A hydraulic excavator optimization method based on engine output power prediction and controller parameter setting

Abstract

A hydraulic excavator optimization method based on engine output power prediction and controller parameter setting belongs to the technical field of engineering machinery excavator control. This method establishes an engine output power time series prediction model and a hydraulic pump load torque multivariate mapping model, indirectly predicts the engine output torque, and directly calculates the hydraulic pump load torque. The gray wolf algorithm is used to optimize the PID controller for parameter setting. When the excavator is working, the controller predicts the engine output torque using the engine output power time series prediction model, calculates the hydraulic pump load torque using the hydraulic pump load torque multivariate mapping model, imports the difference between the engine output torque and the hydraulic pump load torque into the PID controller optimized by the gray wolf algorithm, adjusts the controller output current, changes the hydraulic pump output flow, and then adjusts the hydraulic pump output torque to be dynamically equal to the engine output torque, thereby achieving real-time stable control of the engine speed.

Description

Technical Field

[0001] The invention belongs to the technical field of hydraulic excavator control of engineering machinery, and in particular is a hydraulic excavator optimization method based on engine output power prediction and controller parameter setting. Background Art

[0002] Hydraulic excavator operation has related problems such as complex working conditions, nonlinear parameter constraints, and lack of effective adaptive control methods, which usually cause drastic changes in speed or speed drops during hydraulic excavator operation, indirectly increasing the engine fuel consumption rate and increasing the actual fuel consumption of the operation. At the same time, when the engine speed changes drastically, the hydraulic excavator body usually shakes, seriously affecting the driver's operating comfort. The engine speed drop will also make the hydraulic excavator unable to complete the specified load operation, reducing the operating efficiency of the hydraulic excavator. In order to improve the operating comfort and efficiency of the hydraulic excavator, it is imperative to study the stable control of the engine speed of the hydraulic excavator.

[0003] For stabilizing the engine speed of hydraulic excavators, controlling the dynamic matching of engine output power and hydraulic pump load demand power is the fundamental method. At present, engine control can be optimized in two directions. One is to use the actual output speed of the engine as system feedback and use the speed controller to directly stabilize the speed. The second is to use the load torque of the hydraulic pump as system feedback and use the torque controller to indirectly stabilize the speed. Its essence belongs to the power matching controller of the engine and the hydraulic pump. Obviously, the indirect stable control of the speed using the torque controller is fundamentally stable speed control. At present, many scholars at home and abroad have conducted relevant research, but most of the current torque controllers are negative feedback controllers, that is, delay control. Because the operation of the hydraulic system and the engine requires a certain response time during the actual operation of the hydraulic excavator, it is necessary to study the real-time control of the engine speed stability.

[0004] Therefore, the present invention proposes a hydraulic excavator optimization method based on engine output power prediction and controller parameter setting in response to the demand for real-time stable control of engine speed of the hydraulic excavator. Summary of the invention

[0005] In order to solve the above problems, the present invention proposes an optimization method for a hydraulic excavator. The main optimization methods of the present invention are:

[0006] Collect engine output power data of hydraulic excavator operation. Train engine output power time series prediction model. Use flow data and outlet pressure data of hydraulic dual pump operation to calculate hydraulic pump load torque value, and establish a multivariate mapping model with boom large / small chamber pressure, dipper arm large / small chamber pressure, bucket large / small chamber pressure, hydraulic dual pump outlet pressure as input and load torque as output. Design a PID controller with the difference between engine output torque and hydraulic pump load torque as input and control current change as output. Use the gray wolf algorithm to set parameters of torque control PID controller so that the proportional, integral, and differential coefficients of PID controller can be adjusted in real time according to load changes.

[0007] The engine output power time series prediction model, the hydraulic pump load torque multivariate mapping model, and the parameter-tuned PID controller are embedded in the hydraulic excavator controller. During operation, the controller collects the historical time series data of the engine output power, predicts the engine output power through the engine output power time series prediction model, and calculates the engine predicted torque. At the same time, the controller calculates the real-time hydraulic pump load torque based on the real-time data of the boom large / small chamber pressure, the arm large / small chamber pressure, the bucket large / small chamber pressure, and the hydraulic dual pump outlet pressure, using the load torque multivariate mapping model. The torque data indirectly predicted by the engine and the real-time load torque data of the hydraulic pump are subtracted and imported into the PID controller after parameter tuning optimization, and the change in the control current is output to change the output flow of the hydraulic dual pump, thereby achieving the effect of dynamic matching of the real-time torque of the engine and the hydraulic pump, so as to stabilize the engine speed in real time / ahead of time.

[0008] Control principle Figure 1 shown.

[0009] In order to achieve the above-mentioned purpose, the technical solution of the present invention is:

[0010] A hydraulic excavator optimization method based on engine output power prediction and controller parameter setting, the steps are as follows:

[0011] Step 1: Hydraulic excavator modeling data collection

[0012] The data acquisition instrument (DEWE-43A) is used to collect the engine output power data, boom large / small chamber pressure data, arm large / small chamber pressure data, bucket large / small chamber pressure data, hydraulic dual pump outlet pressure data, hydraulic dual pump output flow data, and controller output current signal data during hydraulic excavator operation.

[0013] Step 2: Establishment of engine output power timing prediction model

[0014] 1) Calculate the autocorrelation of the engine output power signal and determine the input structure of the time series prediction model.

[0015] Autocorrelation is used to describe the random signal X (t) The difference is in t a and t b The value of time and and The calculation process of the autocorrelation coefficient of the engine output power signal is as follows:

[0016] Calculate the mathematical expectation E(X) and variance D(X) of the engine output power, and calculate the autocovariance of the output power based on the mathematical expectation. The autocovariance function is:

[0017] C xx (t a ,t b )=E[(X(t a )-E(X(t a )))(X(t b )-E(X(t b )))](1-1)

[0018] Then calculate the autocorrelation coefficient of the engine output power based on the autocovariance. The formula for calculating the autocorrelation coefficient is as follows:

[0019]

[0020] When t a and t b When the autocorrelation coefficient of the output power at time t is greater than the set threshold (usually 0.8), it indicates that a -t b The output power signal during the time is the same as t b The output power at the moment is highly correlated. The autocorrelation coefficient is greater than the set threshold [t a -t b-1 ] ba-1 engine output power signals are used as the input of the time series prediction model, and the output is t b Predicted power at the moment.

[0021] 2) Select the recurrent neural network algorithm (RNN) that can mine the time series information in the data as the algorithm basis of the time series prediction model. Set the algorithm input structure to t m -t m-n n historical time series output power data within a time period, the output structure is t m+1 The predicted output power at the time.

[0022] 3) According to the input and output structure of the recurrent neural network algorithm (RNN) designed in step 2) and the engine output power data of the hydraulic excavator during operation collected in step 1, a data set is prepared. The input data set is: [t i -t i+n-1] engine sequential output power in the time period [W t(i) -W t(i+n-1) ], the output data is t i+n The engine output power at the time W t(i+n) .

[0023] 4) The data set in step 2 (3) is divided into data sets according to the ratio (7:3), and used as the training data and test data of the engine timing prediction model respectively, and the engine output power prediction model is trained and tested. When the prediction accuracy of the engine output power prediction model reaches the expected effect, the engine output power timing prediction model is established. The calculation formula of the engine output power timing test model accuracy is as follows:

[0024]

[0025] Where:

[0026] acc W The prediction accuracy of the engine output power prediction model;

[0027] is the t predicted by the engine output power timing prediction model i Output power at the moment;

[0028] W t(i) For engine t i The actual output power at the moment.

[0029] Step 3: Calculation of engine predicted output torque

[0030] After the engine output power timing prediction model is established, the engine output torque is calculated according to the conversion formula (1-4) between engine output power and output torque.

[0031]

[0032] Where: T is the engine output torque, Nm;

[0033] P is the predicted engine output power obtained by the engine output power timing prediction model, KW;

[0034] n is the engine speed, rad / s.

[0035] Step 4: Establishment of multivariate mapping model of hydraulic pump load torque

[0036] 1) Based on the hydraulic dual pump outlet pressure data and hydraulic dual pump output flow data collected in step 1, the hydraulic dual pump load torque is calculated using formula (1-5).

[0037]

[0038] Where:

[0039] M: hydraulic double pump load torque, Nm;

[0040] Q s : Hydraulic pump flow, L;

[0041] P: Hydraulic dual pump outlet pressure, bar;

[0042] η m : Mechanical efficiency of hydraulic double pump.

[0043] 2) Arrange the data set, and use the eight pressure data collected in step 1, namely, the boom large / small chamber pressure data, the dipper arm large / small chamber pressure data, the bucket large / small chamber pressure data, and the hydraulic dual pump outlet pressure data, as the input data set of the multivariate mapping model, and use the hydraulic dual pump load torque calculated in step 4 (1) as the output data set of the multivariate mapping model.

[0044] 3) Using BP neural network as the algorithm basis of the hydraulic pump load torque multivariate mapping model, the input / output data set sorted in step 4 (2) is used as the training data and test data of the hydraulic pump load torque multivariate mapping model in a ratio of (7:3), and the training and testing of the hydraulic dual pump load torque multivariate mapping model are carried out. When the hydraulic pump load torque mapping accuracy reaches the expected effect, the hydraulic dual pump load torque multivariate mapping model is established. The calculation formula for the accuracy of the hydraulic pump load torque multivariate mapping model is as follows:

[0045]

[0046] Where:

[0047] acc M is the prediction accuracy of the multivariate mapping model of hydraulic pump load torque;

[0048] t calculated by the multivariate mapping model of hydraulic pump load torque i Load torque at the moment;

[0049] M t(i) For hydraulic pump t i The actual load torque at that moment.

[0050] Step 5: PID controller parameter tuning optimization

[0051] 1) Establish a PID controller model in Matlab / Simulink, such as Figure 2 shown.

[0052] 2) Take the output flow data of the hydraulic dual pump collected in step 1 and the controller output current signal data to identify the hydraulic pump system and obtain the transfer function of the hydraulic pump system. According to formula (1-5), a simulation model of the hydraulic pump is established in Matlab / Simulink. The model input is the load pressure and input current of the two hydraulic pumps, and the output is the overall torque of the hydraulic dual pump.

[0053] 3) Establish a PID controller optimization simulation model. Connect the PID controller and the hydraulic dual-pump simulation model in Matlab / Simulink. Set the engine output torque as time-varying data, the initial control current of hydraulic pump 1 and hydraulic pump 2, and the load pressure of hydraulic pump 1 and hydraulic pump 2 as fixed data, and the input of the PID controller optimization simulation model is the engine output torque, the initial control current of hydraulic pump 1 and hydraulic pump 2, and the load pressure of hydraulic pump 1 and hydraulic pump 2; the output and feedback of the PID controller optimization simulation model are the hydraulic dual-pump output load torque signal, the PID controller input is the difference between the engine output torque and the hydraulic pump load torque, and the PID controller output is the control current change, such as Figure 3 shown.

[0054] 4) Write the Gray Wolf algorithm in Matlab, connect the PID controller optimization simulation model established in step 5) in 3), simulate the PID controller optimization model multiple times through the Gray Wolf algorithm, obtain the proportional, integral, differential coefficients of the PID controller optimized by the Gray Wolf algorithm at different times, and the error between the engine output torque and the hydraulic pump load torque, and take the minimum error as the iterative goal of the Gray Wolf algorithm, and finally obtain the proportional, integral, and differential coefficients of the PID controller that can dynamically balance the engine output torque and the hydraulic pump load torque and change dynamically with the load, and complete the parameter tuning of the PID controller.

[0055] Step 6: Model programming and controller embedding

[0056] The engine output power timing prediction model established in step 2, the engine output torque calculation formula in step 3, the hydraulic pump load torque multivariate mapping model established in step 4, and the PID controller with automatic parameter adjustment optimized by the Grey Wolf algorithm in step 5 are compiled into C language programs, embedded in the hydraulic excavator controller, and put into use.

[0057] The overall optimization control process is as follows Figure 4 shown.

[0058] The beneficial effects of the present invention are:

[0059] 1. By predicting the engine output power, the lag control phenomenon caused by the delayed response of the engine and hydraulic system is reduced, and real-time control is achieved.

[0060] 2. Using the Gray Wolf Algorithm to optimize the parameters of the PID controller can make the proportional, integral, and differential coefficients of the PID change with the load, thereby improving the control accuracy and speed of the PID controller. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 It is a schematic diagram of the optimized control of hydraulic excavators based on engine output power prediction and controller parameter setting;

[0062] Figure 2 This is a PID controller model diagram based on Matlab / Simulink;

[0063] Figure 3 This is a PID controller optimization simulation model diagram based on Matlab / Simulink;

[0064] Figure 4 It is the overall optimization control flow chart of hydraulic excavator.

[0065] Figure 5 This is a schematic diagram of the load torque multivariate mapping model structure. DETAILED DESCRIPTION

[0066] In order to make the technical solutions and advantages of the present invention more clear, the present invention will be described clearly and completely with examples below.

[0067] Example: A hydraulic excavator is required to perform sand and stone excavation, transportation, and unloading work on a plain site for a long time. It is required to prevent the engine of the hydraulic excavator from vigorous shaking or flameout during operation.

[0068] 1. Operate the hydraulic excavator to carry out sand and stone excavation on a plain site for 30 minutes, and collect the engine output power data, boom large / small cavity pressure data, arm large / small cavity pressure data, bucket large / small cavity pressure data, hydraulic dual pump outlet pressure data, hydraulic dual pump output flow data, and controller output current signal data during this period.

[0069] 2. The autocorrelation coefficient is calculated based on the engine output power data collected within 30 minutes. Some of the sampling points, delay time, and autocorrelation coefficients obtained are shown in Table 1.

[0070] Table 1 Engine output power autocorrelation coefficient table

[0071]

[0072] As shown in Table 1, within 11 consecutive sampling points, the autocorrelation coefficient of the engine output power signal is greater than 0.8, that is, the 11 consecutively sampled engine output power data are highly correlated. Considering the autocorrelation coefficient and the data capacity that the controller can store, the data set structure of the engine output power timing prediction model is designed to be the first 10 consecutive output power sampling data, and the output is the predicted engine output power value at the next moment (the 11th).

[0073] The prepared data set is used to train and test the engine output power timing prediction model in a ratio of (7:3). When the prediction accuracy of the engine output power timing prediction model reaches 90%, the engine output power timing prediction model is established. The model algorithm language is compiled into a C language program.

[0074] 3. Calculate the hydraulic pump load torque data based on the hydraulic dual pump output flow data and hydraulic dual pump outlet pressure data collected within 30 minutes of hydraulic excavator operation. Take the collected boom large / small cavity pressure data, dipper arm large / small cavity pressure data, bucket large / small cavity pressure data, hydraulic dual pump outlet pressure data as input, and hydraulic pump load torque data as output. Design the BP neural network with a single hidden layer and 20 neurons, such as Figure 5 As shown, the data set is trained and tested for the load torque multivariate mapping model in a ratio of (7:3). When the load torque multivariate mapping model calculates the load torque with an accuracy of 95%, the load torque multivariate mapping model is established. The model algorithm language is compiled into a C language program.

[0075] 4. Establish a PID controller simulation model in Matlab / Simulink. Design the proportional coefficient, integral coefficient, and differential coefficient of the PID controller as variable parameters as the three optimization variables of the subsequent Grey Wolf algorithm. The input of the PID controller is the error value between the engine output torque and the hydraulic pump load torque, and the output is the change in the controller output current. In order to facilitate the observation of the subsequent PID controller optimization effect, the proportional, integral, and differential coefficients are also output as the PID controller model. After the algorithm is optimized, the proportional, integral, and differential coefficients of the PID controller can be adjusted accordingly when the input error value changes.

[0076] 5. The hydraulic dual pump output flow data Q and controller output current signal I data collected within 30 minutes of hydraulic excavator operation are used for system identification to obtain the hydraulic pump transfer function G(s) of the hydraulic excavator as shown in formula (1-7). Then, according to formula (1-5), a simulation model of the hydraulic pump is established in Matlab / Simulink, with the input being the load pressure and input current of the two hydraulic pumps, and the output being the overall torque of the hydraulic dual pumps.

[0077]

[0078] 6. Connect the PID controller simulation model and the hydraulic pump model in Matlab / Simulink to complete the establishment of the PID controller optimization simulation model. Set the input data of the PID controller optimization simulation model: the engine output torque is a sinusoidal waveform data of 500±80Nm; the basic control current of hydraulic pump 1 and hydraulic pump 2 is 690Ma; the load pressure of hydraulic pump 1 is 144bar; the load pressure of hydraulic pump 2 is 139bar; the output of the PID controller optimization simulation model is the hydraulic pump load torque.

[0079] 7. Write the Gray Wolf optimization algorithm in Matlab, connect the PID controller optimization simulation model, set the optimization variables as the proportional coefficient, integral coefficient, and differential coefficient of the PID controller, perform 100 iterations of the Gray Wolf algorithm to complete the parameter tuning of the PID controller. Compile the optimized PID controller into a C language program.

[0080] 8. The engine output power timing prediction model, the engine output torque calculation formula, the multivariate mapping model of the hydraulic dual pump load torque, and the C language program of the optimized PID controller are embedded in the hydraulic excavator controller and put into use.

[0081] 9. When the optimized hydraulic excavator is working, the controller takes the continuous historical output power data of 10 engines including the current moment, predicts the engine output power at the next moment through the engine output power time series prediction model, and calculates the engine output torque at the next moment using the engine output torque formula. The controller takes the hydraulic dual pump outlet pressure data, the boom large / small chamber pressure data, the dipper arm large / cavity pressure data, the bucket large / small chamber pressure data, and the hydraulic dual pump outlet pressure data, and calculates the load torque through the hydraulic pump load torque multivariate mapping model. The load torque is subtracted from the predicted engine output torque and imported into the PID controller optimized by the Gray Wolf algorithm to calculate the controller current change, change the hydraulic dual pump output flow, make the hydraulic pump load torque dynamically equal to the engine output torque, and complete the real-time stabilization of the engine speed.

[0082] 10. Take the driver's operating experience as feedback. If the jitter / stall phenomenon is reduced, the hydraulic excavator optimization is completed. If the jitter / stall phenomenon has not been reduced or the change is not obvious, repeat steps 2-9 to improve the accuracy of the engine output power timing prediction model and the multivariate mapping model of the hydraulic pump load torque, increase the number of Grey Wolf algorithm optimization iterations, and re-optimize the PID controller until the jitter / stall phenomenon is significantly reduced and the hydraulic excavator optimization is completed.

Claims

1. A hydraulic excavator optimization method based on engine output power prediction and controller parameter setting, characterized in that: Here are the steps: Step 1: Hydraulic excavator modeling data collection Use the data acquisition instrument to collect the engine output power data, boom large / small chamber pressure data, dipper arm large / small chamber pressure data, bucket large / small chamber pressure data, hydraulic dual pump outlet pressure data, hydraulic dual pump output flow data, and controller output current signal data during hydraulic excavator operation; Step 2: Establishment of engine output power timing prediction model 1) Calculate the autocorrelation of the engine output power signal and determine the input structure of the time series prediction model; Autocorrelation is used to describe the random signal X (t) The difference is in t a and t b The value of time and X(ta) and X (tb) The degree of correlation between the two; the calculation process of the autocorrelation coefficient of the engine output power signal is as follows: Calculate the mathematical expectation E(X) and variance D(X) of the engine output power, and calculate the autocovariance of the output power based on the mathematical expectation; the autocovariance function is: C xx (t a ,t b )=E[(X(t a )-E(X(t a )))(X(t b )-E(X(t b )))](1-1) Then the autocorrelation coefficient of the engine output power is calculated based on the autocovariance; the autocorrelation coefficient calculation formula is as follows: When t a and t b When the autocorrelation coefficient of the output power at time t is greater than the set threshold, it indicates that a -t b The output power signal during the time is the same as t b The output power at each moment is highly correlated; the autocorrelation coefficient is greater than the set threshold [t a -t b-1 ] ba-1 engine output power signals are used as the input of the time series prediction model, and the output is t b Predicted power at the moment; 2) Select the recurrent neural network algorithm that can mine the time series information in the data as the algorithm basis of the time series prediction model; set the algorithm input structure to t m -t m-n n historical time series output power data within a time period, the output structure is t m+1 The predicted output power at the time; 3) According to the input and output structure of the recurrent neural network algorithm designed in step 2) and the engine output power data of the hydraulic excavator during operation collected in step 1, a data set is prepared. The input data set is: [t i -t i+n-1 ] engine sequential output power in the time period [W t(i) -W t(i+n-1) ], the output data is t i+n The engine output power at the time W t(i+n) ; 4) The data set in step 2 (3) is divided into data sets according to the ratio (7:3), and used as the training data and test data of the engine timing prediction model respectively, and the engine output power prediction model is trained and tested. When the prediction accuracy of the engine output power prediction model reaches the expected effect, the establishment of the engine output power timing prediction model is completed; the calculation formula of the engine output power timing test model accuracy is as follows: Where: acc W The prediction accuracy of the engine output power prediction model; is the t predicted by the engine output power timing prediction model i Output power at the moment; W t(i) For engine t i The actual output power at the moment; Step 3: Calculation of engine predicted output torque After the engine output power timing prediction model is established, the engine output torque is calculated according to the conversion formula (1-4) between the engine output power and the output torque; Where: T is the engine output torque, Nm; P is the predicted engine output power obtained by the engine output power timing prediction model, KW; n is the engine speed, rad / s; Step 4: Establishment of multivariate mapping model of hydraulic pump load torque 1) According to the hydraulic dual pump outlet pressure data and hydraulic dual pump output flow data collected in step 1, the hydraulic dual pump load torque is calculated using formula (1-5); Where: M: hydraulic double pump load torque, Nm; Q s : Hydraulic pump flow, L; P: Hydraulic dual pump outlet pressure, bar; η m : Mechanical efficiency of hydraulic double pump; 2) Arranging the data set, using the eight pressure data collected in step 1, namely, the large / small cavity pressure data of the boom, the large / small cavity pressure data of the dipper arm, the large / small cavity pressure data of the bucket, and the hydraulic dual pump outlet pressure data, as the input data set of the multivariate mapping model, and using the hydraulic dual pump load torque calculated in step 4 1) as the output data set of the multivariate mapping model; 3) Using the BP neural network as the algorithm basis of the hydraulic pump load torque multivariate mapping model, the input / output data sets sorted in step 4 (2) are used as the training data and test data of the hydraulic pump load torque multivariate mapping model in proportion, and the training and testing of the hydraulic dual pump load torque multivariate mapping model are carried out. When the hydraulic pump load torque mapping accuracy reaches the expected effect, the hydraulic dual pump load torque multivariate mapping model is established; the calculation formula of the hydraulic pump load torque multivariate mapping model accuracy is as follows: Where: acc M is the prediction accuracy of the multivariate mapping model of hydraulic pump load torque; t calculated by the multivariate mapping model of hydraulic pump load torque i Load torque at the moment; M t(i) For hydraulic pump t i The actual load torque at the moment; Step 5: Parameter tuning and optimization of PID controller 1) Establish a PID controller model in Matlab / Simulink; 2) Take the output flow data of the hydraulic dual pumps collected in step 1 and the controller output current signal data to identify the hydraulic pump system and obtain the transfer function of the hydraulic pump system; establish a simulation model of the hydraulic pump in Matlab / Simulink according to formula (1-5), the model input is the load pressure and input current of the two hydraulic pumps, and the output is the overall torque of the hydraulic dual pumps; 3) Establish a PID controller optimization simulation model; connect the PID controller and the hydraulic dual pump simulation model in Matlab / Simulink; set the engine output torque as time-varying data, the initial control current of hydraulic pump 1 and hydraulic pump 2, and the load pressure of hydraulic pump 1 and hydraulic pump 2 as fixed data, and the input of the PID controller optimization simulation model is the engine output torque, as well as the initial control current of hydraulic pump 1 and hydraulic pump 2, and the load pressure of hydraulic pump 1 and hydraulic pump 2; the output and feedback of the PID controller optimization simulation model are the hydraulic dual pump output load torque signal, the PID controller input is the difference between the engine output torque and the hydraulic pump load torque, and the PID controller output is the control current change; 4) Write the Gray Wolf algorithm in Matlab, connect the PID controller optimization simulation model established in step 5) in 3), simulate the PID controller optimization model multiple times through the Gray Wolf algorithm, obtain the proportional, integral, differential coefficients of the PID controller optimized by the Gray Wolf algorithm at different times, and the error between the engine output torque and the hydraulic pump load torque, and take the minimum error as the iteration target of the Gray Wolf algorithm, and finally obtain the proportional, integral, differential coefficients of the PID controller that can dynamically balance the engine output torque and the hydraulic pump load torque and change dynamically with the load, and complete the parameter setting of the PID controller; Step 6: Model programming and controller embedding The engine output power timing prediction model established in step 2, the engine output torque calculation formula in step 3, the hydraulic pump load torque multivariate mapping model established in step 4, and the PID controller with automatic parameter adjustment optimized by the Grey Wolf algorithm in step 5 are compiled into C language programs, embedded in the hydraulic excavator controller, and put into use.

Images