Front and rear stage collaborative optimization control method for vehicle-mounted V2G bidirectional converter

Abstract

The invention discloses a front-stage and back-stage collaborative optimization control method for a vehicle-mounted V2G bidirectional converter, and relates to the technical field of vehicle-mounted bidirectional converter control. According to the invention, a front-stage totem-pole bridgeless PFC double-closed-loop controller and a rear-stage CLLC resonant converter double-closed-loop controller are designed; secondly, establishing a system joint model, reflecting the dynamic behavior of the bus voltage in the charging and discharging process, designing a corresponding bus voltage optimization method, and adopting an improved particle swarm optimization (IPSO) algorithm to optimize the parameters of a front-stage controller and a rear-stage controller by using a comprehensive cost function in the bus voltage optimization method; and the prediction step length and the weight parameter of the MPC link and the gain of the PI link are optimized in real time, so that the front-stage and rear-stage double-closed-loop control strategy can be more flexible and reliable in application. Through the innovation, the system can adapt to different working conditions in real time, front-stage and back-stage collaborative optimization control is achieved, the overall energy transmission efficiency is improved, and energy loss is remarkably reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of on-vehicle bidirectional converter control, and particularly relates to a method for collaborative optimization control of the front and rear stages of an on-vehicle V2G bidirectional converter. Background Art

[0002] As a key component for energy interaction between electric vehicles and the power grid, with the development of V2G technology, on-vehicle bidirectional converters have gradually become a research hotspot in this field. Existing on-vehicle V2G converters generally adopt single-stage or two-stage structures. Among them, the totem-pole bridgeless PFC circuit is widely used in the front-stage AC-DC conversion field due to its simplicity and high power density; while the CLLC resonant converter shows great advantages in the rear-stage DC-DC conversion due to its high efficiency and soft-switching characteristics. The selection of the front and rear stage circuit controllers of the on-vehicle bidirectional converter determines the control accuracy and energy efficiency of the system. Industrially, traditional PI double closed-loop control is mostly used, which makes the system have poor robustness and a small power range.

[0003] In recent years, model predictive control (MPC) has gradually become an important technology for solving control problems due to its fast response and convenient parameter adjustment characteristics. Using MPC as the front and rear stage current inner loop control can predict the system state at multiple future moments, take into account possible disturbances and constraints, and generate efficient control commands. By setting a cost function, MPC evaluates multiple control input schemes in each control cycle, and thus selects the optimal control command. Moreover, MPC can effectively handle the state and input constraints of the system to ensure that the control command is within a safe range.

[0004] Traditional schemes adopt hierarchical independent modeling and control, without considering the overall system efficiency, often resulting in problems such as unreasonable system energy scheduling, insufficient dynamic response, and resonance control imbalance. Too high bus voltage will increase the switching losses of the front and rear stages, and the frequent switching of the rear-stage working state will also bring fluctuations to the control of the front-stage PFC. There is an urgent need for a front and rear stage collaborative optimization control method to make full use of the coupling characteristics of the converter structure to achieve system collaborative optimization.

[0005] At present, there is very little research on the cooperative optimization control strategy for bidirectional converters, and in the traditional control process, the DC bus voltage is usually set to a fixed value or simplified. However, during the operation of the converter, affected by load changes, external environment, and internal dynamic characteristics, the stability of the bus voltage will be challenged. If the bus voltage cannot be maintained within the ideal range, it may lead to a decrease in energy transfer efficiency, equipment damage, or even system failure. The bus voltage optimization method proposed in the present invention can adjust the controller parameters in real time according to the bus voltage to ensure the safe and stable operation of the system. By adaptively adjusting the controller parameters in combination with the bus voltage state under different charge and discharge modes to match the grid connection voltage, load requirements, etc., this method can reduce unnecessary secondary conversion losses during forward / reverse energy transfer, better meet the grid or auxiliary service requirements, enable the subsequent CLLC resonant converter to always operate in the high-efficiency range of the resonant cavity, maintain the zero-voltage switching (ZVS) or zero-current switching (ZCS) characteristics, and at the same time significantly reduce the switching losses and device temperature rise of the front-stage PFC.

[0006] The particle swarm optimization (PSO) algorithm realizes the real-time optimization of parameters by simulating the behavior of biological populations in nature. Although the traditional PSO algorithm has the advantages of simple structure and easy implementation, it has limitations such as insufficient convergence performance, unbalanced exploration and development, and the risk of premature convergence in practical applications. To solve the problems of its fixed inertia weight, fixed learning factor, and lack of diversity maintenance mechanism, an improved particle swarm optimization (IPSO) algorithm is proposed, and a dynamic inertia weight mechanism, an elite learning strategy, and chaotic local search are designed to meet the parameter optimization requirements.

[0007] The present invention proposes a method for cooperative optimization control of the front and rear stages of a vehicle-mounted V2G bidirectional converter, designs a double-loop controller with a current inner loop model predictive control and a voltage outer loop PI control for both the front and rear stages, proposes a bus voltage optimization method, and uses an improved particle swarm optimization algorithm to perform cooperative optimization of the controller parameters for both the front and rear stages to improve the performance and energy management efficiency of the vehicle-mounted V2G bidirectional converter, laying a solid technical foundation for the intelligent development of electric vehicles. Summary of the Invention

[0008] The object of the present invention is to provide a method for cooperative optimization control of the front and rear stages of a vehicle-mounted V2G bidirectional converter in order to solve the above problems.

[0009] To achieve the above object, the technical solution adopted by the present invention is as follows: including the following steps:

[0010] S1: Design a double-loop controller for the front-stage totem-pole bridgeless PFC and a double-loop controller for the rear-stage CLLC resonant converter, where the current inner loop uses model predictive control (MPC) and the voltage outer loop uses proportional-integral (PI) control;

[0011] S2: Establish a system joint model to reflect the dynamic behavior of the bus voltage during the charging and discharging process, design a bus voltage optimization method, and realize the collaborative optimization of the parameters of the front and rear stage controllers through a comprehensive cost function. The comprehensive cost function includes the bus voltage error and the current tracking error of the front and rear stages;

[0012] S3: Use the improved particle swarm optimization (IPSO) algorithm to optimize the prediction step length, weight parameters of the MPC link, and the gain of the PI link in real time, making the front and rear two-stage double closed-loop control strategy more flexible and reliable in application.

[0013] Furthermore, the design of the front-stage totem-pole bridgeless PFC double closed-loop controller is as follows:

[0014] Current inner-loop MPC control:

[0015] Prediction model: Discretize the continuous-time state equation to obtain a discrete-time state space model. Based on the current state and the system state space model, predict the system state at future moments;

[0016] Cost function: Design the cost function to minimize the current tracking error and the change of the control input, and use it to evaluate the performance of the predicted control sequence;

[0017] Rolling optimization: In each control period, solve the optimization problem to obtain the optimal control input sequence, execute the first control input, then update the state and repeat the optimization process;

[0018] Weight matrix design: Set the weighting factor according to the system requirements to balance the state error and the change of the control input;

[0019] Voltage outer-loop PI control: The transfer function of the PI controller is a known formula, and its output is the reference current of the current inner loop. By reasonably selecting the gain, ensure the stability and dynamic performance of the system.

[0020] Furthermore, the design of the rear-stage CLLC resonant converter double closed-loop controller is as follows:

[0021] Current inner-loop MPC control:

[0022] Prediction model: Discretize the continuous-time state equation to obtain a discrete-time state space model. Based on the current state and the system state space model, predict the system state at future moments;

[0023] Cost function: Define the cost function to evaluate the performance of the predicted control sequence, and minimize the current tracking error and the change of the control input;

[0024] Rolling optimization: within each control cycle, solve the optimization problem to obtain the optimal control input sequence, execute the first control input, then update the state and repeat the optimization process;

[0025] Weight matrix design: Set the weighting factor according to system requirements to balance the state error and the change of control input;

[0026] Outer voltage loop PI control: The transfer function of the PI controller is a known formula, and its output serves as the reference current for the inner current loop MPC control. By reasonably selecting the gain, ensure the stability and dynamic performance of the system.

[0027] Furthermore, the bus voltage optimization method includes:

[0028] Establish a system-level mathematical model to clarify the mapping relationship between the output bus voltage of the previous stage and the input port characteristics of the subsequent resonant converter;

[0029] Analyze the dynamic characteristics of the bus voltage and describe the change of the bus voltage through the charge and discharge process of the capacitor;

[0030] Define the bus voltage deviation as the difference between the target voltage and the actual voltage, and the optimization goal is to minimize the bus voltage deviation;

[0031] Comprehensively consider the performance indicators of the system and define a comprehensive cost function, including the DC bus voltage error, the output voltage error of the subsequent stage, the current tracking error of the previous stage, and the current tracking error of the subsequent stage.

[0032] Furthermore, the improved particle swarm optimization (IPSO) algorithm includes the following steps:

[0033] S1: Initialize the position and velocity of the particle swarm. The particle position represents the key parameters of the controller. The velocity of each particle is randomly initialized, the individual historical best position of each particle is initialized to its current position, and the global best position is initialized to the best position in the particle swarm;

[0034] S2: Fitness evaluation: Calculate the fitness value of each particle by simulating the dynamic response of the system. The fitness function is the reciprocal of the cost function;

[0035] S3: Update the position and velocity of the particle according to the individual historical best position and the global best position;

[0036] S4: Introduce a population fitness feedback term on the basis of the traditional linearly decreasing weight to perform dynamic inertia weight adjustment;

[0037] S5: Design an elite learning strategy and adopt differential updates for the top 20% elite particles in terms of fitness;

[0038] S6: When the global optimal solution has not been updated for 5 consecutive generations, activate chaotic local search and dynamically adjust the learning factor;

[0039] S7: Update the individual historical optimum and the global optimum according to the fitness value, and repeat the above steps until the convergence condition is met.

[0040] Furthermore, the parameters of the IPSO algorithm are set as follows: the number of particles is a known quantity, and the maximum number of iterations is set to 150 times.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] (1) The present invention designs a double closed-loop controller for the front-stage totem-pole bridgeless PFC and a double closed-loop controller for the rear-stage CLLC resonant converter; secondly, a system joint model is established to reflect the dynamic behavior of the bus voltage during the charging and discharging process, and a corresponding bus voltage optimization method is designed. This method not only considers the deviation of the bus voltage, but also takes into account the smoothness of the coordinated control of the front and rear stages and the loss of the system, so as to optimize the parameters of the front and rear stage controllers. This method effectively improves the dynamic response ability of the system under load changes and external disturbances, and ensures that the output can be adjusted quickly and accurately when the reference voltage and current change greatly.

[0043] (2) In order to realize the optimization of the parameters of the front and rear stage controllers by the comprehensive cost function in the bus voltage optimization method, the present invention adopts an improved particle swarm optimization (IPSO) algorithm to optimize the prediction step length and weight parameters of the MPC link and the gain of the PI link in real time, so that the double closed-loop control strategies of the front and rear stages can be more flexible and reliable in application. Through this innovation, the system can adapt to different working conditions in real time, realize the coordinated optimization control of the front and rear stages, improve the overall energy transmission efficiency, and significantly reduce the energy loss. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is the structural diagram of the coordinated optimization control of the front and rear stages of the on-vehicle V2G bidirectional converter of the present invention;

[0045] Figure 2 is the flow chart of the coordinated optimization control of the front and rear stages of the on-vehicle V2G bidirectional converter of the present invention;

[0046] Figure 3 is the circuit structure of the two-stage V2G on-vehicle bidirectional converter of the present invention;

[0047] Figure 4 is the double closed-loop control structure of the V2G on-vehicle front-stage converter of the present invention;

[0048] Figure 5 is the double closed-loop control structure of the V2G on-vehicle rear-stage converter of the present invention;

[0049] Figure 6 This is the flow chart of the improved particle swarm optimization algorithm of the present invention. Specific embodiments

[0050] In order to make the technical means, creative features, achieved purposes and effects of the present invention easy to understand, the present invention will be further described below in conjunction with specific embodiments.

[0051] As follows Figure 1-6 As shown, the present invention provides a method for collaborative optimization control of the front and rear stages of an in-vehicle V2G bidirectional converter, a method for collaborative optimization control of the front and rear stages of an in-vehicle V2G bidirectional converter based on bus voltage optimization, aiming to solve the problems of insufficient collaborative control between the front and rear stages, slow dynamic response and low energy transmission efficiency during the charging and discharging process of the bidirectional converter. The specific technical solutions include the following key steps.

[0052] This method first designs a double closed-loop control strategy of current inner-loop MPC control and voltage outer-loop PI control for the front and rear two stages; secondly, a system joint model is established to reflect the dynamic behavior of the bus, a bus voltage optimization method is designed, and the bus voltage deviation and the current tracking error of the front and rear stages are minimized through a comprehensive cost function, and the parameters of the front and rear two-stage controllers are optimized to achieve the purpose of system collaborative optimization control. The key parameters of the front and rear two-stage controllers are optimized by using the improved particle swarm optimization (IPSO) algorithm to ensure that the system can quickly adapt to changes and achieve the best operating state under different load conditions. Through this innovation, the collaborative optimization control of the front and rear stages is realized, significantly improving the efficiency and stability of the bidirectional converter during the charging and discharging process of electric vehicles. The system control structure diagram and the control flow chart are respectively as Figure 1 and Figure 2 shown.

[0053] Design of the front and rear stage controllers of the in-vehicle V2G bidirectional converter

[0054] The circuit structure of the two-stage V2G in-vehicle bidirectional converter is as Figure 3 shown. The front stage AC / DC realizes power factor correction and rectification and boost functions by a totem-pole bridgeless PFC, and its working principle is similar to that of a Boost boost circuit, which is composed of an energy storage inductor L s and power switch devices S1~S4, and is controlled by the pulse width modulation method (PWM); C1 is the intermediate bus capacitor; the rear stage DC / DC adopts a CLLC resonant converter, where S5~S8 form an inverter bridge, and S9~S 12 form a rectifier bridge, the resonant inductor L r1 and L r2 , the resonant capacitors C r1 and C r2 and the transformer excitation inductor L mTogether they form a resonant cavity, and the voltage gain during forward and reverse operation can be adjusted by the pulse frequency modulation method (PFM), which has the advantage of wide-range voltage output. This invention mainly focuses on the collaborative optimization control of the front and rear stages of the V2G vehicle-mounted bidirectional converter, so the complex derivation of the mathematical models of the front and rear stages will not be elaborated too much.

[0055] (1) Double-loop control strategy for the front-stage totem-pole bridgeless PFC circuit

[0056] As Figure 4 shown, a double-loop controller for the front-stage PFC converter is established, with model predictive control (MPC) for the current inner loop and proportional-integral (PI) control for the voltage outer loop.

[0057] 1) Design of model predictive control for the current inner loop

[0058] The goal of model predictive control for the current inner loop is to make the inductor current quickly track the reference current to achieve power factor correction (PFC).

[0059] a Prediction model

[0060] Discretize the continuous-time state equation to obtain the discrete-time state-space model:

[0061] (1)

[0062] In the formula, is the state vector, is the duty-cycle control input, is the output, is the system matrix, which can be calculated in the following way:

[0063] (2)

[0064] In the formula, is the system matrix of the continuous-time state-space model, is the sampling time, .

[0065] In the design of the prediction model section, based on the current state and the system state-space model, predict the system state at the next moments. For the current inner loop, the prediction model can be expressed as:

[0066] (3)

[0067] In the formula, is the sampling period, is the current moment. Considering the presence of the output voltage in the state vector, from equations (1) and (3), predict the next State at each moment:

[0068] (4)

[0069] b Cost function

[0070] The cost function of MPC is designed to minimize the current tracking error and the variation of the control input, and is used to evaluate the performance of the predicted control sequence:

[0071] (5)

[0072] In the formula, is the prediction step length, and are weight matrices, which are used to balance the state error and the variation of the control input respectively; is the adjustment parameter, represents the variation of the duty cycle of the control input.

[0073] c Receding horizon optimization

[0074] The optimization problem is to find the control sequence that minimizes the cost function:

[0075] (6)

[0076] The constraint conditions are:

[0077] (7)

[0078] (8)

[0079] In the design of the receding horizon optimization link, in each control period, the above optimization problem is solved to obtain the optimal control input sequence , and the first control input is executed. In the next control period, the state is updated and the optimization process is repeated.

[0080] (9)

[0081] d Design of the weight matrix

[0082] The design of the weight matrices and has an important impact on the performance of the MPC controller. Generally, is used to balance the state error, is used to balance the variation of the control input. By reasonably designing the weight matrices, it can be ensured that the system operates efficiently under dynamic load and external disturbance conditions.

[0083] (10)

[0084] In the formula, , , , are weighting factors set according to system requirements.

[0085] 2) Design of voltage outer-loop PI control

[0086] The goal of voltage outer-loop PI control is to make the output voltage track the reference voltage . The transfer function of the PI controller is:

[0087] (11)

[0088] In the formula, is the proportional gain, is the integral gain. By reasonably selecting the gains, the stability and dynamic performance of the system can be ensured.

[0089] The closed-loop transfer function of the voltage outer-loop can be derived through the small-signal model. Assuming that the system operates near the steady state, the small-signal model is:

[0090] (12)

[0091] In the formula, is the load resistance, is the capacitance value.

[0092] The output of the PI controller is the reference current of the current inner-loop, and its control law is:

[0093] (13)

[0094] Discretize the PI controller to obtain the discrete-time PI controller:

[0095] (14)

[0096] In the formula, is the voltage error, is the sum of the voltage errors accumulated at the current moment and all previous moments.

[0097] (2) Double-loop control strategy of the post-stage CLLC resonant converter

[0098] As Figure 5 shown, establish a double-loop controller for the post-stage CLLC resonant converter, with MPC control for the current inner-loop and PI control for the voltage outer-loop.

[0099] 1) MPC control for the current inner-loop

[0100] The resonant inductor and the resonant capacitor are used to form a resonant cavity, which determines the resonant frequency, output voltage, and energy transfer characteristics. The resonant frequency is , , and the switching frequency is the operating frequency of the converter. The goal of the current inner-loop MPC control is to make the resonant current track the reference current .

[0101] a Prediction model

[0102] Discretize the continuous-time state equation to obtain the discrete-time state-space model:

[0103] (15)

[0104] In the formula, is the state vector, is the frequency control input, is the output. is the system matrix, which can be calculated in the following way:

[0105] (16)

[0106] In the formula, is the system matrix of the continuous-time state-space model, , , , is the sampling time.

[0107] In the design of the prediction model link, based on the current state and the system state-space model, predict the system state at the next moments.

[0108] (17)

[0109] b Cost function

[0110] Define the cost function to evaluate the performance of the predicted control sequence:

[0111] (18)

[0112] In the formula, is the prediction step; and are the weight matrices, which are used to balance the state error and the change of the control input respectively; is the adjustment parameter; represents the change in the switching frequency of the control input.

[0113] c Receding horizon optimization

[0114] The optimization problem is to find the control sequence that minimizes the cost function:

[0115] (19)

[0116] Constraints:

[0117] (20)

[0118] In the design of the rolling optimization loop, within each control period, the above optimization problem is solved to obtain the optimal control input sequence , and the first control input is executed . In the next control period, the state is updated and the optimization process is repeated.

[0119] (21)

[0120] Design of the d weight matrix

[0121] Weight matrix and have an important impact on the performance of the MPC controller. By reasonably designing the weight matrix, it can be ensured that the system operates efficiently under dynamic load and external disturbance conditions.

[0122] (22)

[0123] Wherein, , , , are weighting factors set according to system requirements.

[0124] 2) Voltage outer loop PI control

[0125] The goal of the voltage outer loop PI control is to make the output voltage track the reference voltage , and the transfer function of the PI controller is:

[0126] (23)

[0127] Wherein, is the proportional gain, is the integral gain.

[0128] The output of the PI controller serves as the reference current for the current inner loop MPC control:

[0129] (24)

[0130] The PI controller is discretized to obtain the discrete-time PI controller:

[0131] (25)

[0132] Wherein, is the voltage error, is the sum of the voltage errors accumulated at the current moment and all previous moments.

[0133] Design of the coordinated optimization control method for the front and rear stages

[0134] After designing the double - closed - loop control of the totem - pole bridgeless PFC and the double - closed - loop control of the CLLC resonant converter, the parameter coordination optimization of the front - and rear - stage controllers is carried out. Considering the intermediate bus voltage and the key control quantities of the front and rear stages, a bus - voltage optimization method is proposed, a comprehensive cost function is designed, and an improved particle swarm optimization algorithm is used to realize the real - time optimization of the parameters of the front and rear - stage controllers.

[0135] (1) Bus - voltage optimization method

[0136] Based on the totem - pole PFC and the CLLC resonant converter, a system - level mathematical model is established to clarify the mapping relationship between the output bus voltage of the front stage and the input - port characteristics of the rear - stage resonant converter. The model can be expressed as:

[0137] (26)

[0138] Wherein, is the input power of the front stage, is the output power of the rear stage, and represent the efficiencies of the front stage and the rear stage respectively.

[0139] Under the energy - conservation constraint, the model can be expressed as:

[0140] (27)

[0141] By adjusting the bus voltage , the operating point with the highest overall efficiency can be found. In the on - vehicle V2G system, the dynamic characteristics of the bus voltage are mainly determined by the energy flow of the front - stage PFC circuit and the rear - stage CLLC converter. After establishing the overall system model, it can be analyzed that the change of the bus voltage can be described by the charging and discharging process of the capacitor. Equation (28) is the expression of the dynamic characteristics of the bus voltage:

[0142] (28)

[0143] Wherein, is the bus capacitor, is the bus voltage, is the input current (from the front - stage PFC circuit), is the output current (flowing from the secondary CLLC converter to the load). During equivalent analysis, the input voltage of the secondary stage is approximately equal to .

[0144] The bus voltage deviation is defined as the difference between the target voltage and the actual voltage :

[0145] (29)

[0146] The optimization goal is to minimize the bus voltage deviation , that is:

[0147] (30)

[0148] Considering the performance indicators of the system comprehensively, a comprehensive cost function is defined. This cost function includes the DC bus voltage error: ; the output voltage error of the secondary stage: ; the primary current tracking error: ; the secondary current tracking error: .

[0149] Finally, the optimized comprehensive cost function is:

[0150] (31)

[0151] In the formula, are the weights of each index.

[0152] The parameters optimized by this method include the proportional gain and integral gain of the primary voltage loop PI controller; the weight matrix and prediction step of the primary current loop MPC controller; the proportional gain and integral gain of the secondary voltage loop PI controller, as well as the weight matrix and prediction step of the secondary current loop MPC controller.

[0153] In summary, the optimization variables can be expressed as:

[0154] (32)

[0155] To ensure the stability and performance of the system, constraint conditions need to be set for the optimization variables.

[0156] (33)

[0157] (2) Improved Particle Swarm Optimization (IPSO) algorithm

[0158] The parameter optimization link of the bus voltage optimization method is implemented by using the Improved Particle Swarm Optimization (IPSO) algorithm. The main steps of the IPSO algorithm are as Figure 6 shown.

[0159] Initialize the position and velocity of the particle swarm. The particle position represents the key parameters of the controller. The position of each particle represents a set of parameters . The velocity of each particle is randomly initialized. The individual historical best position of each particle is initialized to its current position. The global best position is initialized to the best position in the particle swarm.

[0160] The fitness evaluation calculates the fitness value of each particle by simulating the dynamic response of the system. The fitness function is the reciprocal of the cost function . Based on the current particle position , simulate the dynamic response of the system, calculate the bus voltage deviation and the changes in the voltage and current of the front and rear stages. Then calculate the fitness value. The fitness function is the reciprocal of the cost function, that is:

[0161] (34)

[0162] In the formula, is the cost function value of the particle , is equal to the comprehensive cost function .

[0163] Update the position and velocity of the particle according to the individual historical best position and the global best position. As shown in the following formula:

[0164] (35)

[0165] (36)

[0166] In the formula, and are the velocity and position of the particle at time respectively; is the inertia weight, which is used to control the exploration amplitude of the particle; and are the learning factors, which control the influence of individual learning and group learning; and are random numbers, uniformly distributed between . is the historical optimal position of the particle ; is the global optimal position.

[0167] Introduce a population fitness feedback term on the basis of the traditional linearly decreasing weight to perform dynamic inertia weight adjustment:

[0168] (37)

[0169] In the formula, is the maximum number of iterations, is the optimal fitness value of the t-th generation, is the average fitness value of the t-th generation, is the maximum fitness value of the t-th generation.

[0170] Design an elite learning strategy and adopt differential updates for the top 20% of elite particles in terms of fitness:

[0171] (38)

[0172] The elite determination condition is:

[0173] (39)

[0174] In the formula, is the population average fitness, is the fitness standard deviation, is the historical optimum of the elite particle.

[0175] When the global optimal solution has not been updated for 5 consecutive generations, activate chaotic local search:

[0176] (40)

[0177] In the formula, is a chaotic variable, and the initial value is taken from the uniform distribution of (0, 1).

[0178] Dynamically adjust the learning factor as:

[0179] (41)

[0180] According to the fitness value, update the individual historical optimum and the global optimum. If the current fitness value is greater than the historical optimal fitness value, then update the individual historical optimum .

[0181] (42)

[0182] If the current fitness value If it is greater than the global optimal fitness value, update the global optimum. .

[0183] (43)

[0184] Repeat the above steps until the convergence condition is met (reaching the maximum number of iterations or the fitness value meets the requirements).

[0185] The parameter settings of the IPSO algorithm are as follows:

[0186] (44)

[0187] In the formula, is the number of particles, and the maximum number of iterations is set to 150 times.

[0188] During the operation of the system, the IPSO algorithm is used to optimize the key parameters of the front and rear stage controllers in real time to ensure that the control system can operate efficiently under different working conditions. Through real-time optimization, the adaptive ability and robustness of the system are improved.

[0189] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and without departing from the spirit or basic characteristics of the present invention, the present invention can be implemented in other specific forms. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be included in the present invention. Any reference signs in the claims should not be regarded as limiting the claimed rights.

[0190] In addition, it should be understood that although this specification is described according to embodiments, not every embodiment only contains an independent technical solution. This narrative way of the specification is only for clarity. Those skilled in the art should regard the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for collaborative optimization control of the front and rear stages of a vehicle-mounted V2G bidirectional converter, characterized in that, It includes the following steps: S1: Design the double - closed - loop controller for the front - end totem - pole bridgeless PFC and the double - closed - loop controller for the rear - end CLLC resonant converter. Among them, the current inner loop adopts model predictive control (MPC), and the voltage outer loop adopts proportional - integral (PI) control; S2: Establish a combined system model to reflect the dynamic behavior of the bus voltage during the charging and discharging process. Design a bus voltage optimization method to achieve the collaborative optimization of the parameters of the front - end and rear - end controllers through a comprehensive cost function. The comprehensive cost function includes the bus voltage error and the current tracking errors of the front - end and rear - end; S3: Use the improved particle swarm optimization (IPSO) algorithm to optimize the prediction step size and weight parameters of the MPC link and the gains of the PI link in real - time, making the double - closed - loop control strategies of the front - end and rear - end more flexible and reliable in application.

2. The collaborative optimization control method for the front and rear stages of the in-vehicle V2G bidirectional converter according to claim 1, wherein The design of the double - closed - loop controller for the front - end totem - pole bridgeless PFC is as follows: Current inner - loop MPC control: Prediction model: Discretize the continuous - time state equation to obtain a discrete - time state - space model. Based on the current state and the system state - space model, predict the system state at future moments; Cost function: Design the cost function to minimize the current tracking error and the change in the control input, which is used to evaluate the performance of the predicted control sequence; Rolling optimization: In each control cycle, solve the optimization problem to obtain the optimal control input sequence, execute the first control input, then update the state and repeat the optimization process; Weight matrix design: Set the weighting factor according to the system requirements to balance the state error and the change in the control input; Voltage outer - loop PI control: The transfer function of the PI controller is a known formula, and its output is the reference current of the current inner loop. By reasonably selecting the gain, ensure the stability and dynamic performance of the system.

3. The collaborative optimization control method for the front and rear stages of the in-vehicle V2G bidirectional converter according to claim 1, wherein, The design of the double - closed - loop controller for the rear - end CLLC resonant converter is as follows: Current inner - loop MPC control: Prediction model: Discretize the continuous - time state equation to obtain a discrete - time state - space model. Based on the current state and the system state - space model, predict the system state at future moments; Cost function: Define the cost function to evaluate the performance of the predicted control sequence, minimizing the current tracking error and the change in the control input; Rolling optimization: In each control cycle, solve the optimization problem to obtain the optimal control input sequence, execute the first control input, then update the state and repeat the optimization process; Weight matrix design: Set the weighting factor according to the system requirements to balance the state error and the change in the control input; Voltage outer - loop PI control: The transfer function of the PI controller is a known formula, and its output is used as the reference current for the current inner - loop MPC control. By reasonably selecting the gain, ensure the stability and dynamic performance of the system.

4. The collaborative optimization control method for the front and rear stages of the in-vehicle V2G bidirectional converter according to claim 1, characterized in that The bus voltage optimization method includes: Establish a system - level mathematical model to clarify the mapping relationship between the front - end output bus voltage and the input - port characteristics of the rear - end resonant converter; Analyze the dynamic characteristics of the bus voltage, and describe the change of the bus voltage through the charging and discharging process of the capacitor; Define the bus voltage deviation as the difference between the target voltage and the actual voltage, and the optimization goal is to minimize the bus voltage deviation. Considering the performance indicators of the system comprehensively, a comprehensive cost function is defined, including the DC bus voltage error, the output voltage error of the latter stage, the current tracking error of the former stage, and the current tracking error of the latter stage.

5. The collaborative optimization control method for the front and rear stages of the in-vehicle V2G bidirectional converter according to claim 1, wherein The improved particle swarm optimization (IPSO) algorithm includes the following steps: S1: Initialize the positions and velocities of the particle swarm. The particle positions represent the key parameters of the controller. The velocity of each particle is randomly initialized. The individual historical best position of each particle is initialized to its current position, and the global best position is initialized to the best position in the particle swarm. S2: Fitness evaluation: Calculate the fitness value of each particle by simulating the dynamic response of the system. The fitness function is the reciprocal of the cost function. S3: Update the positions and velocities of the particles according to the individual historical best position and the global best position. S4: Introduce a population fitness feedback term on the basis of the traditional linearly decreasing weight to adjust the dynamic inertia weight. S5: Design an elite learning strategy and update the top 20% of the elite particles with differentials according to their fitness. S6: When the global optimal solution has not been updated for 5 consecutive generations, activate the chaotic local search and dynamically adjust the learning factor. S7: Update the individual historical best and the global best according to the fitness value, and repeat the above steps until the convergence condition is met.

6. The collaborative optimization control method for the front and rear stages of the in-vehicle V2G bidirectional converter according to claim 1, characterized in that The parameters of the IPSO algorithm are set as follows: the number of particles is a known quantity, and the maximum number of iterations is set to 150 times.

Images