Signal compensation model algorithm control method for electric vehicle air conditioner driving system

Abstract

The invention discloses a signal compensation model algorithm control method for an electric automobile air conditioner driving system, belongs to the field of advanced control of industrial processes, and overcomes the limitation that a traditional control method cannot realize high-precision control under complex working conditions. The method specifically comprises the following steps that 1, an electric automobile air conditioner driving motor rotating speed control system model is established; 2, establishing a discrete state space model with dynamic characteristics which are not included in the modeling process; 3, designing a signal compensation model algorithm controller; according to the method, the problem of excessive power consumption caused by large fluctuation of the rotating speed of the air conditioner motor is avoided, and the system has better prediction output and higher energy utilization rate under the current working condition.

Description

Technical Field

[0001] The invention relates to a signal compensation model algorithm control method for an electric vehicle air-conditioning drive system, and belongs to the field of advanced industrial process control. Background Art

[0002] The air-conditioning drive motor is a key device in the air-conditioning system of electric vehicles, and its speed control is crucial to the performance and energy efficiency of the system. The air-conditioning drive motor involves many uncertain factors in actual working conditions, such as external load mutation, parameter perturbation, and friction. These factors lead to the uncertainty and nonlinear characteristics of the speed control object, which poses a challenge to the stability and performance of the control system. Due to the existence of uncertainty and nonlinearity, it becomes difficult to establish an accurate control system model. Therefore, for actual control objects with uncertainties, nonlinearities and other characteristics, how to design a controller with good control performance while ensuring the stability of the system has always been a difficulty and focus of research in the control field.

[0003] This patent studies the air conditioning drive control of electric vehicles. Since the sudden load changes and friction of the air conditioning system will have an adverse effect on the control performance of the system, correspondingly, a large uncertainty will be generated, which makes it difficult to accurately establish a control system model. On the other hand, when dealing with the modeling of complex systems with nonlinear characteristics, simplified models are usually used to approximate the characteristics of the complex system. However, this simplification may cause some dynamic performance to be ignored in the modeling process. We call these unmodeled dynamic performances unmodeled dynamics. Directly ignoring the unmodeled dynamics will lead to modeling errors. Therefore, we need to design an advanced control algorithm to offset the impact of unmodeled dynamics on system stability.

[0004] Traditional PID control algorithms are widely used due to their simple structure and easy implementation, but in some complex scenarios, there are problems such as poor following effect and large fluctuations. Compared with traditional PID, model algorithm control replaces the speed loop and current loop with a control loop. It not only has a simple structure and fewer parameters to adjust, but also overcomes the shortcomings of slow dynamic response of cascade structure. It is also more suitable for complex nonlinear control objects such as brushless DC motors. It also has better stability and robustness, and can more accurately control the air conditioning system of electric vehicles. Summary of the invention

[0005] In order to solve the existing technical problems, this patent proposes a signal compensation model algorithm control method for electric vehicle air conditioning drive system. This method adopts a composite control method combining model algorithm control and signal compensation for electric vehicle air conditioning drive system, designs a precise prediction controller based on conventional state space model, an efficient compensation controller for dynamic characteristic items that were not included in the modeling process at the previous moment, and an optimized compensation controller for dynamic characteristic items that were not included in the modeling process, and finally forms a signal compensation state space model algorithm controller;

[0006] In order to make the controller better applied to the actual system, the uncertainty and strong nonlinearity of the system need to be taken into account when designing the controller. This method studies this problem of the brushless DC drive motor speed control system model. First, the electric vehicle air conditioning drive motor speed control system model is established. Secondly, a discrete state space model with dynamic characteristics that are not included in the modeling process is established. An accurate prediction controller based on the conventional state space model is designed without considering the influence of the unmodeled dynamics; then, for the dynamic characteristic items that were not included in the modeling process at the previous moment, an efficient compensation controller is designed to eliminate their influence on the closed-loop system. Finally, an optimized compensation controller for the dynamic characteristic items that are not included in the modeling process is established to eliminate the influence of the rate of change of high-order nonlinear items.

[0007] The method is implemented through the following technical solutions:

[0008] Step 1: Establish the electric vehicle air conditioning drive motor speed control system model;

[0009] The speed of the air-conditioning drive motor is closely related to the current and angular velocity of the motor during operation. During the operation of the motor, the magnitude of the current directly affects the output torque of the motor. The angular velocity is an important representation of the motor speed. Its change will directly affect the working state and performance of the motor, and then affect the motor speed. Controlling the speed of the electric vehicle air-conditioning drive motor is equivalent to controlling the motor current and angular velocity. Therefore, the mathematical model of the electric vehicle air-conditioning drive motor speed control system is established as follows:

[0010]

[0011] Where i is the stator winding current, ω is the angular velocity of the motor, r a is the winding wire resistance, k e is the line back electromotive force, U d is the DC bus voltage, L a is the equivalent line inductance of the winding, K T is the motor torque coefficient, B v is the viscous friction coefficient, T L is the load torque, J is the rotor moment of inertia;

[0012] Step 2: Build a discrete state space model with dynamic characteristics that are not incorporated into the modeling process;

[0013] Select state variable x = [iω] Τ As input, the control quantity u=U d , the feedback output signal y = ω, and the state space equation of the air conditioner drive motor is as follows:

[0014]

[0015] In the model algorithm control, the current and speed are assumed to be constant in each prediction step, so the state space equation of the brushless DC motor needs to be discretized. Formula (2) uses the forward difference discretization method to obtain:

[0016]

[0017] Among them, T S is the sampling time;

[0018] The above discretized state space model can be simplified as:

[0019]

[0020] Among them, x(k+1) represents the system state at discrete time k+1, u(k) represents the control input at discrete time k, v(k) represents the unmodeled dynamics between the discretized model and the original model at discrete time k, y(k) represents the system output at discrete time k+1, A is the state matrix of the system, B is the input matrix of the system, and C is the output matrix of the system;

[0021] Step 3: Design a signal compensation model algorithm controller;

[0022] The signal-compensated state-space model predictive controller can be expressed as:

[0023] u(k)=u1(k)+u2(k)+u3(k) (5)

[0024] Among them, u1(k) is the conventional state space model predictive controller, u2(k) is the unmodeled dynamic compensation controller at the previous moment, and u3(k) is the unmodeled dynamic compensation controller;

[0025] First, an accurate predictive controller based on a conventional state space model is designed, and the dynamic characteristic items that are not included in the modeling process are not considered here. At this time, the prediction model of the drive motor is:

[0026]

[0027] In order to solve the optimal control input sequence in the prediction time domain, the quadratic performance index function is derived, and the accurate predictive controller u1(k) of the traditional conventional state space model can be obtained as:

[0028] u1(k)=-(Y in Τ QY in +R) -1 Y in Τ QY out x(k) (7)

[0029] in,

[0030]

[0031] Y in is the system control input coefficient matrix, Y out is the system output matrix, Y is the system output matrix, x(k) represents the system state at discrete time k, U is the system input matrix, Q is the state compensation matrix of the state space model predictive controller, R is the input compensation matrix of the state space model predictive controller, the prediction stage length is N1, the control stage length is N2, and N1 ≥ N2;

[0032] Secondly, based on the data information of the virtual unmodeled dynamics at the previous moment, an incremental unmodeled dynamic compensation controller u2(k) is designed;

[0033] In order to perform dynamic and static compensation for the unmodeled dynamics at the previous moment and ensure the control performance of the controller u2(k), the following performance indicators are introduced:

[0034] J2=||P(z -1 )x(k+1)+Q(z -1 )u(k)+K(z -1 )v(k-1)-H(z -1 )u3(k)|| 2 (8)

[0035] Among them, z -1 is the backshift operator, P(z -1 ) is x(k+1) with respect to z -1 The weighted polynomial, Q(z -1 ) is the value of u(k) with respect to z -1 The weighted polynomial of K(z -1 ) is v(k-1) with respect to z -1 The weighted polynomial of -1 )=1-z -1 For about z -1The backward compensation parameter, u3(k) is the output of the unmodeled dynamic compensation controller at discrete time k;

[0036] The optimal feedforward control law obtained by minimizing J2 is:

[0037] [B+Q(z -1 )]u(k)=K1x(k)-[1+K(z -1 )]v(k-1)+H(z -1 )u3(k) (9)

[0038] In order to ensure the stability of the system, dynamic and static compensation is performed on v(k-1), and the following is obtained:

[0039]

[0040] Finally, the unmodeled dynamic compensation controller is designed. Define y r is the set value, then the system output error e(k) at discrete time k is:

[0041] e(k)=y r -y(k) (11)

[0042] u3(k) is designed to eliminate output error. The relationship between output error and control input is:

[0043] e(k+1)=e(k)-CAx(k)-CBK1x(k)-CBu3(k)-CΔv(k-1) (12)

[0044] Among them, K1 is the control law gain of u1(k) at discrete time k, and Δv(k-1) is the dynamic increment that is not included in the modeling process;

[0045] In order to design the unmodeled dynamic incremental compensation controller u3, the following performance indicators are introduced:

[0046] J3=min||e Τ (k+1)Me(k+1)+u3 Τ (k+1)Nu3(k+1)|| 2 (13)

[0047] Where, e(k+1) is the system output error at discrete time k+1;

[0048] At this time, the one-step optimal forecast of the tracking error e(k+1|k) is:

[0049] e(k+1|k)=e(k)-CAx(k)-CBK1x(k)-CBu3(k)-CΔv(k-1) (11)

[0050] Depend on The optimal compensation controller of the dynamic characteristic terms that are not included in the modeling process is obtained as follows:

[0051]

[0052] By designing a precise prediction controller u1(k) based on a conventional state-space model, an efficient compensation controller u2(k) for the dynamic characteristic items that were not included in the modeling process at the previous moment, and an optimized compensation controller u3(k) for the dynamic characteristic items that were not included in the modeling process, a signal compensation model algorithm controller is finally constructed.

[0053] Compared with the prior art, the beneficial effects of this method are: a model algorithm control method is proposed for the air-conditioning drive system of an electric vehicle based on signal compensation. On the basis of the traditional model algorithm control method, this method regards the high-order nonlinear terms in the controlled object model as unmodeled dynamics, and designs a feedforward signal compensator based on the unmodeled dynamics to eliminate the impact of model reduction and nonlinear characteristics on the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 is a random noise signal;

[0055] Figure 2 It is the output response simulation diagram of the traditional model algorithm control method;

[0056] Figure 3 This is the output response simulation diagram of the model algorithm control method based on signal compensation. DETAILED DESCRIPTION

[0057] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0058] Example:

[0059] The present invention studies the speed control system of the electric vehicle air conditioner drive motor to solve this problem. The estimated values ​​of the model parameters in formula (4) obtained by alternately identifying the recursive least squares method and the neural network are:

[0060]

[0061] The random noise signal noise(k) superimposed on the simulation model of the controlled object is as follows: Figure 1 shown.

[0062] The control objective can be expressed as:

[0063] e(k)|=|y r (k)-y(k)|<2, 0<k<∞ (16)

[0064] In the formula, the setting value y r(k) = 3200r / min.

[0065] Since the controller parameters designed by this method will be optimized in a rolling manner as the system runs, and the controller parameters will only fluctuate within a small range, only the controller parameters when k = 500s are given here for reference. When k = 500s, the signal compensated robust multi-step predictive controller u(500) = 2797.

[0066] The introduced performance evaluation function is:

[0067]

[0068] Where E(k) is the performance evaluation index at discrete time k.

[0069] The present invention simulates the air conditioning drive system of an electric vehicle. The simulation adopts the traditional model algorithm control method and the proposed signal compensation model algorithm control method. The tracking performance of the two methods is shown in the figure Figure 2 , 3 shown.

[0070] Depend on Figure 2 It can be seen that under the influence of the same external noise, the proposed method can better control the air-conditioning drive motor speed to near the set value, while the traditional model algorithm control method has several large fluctuations, which greatly reduce the controller accuracy of the air-conditioning compressor speed.

[0071] Depend on Figure 3 It can be seen that the absolute value of the tracking error of the traditional model algorithm control method exceeds the specified error upper limit, and the time exceeding the upper limit accounts for 2.4% of the total simulation time. The proposed method controls the speed of the air-conditioning drive motor of electric vehicle B within the control requirement range, and the time exceeding the upper limit accounts for 0% of the total simulation time. The tracking performance of the system is improved.

[0072] In addition, the performance evaluation indicators error absolute value AD and integrated mean square error SD as shown in equations (18) and (19) are used to compare the control effects. The results are shown in Table 1.

[0073]

[0074] Table 1 Performance evaluation table of controlling the speed of electric vehicle air-conditioning compressor using the proposed control method and the traditional model algorithm control method

[0075]

[0076] It can be seen from Table 1 that when the traditional model algorithm control method is used, the AD of the electric vehicle air-conditioning driving motor speed is 15.13, and when the proposed control method is used, the AD of the electric vehicle air-conditioning driving motor speed is 7.14, which is reduced by 49.12% by the proposed method; when the traditional model algorithm control method is used, the SD of the electric vehicle air-conditioning driving motor speed is 18.81, and when the proposed control method is used, the SD of the electric vehicle air-conditioning driving motor speed is 9.57, which is reduced by 48.84% by the proposed method.

[0077] In summary, the signal compensation model algorithm control method can be effectively applied to the speed control system of the electric vehicle air-conditioning drive motor. Through the analysis of simulation results, this method has a good control effect compared with the traditional model algorithm control method. Therefore, the signal compensation model algorithm control method can better meet the actual control needs, overcome the limitations of the traditional control method for complex interference, effectively utilize the system model, improve the overall performance of the system, and propose a new solution for the controller design in the field of advanced control. It has good application value and application prospects, and is of great significance to the rapid development of science and technology in the field of advanced control of my country's industrial processes.

Claims

1. A signal compensation model algorithm control method for an electric vehicle air conditioning drive system, characterized in that: The specific steps are as follows: Step 1: Establish the electric vehicle air conditioning drive motor speed control system model; The speed of the air-conditioning drive motor is closely related to the current and angular velocity of the motor during operation. During the operation of the motor, the magnitude of the current directly affects the output torque of the motor. The angular velocity is an important representation of the motor speed. Its change will directly affect the working state and performance of the motor, and then affect the motor speed. Controlling the speed of the electric vehicle air-conditioning drive motor is equivalent to controlling the motor current and angular velocity. Therefore, the mathematical model of the electric vehicle air-conditioning drive motor speed control system is established as follows: Where i is the stator winding current, ω is the angular velocity of the motor, r a is the winding wire resistance, k e is the line back electromotive force, U d is the DC bus voltage, L a is the equivalent line inductance of the winding, K T is the motor torque coefficient, B v is the viscous friction coefficient, T l is the load torque, J is the rotor moment of inertia; Step 2: Build a discrete state space model with dynamic characteristics that are not incorporated into the modeling process; Select state variable x = [iω] Τ As input, the control quantity u=U d , the feedback output signal y = ω, and the state space equation of the air conditioner drive motor is as follows: In the model algorithm control, the current and speed are assumed to be constant in each prediction step, so it is necessary to discretize the state space equation of the brushless DC motor. Formula (2) uses the forward difference discretization method to obtain: Among them, T s is the sampling time; The above discretized state space model can be simplified as: in, x(k+1) represents the system state at discrete time k+1, u(k) represents the control input at discrete time k, v(k) represents the dynamic characteristic term not included in the modeling process, y(k) represents the system output at discrete time k+1, A is the system state matrix, B is the system input matrix, and C is the system output matrix; Step 3: Design a signal compensation model algorithm controller; When constructing the signal compensation model algorithm controller, we regard the design of the controller u(k) as a fusion of three parts: first, the precise prediction controller u1(k) based on the conventional state space model, second, the efficient compensation controller u2(k) for the dynamic characteristic items that were not included in the modeling process at the previous moment, and finally, the optimized compensation controller u3(k) for the dynamic characteristic items that were not included in the modeling process. This combination makes the control of the system more comprehensive and flexible, and provides a higher level of prediction and adjustment capabilities for the system operation. The signal compensation state space model predictive controller can be expressed as: Where, l = 1, 2, 3, u(k) is the controller based on the signal compensation model algorithm; First, an accurate predictive controller based on a conventional state space model is designed, and the dynamic characteristic items that are not included in the modeling process are not considered here. At this time, the prediction model of the drive motor is: In order to solve the optimal control input sequence in the prediction time domain, the quadratic performance index function is derived, and the accurate predictive controller u1(k) of the traditional conventional state space model can be obtained as: u1(k)=-(Y in Τ QY in +R) -1 Y in Τ QY out x(k) (7) in, Y in is the system control input coefficient matrix, Y out is the system output matrix, Y is the system output matrix, x(k) represents the system state at discrete time k, U is the system input matrix, Q is the state compensation matrix of the state space model predictive controller, R is the input compensation matrix of the state space model predictive controller, the prediction stage length is N1, the control stage length is N2, and N1 ≥ N2; Secondly, for the dynamic characteristic items that were not included in the modeling process at the previous moment, the efficient compensation controller u2(k) is used to perform dynamic and static adjustment compensation on the dynamic characteristic items that were not included in the modeling process at the previous moment to ensure the control performance of the controller u2(k), that is, Finally, the optimal compensation controller for the dynamic characteristic terms that are not included in the modeling process is designed. u3(k) is designed to optimize the compensation of the output error. The relationship between the output error and the control input is: e(k+1)=e(k)-CAx(k)-CBK1x(k)-CBu3(k)-CΔv(k-1) (9) Among them, K1 is the control law gain of u1(k) at discrete time k, and Δv(k-1) is the dynamic increment that is not included in the modeling process; In order to design the optimal compensation controller u3(k) for the dynamic characteristic terms that are not included in the modeling process, the following performance indicators are introduced: J3=min||e Τ (k+1)Me(k+1)+u3 Τ (k+1)Nu3(k+1)|| 2 (10) Wherein, M is the compensation matrix of the optimized compensation controller, N is the input compensation matrix of the optimized compensation controller, and e(k+1) is the error between the actual output and the expected output of the system at discrete time k+1; At this time, the one-step optimal forecast of the tracking error e(k+1|k) is: e(k+1|k)=e(k)-CAx(k)-CBK1x(k)-CBu3(k)-CΔv(k-1) (11) Depend on The optimal compensation controller of the dynamic characteristic terms that are not included in the modeling process is obtained as follows: By designing a precise prediction controller u1(k) based on a conventional state-space model, an efficient compensation controller u2(k) for the dynamic characteristic items that were not included in the modeling process at the previous moment, and an optimized compensation controller u3(k) for the dynamic characteristic items that were not included in the modeling process, a signal compensation model algorithm controller is finally constructed.

Images