scope of the present disclosure is not intended to be limited in this respect.
The equation (8) may be rewritten into a more general form, represented by equation (9):
The constraints of the equation (9) may be the same as those of the equation (8), i.e., α ∈ (0,1000], b ∈ [−1000,1000], γ ∈[0,0.45], d ∈ [0,1].
Since the parameters a, b, γ, and d are undetermined parameters, the above model needs to be trained in advance to determine these parameters.
As shown in
In order to determine the parameters a, b, γ, and d, the model needs to be trained. To this end, the method 200 further includes: at 220, obtaining n+1 sets of physical quantities in relation to the heating element. Each set of physical quantities is collected at a corresponding one of n+1 time points t0, t1, . . . , tn spaced apart from each other by a time step in a normal operation state of the heating element. Each set of physical quantities includes a current and an actual temperature of the heating element and an ambient temperature. Thus, the n+1 sets of physical quantities include the current I0, I1, . . . , In, the actual temperature Tact0, Tact1, . . . , Tactn of the heating element, and the ambient temperature Tamb0, Tamb1, . . . , Tambn.
The method further includes: at 230, converting the actual temperature Tact0, Tact1, . . . , Tactn in each set of physical quantities into an actual temperature rise Tact0, Tact1, . . . , τactn based on the corresponding ambient temperature Tamb0, Tamb1, . . . , Tambn. Specifically, the actual temperature rise τact0=Tact0−Tamb0, the actual temperature rise τact1=Tact1−Tamb1, . . . , and the actual temperature rise τactn=Tactn−Tambn.
Then, at 240, the model may be trained with the current I0, I1, . . . , In, the actual temperature rise τact0, Tact1, . . . , Tactn, and the ambient temperature Tamb0, Tamb1, . . . , Tambn, to determine the parameters a, b, γ, and d.
In some embodiments, as shown in
In some embodiments, at 2401, each of the n equations is created by using a discrete approximate scheme. In some embodiments, at 2402, the n equations are solved by a least square method.
After the undetermined parameters a, b, γ, and d have been determined, the trained model represented by the equation (9) can be used to determine the temperature rise of the heating element step by step.
As shown in
Specifically, at 4102, during the time period from t0 to t1, the temperature rise τ1 can be calculated by using the equation (9) based on the real-time current I0, the temperature rise τ0 and the ambient temperature Tamb0; during the time period from t1 to t2, the temperature rise τ2 can be calculated by using the equation (9) based on the real-time current I1, the temperature rise τ1 and the ambient temperature Tamb1; and during the time period from tn−1 to tn, the temperature rise τn can be calculated by using the equation (9) based on the real-time current In−1, the temperature rise τ−1 and the ambient temperature Tambn−1. The temperature rises τ1, τ2, and τn can be represented by equations (10), (11), and (12) respectively:
The model trained by the method 200 can be used not only to predict the transient temperature rise of the heating element at any time point, but also to predict the steady state temperature rise of the heating element at any current. According to the equation (1), when the temperature rise of the heating element becomes stable at any current, dτ will equal to 0. Thus, the steady state temperature rise τW at any current I may be represented by an equation (13):
According to embodiments of the approach I, time constant as a function of temperature rise and dynamic current loads are both taken into consideration when building the model. The model trained by the method 200 may be used to predict transient or steady state temperature rise under any current load. The model is designed with strong physical constraints and few parameters which could prevent over fitting on most switchgear workloads. With these embodiments, the outage time of the switchgear can be reduced and the reliability of the switchgear can be improved.
Referring back to
Testing currents are listed in the following Table 1. In some embodiments, the currents of Flag 3 to Flag 16 except Flags 6, 7, 11, and 12 as shown in the Table 1 are selected as a training set, and the currents of Flag 17 to Flag 22 are selected as a validation set. There are 43% of the data in the training set and 57% in the validation set.
Based on the detected currents, the time intervals, and the ambient temperature, the prediction results of the temperature rise are as shown in
Table 2 illustrates prediction errors of the temperature rise of the busbar contact, the circuit breaker upper contact finger, the circuit breaker lower contact finger, and the cable contact of phases A, B, and C on the panels 610 and 620. As can be seen from table 2, the absolute prediction error is less than 2.90° C. in all cases. The overall mean prediction errors are 0.94° C. and 0.52° C. for the panel 610 and the panel 620 respectively. These errors are acceptable in practical applications.
Approach II provides a different manner to solve the equation (1). Specifically, through plugging the equation (2) into the equation (1), an equation (14) is obtained:
pdt=λAτ
1+γ
dt+cmdτ (14).
Then, through plugging the equation (6) into the equation (14), an equation (15) is obtained:
The equation (15) may be rewritten as an equation (16) through letting
g, f, and γ are undetermined parameters.
The model represented by the equation (16) may be trained by the method 200 as shown in
After the undetermined parameters g, f, and γ have been determined, the trained model represented by the equation (16) can be used to determine the temperature rise of the heating element step by step, for example by using the method 400 as shown in
Moreover, the model represented by the equation (16) can be used not only to predict the transient temperature rise of the heating element at any time point, but also to predict the steady state temperature rise of the heating element at any current. According to the equation (16), when the temperature rise of the heating element becomes stable at any current,
will equal to 0. Thus, the steady state temperature rise Tw at any current I may be represented by an equation (17):
In some embodiments of the present disclosure, as shown in
According to embodiments of the present disclosure, the methods 200 and 400 are not limited to be implemented by a local processor, but may be implemented by a remote processor. For example, the methods 200 and 400 may be implemented at a cloud computing node. Specifically, the data collected at site can be transferred to the cloud computing node and processed at the cloud computing node. Then, the calculation results can be returned to local devices or systems from the cloud computing node. In this way, there is no need to provide additional calculation devices at site.
In some embodiments, each of the methods 200 and 400 may be implemented by one processor, for example a MCU or a cloud computing node. In other embodiments, each of the methods 200 and 400 may be implemented by several processors. For example, some acts of the methods 200 and 400 may be implemented by the MCU locally, and other acts of the methods 200 and 400 may be implemented the cloud computing node remotely. The scope of the present disclosure is not intended to be limited in this respect.
In some embodiments of the present disclosure, a computer readable medium is provided. The computer readable medium has instructions stored thereon, and the instructions, when executed on at least one processor, may cause at least one processor to perform the methods 200 and 400 as described in the preceding paragraphs, and details will be omitted hereinafter.
In some embodiments of the present disclosure, a computer program product is provided. The computer program product comprises instructions which, when executed on at least one processor, may cause at least one processor to perform the methods 200 and 400 as described in the preceding paragraphs, and details will be omitted hereinafter.
Generally, various embodiments of the present disclosure may be implemented in hardware or special purpose circuits, software, logic or any combination thereof. Some aspects may be implemented in hardware, while other aspects may be implemented in firmware or software which may be executed by a controller, microprocessor or other computing device. While some aspects of embodiments of the present disclosure are illustrated and described as block diagrams, flowcharts, or using some other pictorial representation, it will be appreciated that the blocks, apparatus, systems, techniques or methods described herein may be implemented in, as non-limiting examples, hardware, software, firmware, special purpose circuits or logic, general purpose hardware or controller or other computing devices, or some combination thereof.
Program code for carrying out methods of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the program codes, when executed by the processor or controller, cause the functions/operations specified in the flowcharts and/or block diagrams to be implemented. The program code may execute entirely on a machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.
The above program code may be embodied on a machine readable medium, which may be any tangible medium that may contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine readable medium may be a machine readable signal medium or a machine readable storage medium. A machine readable medium may include but not limited to an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the machine readable storage medium would include an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
Further, while operations are depicted in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. In certain circumstances, multitasking and parallel processing may be advantageous. Likewise, while several specific implementation details are contained in the above discussions, these should not be construed as limitations on the scope of the present disclosure, but rather as descriptions of features that may be specific to particular embodiments. Certain features that are described in the context of separate embodiments may also be implemented in combination in a single embodiment. On the other hand, various features that are described in the context of a single embodiment may also be implemented in multiple embodiments separately or in any suitable sub-combination.
Although the subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims.
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/CN2021/072682 | 1/19/2021 | WO |