阅读说明：本技术 一种基于负载变化率主动观测的自学习转速控制方法 (Self-learning rotating speed control method based on load change rate active observation ) 是由 宋康 谢辉 邵灿 于 2020-06-15 设计创作，主要内容包括：本发明公开了一种基于负载变化率主动观测的自学习转速控制方法,包括以下步骤：步骤1,通过反馈控制计算转动惯性力矩；利用摩擦扭矩模型估计当前的摩擦扭矩,得到摩擦扭矩；步骤2,在发动机的转速动态变化的基础上,增加负载扭矩和负载扭矩变化率两个“扩张状态”；步骤3,通过观测器进行在线迭代,对转速、负载扭矩及负载扭矩变化率进行在线观测；步骤4,在转动惯性力矩的基础上,利用负载扭矩的估计值做补偿,得到有效扭矩；在有效扭矩的基础上叠加步骤1的摩擦扭矩,获得指示扭矩；步骤5,通过发动机的指示扭矩模型计算得到喷油量,喷油控制系统根据喷油量控制转速。解决了造成发动机转速波动的原因,显著提高了转速控制的抗干扰能力。(The invention discloses a self-learning rotating speed control method based on load change rate active observation, which comprises the following steps of: step 1, calculating a rotation inertia moment through feedback control; estimating the current friction torque by using a friction torque model to obtain the friction torque; step 2, on the basis of the dynamic change of the rotating speed of the engine, two expansion states of load torque and load torque change rate are increased; step 3, carrying out online iteration through an observer, and carrying out online observation on the rotating speed, the load torque and the load torque change rate; step 4, on the basis of the rotational inertia moment, compensating by using the estimated value of the load torque to obtain an effective torque; superposing the friction torque of the step 1 on the basis of the effective torque to obtain an indicated torque; and 5, calculating to obtain the fuel injection quantity through an indicated torque model of the engine, and controlling the rotating speed by the fuel injection control system according to the fuel injection quantity. The problem of causing the fluctuation of the rotating speed of the engine is solved, and the anti-interference capability of rotating speed control is obviously improved.)

1. An engine rotating speed self-learning control method based on load change rate active observation is characterized by comprising the following steps:

step 1, calculating a rotational inertia moment through feedback control according to the deviation of the target rotating speed of the engine and the actual rotating speed of the engine; estimating the current friction torque by using a friction torque model to obtain the friction torque;

step 2, on the basis of the dynamic change of the rotating speed of the engine, two expansion states of load torque and load torque change rate are added, and a rotating speed dynamic model with the expansion states is constructed;

step 3, aiming at the rotating speed dynamic model with the expansion state, performing online iteration through an observer, and performing online observation on the rotating speed, the load torque and the load torque change rate by combining the friction torque obtained in the step 1 to obtain an estimated value of the load torque;

step 4, on the basis of the rotational inertia moment obtained in the step 1, compensating by using the estimated value of the load torque obtained in the step 3 to obtain an effective torque; superposing the friction torque of the step 1 on the basis of the effective torque to obtain an indicated torque;

and 5, combining the indicated thermal efficiency and the indicated torque, calculating by using an indicated torque model of the engine to obtain the fuel injection quantity, and controlling the rotating speed by the fuel injection control system according to the fuel injection quantity.

2. The self-learning engine speed control method based on active load change rate observation as claimed in claim 1, wherein in step 1, the rotational inertia moment u₀＝k_p(ω^ref-ω)，ω^refFor the target engine speed, ω is the actual engine speed, k_pIs a scaling factor.

3. The engine speed self-learning control method based on load change rate active observation according to claim 1, wherein in step 2, the speed dynamic model with the expansion state is:

where, ω is the actual engine speed,a derivative representing the actual engine speed;

j is the moment of inertia of the crankshaft rotational system, M_iIs an indicator torque;

4. The self-learning engine speed control method based on active load change rate observation as claimed in claim 3, wherein the observer in step 3 is:

where, sum ξ is an intermediate variable, β₁And β₂For observer gain, ω is the actual engine speed, ω_oIn order to be the bandwidth of the observer,for equivalent load torqueUsing equivalent load torque

5. The method as claimed in claim 1The engine rotating speed self-learning control method with the load change rate actively observed is characterized in that in the step 4,

6. The self-learning engine speed control method based on active load change rate observation as claimed in claim 1, wherein in step 5, the indicated torque model is:

wherein the content of the first and second substances,for the quantity of injected fuel, M_iIs the indicated torque, H, from step 4_LHVη for lower heating value of diesel engines_iIs indicative of the thermal efficiency, n_cylIs the number of engine cylinders and ω is the actual engine speed.

7. The self-learning engine speed control method based on active observation of load change rate as claimed in claim 1, wherein the indicated thermal efficiency in step 5 is artificially constant between 0 and 1.

8. The self-learning engine speed control method based on active load change rate observation as claimed in claim 1, wherein the indicated thermal efficiency in step 5 is obtained by means of model parameter online learning to obtain a value between 0 and 1.

9. The self-learning engine speed control method based on active load change rate observation as claimed in claim 8, wherein if the current condition is determined to be a steady state condition, a 1-10% sine disturbance signal is added to the original fuel injection amount signal and injected into the engine; the engine speed slightly fluctuates under the action of the sine disturbance signal; and performing online calculation on the indicated thermal efficiency by utilizing an online estimation algorithm according to the current fuel injection quantity of the engine, the actual rotating speed of the engine and the friction torque.

10. The self-learning engine speed control method based on active load rate observation as claimed in claim 8 wherein the indicated thermal efficiency is learned on-line using recursive least squares to obtain η_iIs estimated value of

by

Definition of

Y ═ Y (1) Y (2) Y (3.) for multiple samples,. Y (n)]^T,Comprises the following steps:

Y＝φη_iand performing online iteration to obtain:

Technical Field

The invention relates to the technical field of engine rotating speed control, in particular to a self-learning rotating speed control method based on load change rate active observation.

Background

Speed control is one of the important functions of engine control. The control quality of the rotating speed has obvious influence on the oil consumption and the comfort of the idling working condition of the engine, the stability of the voltage and the power of the generator for power generation and the smoothness of mode transition in a hybrid power system. Although engine speed control is not a new problem, the problem of unknown load torque is not solved well, and the quality of speed control is also influenced.

Proportional-derivative-integral (PID) control is the most commonly used algorithm for controlling the rotation speed, and complicated parameter calibration is usually required to ensure the control quality. Robust Control is a controller with relatively stable performance, and is also tried to be applied to rotational speed Control, as shown in the literature (Hrovat, device, and sting Sun. "Models and Control methods for IC engine idle speed Control design." Control Engineering practice5.8(1997): 1093-. However, the design of a robust controller is conservative, limiting its transient performance. Song et al propose a linear variable parameter (LPV) model-based rotational speed controller, however, the design process of such LPV model is relatively complicated (Song, Qingwen, and Karolos M.Gridiroadis. "Diesel engine speed regulation using linear parameter varying control." Proceedings of the 2003American control Conference,2003.. Vol.1.IEEE, 2003.). Sun et al propose an Optimal Control algorithm for rotational speed, however Optimal Control has limitations in robustness, limiting its engineering applications (Sun, Pu, B. Powell, and Davor Hrovat. "Optimal idle speed Control of an automatic engine." Proceedings of the2000American Control reference. ACC. (IEEE Cat. No.00CH36334). Vol.2.IEEE, 2000.). Yin et al propose a Fuzzy logic-based speed control algorithm, but the design rules of Fuzzy logic are relatively complex (Yin, Xiaoofeng, Dianlun Xue, and Yun Cai. "Application of time-optimal geometry and Fuzzy logic to the engine speed control circuit of the gear-shifting process of AMT." Fourth International Conference on Fuzzy Systems and Knowledge Discovery (FSKD 2007). Vol.4.IEEE, 2007.). Shu et al developed speed control using a non-linear model predictive control (NMPC) approach, but NMPC has a large computational burden, has a high demand for model accuracy, and has applications in embedded systemsTo some extent (Li, Shu, Hong Chen, and Shuyou Yu., "Nonlinear model predictive controlled for ideal speed controlled of SI engine," Proceedings of the 48h IEEE Conference on decision and Control (CDC) old joint with 200928 th Chinese Control IEEE, 2009.). Feng et al propose a rotating speed control method based on an adaptive algorithm (Feng, Meiyu, and Xiaohohong Jiao. "Double closed-loop control with adaptive strategy for automatic engine speed tracking system." International Journal of adaptive control and Signal processing31.11(2017):1623 1635.), but the algorithm does not directly solve the uncertainty problem of load torque. Stotsky et al propose variable structure idle speed control algorithms for unknown disturbances, such as the references (Stotsky, Alexander, Bo Egardt, and)"variable structure Control of engine idle speed with estimation of unmeasurable disorders," J.Dyn.Sys., Meas., Control 122.4(2000): 599-. However, the problem of chattering in sliding mode control has not been well solved.

In order to improve the control quality of the engine speed, a control algorithm which is simple in calibration, small in calculation amount, capable of directly estimating the load torque and adaptive is needed.

Disclosure of Invention

The invention aims to provide a self-learning rotating speed control method based on active load change rate observation, aiming at the problem of poor rotating speed control quality caused by unknown load torque in the rotating speed control of an engine in the prior art.

The technical scheme adopted for realizing the purpose of the invention is as follows:

an engine rotating speed self-learning control method based on load change rate active observation is characterized by comprising the following steps:

step 1, calculating a rotational inertia moment through feedback control according to the deviation of the target rotating speed of the engine and the actual rotating speed of the engine; estimating the current friction torque by using a friction torque model to obtain the friction torque;

step 2, on the basis of the dynamic change of the rotating speed of the engine, two expansion states of load torque and load torque change rate are added, and a rotating speed dynamic model with the expansion states is constructed;

step 3, aiming at the rotating speed dynamic model with the expansion state, performing online iteration through an observer, and performing online observation on the rotating speed, the load torque and the load torque change rate by combining the friction torque obtained in the step 1 to obtain an estimated value of the load torque;

step 4, on the basis of the rotational inertia moment obtained in the step 1, compensating by using the estimated value of the load torque obtained in the step 3 to obtain an effective torque; superposing the friction torque of the step 1 on the basis of the effective torque to obtain an indicated torque;

and 5, combining the indicated thermal efficiency and the indicated torque, calculating by using an indicated torque model of the engine to obtain the fuel injection quantity, and controlling the rotating speed by the fuel injection control system according to the fuel injection quantity.

In the above technical solution, in the step 1, the moment of inertia u is rotated₀＝k_p(ω^ref-ω)，ω^refFor the target engine speed, ω is the actual engine speed, k_pIs a scaling factor.

In the above technical solution, in the step 2, the rotation speed dynamic model with the expansion state is:

where, ω is the actual engine speed,

a derivative representing the actual engine speed;

j is the moment of inertia of the crankshaft rotational system, M_iIs an indicator torque;as the equivalent friction torque, there is a torque,

M_Friis the friction torque described in step 1;as the equivalent load torque, there is,M_loadis the load torque;

is that

The derivative of (a) of (b),is the rate of change of the equivalent load torque and h is the derivative of the rate of change of the equivalent load torque.

In the above technical solution, the observer in step 3 is:

where, sum ξ is an intermediate variable, β₁And β₂For observer gain, ω is the actual engine speed, ω_oIn order to be the bandwidth of the observer,for equivalent load torqueIs estimatedEvaluating, using equivalent load torqueIs divided by the estimated value ofObtaining a load torque M_loadAn estimate of (d).

In the above technical solution, in the step 4,u₀for the rotational moment of inertia obtained in step 1,

for the estimated value of the load torque obtained in step 3The number of times of the total number of the parts,

of friction torque obtained in step 1

Multiple, indicated torque of

In the above technical solution, in the step 5, the indicated torque model is:

wherein the content of the first and second substances,

for the quantity of injected fuel, M_iIs the indicated torque, H, from step 4_LHVη for lower heating value of diesel engines_iIs indicative of the thermal efficiency, n_cylIs the number of cylinders of the engine, omega is the engineThe actual rotational speed.

In the above technical solution, the indication of the thermal efficiency in step 5 is a constant value between 0 and 1 (excluding 0 and 1) artificially assigned or a value between 0 and 1 (excluding 0 and 1) obtained by online learning of model parameters.

In the above technical solution, when the indicated thermal efficiency is obtained by means of online learning of model parameters, the steps are as follows:

if the current condition is judged to be a steady-state working condition, adding 1-10% of sine disturbance signals in the original fuel injection quantity signals, and injecting the signals into the engine; the engine speed slightly fluctuates under the action of the sine disturbance signal; and performing online calculation on the indicated thermal efficiency by utilizing an online estimation algorithm according to the current fuel injection quantity of the engine, the actual rotating speed of the engine and the friction torque.

In the technical scheme, the indicated thermal efficiency is learned on line by adopting a recursive least square method to obtain η_iIs estimated value of

The calculation process of (2) is as follows:

by

And

to obtain

Definition of

Y ═ Y (1) Y (2) Y (3.) for multiple samples,. Y (n)]^T,

Comprises the following steps:

Y＝φη_iis carried out on-lineIteration to obtain:

compared with the prior art, the invention has the beneficial effects that:

1. through the active observation of the load torque, the reason of the fluctuation of the rotating speed of the engine is fundamentally solved, and the anti-interference capability of the rotating speed control is obviously improved.

2. The change rate of the load is actively observed in the observer, the speed is higher than that of the traditional load observation method, and the control quality of the rotating speed is further improved.

3. By designing an online learning algorithm for indicating the thermal efficiency, the controller can actively adapt to the change of the operating characteristics of the engine caused by aging and faults, and the reduction of the control performance is avoided.

4. By using the extended state observer, the robustness of the controller is obviously improved, and only one set of control parameters is needed under all working conditions. Compared with the traditional PID controller, the calibration workload is reduced by more than 80%.

5. The algorithm is simple to calculate, the memory occupies less than 2kBytes, and the running time of the algorithm on a 200MHz single chip microcomputer is about 10 microseconds. The method is more suitable for embedded systems than traditional MPC and other model-based control algorithms.

Drawings

FIG. 1 is a control block diagram of the present invention.

Fig. 2 is a block diagram of a learning algorithm indicating thermal efficiency in the present invention.

Fig. 3 is a diagram comparing the active disturbance rejection controller of the present invention with a conventional PID controller.

Detailed Description

The present invention will be described in further detail with reference to specific examples. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.