Patent Application
20160333902
The present disclosure relates generally to a system for determining displacement of a hydraulic cylinder and, more particularly, improved accuracy in determination of hydraulic cylinder displacement.
Machines such as haul trucks, dozers, motor graders, wheel loaders, and other types of heavy equipment are used to perform a variety of tasks. Many of these machines use hydraulic actuators to move various components of the machines. The components may include, for example, a boom, a stick, a bucket, a blade, a shovel, a drill, an impact tool, or any other member of the machine. These machines can be manually controlled, autonomously controlled, semi-autonomously controlled, or remotely controlled. In any of these situations, control of the machines may be dependent on positional data provided by different sensors mounted onboard the machines. Positional data may include data that is indicative of the amount of hydraulic cylinder displacement for each of the hydraulic actuators connected to the various components or members of the machine. Hydraulic cylinder displacement is the amount by which a distal end of a piston rod of the hydraulic actuator has either extended from the hydraulic cylinder or retracted into the cylinder from a fixed reference point. The proximal end of the piston rod is connected to a piston that slides within the inner bore of the hydraulic cylinder as hydraulic fluid is forced into either a head end of the cylinder on one side of the piston or a rod end of the cylinder on the opposite, piston rod side of the piston. The piston rod extends from the cylinder as pressurized hydraulic fluid enters the head end of the cylinder and exits the rod end of the cylinder. The piston rod retracts into the cylinder as pressurized hydraulic fluid enters the rod end of the cylinder and exits the head end of the cylinder.
Control of the machines based upon this positional data requires the information provided by the sensors to be accurate at all times. And even though most machine sensor systems are calibrated when first commissioned, vibrations, collisions, interference, range issues, damage, environmental contamination, and normal wear and tear on the machine during operation can reduce the quality of information provided by the sensors.
An exemplary system for determining the position of a piston within a hydraulic cylinder is disclosed in U.S. Pat. No. 6,817,252 (“the '252 patent”) of Wiklund et al. that issued on Nov. 16, 2004. The system of the '252 patent discloses a measuring device that determines the position of the piston as a function of a hydraulic fluid flow. The measuring device includes a bi-directional flow sensor positioned inline with the hydraulic fluid flow and a piston position module. The flow sensor is configured to produce a flow rate signal that is indicative of the flow rate of the hydraulic fluid flow into or out of cavities on the head end side or the rod end side of the piston in the hydraulic cylinder. A piston position module is configured to receive the flow rate signal from the flow sensor and produce a piston position signal that is indicative of the position of the piston within the hydraulic actuator as a function of the flow rate signal.
Although the system of the '252 patent may be useful in determining the displacement of a hydraulic cylinder, there is still room for improvement. In particular, the system of the '252 patent may be subject to errors caused by damage to the flow sensor that measures the hydraulic fluid flow rate. The hardware necessary to measure the hydraulic fluid flow rate into and out of the cylinder is also expensive. Moreover, the flow sensor may introduce undesirable effects on the operation of the hydraulic cylinder since it measures flow rate by introducing a discontinuity within the hydraulic fluid flow and measuring a pressure drop across the discontinuity.
The systems and methods of the present disclosure are directed toward solving one or more of the problems set forth above and/or other problems of the prior art.
In one aspect, the present disclosure is directed to a hydraulic cylinder displacement measuring system. The system may include a sensor configured to generate a signal indicative of a hydraulic fluid flow velocity into or out of the hydraulic cylinder, and a hydraulic fluid flow model configured to scale the hydraulic fluid flow velocity to a predicted rate of change of cylinder displacement for the hydraulic cylinder. The system may further include a Kalman filter configured to integrate the predicted rate of change of cylinder displacement during a predict phase in order to determine a series of predicted cylinder displacement values over time during operation of the hydraulic cylinder. The Kalman filter may be further configured to periodically receive values of cylinder displacement measured optically during operation of the cylinder, and to periodically update the predicted cylinder displacement values using the optically measured cylinder displacement values during an update phase. The Kalman filter may be still further configured to fuse the optically measured cylinder displacement values with the predicted cylinder displacement values over time, and output an estimated time series of cylinder displacement values based on the fused values during operation of the hydraulic cylinder.
In another aspect, the present disclosure is directed to a method of determining a hydraulic cylinder displacement. The method may include generating a signal indicative of a hydraulic fluid flow velocity into or out of the hydraulic cylinder, and mapping the hydraulic fluid flow velocity to a predicted rate of change of cylinder displacement for the hydraulic cylinder. The method may further include integrating the predicted rate of change of cylinder displacement during a predict phase of a Kalman filter in order to determine a series of predicted cylinder displacement values over time during operation of the hydraulic cylinder, periodically receiving values of cylinder displacement measured optically during operation of the cylinder, periodically updating the predicted cylinder displacement values using the optically measured cylinder displacement values during an update phase of the Kalman filter, fusing the optically measured cylinder displacement values with the predicted cylinder displacement values over time, and outputting an estimated time series of cylinder displacement values based on the fused values during operation of the hydraulic cylinder.
In yet another aspect, the present disclosure is directed to a non-transitory computer-readable medium including instructions, that, when executed by one or more processors, cause the one or more processors to perform a method of determining a hydraulic cylinder displacement. The method may include generating a signal indicative of a hydraulic fluid flow velocity into or out of the hydraulic cylinder, and mapping the hydraulic fluid flow velocity to a predicted rate of change of cylinder displacement for the hydraulic cylinder. The method may further include integrating the predicted rate of change of cylinder displacement during a predict phase of a Kalman filter in order to determine a series of predicted cylinder displacement values over time during operation of the hydraulic cylinder, periodically receiving values of cylinder displacement measured optically during operation of the cylinder, periodically updating the predicted cylinder displacement values using the optically measured cylinder displacement values during an update phase of the Kalman filter, fusing the optically measured cylinder displacement values with the predicted cylinder displacement values over time, and outputting an estimated time series of cylinder displacement values based on the fused values during operation of the hydraulic cylinder.
The system 100 may comprise one or more processors, processing modules, software subroutines, hardware, circuits, sensors, firmware, and/or various combinations of these and other components and devices configured to provide accurate indications of the hydraulic cylinder displacement for one or more hydraulic actuators while they are in use. Throughout the description that follows, reference to various “modules” will be understood to encompass any of the various components and devices referenced above. For example, a module may embody one or more processors programmed with one or more software subroutines or subprograms in order to perform the desired function encompassed by the module. The various modules may be combined together in various ways, or provided separately. The processor(s) may include any known processing device(s), such as a microprocessor from the Pentium™ or Xeon™ family manufactured by Intel™ the Turion™ family manufactured by AMD™, or any other type of processor. In various alternative embodiments, the modules may be encompassed on one or more Application Specific Integrated Circuits (ASIC), or on Field-Programmable Gate Arrays (FPGA) designed to be configured by a customer or a designer after the FPGA is manufactured. Memory may be provided in conjunction with the modules, and may include one or more storage devices configured to store information used by each processing module to perform certain functions related to disclosed embodiments. The storage devices may include volatile or non-volatile, magnetic, semiconductor, tape, optical, removable, non-removable, or another type of non-transitory computer-readable medium. The storage devices may store programs, subroutines, and/or other information, such as information related to processing data received from one or more sensors.
The hydraulic actuators may include actuators configured to move various components of a mobile machine relative to each other, or actuators associated with other machinery or equipment. Mobile machines that use hydraulic actuators may include, for example, a haul truck, a dozer, a motor grader, or a loader. A mobile machine may be a manned machine configured to traverse a worksite under the control of an operator, a self-directed machine configured to autonomously traverse the worksite, or a hybrid machine configured to perform some functions autonomously and other functions under the control of a local or remote operator. One of ordinary skill in the art will recognize that the same principles and implementations disclosed herein for accurate measurement of hydraulic cylinder displacement may also be applied on stationary equipment having hydraulic actuators used to move various components or members of the machine while the machine remains in a fixed position. The machine may include an engine or other power source functional to provide a power output for various operations of the machine. The power produced by the engine may be directed to various electrical and mechanical systems of the machine such as, for example, a transmission for transmitting the power to one or more traction devices, to a hydraulic system for driving one or more of the hydraulic actuators mounted on the machine, to an implement system for actuating one or more work tools, or to any other electrical or mechanical system configured to perform an operation associated with the machine.
As shown in
A number of different factors including temperatures, viscosities of the hydraulic fluid, impurities in the fluid, and movement of the hydraulic actuator can affect the measurement of hydraulic fluid flow velocity by introducing high frequency noise or fluctuations in the output from module 110. Therefore, the system 100 may include a module 112 configured to receive the flow velocity signals from the module 110, with the module 112 being configured to filter high frequency noise in the flow velocity signals using, for example, a Finite Impulse Response Low Pass filter (FIR LP). The resulting signal output from the module 112 may be indicative of a smoother, longer term trend in the hydraulic fluid flow velocity x entering and exiting the hydraulic cylinder cavities.
Alternative implementations of this disclosure may include determining the hydraulic fluid flow velocity with module 110 indirectly, rather than by actually measuring fluid flow rates with flow sensors. For example, in applications where an electrical solenoid controls the amount of hydraulic fluid provided to the hydraulic actuator, the module 110 may be configured to produce a signal indicative of the amount of current flowing to the electrical solenoid. This current signal may correlate indirectly through a functional mapping to the hydraulic fluid flow velocity entering and exiting the hydraulic cylinder.
The system 100 may further include modules configured to receive the smoothed output from the module 112, indicative of a longer term trend in the fluid flow velocity x, and scale that smoothed fluid flow velocity x to a predicted rate of change of hydraulic cylinder displacement over time. As shown in
The linear regression module 116 may be configured to receive data during an initial calibration phase that enables the module 116 to determine the most appropriate analysis and solution for accurately correlating the hydraulic fluid flow velocity to a predicted rate of change of cylinder displacement for a particular hydraulic actuator. In some alternative implementations, the module 116 may determine that a linear regression analysis is not the most appropriate analysis for correlation, and may instead use other algorithms for establishing a best fit model. The exemplary system 100 shown in
The system 100 may also include a Kalman filter, which is illustrated in the embodiment of
The system 100 may further include a clipping module such as clip floor/ceiling module 120 configured to receive the time series of predicted values for hydraulic cylinder displacement resulting from the integration performed by the Kalman filter predict phase module 118. The clip floor/ceiling module 120 may be configured to set boundaries on the values of predicted cylinder displacement that are designed to keep the predicted values within the realm of believability. For example, the clip floor/ceiling module 120 may be configured to recognize that a value for actual cylinder displacement cannot go beyond an amount corresponding to the piston of the hydraulic actuator reaching a stop at the head end or the rod end of the hydraulic cylinder. The system 100 may include another module 121 configured to receive the data that falls within the range established by the clip floor/ceiling module 120. The module 121 may be configured to receive only the realistic values of cylinder displacement that have been predicted by the Kalman filter predict phase module 118 during operation of the hydraulic actuator.
As mentioned above, the Kalman filter's general operation involves two phases, a “predict” phase and a measurement or “update” phase. In the predict phase, the value estimate from the previous timestep in the time series of predicted cylinder displacement values is used to generate an a priori value estimate. In the update phase, the a priori estimate calculated in the predict phase is combined with an estimate of the accuracy (e.g., the variance) of the a priori estimate, and a current measurement value to produce a refined a posteriori estimate. In the exemplary embodiment shown in
In certain exemplary embodiments, the controller 250 may be configured to include the Kalman filter predict phase module and the Kalman filter update phase module in the one or more processors 251. The Kalman filter modules may utilize the following equations in the calculations performed by each module. For the predict phase module, the Kalman filter may be configured to utilize the following generic equations:
{circumflex over (x)}
k
−
=F
k-1
{circumflex over (x)}
k-1
+G
k-1
u
k-1 (1)
P
k
−
=F
k-1
P
K-1
+
F
k-1
T
+Q
k-1 (2)
For the measurement or “update” phase, the Kalman filter may be configured to utilize the following generic equations:
K
k
=P
k-1
−
H
T(HkPk-1−HkT+Rk)−1 (3)
{circumflex over (x)}
k
+
={circumflex over (x)}
k
−
+K
k(yk−Hk{circumflex over (x)}k−) (4)
P
k
+=(I−KkHk)Pk− (5)
In the above equations, {circumflex over (x)}k− may be the a priori state estimate of the predicted hydraulic cylinder displacement that is calculated based on the a posteriori state estimate ({circumflex over (x)}k-1) from an immediately preceding time step. The Kalman filter predict phase module 118 discussed above may be configured to calculate the time series of predicted hydraulic cylinder displacement values by integrating the rate of change of cylinder displacement over time. F, G, and H may be appropriate state transition matrices.
In the measurement or “update” phase, the Kalman filter may calculate the Kalman gain Kk utilizing equation (3), in which P is an error covariance matrix and R is a matrix setting forth the variance associated with the different state variables. For example, the values in the R matrix may specify the uncertainty associated with the measurement of hydraulic fluid flow velocity. In the measurement or update phase, the Kalman filter may also obtain an independent measure of the hydraulic cylinder displacement, such as by optically measuring the cylinder displacement, and set the independent measurement as yk. Utilizing the a priori estimate {circumflex over (x)}k− from the “predict” phase, the updated measurement yk, and the Kalman gain Kk, the controller 250 may calculate the a posteriori state estimate {circumflex over (x)}k+ utilizing equation (4). The result output by the Kalman filter is the best estimate of the hydraulic cylinder displacement values during operation of the hydraulic actuator.
Following the predict phase 301, the Kalman filter 300 may periodically implement the update phase 302 to calculate the a posteriori state estimate {circumflex over (x)}k+ utilizing, for example, equation (4). For example, the Kalman filter 300 may calculate the a posteriori state estimate xk+ using the a priori estimate {circumflex over (x)}k− from the predict phase 301, the update measurement yk obtained from the optically measured cylinder displacement 330, and the Kalman gain Kk. The Kalman filter 300 may receive as input in the update phase 302, the optically measured cylinder displacement 330, and set this input as the update measurement yk in equation (4). Because the certainty of the measurements is often difficult to measure precisely, it is common to discuss the filter's behavior in terms of gain. The Kalman gain Kk is a function of the relative certainty of the measurements and current state estimate, and can be “tuned” to achieve particular performance. With a high gain, the filter places more weight on the measurements, and thus follows them more closely. With a low gain, the filter follows the model predictions more closely, smoothing out noise but decreasing the responsiveness. At the extremes, a gain of one causes the filter to ignore the state estimate entirely, while a gain of zero causes the measurements to be ignored.
In determining the uncertainty associated with the measurements of hydraulic fluid flow velocity using the Kalman filter 300, the controller 250 may also consider the operation state of the machine on which the hydraulic actuator is mounted. For example, if the machine is older, and is operating in an environment where flow sensors are likely to have experienced a lot of shocks and other harsh conditions, the controller 250 may change certain parameters of the Kalman filter 300, such as the Kalman gain Kk, the error covariance matrix P, or the state variable variance matrix R, to reflect the accuracy or confidence in certain input parameters.
The disclosed hydraulic cylinder displacement measurement system may be applicable to any machine where accurate detection of the amount of hydraulic cylinder displacement is desired. The disclosed system may provide for improved prediction and measurement of the hydraulic cylinder displacement by utilizing a Kalman filter to fuse predicted cylinder displacement values based on hydraulic fluid flow velocities and optically measured cylinder displacement values. The physics-based calculations used to establish a correlation between a hydraulic fluid flow velocity into or out of the hydraulic cylinder and time series values for predicted cylinder displacement may result in the predicted cylinder displacement values drifting away from the true values of cylinder displacement. This drift may be caused by errors in the flow velocities measured by the flow sensor, changes to the internal configuration of the hydraulic cylinder caused by wear or impurities in the hydraulic fluid, and integration errors introduced during integration of the predicted rate of change of cylinder displacements obtained from a flow model for the hydraulic cylinder. The Kalman filter helps to reduce errors by periodically updating the predicted values of cylinder displacement with measured values that may be obtained, for example, by optical imaging of the hydraulic cylinder during operation. The Kalman filter may average the predictions of cylinder displacement with each new measured value using a weighted average. The purpose of the weighting applied to the predicted and/or measured values is that the values with better (i.e., smaller) estimated uncertainty are “trusted” more. The weights may be calculated from the covariance, which is a measure of the estimated uncertainty of the predictions of the cylinder displacements. The result of the weighted average is a new estimate of cylinder displacement that lies between the predicted and measured values, and has a better estimated uncertainty than either alone. This process may be repeated every time step, with the new estimate and its covariance informing the prediction used in the following iteration. The Kalman filter works recursively and requires only the last best estimate, rather than the entire history of predicted cylinder displacements to calculate a new cylinder displacement.
As shown in
At Step: 420, a flow sensor may be used to measure hydraulic fluid flow velocity into or out of the hydraulic cylinder. In various alternative implementations, the hydraulic fluid flow velocity may be determined indirectly from other correlated measurements, such as the amount of voltage or current being applied to a solenoid that controls the fluid flow into or out of the hydraulic cylinder. The predicted cylinder displacement values may be calculated from the hydraulic fluid flow velocity taken together with other information specific to the particular hydraulic cylinder, such as the internal dimensions of the cylinder, the current temperature, the pressure of the hydraulic fluid, and the viscosity of the hydraulic fluid.
At Step: 430, a model may be used to scale a predicted rate of change of cylinder displacement vs. the hydraulic fluid flow velocity. Information obtained during the calibration and modeling step (Step: 410) may be used to verify the physics-based calculations that correlate the predicted rate of change of cylinder displacement to the hydraulic fluid flow velocity. The model may consist of a mapping between hydraulic fluid flow velocities into or out of a particular hydraulic cylinder and the predicted hydraulic cylinder displacement over time.
At Step: 440, a controller may cause a Kalman filter to integrate the rate of change of cylinder displacement during the Kalman filter predict phase to arrive at a time series of predicted cylinder displacement values. As discussed above, the Kalman filter module of the controller may be configured to fuse the predicted values for cylinder displacement with updated measured values. At Step: 450, the controller may cause imaging devices to periodically optically measure cylinder displacement during operation of the cylinder using various imaging techniques discussed above, such as LIDAR and RADAR. At Step: 460, the controller may receive input from the imaging devices and periodically provide the optically measured cylinder displacement values to the Kalman filter during the Kalman filter update phase.
At Step: 470, the controller may cause feedback of the updated cylinder displacement data and fusing of the updated data with the predicted data that was determined during the Kalman filter predict phase. As discussed above, the Kalman filter may average the predicted cylinder displacement values with the updated cylinder displacement values using a weighted average. In some implementations, for example, the estimated uncertainty associated with hydraulic fluid flow velocities may be high as a result of environmental conditions, the age of the flow sensors, or the quality of the hydraulic fluid. In this situation, the Kalman filter may give more weight to the optically measured cylinder displacement values. However, in other situations where the conditions during operation make it difficult for an optical imaging device to obtain an accurate indication of cylinder displacement, the Kalman filter may give more weight to the predicted cylinder displacement values based on hydraulic fluid flow velocities into or out of the cylinder.
At Step: 480, the controller may set boundaries on the Kalman filter output in order to maintain the estimates of cylinder displacement within the realm of believability. A clip floor/ceiling module may set the boundaries based on specific information for a particular hydraulic actuator, such as the physical limits for extension or retraction of the piston rod based on known stops within the hydraulic cylinder, or other physical limits on movement of the hydraulic actuator imposed by the configuration of the machine on which the hydraulic actuator is mounted. At Step: 490, the output from the fusion of updated measured values of cylinder displacement over time and predicted values for the cylinder displacement provide a best estimate of cylinder displacement output during operation of the hydraulic actuator.
It will be apparent to those skilled in the art that various modifications and variations can be made to the disclosed system and method for improved accuracy in determination of hydraulic cylinder displacement. Other embodiments will be apparent to those skilled in the art from consideration of the specification and practice of the disclosed system. It is intended that the specification and examples be considered as exemplary only, with a true scope being indicated by the following claims and their equivalents.
