Transfer Function Identification Method for Internal Combustion Engine Based on Torque Estimation

By using torque estimation method in the identification of transfer function of internal combustion engines and combining least squares optimization, the problems of large calculation workload and low recognition accuracy caused by manual estimation of initial values ​​are solved, and higher recognition accuracy and response speed are achieved, avoiding closed loops Control the accumulation of errors.

CN115509133BActive Publication Date: 2024-11-15CHONGQING HONGJIANG MACHINERY CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202211253320.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-10-13
Publication Date
2024-11-15
Estimated Expiration
2042-10-13

AI Technical Summary

Technical Problem

In the prior art, the transfer function identification method for the open-loop control of internal combustion engines requires manual estimation of the initial value, resulting in large calculation workload and low recognition accuracy, making it difficult to effectively avoid parameter compatibility and error accumulation caused by closed-loop control.

Method used

The torque estimation method is used to estimate the actual torque through the proportional relationship between the target torque and the calibrated torque. Combined with the least squares method optimization, the operation of manual estimation of the initial value is reduced, the identification accuracy and response speed are improved, and parameter compatibility and error accumulation are avoided.

Benefits of technology

The step of manually estimating the initial value is effectively avoided, the identification accuracy of the transfer function and the engine open-loop accuracy are improved, the solution time is shortened, the response speed is enhanced, and the error caused by closed-loop control is reduced.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115509133B_ABST
    Figure CN115509133B_ABST
Patent Text Reader

Abstract

The present invention provides a method for identifying the transfer function of an internal combustion engine based on torque estimation, which derives a corresponding transfer function model according to the mechanical and electrical structure characteristics of an electronically controlled engine, and introduces a second-order link in the transfer function; then, the system parameters to be identified are determined according to the transfer function model, and a feasible optimization solution close to the truth is obtained through a torque estimation method, and this feasible optimization solution is used as the initial value of a nonlinear least square method for iterative optimization, and finally an optimal solution is obtained; the optimal solution is substituted into the transfer function, and the error of the identification and measurement data is calculated. This method can not only remove the step of manually estimating the initial value, but also ensure the identification accuracy of the transfer function, which is beneficial to improving the open-loop accuracy and response speed of the engine, and avoiding the instability caused by parameter compatibility and error accumulation brought about by closed-loop control.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of engine control, and in particular relates to an identification method for an open-loop control transfer function of an internal combustion engine. Background Art

[0002] With the deepening of control theory research and the development of artificial intelligence technology, using some advanced technical means to simplify complex technologies, or to achieve some previously unrealized functional improvements, or to improve control accuracy are all hot research directions. The precision of the controlled object model makes it possible to use the control model to correct the control output, and at the same time, it can offset the steady-state fluctuations caused by closed-loop regulation and achieve better control effects.

[0003] In order to meet the higher requirements of internal combustion engine emissions, energy saving and improved performance, it is necessary to design a diesel engine control system with better performance. And it is very important to reasonably determine a simplified diesel engine control model for the analysis and design of diesel engine control systems. The internal combustion engine is a time-varying, nonlinear controlled object and the system has pure lag. A real diesel engine dynamic model has dozens of state variables. For such a complex system, it is difficult to establish a mathematical model to describe its nonlinear characteristics. The difficulty in identifying the transfer function of an internal combustion engine lies in the abstraction of the controlled object, that is, the correct identification of the transfer function. However, by solving the optimal value through theoretical calculation of the nonlinear least squares method and optimizing the initial value by torque estimation, the transfer function solution time can be shortened and the identification accuracy can be improved. Summary of the invention

[0004] The present invention aims at the problem that the current identification method of the open-loop control transfer function of an internal combustion engine based on the principle of least squares method requires manual estimation of initial values, and provides an internal combustion engine transfer function identification method based on torque estimation. The method uses the least squares method combined with torque estimation to identify the open-loop control transfer function of the internal combustion engine, and estimates the actual torque through the proportional relationship between the target torque and the calibrated torque. Combining this method can effectively avoid the operation of manually estimating the initial value, ensure the identification accuracy of the transfer function, improve the open-loop accuracy and response speed of the engine, and avoid parameter compatibility and error accumulation caused by closed-loop control.

[0005] The technical solution of the present invention is as follows:

[0006] To achieve the purpose of the present invention, a method for identifying the transfer function of an internal combustion engine based on torque estimation is provided, and the specific steps are as follows:

[0007] Step 1: According to the electrical structure and mechanical structure of the internal combustion engine electronic control system, the corresponding transfer function model is derived. At the same time, considering that the rigid body flexibility problem in the system will cause high-frequency mechanical resonance, it is necessary to introduce one or more double second-order links in the transfer function.

[0008] Step 2: Solve the transfer function of the target model and set the initial values ​​of the undetermined parameters of the reference models of the two second-order transfer functions to 1.

[0009] Step 3: Complete the torque test on the target model to test the relationship between the target torque and the calibrated torque under different working conditions, so as to determine the proportional coefficient between the actual torque and the predicted torque, and correct the transfer function parameters, which will be used as the reference object initial value for model parameter identification.

[0010] Step 4: According to the derived transfer function model, determine the system parameters that need to be identified. First, use the torque estimation method to calculate an approximately feasible optimization solution, and then use this optimization solution as the initial value of the non-least squares method, and further iterate and optimize to obtain the optimal solution.

[0011] Step 5: Verify the optimal solution, substitute the optimal solution into the transfer function model, calculate the error between the identification and measurement data to verify the final identification accuracy.

[0012] In step 3, for different target models, the corresponding torque estimation proportional coefficients under different working conditions are all different. For the same target model, the torque proportional coefficient estimation needs to be performed under at least four working conditions: starting, idling, rated, and 100% load. Only when the proportional coefficients verified under these working conditions are within a certain range can they be used as the initial value of the transfer function of the reference object.

[0013] In step 4, the torque estimation method is first used to obtain a feasible optimization solution, and then the optimization solution is used as the initial value of the nonlinear least squares method. The nonlinear least squares method can use the principle of iterative least squares method to quickly and accurately obtain the optimal solution within a given range.

[0014] It can be seen from the above technical scheme that this method uses the least squares method combined with torque estimation to identify the open-loop control transfer function of the internal combustion engine. The core of this method is: according to the mechanical structure and electrical structure of the electronically controlled engine, the corresponding transfer function model is derived, and the problem of rigid body flexibility in the system is considered to cause high-frequency mechanical resonance. A second-order link is introduced into the transfer function and presented in the transfer function in the form of parameters. The proportional coefficient of the engine torque estimation is used as an identification reference; then the system parameters that need to be identified are determined according to the transfer function model, and a feasible optimization solution that approximates the truth is obtained through the torque estimation method. This feasible optimization solution is used as the initial value of the nonlinear least squares method for iterative optimization, and finally the optimal solution is obtained; the optimal solution is substituted into the transfer function, and the error of the identification and measurement data is calculated to verify the final identification accuracy.

[0015] Compared with the prior art, the present invention has the following advantages:

[0016] 1. In the current method of using the nonlinear least squares algorithm to identify system parameters, an initial value close to the exact value must be manually set. Directly setting all the initial coefficient parameters to 1 will result in a large error and slow convergence. The present invention uses the torque estimation method to find a feasible solution close to the exact value, which can greatly reduce the computational workload.

[0017] 2. When there are many parameters that need to be identified in the transfer function, it is very difficult to use the nonlinear least squares method for system identification in engineering practice. Using the initial value of the torque estimate greatly improves the feasibility of the least squares method.

[0018] 3. Using torque estimation to calculate the initial value can ensure the reliability of the solution. Compared with the manually estimated initial value, it can better approximate the accurate value, and the identification accuracy can be accurately calculated based on the test.

[0019] In summary, the present invention can not only effectively eliminate the step of manually estimating the initial value, but also ensure the identification accuracy of the transfer function, which is beneficial to improving the open-loop accuracy and response speed of the engine and avoiding parameter compatibility and error accumulation caused by closed-loop control. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 A flow chart of the transfer identification method;

[0021] Figure 2 Flowchart implemented for torque estimation. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] Figure 1 This is a flow chart of the internal combustion engine transfer function identification method proposed in the embodiment of the present invention. Obtaining an accurate internal combustion engine transfer function model is the basis for designing a controller. The following are specific implementation steps taking a high-pressure common rail electronically controlled diesel engine as an example:

[0024] Step 1: According to the electrical and mechanical structures of the traditional engine, the corresponding basic transfer function model is selected. Considering the rigid body flexibility problem in the system, which will lead to high-frequency mechanical resonance, it is necessary to introduce one or more double second-order links in the transfer function and present them in the transfer function in the form of parameters. The model reference is as follows:

[0025]

[0026] Among them, k1, k2, k3, k4, k5, k6, k7, and k8 are unknown parameters that need to be fitted in the transfer function of the internal combustion engine control system model.

[0027] Step 2: Solve the transfer function of the target model, set the initial values ​​of the undetermined parameters of the reference model of the two second-order transfer functions to 1, and the basic transfer function model is as follows;

[0028]

[0029] Step 3: Complete the torque test on the target model, test the relationship between the target torque and the calibrated torque under different working conditions, and determine the proportional coefficient between the actual torque and the predicted torque, which is used as the reference object initial value for model parameter identification. According to the control method of the high-pressure common rail electronically controlled diesel engine, the fuel injection amount, injection pressure and injection angle can be flexibly controlled, which can offset the high-frequency mechanical resonance caused by some rigid bodies, and at the same time bring in the action hysteresis of the electrical actuator. The transfer function of the high-pressure common rail electronically controlled diesel engine can be reduced in order on the transfer function of the traditional mechanical diesel engine, that is, combined with the internal combustion engine modeling and experimental experience, the transfer function of the internal combustion engine electronic control system will be k5=0, k6=0, k7=0, k8=0 in the second-order link model, and the transfer function model becomes:

[0030]

[0031] This transfer function model is used as the initial transfer function model for iteration, and the initial parameter is k s1 , k s2 , k s3 , k s4 .

[0032] In this step, the torque test is completed on the target model. Since different target models have different torque estimation proportional coefficients under different operating conditions, the torque proportional coefficient estimation is required under at least four operating conditions: starting, idling, rated, and 100% load for the same target model. Only when the proportional coefficients verified under these operating conditions are within a certain range can they be used as the initial value of the transfer function of the reference object.

[0033] Figure 2 In order to implement the torque estimation roadmap, the target torque T1 is set by the engine calibration system on the controller side, and the calibrated torque T2 is calculated in real time. The difference between T1 and T2 is used to infer the internal torque consumption of the engine. The internal consumption of the engine under different working conditions can be fitted into the torque consumption curve R1, and the predicted torque PT is derived in combination with the calibrated torque T2. The dynamometer side DM can measure the dynamometer torque output AT, and the torque estimation proportional coefficient table E is solved according to the ratio between the predicted torque PT and the actual torque output AT.

[0034] Step 4: Based on the initial parameters of step 3, the initial value of the transfer function model of the least squares iteration is obtained, the system parameters that need to be identified are determined, and an approximately feasible optimization solution is obtained. The optimization solution is then used as the initial value of the non-least squares method, and the optimal solution is further iterated and optimized to obtain the optimal solution. The transfer function model becomes:

[0035]

[0036] The optimal transfer function parameter is k f1 , k f2 , k f3 , k f4 .

[0037] Step 5: Verify the optimal solution, substitute the optimal solution into the transfer function model, calculate the error between the identified torque and the actual measured data, and verify the final identification accuracy. The identified transfer function is placed in the closed-loop control model for verification, and the error between the transfer function simulation control quantity and the measured open-loop control quantity for the same set of control parameters.

[0038] In this step, the size of the error determines the level of identification accuracy. When the error between the closed-loop simulation control quantity and the test open-loop control quantity is within 20%, it can be determined that the identification accuracy requirement is met.

[0039] The embodiments are preferred implementations of the present invention, but the present invention is not limited to the above-mentioned implementations. Any obvious improvements, substitutions or modifications that can be made by those skilled in the art without departing from the essential content of the present invention belong to the protection scope of the present invention.

Claims

1. A method for identifying transfer function of an internal combustion engine based on torque estimation, characterized in that: The steps include: Step 1: According to the electrical structure and mechanical structure of the internal combustion engine electronic control system, derive the corresponding transfer function model, and introduce one or more double second-order links into the transfer function; Step 2: Solve the transfer function of the target model, set the initial values ​​of the undetermined parameters of the reference model of the transfer function to 1, and obtain the basic transfer function model; Step 3: Complete the torque test on the target model, test the relationship between the target torque and the calibrated torque under different working conditions, determine the proportional coefficient between the actual torque and the predicted torque, and modify the transfer function parameters based on the basic transfer function model derived in step 2, which will be used as the reference object initial value for model parameter identification; Step 4: According to the result of step 3, firstly obtain an approximate feasible optimization solution through torque estimation, then use the optimization solution as the initial value of the nonlinear least squares method, and further iterate and optimize to obtain the optimal solution; Step 5: Verify the optimal solution, substitute the optimal solution into the transfer function model, calculate the error between the identification and measurement data to verify the final identification accuracy.

2. The internal combustion engine transfer function identification method based on torque estimation according to claim 1 is characterized in that : The transfer function model derived in step 1 is as follows: Among them, kl, k2, k3, k4, k5, k6, k7, and k8 are unknown parameters that need to be fitted in the internal combustion engine transfer function.

3. The internal combustion engine transfer function identification method based on torque estimation according to claim 2 is characterized in that : The basic transfer function model in step 2 is:

4. The internal combustion engine transfer function identification method based on torque estimation according to claim 1, 2 or 3, characterized in that In step 3, different target models and different working conditions have different corresponding torque estimation proportional coefficients. For the same target model, the torque proportional coefficient estimation needs to be performed under at least four working conditions: starting, idling, rated, and 100% load.

5. The internal combustion engine transfer function identification method based on torque estimation according to claim 1, 2 or 3, characterized in that: In step 3, the torque proportional coefficient is estimated by setting the target torque T1 with the engine calibration system on the controller side, calculating the calibrated torque T2 in real time, and calculating the internal torque consumption of the engine according to the difference between T1 and T2. The torque consumption curve R1 is fitted through the internal consumption of the engine under different working conditions, and the predicted torque PT is deduced in combination with the calibrated torque T2. The torque output AT of the dynamometer is measured, and the torque estimation proportional coefficient table E is solved according to the ratio between the predicted torque PT and the actual torque output AT.

6. The internal combustion engine transfer function identification method based on torque estimation according to claim 1, 2 or 3, characterized in that In step 3, combining the internal combustion engine modeling and experimental experience, the internal combustion engine transfer function is set to k5=0, k6=0, k7=0, k8=0 in the second-order link model, and the transfer function model becomes: This transfer function model is used as the initial transfer function model for iteration, and the initial parameter is k s1 , k s2 , k s3 , k s4 .

7. The internal combustion engine transfer function identification method based on torque estimation according to claim 1, 2 or 3, characterized in that : The optimal solution calculated iteratively can derive the optimal transfer function model, which becomes: The optimal transfer function parameter is k f1 , k f2 , k f3 , k f4 .

8. The internal combustion engine transfer function identification method based on torque estimation according to claim 1, 2 or 3, characterized in that : In step 5, the simulation calculation after transfer function identification must be verified in closed-loop control.

9. The internal combustion engine transfer function identification method based on torque estimation according to claim 1, 2 or 3, characterized in that In step 5, the error analysis of identification accuracy is to analyze the error between the transfer function simulation control quantity and the measured open-loop control measurement data of the same set of control parameters.

Citation Information

Patent Citations

  • Electric drive system rotation speed ring debugging method based on rotational inertia

    CN110077241A

  • Method for determining stability boundary of electromechanical coupling dynamic system of ship power device

    CN112987570A

  • Robot load vibration prediction method based on robot joint kinetic model

    CN114967438A

  • Transfer function identification method based on iterative optimization

    CN116974193A