Automobile thermal management system control method based on multi-target model predictive control

Abstract

The invention relates to an automobile thermal management system control method based on multi-target model predictive control, and belongs to the technical field of automobile thermal management systems. The method comprises the steps of establishing a discrete state space model of the thermal management system, setting a state variable, a control variable, a disturbance variable and a state space equation of the discrete state space model of the thermal management system, and then establishing a multi-target cost function and constraint conditions with stability, safety and low energy consumption of the thermal management system as targets; and then, based on fuzzy control logic, combining with an actual working condition, dynamically calculating a weight of each cost in the multi-target cost function, and finally solving an optimization problem according to the multi-target cost function under the determined weight in a discrete state space model of the thermal management system to obtain an optimal control input. And the optimal control input is transmitted to an actuator to regulate and control the automobile thermal management system. The system can adapt to various working conditions and environments, battery safety and comfort of a passenger compartment can be guaranteed, and energy efficiency and the service life of the actuator can be both considered.

Description

Technical Field

[0001] This invention belongs to the technical field of automotive thermal management systems, and relates to a control method for automotive thermal management systems based on multi-objective model predictive control. Background Technology

[0002] With the rapid development of new energy vehicles, the vehicle thermal management system (TMS) plays an increasingly important role in ensuring the comfort of the power battery, drive motor, power electronics, and passenger compartment. Existing thermal management systems generally employ rule-based or PID control strategies. While these methods can achieve basic temperature regulation under single operating conditions, they often suffer from low energy efficiency, slow response speed, and insufficient control precision under complex and changing road environments, climate conditions, and driving conditions, making it difficult to meet the comprehensive requirements of high performance and long range for the entire vehicle.

[0003] Meanwhile, significant coupling effects exist between thermally managed objects. For example, energy distribution between the power battery and the electric drive system during heat dissipation, and the mutual influence between the passenger compartment air conditioning and battery cooling circuits, further increase the complexity of system control. Traditional control methods lack the ability to effectively model and predict multivariable coupling relationships, making it difficult to achieve global optimization under multiple operating conditions.

[0004] Model Predictive Control (MPC), as an advanced control method, has the advantage of simultaneously handling multiple input and output variables, constraints, and optimization objectives within a finite prediction time domain, and has been widely researched and applied in process industries, energy systems, and other fields. However, the application of MPC in automotive thermal management systems is still in the exploratory stage, especially in handling real-time optimization and control under multi-condition operating environments, where technical challenges such as high computational complexity, high model accuracy requirements, and difficult hardware implementation remain. Therefore, it is necessary to propose an MPC-based controller for automotive thermal management systems that can adapt to multiple operating conditions to improve the system's energy efficiency, stability, and intelligence level. Summary of the Invention

[0005] In view of this, the purpose of this invention is to provide a control method for an automotive thermal management system based on multi-objective model predictive control, which can coordinate the control of the power battery, drive motor, electronic control unit and passenger compartment air conditioning under different ambient temperatures, driving conditions and load conditions, thereby achieving efficient heat dissipation, energy optimization and stable operation of the whole vehicle, which falls within the scope of automotive thermal management system design and control engineering applications.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A control method for an automotive thermal management system based on multi-objective model predictive control, the method comprising:

[0008] S1. Establish a discrete state-space model of the thermal management system, and set the state variables, control variables, disturbance variables, and state-space equations of the discrete state-space model of the thermal management system.

[0009] S2. Establish a multi-objective cost function with the goals of stability, safety, and low energy consumption of the thermal management system;

[0010] S3. Establish constraints for the multi-objective cost function;

[0011] S4. Based on fuzzy control logic and combined with actual working conditions, dynamically calculate the weight of each cost in the multi-objective cost function;

[0012] S5. Solve the optimization problem in the discrete state space model of the thermal management system according to the determined multi-objective cost function under the determined weights to obtain the optimal control input. The optimal control input is then transmitted to the actuator to regulate the automotive thermal management system.

[0013] Furthermore, in S1, the state variables of the established discrete state-space model of the thermal management system are represented as follows:

[0014] x = [T] cab ,T bat ] T

[0015] In the formula, T cab ,T bat These represent the passenger compartment temperature and the average temperature of the power battery, respectively.

[0016] The control variables of the established discrete state-space model of the thermal management system are represented as follows:

[0017] u = [n com ,n fan ] T

[0018] In the formula, n com ,n fan These represent the rotational speeds of the compressor and the cooling fan, respectively.

[0019] The disturbance variables of the established discrete state-space model of the thermal management system are represented as follows:

[0020] v = [SOC,S1,S2,N] pump ,V] T

[0021] Wherein, SOC, S1, S2, N pumpV represents the battery SOC, the three-way valve status, the Chier circuit solenoid valve status, the water pump speed, and the vehicle speed, respectively.

[0022] Furthermore, in S1, the state-space model combines the control variables and disturbance variables as inputs, and performs calculations with the correlation coefficient matrix to predict the state variables of the controlled system model. The input variables are:

[0023] w = [SOC,n] com ,n fan ,S1,S2,N pump ,V] T

[0024] The output variables of the state-space equations are the same as the state variables, taking the passenger compartment temperature and the average temperature of the power battery as the values:

[0025] y = [T cab ,T bat ] T

[0026] When performing black-box modeling of the integrated battery thermal management system, the output variables are the same as the state variables, the output matrix C is the identity matrix, the direct transfer matrix D is the zero matrix, and the system matrix A and input matrix B are the parameters to be identified.

[0027]

[0028]

[0029] In a state prediction process, the state equation X(k+1)=AX(k)+BU(k) is executed first, followed by the output equation Y(k)=CX(k)+DU(k). The logical relationship between the two and the function of each matrix are as follows:

[0030] System matrix A: describes the degree of influence of the current state variables on the state at the next time step. Where, a 11 This represents the influence coefficient of the current cabin temperature on the cabin temperature at the next moment X(k+1), reflecting the thermal inertia of the cabin; a 12 This represents the coefficient indicating the influence of the current average temperature of the power battery on the temperature of the passenger compartment at the next moment, reflecting the thermal coupling between the two; a 21 This represents the coefficient representing the influence of the current cabin temperature on the average temperature of the power battery at the next moment; a 22 This represents the influence coefficient of the current average temperature of the power battery on the average temperature of the power battery at the next moment, reflecting the thermal inertia of the battery itself. These coefficients need to be obtained through system identification experiments combined with algorithms such as the least squares method. Input matrix B: quantifies the effect of the input variable U(k) on the state X(k+1) at the next moment. The element b in the matrix corresponding to the control variable U(k) 11 ,b12 ,b 21 ,b 22 To control the gain, for example b 11 This represents the adjustment coefficient of the compressor speed change on the cabin temperature at the next moment; a negative value indicates that an increase in compressor speed will decrease the cabin temperature. The element b corresponds to the disturbance variable D. 13 -b 17 ,b 23 -b 27 For disturbance gain. Output matrix C: Since the output variable Y(k) is completely consistent with the state variable X(k), an identity matrix is ​​used to ensure that Y(k) can completely and without distortion reflect the current state of X(k), providing accurate feedback information to the controller without the need for additional signal conversion or correction. Direct transfer matrix D: Temperature changes in the thermal management system have a hysteresis effect; the control variable or disturbance variable needs to go through a heat exchange process to change the temperature state. There is no instantaneous response where "the input variable directly acts on the output variable." Therefore, D is set as a zero matrix, which conforms to the physical characteristics and dynamic response law of the system.

[0031] Furthermore, in S2, the established multi-objective cost function is expressed as:

[0032] J = αJ y +βJ u +λJ Δu

[0033] Among them, J y The cost function, representing the performance of the thermal management system, is essentially a function of the difference between the predicted model output and the setpoint, characterizing the magnitude of the deviation in system output, including deviations in cabin and battery temperatures; J u The cost function representing the energy consumption of a thermal management system is essentially a function related to the system's control variables; the smaller the control variables, the lower the system's energy consumption. Δu The cost function used to measure the stability of a thermal management system represents the change in the controller's output control variable; the smaller the change, the more stable the system operation. α is the output error cost J. y The weights are used to adjust the degree of emphasis on the tracking accuracy of the reference trajectory, and β is the control input cost J. u The weights are used to limit the magnitude of the control variable itself, and λ is the control increment cost J. Δu The weights are used to smooth the controller's actions.

[0034] Furthermore, in S2, the cost function J y J u and J Δu They are represented as follows:

[0035]

[0036] Where, n y and n u These represent the number of system output variables and control variables, respectively; p and c represent the prediction time domain and control time domain set by the model predictive controller, respectively. r is the weighting coefficient of the j-th output variable of the prediction model at time k for the i-th value; j (k+i|k) is the reference value of the j-th output variable of the controller; y j (k+i|k) is the prediction output of the prediction model at time k; The weighting coefficients represent the weights of the j-th control variable sequence output by the model predictive controller at time k, specifically the i-th value. j (k+i|k) represents the control variable of the controller at time k; The weighting coefficient is used to determine the rate of change of the i-th value of the j-th control variable sequence output by the model predictive controller at time k.

[0037] Furthermore, in S3, considering that the compressor and cooling fan have certain limitations in terms of speed and rate of change of speed under actual operating conditions, the following constraints are established:

[0038]

[0039] in, This indicates the compressor's maximum speed. This represents the maximum threshold for the rate of change of compressor speed. This indicates the fan's maximum speed. This represents the maximum threshold for the rate of change of fan speed.

[0040] Furthermore, in S4, by flexibly adjusting the weights of these three factors under different operating conditions, the optimal balance between safety, energy efficiency, and stability of the thermal management system can be achieved.

[0041] When the battery or passenger compartment temperature is significantly different from the suitable target temperature, α needs to be increased to ensure precise control of the battery and passenger compartment temperature.

[0042] When the vehicle's battery is low or it is in a range-sensitive scenario, β needs to be increased to limit energy consumption output.

[0043] When there are significant fluctuations in the control, it is necessary to increase λ to avoid excessively frequent and drastic actuator control commands.

[0044] Furthermore, in S4, the specific process of adjusting weights based on operating conditions is as follows:

[0045] The input vector of the fuzzer is in:

[0046]

[0047] All the above inputs are normalized to the interval [0,1], where x1, x2, and x3 correspond to the magnitude of the temperature difference, the level of power shortage, and the degree of fluctuation in the control quantity, respectively; T bat T represents the actual temperature of the battery. battar Set the temperature for the battery, T pas The actual temperature in the crew cabin, T pastar To set the temperature in the crew cabin, T max The maximum allowable temperature difference; Δu is the rate of change of the control variable, Δu max This refers to the maximum permissible variation of the control quantity.

[0048] Each input variable uses a triangular membership function, divided into three linguistic variables: Low (L), Medium (M), and High (H). A zero-order Sugeno model is used. The rule form is as follows:

[0049] R k :IF x1 is AND x2 is AND x3 is

[0050] THEN(r y ,r u ,r Δu )=(α (k) ,β (k) ,λ (k) )

[0051] Let r represent the linguistic variables that indicate the magnitude of the temperature difference; y ,r u ,r Δu α represents the output of the fuzzy rule. (k) ,β (k) ,λ (k) These represent the weights corresponding to the k-th rule;

[0052] Then, reasoning and deblurring are performed, and the activation degree of the k-th rule is defined as:

[0053]

[0054] The output score of the fuzzer is:

[0055]

[0056] Finally, the corresponding weight values ​​are calculated based on the rule table. The rule table needs to be set separately in advance, and its formulation should be combined with the operating characteristics and control requirements of the thermal management system. Specific basis includes: engineering experience summary, based on the actual operating rules of the automotive thermal management system; simulation and experimental optimization: a simulation model of the thermal management system is built using MATLAB / Simulink to simulate typical operating conditions and quantitatively evaluate the system performance under different rule combinations, and finally determine the weight values ​​of the rules to ensure that the rule table can cover multiple operating scenarios and achieve the optimal balance between temperature, energy consumption, and stability.

[0057] Furthermore, in S5, the controller collects vehicle operating data in real time, combines the typical characteristics of each operating condition, and uses the MPC algorithm to predict temperature changes and energy consumption in the future prediction time domain. Through rolling optimization, the optimal control input is continuously calculated and transmitted to the actuator in real time. Under the premise of ensuring battery safety and passenger cabin comfort, the compressor energy consumption is minimized, and the dynamic thermal balance and energy efficiency improvement of the whole vehicle under high temperature conditions are achieved.

[0058] The beneficial effects of this invention are as follows:

[0059] Compared to existing MPCs, this invention introduces a multi-objective MPC with weighting coefficients, which can achieve a flexible balance between output accuracy, energy consumption optimization, and control stability, thereby improving the overall performance of the system under different operating conditions. Without weighting coefficients, the controller can only optimize under a single objective, which can easily lead to results biased towards one aspect. However, by setting weighting coefficients, the optimization focus can be dynamically adjusted according to the vehicle's operating environment and needs, enabling the thermal management system to adapt to various operating conditions. This ensures both battery safety and passenger cabin comfort, while also considering energy efficiency and actuator lifespan, better meeting the complexity and diversity of actual vehicle operation.

[0060] The multi-objective MPC of this invention helps to overcome the problems of low energy efficiency, slow response speed and insufficient control accuracy of existing rule-based or PID-based control strategies under complex and changing road environments, climate conditions and driving conditions. It can adapt to the multi-condition automotive thermal management system controller to improve the system's energy efficiency, stability and intelligence level.

[0061] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0062] To make the objectives, technical solutions, and advantages of the present invention clearer, the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, wherein:

[0063] Figure 1 This is a schematic diagram of the overall process of the automotive thermal management system control method based on multi-objective model predictive control according to an embodiment of the present invention. Detailed Implementation

[0064] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0065] The accompanying drawings are for illustrative purposes only and are schematic diagrams, not actual pictures. They should not be construed as limiting the invention. To better illustrate the embodiments of the invention, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0066] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components. In the description of the present invention, it should be understood that if terms such as "upper," "lower," "left," "right," "front," and "rear" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the drawings are only for illustrative purposes and should not be construed as limiting the present invention. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0067] Please see Figure 1 This is a control method for an automotive thermal management system based on multi-objective model predictive control.

[0068] Example

[0069] This embodiment provides a detailed implementation scheme for a vehicle thermal management system control method based on multi-objective model predictive control, such as... Figure 1 As shown, the overall steps include:

[0070] Step 1: Establish a discrete state-space model of the thermal management system, and set the state variables, control variables, disturbance variables, and state-space equations of the discrete state-space model of the thermal management system, as follows:

[0071] In this embodiment, the purpose of the established model predictive controller is to coordinate the control of the compressor and cooling fan while meeting the cooling requirements of the power battery and passenger compartment under different operating conditions. Therefore, the passenger compartment temperature and the average temperature of the power battery are selected as state variables:

[0072] x = [T] cab ,T bat ] T

[0073] In the formula, T cab ,T bat These represent the passenger compartment temperature and the average temperature of the power battery, respectively.

[0074] In this embodiment, the compressor and cooling fan are shared by the battery and passenger compartment temperature control systems, and their operating states directly affect the temperature performance of the battery and passenger compartment. The compressor and cooling fan speeds are used as control variables for the model predictive controller.

[0075] u = [n com ,n fan ] T

[0076] In the formula, n com ,n fan These represent the rotational speeds of the compressor and the cooling fan, respectively.

[0077] In this embodiment, besides its own temperature, the battery's SOC and charging / discharging current are also important influencing factors at the heat-generating end of the battery; the power battery's heat dissipation end is mainly affected by the cooling mode and the water pump duty cycle. Passenger compartment temperature control mainly considers the compressor and cooling fan, while vehicle speed also affects the passenger compartment's thermal load. Therefore, this embodiment uses battery SOC, three-way valve status, Chier circuit solenoid valve status, water pump speed, and vehicle speed as perturbation variables for the model predictive controller.

[0078] v = [SOC,S1,S2,N] pump ,V] T

[0079] Wherein, SOC, S1, S2, N pump V represents the battery SOC, the three-way valve status, the Chier circuit solenoid valve status, the water pump speed, and the vehicle speed, respectively.

[0080] The state-space model takes the control variables and disturbance variables as input, and performs calculations with the correlation coefficient matrix to predict the state variables of the controlled system model. The input variables are:

[0081] w = [SOC,n] com ,n fan,S1,S2,N pump ,V] T

[0082] The output variables of the state-space equations are the same as the state variables, taking the passenger compartment temperature and the average temperature of the power battery as the values:

[0083] y = [T cab ,T bat ] T

[0084] When performing black-box modeling of an integrated battery thermal management system, if the output variables are the same as the state variables, then the output matrix C is the identity matrix, and the direct transfer matrix D is the zero matrix. As shown in the following equation, the system matrix A and the input matrix B are the parameters to be identified.

[0085]

[0086] In a state prediction process, the state equation X(k+1)=AX(k)+BU(k) is executed first, followed by the output equation Y(k)=CX(k)+DU(k). The logical relationship between the two and the function of each matrix are as follows:

[0087] System matrix A: describes the degree of influence of the current state variables on the state at the next time step. Where, a 11 This represents the influence coefficient of the current cabin temperature on the cabin temperature at the next moment X(k+1), reflecting the thermal inertia of the cabin; a 12 This represents the coefficient indicating the influence of the current average temperature of the power battery on the temperature of the passenger compartment at the next moment, reflecting the thermal coupling between the two; a 21 This represents the coefficient representing the influence of the current cabin temperature on the average temperature of the power battery at the next moment; a 22 This represents the influence coefficient of the current average temperature of the power battery on the average temperature of the power battery at the next moment, reflecting the thermal inertia of the battery itself. These coefficients need to be obtained through system identification experiments combined with algorithms such as the least squares method. Input matrix B: quantifies the effect of the input variable U(k) on the state X(k+1) at the next moment. The element b in the matrix corresponding to the control variable U(k) 11 ,b 12 ,b 21 ,b 22 To control the gain, for example b 11 This represents the adjustment coefficient of the compressor speed change on the cabin temperature at the next moment; a negative value indicates that an increase in compressor speed will decrease the cabin temperature. The element b corresponds to the disturbance variable D. 13 -b 17 ,b 23 -b 27For disturbance gain. Output matrix C: Since the output variable Y(k) is completely consistent with the state variable X(k), an identity matrix is ​​used to ensure that Y(k) can completely and without distortion reflect the current state of X(k), providing accurate feedback information to the controller without the need for additional signal conversion or correction. Direct transfer matrix D: Temperature changes in the thermal management system have a hysteresis effect; the control variable or disturbance variable needs to go through a heat exchange process to change the temperature state. There is no instantaneous response where "the input variable directly acts on the output variable." Therefore, D is set as a zero matrix, which conforms to the physical characteristics and dynamic response law of the system.

[0088] Step 2: Construct the multi-objective cost function:

[0089] In an integrated battery thermal management system, the control objective for the compressor and cooling fan speeds is to reduce the energy consumption of the thermal management system while meeting the temperature requirements of the power battery and passenger compartment. Furthermore, a gradual change in the control variables can make the controlled system operate more smoothly. Therefore, the objective function of the model predictive controller's rolling optimization designed in this embodiment mainly consists of three parts:

[0090] J = αJ y +βJ u +λJ Δu

[0091]

[0092] In the formula, J y J is a function representing the difference between the predicted model output and the setpoint, characterizing the magnitude of the system output deviation, including deviations in cabin and battery temperatures; u It is a function related to the system's control variables; the smaller the control variables, the lower the system's energy consumption. Δu The minimum value of the change in the controller output variable represents the stability of the system. y and n u These represent the number of system output variables and control variables, respectively; p and c represent the prediction time domain and control time domain set by the model predictive controller, respectively. r is the weighting coefficient of the j-th output variable of the prediction model at time k for the i-th value; j (k+i|k) is the reference value of the j-th output variable of the controller; y j (k+i|k) is the prediction output of the prediction model at time k; The weighting coefficients represent the weights of the j-th control variable sequence output by the model predictive controller at time k, specifically the i-th value. j (k+i|k) represents the control variable of the controller at time k; The weighting coefficient for the rate of change of the j-th control variable sequence output by the model predictor at time k is the weighting coefficient for the i-th value.y The weights are used to adjust the degree of emphasis on the tracking accuracy of the reference trajectory, and β is the control input cost J. u The weights are used to limit the magnitude of the control variable itself, avoiding excessively large inputs. λ is the control increment cost J. Δu The weights are used to smooth the controller's actions and prevent excessive jumps in the input signal.

[0093] Step 3: Set constraints:

[0094] Considering that the compressor and cooling fan have certain limitations in terms of speed and rate of change of speed under actual operating conditions, the following constraints are established:

[0095]

[0096] in, This indicates the compressor's maximum speed. This represents the maximum threshold for the rate of change of compressor speed, in this embodiment. The speed is 6000 r / min. It is 70. This indicates the fan's maximum speed. This represents the maximum threshold for the rate of change of fan speed; in this embodiment, the maximum fan speed is 4300 r / min. The value is 70; therefore, the control variable constraints are as follows when solving for the optimal control sequence:

[0097] 0≤N com ≤6000

[0098] -70≤ΔN com ≤70

[0099] 0≤N fan ≤4300

[0100] -70≤ΔN fan ≤70

[0101] Step 4: Calculate the weighting coefficients for different control objectives under different actual vehicle operating conditions:

[0102] Under different real-world vehicle operating conditions, the control objectives differ. In model predictive control, the weighting coefficients reflect the trade-offs between different objectives. Adjusting these coefficients under different operating conditions is crucial: when the battery or passenger compartment temperature deviates significantly from the target temperature, α should be appropriately increased to ensure precise temperature control of the battery and passenger compartment, prioritizing safety and comfort; when the vehicle's battery is low or in a range-sensitive scenario, β needs to be increased to limit the energy consumption output of components such as the compressor and water pump, thereby improving overall vehicle energy efficiency and driving range; and when control fluctuations are large, λ should be increased to avoid excessively frequent and drastic actuator control commands, thus achieving smooth adjustment and extending hardware lifespan. By flexibly adjusting the weights of these three factors under different operating conditions, the optimal balance between safety, energy efficiency, and stability in the thermal management system can be achieved. Table 1 shows the demand relationship between typical operating conditions and control objectives:

[0103] Table 1

[0104]

[0105]

[0106] To calculate the weighting coefficients suitable for various operating conditions, fuzzy control logic is used to automatically calculate each weight:

[0107] To achieve adaptive weight allocation, a weight regulator based on fuzzy logic is introduced. This regulator takes battery temperature difference, SOC, and control variable change rate as inputs, outputs three weight coefficients through fuzzy inference, and obtains the final weights through normalization.

[0108] The input vector of the weight regulator based on fuzzy control is in:

[0109]

[0110] All the above inputs are normalized to the interval [0,1], where x1, x2, and x3 correspond to the magnitude of the temperature difference, the level of power shortage, and the degree of fluctuation in the control quantity, respectively; T bat T represents the actual temperature of the battery. battar Set the temperature for the battery, T pas The actual temperature in the crew cabin, T pastar To set the temperature in the crew cabin, T max The maximum allowable temperature difference; Δu is the rate of change of the control variable, Δu max This refers to the maximum permissible variation of the control quantity.

[0111] Each input variable uses a triangular membership function, divided into three linguistic variables: Low (L), Medium (M), and High (H). A zero-order Sugeno model is used. The rule form is as follows:

[0112] R k :IF x1 is AND x2 is AND x3 is

[0113] THEN(r y ,r u ,r Δu )=(α (k) ,β (k) ,λ (k) )

[0114] Let r represent the linguistic variables that indicate the magnitude of the temperature difference; y ,r u ,r Δu α represents the output of the fuzzy rule. (k) ,β (k) ,λ (k) These represent the weights corresponding to the k-th rule;

[0115] Then, reasoning and deblurring are performed, and the activation degree of the k-th rule is defined as:

[0116]

[0117] The output score of the fuzzer is:

[0118]

[0119] The complete rule table is shown in Table 2 below:

[0120] Table 2

[0121]

[0122]

[0123] By substituting the weights into the cost function, the optimization problem can be solved in real time.

[0124] Step 5: Predictive Control and Rolling Optimization Execution:

[0125] The controller collects vehicle operating data in real time, combines the typical characteristics of each operating condition, and uses the MPC algorithm to predict temperature changes and energy consumption in the future prediction time domain.

[0126] Based on this, the optimal control input is continuously calculated through rolling optimization. The optimization results are transmitted to the actuators in real time, thereby minimizing compressor energy consumption while ensuring battery safety and passenger cabin comfort, and achieving dynamic thermal balance and energy efficiency improvement for the entire vehicle under high-temperature conditions.

[0127] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for controlling an automotive thermal management system based on multi-objective model predictive control, characterized in that: The method comprises: S1, a thermal management system discrete state space model is established, state variables, control variables, disturbance variables and state space equations of the thermal management system discrete state space model are set; S2, a multi-objective cost function taking the stability, safety and low energy consumption of the thermal management system as the target is established; S3, a constraint condition of the multi-objective cost function is established; S4, weights of each cost in the multi-objective cost function are dynamically calculated based on fuzzy control logic and in combination with actual working conditions; S5, an optimal control input is obtained by solving an optimization problem in the thermal management system discrete state space model according to the multi-objective cost function under the determined weights, and the optimal control input is transmitted to an actuator to regulate and control the automobile thermal management system.

2. The control method of the automobile thermal management system based on multi-objective model predictive control according to claim 1, characterized in that: In S1, the state variables of the established thermal management system discrete state space model are represented as: x = [T cab , T bat ] T In the formula, T cab bat respectively represent the passenger cabin temperature and the average temperature of the power battery.​ The control variables of the established thermal management system discrete state space model are represented as: u = [n com ,n fan ] T In the formula, n com ,n fan respectively represent the rotational speed of the compressor and the cooling fan. The disturbance variables of the established thermal management system discrete state space model are represented as: v = [SOC, S1, S2, N pump T ​ Wherein, SOC, S1, S2, N pump V respectively represent battery SOC, three-way valve state, Chier circuit solenoid valve state, water pump speed and vehicle speed.

3. The control method of the automobile thermal management system based on multi-objective model predictive control according to claim 2, characterized in that: In S1, the state space model combines the control variables and the disturbance variables for input, and performs operation on the related coefficient matrix to predict the state variables of the controlled system model, wherein the input variables are: w = [SOC, n com , n fan , S1, S2, N pump , V] T The output variables and the state variables of the state space equation are the same, and the passenger cabin temperature and the average temperature of the power battery are taken: y = [T cab ,T bat ] T When the integrated battery thermal management system black box modeling is performed, the output variables and the state variables are the same, the output matrix C is a unit matrix, the direct transmission matrix D is a zero matrix, and the system matrix A and the input matrix B are to-be-identified parameters: In a state prediction process, the state equation X(k+1)=AX(k)+BU(k) is executed first, and then the output equation Y(k)=CX(k)+DU(k) is executed, and the logical relationship and the action of each matrix of the two are as follows: The system matrix A describes the influence degree of the current time state variable on the next time state, wherein, a 11 represents the influence coefficient of the current passenger compartment temperature on the next time X(k+1) passenger compartment temperature, reflecting the thermal inertia of the passenger compartment; a 12 represents the influence coefficient of the current average temperature of the power battery on the next time passenger compartment temperature, embodying the thermal coupling between the two; a 21 represents the influence coefficient of the current passenger compartment temperature on the next time average temperature of the power battery; a 22 represents the influence coefficient of the current average temperature of the power battery on the next time average temperature of the power battery, reflecting the thermal inertia of the battery itself; The input matrix B quantifies the effect of the input variable U(k) on the next state X(k+1). The element b 11 12 21 22 is the control gain and the element b 13 17 23 27 is the disturbance gain.​​​​​​ The output matrix C adopts the unit matrix to ensure that Y(k) can completely reflect the current state of X(k), thereby providing accurate feedback information for the controller; The direct transmission matrix D reflects the physical characteristics and dynamic response law of the system through the zero matrix.

4. The control method of the automobile thermal management system based on multi-objective model predictive control according to claim 1, characterized in that: In S2, the established multi-objective cost function is represented as: J = aJ y + βJ u + λJ Δu wherein J y represents a cost function measuring the thermal management performance, which is essentially a function of the difference between the predicted model output value and the set value, representing the size of the system output deviation, including the passenger cabin and battery temperature deviation; J u represents a cost function measuring the energy consumption of the thermal management system, which is essentially a function related to the system control variable, the smaller the control variable, the lower the system energy consumption; J Δu is a cost function measuring the stability of the thermal management system, which represents the change amount of the controller output control variable, the smaller the change amount, the more stable the system runs; α is the weight of the output error cost J y , used to adjust the importance of the reference trajectory tracking accuracy, β is the weight of the control input cost J u , used to limit the amplitude of the control variable itself, and λ is the weight of the control increment cost J Δu , used to smooth the action of the controller.

5. The control method of the automobile thermal management system based on multi-objective model predictive control according to claim 4, characterized in that: In S2, the cost function J y , J u , and J Δu are respectively expressed as: where n y and n u represent the number of system output variables and control variables, respectively; p and c represent the prediction horizon and control horizon set by the model predictive controller, respectively; is the weight coefficient of the jth output variable of the predictive model at the kth time and the ith value; r j (k+i|k) is the reference value of the jth output variable of the controller; y j (k+i|k) is the predicted output of the predictive model at the kth time; is the weight coefficient of the jth control variable sequence output by the model predictive controller at the kth time and the ith value; u j (k+i|k) is the control variable of the controller at the kth time; is the weight coefficient of the rate of change of the jth control variable sequence output by the model predictive controller at the kth time and the ith value.

6. The control method of the automobile thermal management system based on multi-objective model predictive control according to claim 1, characterized in that: In S3, considering that the speed and the speed change rate of the compressor and the cooling fan exist certain limitations in the actual running working conditions, the following constraint conditions are established: wherein, represents the maximum rotational speed of the compressor, represents the maximum threshold of the rate of change of the rotational speed of the compressor; represents the maximum rotational speed of the fan, represents the maximum threshold of the rate of change of the rotational speed of the fan.

7. The control method of the automobile thermal management system based on multi-objective model predictive control according to claim 4, characterized in that: In S4, by flexibly adjusting the weights of the three under different working conditions, the optimal balance among the safety, energy efficiency and stability of the thermal management system can be realized, wherein, When the battery or the passenger cabin temperature is larger than the appropriate target temperature, alpha needs to be increased to ensure the accurate control of the battery and the passenger cabin temperature; When the vehicle power is insufficient or is in a range anxiety scenario, beta needs to be increased to limit the energy consumption output; When the control fluctuation changes obviously, lambda needs to be increased to avoid that the actuator control instruction is too frequent and intense.

8. The control method of the automobile thermal management system based on multi-objective model predictive control according to claim 4, characterized in that: In S4, the specific process of adjusting the weights according to the working conditions is as follows: The input vector to the fuzzer is x in = [x1, x2, x3] T where: The above inputs are all normalized to the interval [0, 1], x1, x2, x3 respectively correspond to the temperature difference size, the power shortage degree, and the control quantity fluctuation degree; T bat is the actual temperature of the battery, T battar is the set temperature of the battery, T pas is the actual temperature of the passenger compartment, T pastar is the set temperature of the passenger compartment, T max is the maximum allowed temperature difference; Δu is the control quantity change rate, Δu max is the maximum allowed control quantity change; Each input variable adopts a triangular membership function, is divided into three language variables of Low (L), Medium (M) and High (H), adopts a zero-order Sugeno model, and the rule form is: Let r represent the linguistic variables that indicate the magnitude of the temperature difference; y ,r u ,r Δu α represents the output of the fuzzy rule. (k) ,β (k) ,λ (k) These represent the weights corresponding to the k-th rule; Then reasoning and defuzzification are performed, the activation degree of the kth rule is defined as: The output score of the fuzzy ware is: Finally, the corresponding weight value is calculated according to the rule table, which is set separately in advance and is formulated in combination with the operating characteristics and control requirements of the thermal management system. The specific basis includes: engineering experience summary, actual operation law based on the automotive thermal management system; simulation and experimental optimization: a thermal management system simulation model is built through MATLAB / Simulink, the system performance under different rule combinations is quantitatively evaluated under typical working conditions, and finally the weight value of the rule is determined to ensure that the rule table can cover multiple working condition scenarios and achieve the optimal balance of temperature-energy consumption-stability.

9. The method of claim 1, wherein the method is based on a multi-objective model predictive control for an automotive thermal management system. In S5, the controller collects vehicle operating data in real time, combines the typical characteristics of each working condition, uses the MPC algorithm to predict the temperature change and energy consumption in the future prediction time domain, continuously calculates the optimal control input through rolling optimization, and the optimal control input is transmitted to the actuator in real time. Under the premise of ensuring the safety of the battery and the comfort of the passenger compartment, the compressor energy consumption is minimized to achieve dynamic thermal balance and energy efficiency improvement of the whole vehicle under high temperature working conditions.

Images