Thermal management control method for fuel cell hybrid vehicles

Abstract

To disclose a fuel cell hybrid electric vehicle thermal management control method, relating to a technical field of fuel cell hybrid electric vehicle thermal management.SOLUTION: A thermal management control method for a fuel cell hybrid electric vehicle optimizes a flow of coolant liquid using an MPC in a multi-input multi-output complex nonlinear system of the fuel cell hybrid electric vehicle to ensure that respective parts are in a normal working temperature range. Then, the method takes inlet and outlet temperatures of the respective parts as state variables and a flow of coolant liquid of a pump of each component as control quantity to cooperatively control the temperatures of the respective parts using an MPC controller and considers relevance of heat management systems of respective circuits to improve the overall performance of the fuel cell hybrid electric vehicle.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present application relates to the technical field of thermal management of fuel cell hybrid vehicles, and more particularly to a thermal management control method for fuel cell hybrid vehicles. [Background technology]

[0002] Fuel cell hybrid vehicles (FVs) offer advantages such as fast hydrogen refueling, long driving range, high energy conversion efficiency, and environmental friendliness, making them the ultimate transportation solution. Unfortunately, excessively high temperatures can significantly affect the performance, operating efficiency, and lifespan of fuel cells, traction batteries, and electric motors. As driving conditions change, heat generation is inevitable, and each component has its own normal operating temperature range, making heat dissipation even more difficult. The control accuracy of the thermal management system significantly affects the steady-state and transient power output characteristics of FVs. Therefore, an efficient and reliable thermal management control policy is crucial for the operation and safety of FVs.

[0003] Currently, it is difficult to build a comprehensive and accurate control system for the complex multi-input, multi-output, nonlinear systems of the entire vehicle. Summary of the Invention [Problem to be solved by the invention]

[0004] In contrast to the shortcomings of the prior art, the present application provides a thermal management control method for a fuel cell hybrid vehicle, which has advantages such as accurate temperature control, and solves the above technical problems. [Means for solving the problem]

[0005] To achieve the above object, the present application proposes the following technical solution: A thermal management control method for a fuel cell hybrid vehicle includes the following steps.

[0006] In step S1, performance data of the fuel cell hybrid vehicle is acquired, and a cycle mode test is performed on the entire vehicle. Cell stack The system records current, cell stack output voltage, fuel cell coolant inlet / outlet temperatures, cell stack coolant flow rate, drive battery inlet / outlet temperatures, battery coolant flow rate, chiller temperature, drive battery output current, drive battery voltage and drive battery internal temperature, electric motor coolant flow rate, cabin inlet / outlet temperatures, air conditioner refrigerant flow rate, electric motor inlet / outlet temperatures, electric motor pump flow rate, total electric motor power demand, electric motor and power control unit equivalent efficiency, air flow rate, and evaporator air flow rate data in real time, and obtains solenoid valve opening / closing signals, such as the cell stack coolant circulation opening / closing signal and the drive battery chiller circuit opening / closing signal, as well as cabin-related parameters, including ambient temperature, window solar transmittance, solar radiation intensity, windshield, rear window, and side window areas, shading coefficient, angle between the window and the vertical direction, convective heat exchange coefficient between the body shell and the cabin, heat transfer area of ​​the roof panel, windshield, rear window, and side windows, number of people in the cabin, cabin air temperature, air flow rate passing through the evaporator, and specific heat capacity of the cabin air.

[0007] In step S2, a thermodynamic model of the vehicle is constructed, which specifically includes the following steps:

[0008] In step S2.1, the cell stack heat generation power Q is calculated from the cell stack output voltage and cell stack current in step S1. gen The specific formula is: JPEG0007817511000001.jpg14168, Here, Q gen represents the cell stack heat generation power, and E nernst represents the Nernst voltage, which is typically 1.2 V, and N FC Indicates the number of cells in the fuel cell stack, V st is a cell stack output represents the voltage, and I st is a cell stack Kuden Represents flow.

[0009] In step S2.2, the battery heat generation power Q is calculated from the drive battery output current, drive battery voltage, and drive battery internal temperature in step S1. bat The specific formula is: JPEG0007817511000002.jpg14165 Here, Q bat represents the battery heat generation power, and I bat is the drive battery output represents the current, and R esistance represents the electrical resistance, and T amb represents the ambient temperature, JPEG0007817511000003.jpg2228 represents the temperature effect coefficient.

[0010] In step S2.3, the conductive heat load Q is calculated from the cabin-related parameters already obtained in step S1. rb , radiant heat load Q cb , the heat generated by the passengers in the car Q h and ventilation system heat load Q n Calculate the cabin heat generation power Q cab The specific formula is JPEG0007817511000004.jpg59162, Here, Q rb represents the conductive heat load, ξ represents the solar transmittance of the window, I represents the solar radiation intensity, and F g,j represents the area of ​​the roof panel, windshield, rear window, and side window, C represents the shading coefficient, and θ j represents the angle between the window and the vertical direction, JPEG0007817511000005.jpg21164 represents the sequential addition of calculations for the roof panel, windshield, rear window and side window, and Q cb represents the radiant heat load, and h b represents the convective heat exchange coefficient between the body shell and the cabin, F represents the heat transfer area of ​​the roof panel, windshield, rear window, and side window, and T amb represents the ambient temperature, and T cabin,inrepresents the cabin air temperature, and Q h represents the heat generated by the passengers inside the vehicle, and n m represents the number of people in the cabin, and Q n represents the ventilation system heat load, and m e represents the air flow rate passing through the evaporator, and C cabin represents the specific heat capacity of the cabin air.

[0011] In step S2.4, the electric motor heat generation power is calculated from the electric motor parameters obtained in step S1, and the specific formula is: JPEG0007817511000006.jpg14165, Here, Q mot represents the heat generation power of the electric motor, P represents the total demand power of the electric motor, and η mot / PCU represents the equivalent efficiency of the electric motor and PCU, where PCU is the power control unit.

[0012] In step S2.5, a thermodynamic model of the vehicle is constructed based on the first law of thermodynamics from the heat generation power of each component obtained in steps S2.1 to S2.4. The specific formula is: JPEG0007817511000007.jpg21167, Here, Q coolant represents the heat output of the fluid per unit time, and C p is the fluid specific heat capacity, ρ is the fluid density, q is the fluid flow rate, and T in represents the temperature of the inlet fluid, and T out represents the temperature of the outlet fluid, and m Fi represents the mass of part Fi, and C Fi represents the specific heat capacity of the component Fi, JPEG0007817511000008.jpg2630 represents the temperature of the component Fi, and Q Fi is the heat generated by the components electric power represents the fuel cell. Cell stack , a drive battery, an electric motor, and a cabin.

[0013] In step S3, a nonlinear temperature dynamic model is constructed, and the heat generation and heat dissipation of each important part during vehicle operation are taken into consideration, and the flow rate of each pump is adjusted based on the constructed vehicle thermodynamic model, and the desired temperature is output. The specific steps are as follows:

[0014] In step S3.1, a state variable x is configured using the fuel cell coolant inlet / outlet temperatures, drive battery inlet / outlet temperatures, electric motor inlet / outlet temperatures, chiller temperature, and cabin inlet / outlet temperatures obtained by testing in step S1.

[0015] In step S3.2, the input variable u is configured with the cell stack coolant flow rate, battery coolant flow rate, electric motor coolant flow rate, air conditioner refrigerant flow rate, and air flow rate obtained by the test in step S1.

[0016] In step S3.3, the cell stack heat generation power Q in step S2.1 gen , the heat generation power Q of the driving battery in step S2.2 bat , the cabin heat generation power Q in step S2.3 cab , the heat generation power Q of the electric motor in step S2.4 mot , ambient temperature degree, The disturbance variable d is composed of the cell stack coolant liquid circulation open / close signal and the drive battery chiller circuit open / close signal.

[0017] In step S4, the constructed nonlinear temperature dynamic model is linearized, that is, the nonlinear system is Taylor expanded at the reference operating point ref, and linear processing is performed while ignoring higher-order terms. Then, the disturbance variable d in S3.3 is expanded into the system of the nonlinear temperature dynamic model in the form of a state variable, and its specific formula is: JPEG0007817511000009.jpg15164, JPEG0007817511000010.jpg1418 represents the linearized state variables, JPEG0007817511000011.jpg9163, JPEG0007817511000012.jpg1525 represents the reference state variable, y represents the output variable of the system, JPEG0007817511000013.jpg1317 represents the state variable, JPEG0007817511000014.jpg1314 represents the input variable, JPEG0007817511000015.jpg9166, and u ref represents the criterion input variable, u represents the input variable, JPEG0007817511000016.jpg8164, where the superscript T denotes transpose, and A denotes the system state matrix obtained by partially differentiating the equation established in step S2.5, JPEG0007817511000017.jpg10166, JPEG0007817511000018.jpg1418 represents the number of system state variables, B represents the system input matrix obtained by partially differentiating the equation established in step S2.5, JPEG0007817511000019.jpg8165, JPEG0007817511000020.jpg1522 represents the number of system input variables, C represents the constructed MPC system output matrix, JPEG0007817511000021.jpg9167, n y represents the number of system output variables.

[0018] In step S5, the model linearized in step S4 is discretized. Specifically, the discretization is performed using the forward Euler method. The specific formula is: JPEG0007817511000022.jpg28164, where A k,t and B k,t respectively represent the discretized system matrices, JPEG0007817511000023.jpg1761 represents the predicted value for the next time, JPEG0007817511000024.jpg1869 represents the state variables of the system at the kth time, the input variables of the system at the kth time, and the identity matrix, respectively, and T ime represents the sampling time, A represents the system state matrix, JPEG0007817511000025.jpg9167, JPEG0007817511000026.jpg1520 represents the number of system state variables, B represents the system input matrix, JPEG0007817511000027.jpg9165, JPEG0007817511000028.jpg1323 represents the number of system input variables, C represents the system output matrix, and k represents the system operation time. JPEG0007817511000029.jpg8162.

[0019] In step S6, the MPC control system is designed, and the currently required state variable signals and target value signals are given to the MPC control system to obtain the corresponding input variable u, and the flow rate of the pumps of each component is adjusted to control the temperature.

[0020] In step S6.1, an objective function J(k) is set for the MPC control system constructed in step S6.

[0021] In step S6.2, constraint conditions are set for the limit constraint on the input variable u and the control increment Δu constraint in the control process of the MPC control system.

[0022] In step S6.3, the objective function in step S6.1 is transformed once to a form that can be solved by quadratic programming. The specific formula is: JPEG0007817511000030.jpg10165, where: JPEG0007817511000031.jpg15112 represent the objective function, the state variable matrix in the prediction time domain, the transpose of the state variable matrix in the prediction time domain, the transpose of the output variable matrix in the prediction time domain, and the output variable matrix in the prediction time domain, respectively. JPEG0007817511000032.jpg28163, JPEG0007817511000033.jpg16137 are the time periods from the kth to the k+Nth predicted by the system. p represents the discretized state variables up to time, JPEG0007817511000034.jpg17147 are the images from the kth time to the k+Nth time. c represents the discretized input variables up to time -1, JPEG0007817511000035.jpg1653 is N p ×N p and N c ×N c represents the identity matrix of JPEG0007817511000036.jpg1720 represents the Kronecker product.

[0023] In step S6.4, the solvable quadratic programming solution form obtained based on step S6.3 is JPEG0007817511000037.jpg15162, where: JPEG0007817511000038.jpg1674 respectively represent the transpose of the output variable matrix in the forecast time domain, the quadratic term matrix in the quadratic programming, the output variable matrix in the forecast time domain, and the linear term matrix in the quadratic programming. JPEG0007817511000039.jpg30166JPEG0007817511000040.jpg37164, In step S7, a solution is found based on the objective function and constraints in step S6 to achieve optimal temperature control. Specifically, the process includes the following steps.

[0024] In step S7.1, the prediction model, the objective function and the constraints are integrated to find the optimal solution.

[0025] In step S7.2, the optimal control sequence at the current time is obtained and input to the fuel cell hybrid vehicle, and the fuel cell hybrid vehicle executes the received control instructions and feeds them back to the MPC control layer.

[0026] In step S7.3, the state quantity at the current time is input to the MPC controller, and the optimal solution is obtained again to obtain the flow control variables required for the fuel cell hybrid vehicle at the next time.

[0027] As a preferred technical solution of the present application, the specific formula of the objective function in step S4 above is: JPEG0007817511000041.jpg21166, where N p is the predicted time domain, and N c is the control time domain, τ is a weighting coefficient, ε is a relaxation coefficient, ΔU is a control change amount, k+i represents the prediction time, t represents time, and x(k+i|t) represents the prediction of the state variable at time k+i at time t, JPEG0007817511000042.jpg18165 represents the system's tracking ability against the reference target, JPEG0007817511000043.jpg19163 represents the system constraints on the control change amount, and τε 2 represents a soft constraint, Q and Z are weight matrices, ΔU(k+i|t) represents a prediction of the control change amount at time k+i at time t, TIFF0007817511000044.tif1235 represents the L2 norm adjusted by the Q weighting matrix and the L2 norm adjusted by the Z weighting matrix.

[0028] In a preferred technical solution of the present application, the specific formula of the constraint condition in step S5 is as follows: JPEG0007817511000045.jpg14161, where u(t+k), Δu(t+k), u min(t+k) , u max(t+k) , Δu min(t+k) and Δu max(t+k) represent the input variable value at time t+k, the control increment value at time t+k, the lower limit constraint of the input variable, the upper limit constraint of the input variable, the lower limit constraint of the control increment, and the upper limit constraint of the control increment, respectively, where k=0, 1, ..., Nc-1.

[0029] Compared with the prior art, the present application provides a thermal management control method for a fuel cell hybrid vehicle, which has the following beneficial effects:

[0030] This application uses an MPC to optimize the coolant flow rate in the complex, multi-input, multi-output, nonlinear system of a fuel cell hybrid vehicle, maintaining each component within its normal operating temperature range. Using the inlet and outlet temperatures of each component as state variables and the coolant flow rate of each component's pump as a controlled variable, the MPC controller coordinates the temperature of each component, while also considering the relationship between the thermal management systems of each circuit, thereby improving the overall performance of the fuel cell hybrid vehicle. [Brief explanation of the drawings]

[0031] [Figure 1] FIG. 10 is a diagram showing the cell stack temperature effect due to flow rate control according to the present application. [Figure 2] FIG. 10 is a diagram showing a battery temperature curve when the present application is implemented. [Figure 3] FIG. 1 shows an electric motor temperature curve during implementation of the present application. [Figure 4] FIG. 10 shows a cabin temperature comparison curve obtained by implementing the present application. [Figure 5] FIG. 1 is a schematic diagram of an MPC fuel cell hybrid vehicle control system according to an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0032] The following will clearly and completely describe the technical solutions in the embodiments of the present application in combination with the accompanying drawings in the embodiments of the present application. It is clear that the described embodiments are only some of the embodiments of the present application, and not all of the embodiments of the present application. Based on the embodiments of the present application, all other embodiments that a person skilled in the art can obtain without creative work fall within the scope of protection of the present application.

[0033] 1 to 5, the present application provides the following technical solution: A thermal management control method for a fuel cell hybrid vehicle includes the following steps:

[0034] In step S1, performance data of the fuel cell hybrid vehicle is acquired, and a cycle mode test is performed on the entire vehicle. Cell stack The system records current, cell stack output voltage, fuel cell coolant inlet / outlet temperatures, cell stack coolant flow rate, drive battery inlet / outlet temperatures, battery coolant flow rate, chiller temperature, drive battery output current, drive battery voltage and drive battery internal temperature, electric motor coolant flow rate, cabin inlet / outlet temperatures, air conditioner refrigerant flow rate, electric motor inlet / outlet temperatures, electric motor pump flow rate, total electric motor power demand, electric motor and power control unit equivalent efficiency, air flow rate, and evaporator air flow rate data in real time, and obtains solenoid valve opening / closing signals, such as the cell stack coolant circulation opening / closing signal and the drive battery chiller circuit opening / closing signal, as well as cabin-related parameters, including ambient temperature, window solar transmittance, solar radiation intensity, windshield, rear window, and side window areas, shading coefficient, angle between the window and the vertical direction, convective heat exchange coefficient between the body shell and the cabin, heat transfer area of ​​the roof panel, windshield, rear window, and side windows, number of people in the cabin, cabin air temperature, air flow rate passing through the evaporator, and specific heat capacity of the cabin air.

[0035] In step S2, a thermodynamic model of the vehicle is constructed, which specifically includes the following steps:

[0036] In step S2.1, the cell stack heat generation power Q is calculated from the cell stack output voltage and cell stack current in step S1. gen The specific formula is: JPEG0007817511000046.jpg14168, Here, Q gen represents the cell stack heat generation power, and E nernst represents the Nernst voltage, which is typically 1.2 V, and N FC Indicates the number of cells in the fuel cell stack, V st is a cell stack output represents the voltage, and I st is a cell stack Kuden Represents flow.

[0037] In step S2.2, the battery heat generation power Q is calculated from the drive battery output current, drive battery voltage, and drive battery internal temperature in step S1. bat The specific formula is: JPEG0007817511000047.jpg14165 Here, Q bat represents the battery heat generation power, and I bat is the drive battery output represents the current, and R esistance represents the electrical resistance, and T amb represents the ambient temperature, JPEG0007817511000048.jpg2228 represents the temperature effect coefficient.

[0038] In step S2.3, the conductive heat load Q is calculated from the cabin-related parameters already obtained in step S1. rb , radiant heat load Q cb , the heat generated by the passengers in the car Q h and ventilation system heat load Q n Calculate the cabin heat generation power Q cab The specific formula is JPEG0007817511000049.jpg59162, Here, Q rb represents the conductive heat load, ξ represents the solar transmittance of the window, I represents the solar radiation intensity, and Fg,j represents the area of ​​the roof panel, windshield, rear window, and side window, C represents the shading coefficient, and θ j represents the angle between the window and the vertical direction, JPEG0007817511000050.jpg21164 represents the sequential addition of calculations for the roof panel, windshield, rear window and side window, and Q cb represents the radiant heat load, and h b represents the convective heat exchange coefficient between the body shell and the cabin, F represents the heat transfer area of ​​the roof panel, windshield, rear window, and side window, and T amb represents the ambient temperature, and T cabin,in represents the cabin air temperature, and Q h represents the heat generated by the passengers inside the vehicle, and n m represents the number of people in the cabin, and Q n represents the ventilation system heat load, and m e represents the air flow rate passing through the evaporator, and C cabin represents the specific heat capacity of the cabin air.

[0039] In step S2.4, the electric motor heat generation power is calculated from the electric motor parameters obtained in step S1, and the specific formula is: JPEG0007817511000051.jpg14165, Here, Q mot represents the heat generation power of the electric motor, P represents the total demand power of the electric motor, and η mot / PCU represents the equivalent efficiency of the electric motor and PCU, where PCU is the power control unit.

[0040] In step S2.5, a thermodynamic model of the vehicle is constructed based on the first law of thermodynamics from the heat generation power of each component obtained in steps S2.1 to S2.4. The specific formula is: JPEG0007817511000052.jpg21167, Here, Q coolant represents the heat output of the fluid per unit time, and C pis the fluid specific heat capacity, ρ is the fluid density, q is the fluid flow rate, and T in represents the temperature of the inlet fluid, and T out represents the temperature of the outlet fluid, and m Fi represents the mass of part Fi, and C Fi represents the specific heat capacity of the component Fi, JPEG0007817511000053.jpg2630 represents the temperature of the component Fi, and Q Fi is the heat generated by the components electric power represents the fuel cell. Cell stack , a drive battery, an electric motor, and a cabin.

[0041] In step S3, a nonlinear temperature dynamic model is constructed, and the heat generation and heat dissipation of each important part during vehicle operation are taken into consideration, and the flow rate of each pump is adjusted based on the constructed vehicle thermodynamic model, and the desired temperature is output. The specific steps are as follows:

[0042] In step S3.1, a state variable x is configured using the fuel cell coolant inlet / outlet temperatures, drive battery inlet / outlet temperatures, electric motor inlet / outlet temperatures, chiller temperature, and cabin inlet / outlet temperatures obtained by testing in step S1.

[0043] In step S3.2, the input variable u is configured with the cell stack coolant flow rate, battery coolant flow rate, electric motor coolant flow rate, air conditioner refrigerant flow rate, and air flow rate obtained by the test in step S1.

[0044] In step S3.3, the cell stack heat generation power Q in step S2.1 gen , the heat generation power Q of the driving battery in step S2.2 bat , the cabin heat generation power Q in step S2.3 cab , the heat generation power Q of the electric motor in step S2.4 mot , ambient temperature degree, The disturbance variable d is composed of the cell stack coolant liquid circulation open / close signal and the drive battery chiller circuit open / close signal.

[0045] In step S4, the constructed nonlinear temperature dynamic model is linearized, that is, the nonlinear system is Taylor expanded at the reference operating point ref, and linear processing is performed while ignoring higher-order terms. Then, the disturbance variable d in S3.3 is expanded into the system of the nonlinear temperature dynamic model in the form of a state variable, and its specific formula is: JPEG0007817511000054.jpg15164, JPEG0007817511000055.jpg1418 represents the linearized state variables, JPEG0007817511000056.jpg9163, JPEG0007817511000057.jpg1525 represents the reference state variable, y represents the output variable of the system, JPEG0007817511000058.jpg1317 represents the state variable, JPEG0007817511000059.jpg1314 represents the input variable, JPEG0007817511000060.jpg9166, and u ref represents the criterion input variable, u represents the input variable, JPEG0007817511000061.jpg8164, where the superscript T stands for transpose and A stands for the equation established in step S2.5. JPEG0007817511000062.jpg2970 represents the system state matrix obtained by JPEG0007817511000063.jpg10166, JPEG0007817511000064.jpg1418 represents the number of system state variables, and B represents the equation established in step S2.5. JPEG0007817511000065.jpg3071 represents the system input matrix obtained by JPEG0007817511000066.jpg8165, JPEG0007817511000067.jpg1522 represents the number of system input variables, C represents the constructed MPC system output matrix, JPEG0007817511000068.jpg9167, n y represents the number of system output variables.

[0046] In step S5, the model linearized in step S4 is discretized. Specifically, the discretization is performed using the forward Euler method. The specific formula is: JPEG0007817511000069.jpg28164, where A k,t and B k,t respectively represent the discretized system matrices, JPEG0007817511000070.jpg1761 represents the predicted value for the next time, JPEG0007817511000071.jpg1869 represents the state variables of the system at the kth time, the input variables of the system at the kth time, and the identity matrix, respectively, and T ime represents the sampling time, A represents the system state matrix, JPEG0007817511000072.jpg9167, JPEG0007817511000073.jpg1520 represents the number of system state variables, B represents the system input matrix, JPEG0007817511000074.jpg9165, JPEG0007817511000075.jpg1323 represents the number of system input variables, C represents the system output matrix, and k represents the system operation time. JPEG0007817511000076.jpg8162.

[0047] In step S6, the MPC control system is designed, and the currently required state variable signals and target value signals are given to the MPC control system to obtain the corresponding input variables u, and the flow rate of the pumps of each component is adjusted to control the temperature. In the control process, there is a desired reference trajectory.

[0048] The core concept is to completely base the thermal management control of fuel cell hybrid vehicles on predictive control of an MPC model, rather than on traditional PID control or algorithm optimization. It consists of three parts: a predictive model, rolling optimization, and feedback correction, each of which plays a different role. The control system predicts the future state of the thermal management temperature control of the fuel cell hybrid vehicle based on the current temperature status information of each system component provided by the fuel cell hybrid vehicle in the control system, as well as the future control input pump flow variable. Then, at each sampling time, the optimal control rate within the control time domain from that time is determined according to the objective function and constraints. The output quantity is compared with the model prediction value to determine the model prediction error, and the model prediction error is used to correct the model prediction value, resulting in a more accurate future output prediction value.

[0049] In step S6.1, the objective function J(k) of the model is set, and its specific formula is JPEG0007817511000077.jpg21166, where N p is the predicted time domain, and N c is the control time domain, τ is a weighting coefficient, ε is a relaxation coefficient, ΔU is a control change amount, k+i represents the prediction time, t represents time, and x(k+i|t) represents the prediction of the state variable at time k+i at time t, JPEG0007817511000078.jpg18165 represents the system's tracking ability against the reference target, JPEG0007817511000079.jpg19163 represents the system constraints on the control change amount, and τε 2 represents a soft constraint, Q and Z are weight matrices, ΔU(k+i|t) represents a prediction of the control change amount at time k+i at time t, TIFF0007817511000080.tif1235 represents the L2 norm adjusted by the Q weighting matrix and the L2 norm adjusted by the Z weighting matrix.

[0050] In step S6.2, constraint conditions are set for the limit constraint on the input variable u and the control increment Δu constraint in the control process of the MPC control system. The specific formula is as follows: JPEG0007817511000081.jpg14161, where u(t+k), Δu(t+k), u min(t+k) , u max(t+k) , Δu min(t+k) and Δu max(t+k) represent the input variable value at time t+k, the control increment value at time t+k, the lower limit constraint of the input variable, the upper limit constraint of the input variable, the lower limit constraint of the control increment, and the upper limit constraint of the control increment, respectively, where k=0, 1, ..., Nc-1.

[0051] In step S6.3, the objective function in step S6.1 is transformed once to a form that can be solved by quadratic programming. The specific formula is: JPEG0007817511000082.jpg10165, where: JPEG0007817511000083.jpg15112 represent the objective function, the state variable matrix in the prediction time domain, the transpose of the state variable matrix in the prediction time domain, the transpose of the output variable matrix in the prediction time domain, and the output variable matrix in the prediction time domain, respectively. JPEG0007817511000084.jpg28163, JPEG0007817511000085.jpg16137 are the time periods from the kth to the k+Nth predicted by the system. p represents the discretized state variables up to time, JPEG0007817511000086.jpg17147 are the images from the kth time to the k+Nth time respectively. c represents the discretized input variables up to time -1, JPEG0007817511000087.jpg1653 is N p ×N p and N c ×N c represents the identity matrix of JPEG0007817511000088.jpg1720 represents the Kronecker product.

[0052] In step S6.4, the solvable quadratic programming solution form obtained based on step S6.3 is JPEG0007817511000089.jpg15162, where: JPEG0007817511000090.jpg1674 respectively represent the transpose of the output variable matrix in the forecast time domain, the quadratic term matrix in the quadratic programming, the output variable matrix in the forecast time domain, and the linear term matrix in the quadratic programming. JPEG0007817511000091.jpg30166JPEG0007817511000092.jpg37164, JPEG0007817511000093.jpg12164 is 1st, 2nd, ···, N p represents the discretized system matrix, JPEG0007817511000094.jpg1316Represents the 7th-order discretized system matrix.

[0053] In step S7, a solution is found based on the objective function and constraints in step S6 to achieve optimal temperature control. Specifically, the process includes the following steps.

[0054] In step S7.1, the prediction model, the objective function and the constraints are integrated to find the optimal solution.

[0055] In step S7.2, the optimal control sequence at the current time is obtained and input to the fuel cell hybrid vehicle, and the fuel cell hybrid vehicle executes the received control instructions and feeds them back to the MPC control layer.

[0056] In step S7.3, the state quantity at the current time is input to the MPC controller, and the optimal solution is obtained again to obtain the flow control variables required for the fuel cell hybrid vehicle at the next time.

[0057] Although the embodiments of the present application have been illustrated and described, those skilled in the art will understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is limited by the appended claims and their equivalents.

Claims

1. A thermal management control method for a fuel cell hybrid vehicle, comprising: In step S1, performance data of the fuel cell hybrid vehicle is acquired, and a cycle mode test is performed on the entire vehicle to obtain the fuel cell cell stack current, cell stack output voltage, fuel cell coolant inlet / outlet temperatures, cell stack cooling water flow rate, drive battery inlet / outlet temperatures, battery cooling water flow rate, chiller temperature, drive battery output current, drive battery voltage and drive battery internal temperature, electric motor cooling water flow rate, cabin inlet / outlet temperatures, air conditioner refrigerant flow rate, electric motor inlet / outlet temperatures, electric motor pump flow rate, electric motor total power demand, electric motor and power control unit equivalent efficiency, air flow rate, Evaporator air flow rate data is recorded in real time, and solenoid valve open / close signals, such as a cell stack coolant liquid circulation open / close signal and a drive battery chiller circuit open / close signal, are acquired, and cabin-related parameters including the ambient temperature, window solar radiation transmittance, solar radiation intensity, windshield, rear window, and side window areas, shading coefficient, angle between the window and the vertical direction, convective heat exchange coefficient between the body shell and the cabin, heat transfer area of ​​the roof panel, windshield, rear window, and side window, number of passengers in the cabin, cabin air temperature, air flow rate passing through the evaporator, and specific heat capacity of the cabin air are acquired. In step S2, a thermodynamic model of the vehicle is constructed, specifically, In step S2.1, the cell stack heat generation power Q is calculated from the cell stack output voltage and cell stack current in step S1. ｇｅｎ The specific formula is: and Here, Q ｇｅｎ represents the cell stack heat generation power, and E ｎｅｒｎｓｔ represents the Nernst voltage, which is typically 1.2 V, and N ＦＣ represents the number of cells in the fuel cell stack, and V ｓｔ represents the cell stack output voltage, and I ｓｔ represents the cell stack current. In step S2.2, the battery heat generation power Q is calculated from the drive battery output current, drive battery voltage, and drive battery internal temperature in step S1. ｂａｔ The specific formula is: and Here, Q ｂａｔ represents the battery heat generation power, and I ｂａｔ represents the drive battery output current, and R ｅｓｉｓｔａｎｃｅ represents the electrical resistance, and T ａｍｂ represents the ambient temperature, represents the temperature influence coefficient, In step S2.3, the conductive heat load Q is calculated from the cabin-related parameters already obtained in step S1. ｒｂ , radiant heat load Q ｃｂ , the heat generated by the passengers in the vehicle Q ｈ and ventilation system heat load Q ｎ Calculate the cabin heat generation power Q ｃａｂ The specific formula is and Here, Q ｒｂ represents the conductive heat load, ξ represents the solar transmittance of the window, I represents the solar radiation intensity, and F ｇ，ｊ represents the area of ​​the roof panel, windshield, rear window, and side window, C represents the shading coefficient, and θ ｊ represents the angle between the window and the vertical direction, represents the sequential accumulation of calculations for the roof panel, windshield, rear window and side window, and Q ｃｂ represents the radiant heat load, and h ｂ represents the convective heat exchange coefficient between the body shell and the cabin, F represents the heat transfer area of ​​the roof panel, windshield, rear window and side window, and T ａｍｂ represents the ambient temperature, and T ｃａｂｉｎ，ｉｎ represents the cabin air temperature, and Q ｈ represents the heat generated by the passengers in the vehicle, and n ｍ represents the number of people in the cabin, and Q ｎ represents the ventilation system heat load, and m ｅ represents the air flow rate passing through the evaporator, and C ｃａｂｉｎ represents the specific heat capacity of the cabin air, In step S2.4, the electric motor heat generation power is calculated from the electric motor parameters obtained in step S1, and the specific formula for this is: and Here, Q ｍｏｔ represents the heat generation power of the electric motor, P represents the total demand power of the electric motor, and η ｍｏｔ／ＰＣＵ represents the equivalent efficiency of the electric motor and PCU, where PCU is the power control unit; In step S2.5, a thermodynamic model of the vehicle is constructed based on the first law of thermodynamics from the heat generation power of each component obtained in steps S2.1 to S2.

4. The specific formula for this model is: and Here, Q ｃｏｏｌａｎｔ represents the heat output of the fluid per unit time, and C ｐ is the fluid specific heat capacity, ρ is the fluid density, q is the fluid flow rate, and T ｉｎ represents the temperature of the inlet fluid, and T ｏｕｔ represents the temperature of the outlet fluid, m Ｆｉ represents the mass of part Fi, and C Ｆｉ represents the specific heat capacity of part Fi, represents the temperature of the component Fi, and Q Ｆｉ represents the heat generation power of the component, and the component Fi includes the fuel cell stack, the drive battery, the electric motor, and the cabin, In step S3, a nonlinear temperature dynamic model is constructed, and the heat generation and heat dissipation of each important part during vehicle running is taken into consideration, and the flow rate of each pump is adjusted based on the constructed vehicle thermodynamic model, and the desired temperature is output. In step S3.1, a state variable x is configured using the fuel cell coolant inlet / outlet temperatures, drive battery inlet / outlet temperatures, electric motor inlet / outlet temperatures, chiller temperature, and cabin inlet / outlet temperatures obtained by testing in step S1. In step S3.2, an input variable u is configured using the cell stack coolant flow rate, battery coolant flow rate, electric motor coolant flow rate, air conditioner refrigerant flow rate, and air flow rate obtained by the test in step S1. In step S3.3, the cell stack heat generation power Q in step S2.1 ｇｅｎ , the heat generation power Q of the driving battery in step S2.2 ｂａｔ , the cabin heating power Q in step S2.3 ｃａｂ , the heat generation power Q of the electric motor in step S2.4 ｍｏｔ , the ambient temperature, the cell stack coolant liquid circulation open / close signal, and the drive battery chiller circuit open / close signal constitute a disturbance variable d, In step S4, the constructed nonlinear temperature dynamic model is linearized, that is, the nonlinear system is Taylor expanded at the reference operating point ref, and linear processing is performed while ignoring higher-order terms. Then, the disturbance variable d in S3.3 is expanded into the system of the nonlinear temperature dynamic model in the form of a state variable, and its specific formula is: and represents the linearized state variables, and represents the reference state variable, y represents the output variable of the system, represents the state variables, represents the input variables, and u ref represents the criterion input variable, u represents the input variable, where the superscript T denotes transpose, and A denotes the system state matrix obtained by partially differentiating the equation established in step S2.

5. and represents the number of system state variables, B represents the system input matrix obtained by partially differentiating the equation established in step S2.5, and represents the number of system input variables, C represents the constructed MPC system output matrix, and n ｙ represents the number of system output variables, In step S5, the model linearized in step S4 is discretized. Specifically, the discretization is performed using the forward Euler method. The specific formula is: and Here, A ｋ，ｔ and B ｋ，ｔ respectively represent the discretized system matrices, represents the predicted value at the next time, represent the state variable of the system at the kth time, the input variable of the system at the kth time, and the identity matrix, respectively, and T ｉｍｅ represents the sampling time, A represents the system state matrix, and represents the number of system state variables, B represents the system input matrix, and represents the number of system input variables, C represents the system output matrix, and k represents the system operation time. and In step S6, the MPC control system is designed, and the currently required state variable signals and target value signals are provided to the MPC control system to obtain the corresponding input variable u, and the flow rate of the pumps of each component is adjusted to control the temperature; In step S6.1, an objective function J(k) is set for the MPC control system constructed in step S6. A specific formula of the objective function in step S6.1 is: and Here, N ｐ is the predicted time domain, and N ｃ is the control time domain, τ is a weighting coefficient, ε is a relaxation coefficient, ΔU is a control change amount, k+i represents the prediction time, t represents time, and x(k+i|t) represents the prediction of the state variable at time k+i at time t, represents the system's ability to track the reference target, represents the system constraint on the control change amount, and τε ２ represents a soft constraint, Q and Z are weight matrices, ΔU(k+i|t) represents a prediction of the control change amount at time k+i at time t, represents the L2 norm adjusted by the Q weight matrix and the L2 norm adjusted by the Z weight matrix, In step S6.2, a limit constraint on the input variable u in the control process of the MPC control system and a constraint condition for a control increment Δu are set. The specific formula of the constraint condition in step S6.2 is as follows: and where u(t+k), Δu(t+k), u ｍｉｎ（ｔ＋ｋ） , u ｍａｘ（ｔ＋ｋ） , Δu ｍｉｎ（ｔ＋ｋ） and Δu ｍａｘ（ｔ＋ｋ） represent the input variable value at time t+k, the control increment value at time t+k, the lower limit constraint of the input variable, the upper limit constraint of the input variable, the lower limit constraint of the control increment, and the upper limit constraint of the control increment, respectively, where k=0, 1, ..., Nc-1, and Nc is the control time domain. In step S7, a solution is found based on the objective function and constraints in step S6 to realize optimal temperature control, specifically, In step S7.1, the predictive model, the objective function and the constraints are integrated to find an optimal solution; In step S7.2, the optimal control sequence at the current time is obtained and input to the fuel cell hybrid vehicle, and the fuel cell hybrid vehicle executes the received control instructions and feeds them back to the MPC control system; In step S7.3, the state quantity at the current time is input to the MPC control system, and the optimal solution is obtained again to obtain the flow control variables required for the fuel cell hybrid vehicle at the next time. A thermal management control method for a fuel cell hybrid vehicle.

Images