An intelligent fuel cell vehicle energy management strategy for complex dynamic urban traffic environments

Abstract

This invention relates to an energy management strategy for intelligent fuel cell vehicles in complex dynamic urban traffic environments. The speed optimization module updates the state at time k according to a time series, using GRAMPC based on the fast projection gradient method to perform real-time speed optimization by detecting the speeds of vehicles ahead and behind and the status of traffic lights. Simultaneously, fuzzy regenerative braking energy management is implemented due to the influence of a complex nonlinear system with multiple variables, including the external environment and the vehicle's own state. A multi-objective performance function is designed, and MPC control weight coefficients are optimized to achieve speed tracking and energy allocation. Finally, the actual output speed of the vehicle is used as the input at time k+1 to re-optimize the speed curve, and this process is iterated repeatedly. This invention utilizes the Pontryagin minimum principle, MPC control algorithm, and fuzzy control to collaboratively control the entire vehicle system, combining SPAT information to plan feasible speeds for passing traffic lights, thereby improving energy utilization and battery life.

Description

Technical Field

[0001] This invention belongs to the field of energy management technology for hybrid power systems, and specifically relates to an energy management strategy for intelligent fuel cell vehicles in complex and dynamic urban traffic environments. Background Technology

[0002] With the increasing intensity of energy consumption and the environmental damage caused by fossil fuels, the development of clean energy has received widespread attention from all countries. In the automotive field, intelligent fuel cell hybrid vehicles (IFCHVs), powered by fuel cells and lithium batteries, have solved the problems of short driving range and long charging time of electric vehicles. With their advantages of zero pollution, high energy density, and high energy conversion efficiency, they have become a hot topic of research in recent years.

[0003] The complexity of urban road traffic environments presents significant challenges to the implementation of intelligent driving and energy management. Previous research has largely relied on standard driving cycles for offline or global optimization, achieving good speed prediction accuracy and fuel economy. However, real-world traffic environments are filled with uncertainties, rendering these intelligent driving strategies inadequate for practical applications. Implementing intelligent driving and energy management for intelligent fuel cell vehicles in urban traffic environments faces the following challenges: 1. The limited availability of intelligent vehicles in current traffic conditions, with traditional vehicles dominating; 2. Traffic flow bottlenecks caused by signalized intersections; 3. Dynamically changing road environments, such as dynamic interference from surrounding vehicles, road gradient, and road surface adhesion coefficient; 4. Additional energy loss due to frequent braking. Therefore, incorporating these factors to achieve dynamic real-time optimization is crucial for the intelligent driving and energy management of intelligent fuel cell vehicles. Summary of the Invention

[0004] To address the shortcomings of the existing technologies, this invention proposes an intelligent fuel cell vehicle energy management strategy for complex dynamic urban traffic environments. By combining dynamic information such as traffic light status, the speed of vehicles in front and behind (traditional vehicles), and road gradient, intelligent driving and energy management of intelligent fuel cell vehicles are implemented.

[0005] To achieve the above objectives, this invention proposes an intelligent fuel cell vehicle energy management strategy for complex dynamic urban traffic environments. This invention constructs a GRAMPC speed planning module based on the fast projection gradient method, achieving real-time speed reference curves within the prediction range. Simultaneously, a multi-objective optimization energy management strategy based on MPC is established to efficiently allocate energy among various systems according to the vehicle's power demand. To better adapt to driving behavior in urban traffic environments and further improve energy utilization, an energy recovery calculation model is incorporated into the speed planning module, and a fuzzy control-based braking energy recovery strategy is designed in the vehicle control unit. Specifically, the following are included:

[0006] 1) Vehicle longitudinal dynamics model:

[0007] In the time-domain model, the state variable is x = [sv] T Where s is the vehicle's displacement over time and v is the vehicle's speed, the longitudinal dynamics state equation of the vehicle can be described as follows:

[0008]

[0009] Where u = F t F t For vehicle driving force, F t =F f +F i +F w +F j F f For rolling resistance, F i For slope resistance, F w For air resistance, F j To increase resistance.

[0010] 2) Target velocity model

[0011] In order to enable vehicles to pass through signalized intersections without stopping, thereby reducing the additional energy consumption caused by braking, it is necessary to obtain a feasible target speed v for passing through the green light window. traget Its expression is as follows:

[0012]

[0013]

[0014] Make:

[0015] v min ≤v target (k)≤v max

[0016] t cycle =tr +t g

[0017]

[0018] Where d a (k) represents the position between the target vehicle and the traffic light, t r and t g Let t be the window time for red and green lights, respectively. cycle K represents the total cycle time of the traffic light. w The mod() function represents the number of cycles of a traffic light; k divided by t is generated by the mod() function. cycle The remainder. min As the lower limit of the road speed constraint, v max This is the upper limit of the road speed constraint.

[0019] When a vehicle passes through a traffic light within the current green light window, it means that a speed constraint exists [v]. lb (k)v ub (k)] Allows vehicles to avoid stopping at red light windows, with the speed range selected as follows:

[0020]

[0021] v ub (k)=v target (k)

[0022] Where v lb (k) is the lower limit of velocity, v ub (k) is the upper limit of speed.

[0023] 3) Minimum safe distance speed model

[0024] During vehicle movement, after the preceding and following vehicles brake to a stop, to ensure that a rear-end collision does not occur, ΔS1 and ΔS2 must satisfy ΔS1 > 0 and ΔS2 > 0. Therefore:

[0025]

[0026] Where ΔS1 is the stopping distance between the preceding vehicle and the target vehicle, and ΔS2 is the stopping distance between the following vehicle and the target vehicle. L1 is the minimum safe driving distance between the target vehicle and the preceding vehicle, L0 is the minimum safe driving distance between the target vehicle and the following vehicle, and S... A S is the braking distance of the vehicle in front. X S is the braking distance of the target vehicle. B This is the braking distance of the vehicle behind.

[0027] 4) Vehicle system energy model

[0028] Because regenerative braking has a significant impact on overall energy consumption, it is incorporated into the optimization objective function. Therefore, over a period T, the energy consumption E... t This can be expressed as:

[0029]

[0030] in:

[0031]

[0032]

[0033] E brake =E brake_rege +E brake_fric

[0034]

[0035] Among them, E drive E is the energy passed through the wheels during driving. brake It is the energy passing through the wheels during braking, where E brake_rege It is the energy flowing towards regenerative braking, E brake_fric It is the energy flowing towards friction braking. E brake_rege and E brake_fric All are negative values. P W It is the instantaneous power of air resistance, P f It is the instantaneous power of rolling resistance, P i It is the instantaneous power of the slope resistance, P loss η is the motor power loss, δ is the motor efficiency, m is the rotational mass conversion factor, and m is the vehicle mass.

[0036] 5) Establish the vehicle speed optimization objective function and solve the optimal control problem.

[0037] This invention employs the Pontryagin minimum principle, thus transforming the time-based MPC optimization control problem into the following form:

[0038]

[0039] st x'(t)=f(x,u)

[0040] T min ≤T w.k ≤T max

[0041] Where L is the integral cost, P(t) is the output power, and λ P and λ V v is the weighting coefficient. des For the desired speed, T wkFor output torque, T min For minimum torque, T max Maximum torque

[0042] In solving the optimal control problem (OCP), the projection gradient method is used based on the first-order optimality conditions of each problem. Therefore, it is necessary to define the Hamiltonian:

[0043] H(t,x(t),u(t),λ(t))=L(t,x(t),u(t))+λ T (t)f(t+t k ,x(t),u(t))π

[0044] Where λ is the adjoint state, t k This is a sample instance.

[0045] 6) Regenerative braking energy recovery strategy

[0046] During vehicle braking, the distribution of braking force is influenced by many factors, such as State of Charge (SOC), vehicle speed (v), and braking intensity (z). Therefore, when designing a braking control strategy, their impact on braking force distribution should be comprehensively considered. Fuzzy control has unique advantages in handling such nonlinear, multivariable, and complex systems. Therefore, this invention adopts a multivariable fuzzy controller structure. The controller's inputs are State of Charge (SOC), vehicle speed (v), and braking intensity (z), and the output variable is the regenerative braking force distribution coefficient K. d Based on the constructed fuzzy control structure, five linguistic values ​​are used to describe the input and output variables, namely, minimum (NB), small (NS), medium (M), large (PS), and maximum (PB). The physical domains of the input and output variables in this invention are as follows: SoC: [0,1], vehicle speed: [0,30], braking intensity z: [0,1], and regenerative braking distribution coefficient K. d [0.5,1]. Furthermore, based on the above rules, 125 fuzzy rules consisting of fuzzy statements are formulated.

[0047] 7) Energy Management and Control Strategies

[0048] Based on the MPC algorithm, this invention uses a discrete-time system state-space model as its model prediction, and the state variable x(k) at time k is defined as... Control quantity is Where i fc This represents the fuel cell current, SoC represents the state of charge, and u fc For fuel cell voltage, i bat Let be the lithium battery current; in this model, all states of the vehicle system are measurable, and there is a state measurement value x(k) at time k. According to the basic principles of predictive control, the constrained MPC optimization problem can be described as:

[0049]

[0050] Where ΔU(k) ​​is the control increment sequence, the solved ΔU(k) ​​is used as the control increment at the next time step, i.e., time k+1, and the energy distribution between the fuel cell and the lithium battery at time k+1 is completed through the DC converter.

[0051] The constrained MPC optimization problem is further transformed into the following QP problem for solution:

[0052]

[0053] satisfy:

[0054] C u ΔU(k)≥b(k+1|k)

[0055] Where H represents the Hessian matrix, ΔU(k) ​​represents the sequence of control increments, G is the intermediate matrix variable, and C u This is the constraint matrix.

[0056] This invention proposes an intelligent fuel cell vehicle energy management strategy for complex dynamic urban traffic environments. It considers factors such as traffic light status, dynamic changes of vehicles ahead and behind, and road gradient, as well as safe driving distances between vehicles. This invention incorporates the driving behavior of traditional vehicles into the constraints of intelligent driving and establishes more realistic driving scenarios, thus possessing good universality.

[0057] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0058] 1. This invention is an energy management strategy that integrates velocity planning, model prediction and energy recovery. It designs a sequence velocity optimization method based on the Pontryagin minimum principle, a multi-objective optimization control strategy based on MPC and an energy recovery method based on fuzzy control.

[0059] 2. This invention uses the GRAMPC method based on efficient projection gradient algorithm and adaptive linear search, which can effectively solve nonlinear optimal control problems. Its fast response is particularly suitable for real-time velocity trajectory optimization.

[0060] 3. This invention enables vehicles to pass through signalized intersections without stopping under the guidance of the target speed, can efficiently allocate energy according to the power demand of vehicles, and the designed energy recovery strategy improves the energy utilization rate.

[0061] 4. By establishing a complex and dynamic urban traffic environment, the factors affecting vehicle driving are taken into account, providing a more realistic and universal simulation environment for experimental simulation. Attached Figure Description

[0062] Figure 1 This is a schematic diagram of the present invention.

[0063] Figure 2 This is a schematic diagram illustrating the principle of calculating the target velocity.

[0064] Figure 3 This is a schematic diagram of the minimum safe distance model.

[0065] Figure 4 This is a membership function graph for fuzzy control.

[0066] Figure 5 This is a diagram of an embodiment of the present invention.

[0067] Figure 6 This is a speed optimization diagram based on an embodiment of the present invention.

[0068] Figure 7 This is a vehicle charge state diagram in an embodiment of the present invention.

[0069] Figure 8 This is a battery degradation diagram from an embodiment of the present invention. Detailed Implementation

[0070] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0071] Example 1: The schematic diagram of the present invention is as follows Figure 1 As shown, the core mechanism involves the speed optimization control module obtaining a real-time speed reference curve within the predicted range. The energy management module uses this speed for efficient power allocation, while the energy recovery module optimizes the braking force distribution coefficient in real time. Specifically, the speed optimization module combines information such as traffic light status, the speeds of vehicles ahead and behind, its own status, and road gradient to calculate a safe speed in real time within the predicted range. The energy management control module then uses this speed, along with the battery and traction systems, to allocate the required power in real time, while simultaneously coordinating with the energy recovery module for efficient braking force distribution. Finally, the intelligent fuel cell vehicle outputs driving force in the form of torque, and the real-time speed output is fed back to the speed optimization module to continuously optimize the current speed trajectory.

[0072] The principle diagram of this invention for calculating the target speed through the green window of a traffic light is shown below. Figure 2 As shown.

[0073] The minimum safe distance model of this invention is as follows: Figure 3 As shown.

[0074] The fuzzy control membership function proposed in this invention is as follows: Figure 4 As shown, the fuzzy control rules established accordingly are shown in Table 1.

[0075] Table 1 Fuzzy Control Rules

[0076]

[0077]

[0078] This invention proposes an energy management strategy for intelligent fuel cell vehicles with dynamic traffic constraints. The speed optimization module updates state changes according to a time series, combining traffic light status, the vehicle's own state, and the speed changes of vehicles ahead and behind. It then uses GRAMPC based on the fast projection gradient method to obtain the safe speed v at the current time k. Simultaneously, it assesses the current traffic environment and avoids collisions during driving based on a minimum safe distance model. Due to the influence of a complex nonlinear system with multiple variables such as the external environment and the vehicle's own state, fuzzy regenerative braking energy management is implemented. Furthermore, based on a designed multi-objective performance function, MPC control weight coefficients are optimized to achieve speed tracking and energy allocation. Finally, the vehicle's output speed is used as the input at time k+1 to recalculate the speed curve, and this process is iterated repeatedly.

[0079] Step 1: Obtain the reference velocity trajectory

[0080] First, based on the vehicle's position, the speeds of vehicles in front and behind, and the current traffic light status, calculate the target speed for passing through the signalized intersection during the green light window, as shown below:

[0081]

[0082]

[0083] Make:

[0084] v min ≤v target (k)≤v max

[0085] t cycle =t r +t g

[0086]

[0087] Where d a (k) represents the position between the target vehicle and the traffic light, t r and t g Let t be the window time for red and green lights, respectively. cycle K represents the total cycle time of the traffic light. w The mod() function represents the number of cycles of a traffic light; k divided by t is generated by the mod() function. cycle The remainder. min As the lower limit of the road speed constraint, vmax This is the upper limit of the road speed constraint.

[0088] Then, in the speed planning module, to reduce energy output, braking energy recovery is introduced into the optimization objective function. Therefore, the energy consumption model within the current prediction range is described as follows:

[0089]

[0090] in:

[0091]

[0092]

[0093] E brake =E brake_rege +E brake_fric

[0094]

[0095] Where E brake It is the energy passing through the wheels during braking, where E brake_rege It is the energy flowing towards regenerative braking, E brake_fric It is the energy flowing towards friction braking. E brake_rege and E brake_fric All are negative values. P W It is the instantaneous power of air resistance, P f It is the instantaneous power of rolling resistance, P i It is the instantaneous power of the slope resistance, P loss η is the motor power loss, δ is the motor efficiency, m is the rotational mass conversion factor, and m is the vehicle mass.

[0096] After determining the energy output at time k, the next step is to calculate the matching of safe speed and driving force. Based on Pontryagin's minimum principle, the above time-based optimization control problem (OCP) is transformed into the following form:

[0097]

[0098] st x'(t)=f(x,u)

[0099] T min ≤T w.k ≤T max

[0100] Where L is the integral cost, P(t) is the output power, and λ P and λ V v is the weighting coefficient. des For the desired speed, T wk For output torque, Tmin For minimum torque, T max Maximum torque;

[0101] During the speed optimization process described above, the current road conditions are assessed in real time, such as road gradient and pavement adhesion coefficient, while also taking into account the constraints of the road's speed limits. By considering these factors, a speed reference trajectory is obtained.

[0102] Step 2: Energy Recovery and Energy Management

[0103] During vehicle braking, the distribution of braking force is influenced by many factors, such as State of Charge (SOC), vehicle speed (v), and braking intensity (z). Therefore, when designing a braking control strategy, their impact on braking force distribution should be comprehensively considered. Fuzzy control has unique advantages in handling such nonlinear, multivariable, and complex systems. Therefore, this invention adopts a multivariable fuzzy controller structure. The controller's inputs are State of Charge (SOC), vehicle speed (v), and braking intensity (z), and the output variable is the regenerative braking force distribution coefficient K. d Based on the constructed fuzzy control structure, five linguistic values ​​are used to describe the input and output variables, namely, minimum (NB), small (NS), medium (M), large (PS), and maximum (PB). The physical domains of the input and output variables in this invention are as follows: SoC: [0,1], vehicle speed: [0,30], braking intensity z: [0,1], and regenerative braking distribution coefficient K. d [0.5,1]. Based on the above rules, 125 fuzzy rules were formulated as shown in Table 1, and their membership function graph is shown below. Figure 4 As shown.

[0104] Based on the MPC algorithm, this invention uses a discrete-time system state-space model as its model prediction, and the state variable x(k) at time k is defined as... Control quantity is Where i fc This represents the fuel cell current, SoC represents the state of charge, and u fc For fuel cell voltage, i bat Let be the lithium battery current; in this model, all states of the vehicle system are measurable, and there is a state measurement value x(k) at time k. According to the basic principles of predictive control, the constrained MPC optimization problem can be described as:

[0105]

[0106] Where ΔU(k) ​​is the control increment sequence, the solved ΔU(k) ​​is used as the control increment at the next time step, i.e., time k+1, and the energy distribution between the fuel cell and the lithium battery at time k+1 is completed through the DC converter.

[0107] The constrained MPC optimization problem is further transformed into the following QP problem for solution:

[0108]

[0109] satisfy:

[0110] C u ΔU(k)≥b(k+1|k)

[0111] Where H represents the Hessian matrix, ΔU(k) ​​represents the sequence of control increments, G is the intermediate matrix variable, and C u This is the constraint matrix.

[0112] Finally, in order to evaluate the overall performance of the control strategy, this invention mainly evaluates it from three performance indicators: battery degradation, hydrogen consumption, and battery state of charge.

[0113] $ global =$ Δfc +$ H2 +$ Δbat +$ charge

[0114] Among them, $ Δfc For fuel cell degradation costs, $ H2 For the cost of hydrogen consumption, $ Δbat The cost of lithium battery degradation, $ charge It's the cost of charging penalties, $ global It is the overall cost.

[0115] Figure 5 The ST diagram in this embodiment illustrates the time displacement changes of each vehicle throughout the entire journey. It can be clearly observed that during the distance between traffic lights 1, 2, and 3, the vehicle is primarily influenced by the vehicle in front, and its speed changes are consistent with the speed of the vehicle in front, and it cannot exceed the speed of the vehicle in front. Starting from traffic light 5, the following vehicle gradually approaches the target vehicle. When the distance between the two vehicles shrinks to the minimum safe distance, the speed of the target vehicle changes synchronously with that of the following vehicle, ensuring that it does not fall below this speed. At traffic light 7, influenced by the distance between the traffic light and the surrounding space, the planned target speed can pass through the current traffic light without exceeding the maximum speed limit, and then gradually shake off the interference from the following vehicle. However, for adaptive cruise control (ACC), this process continues until after traffic light 9 before gradually shaking off the interference from the following vehicle.

[0116] Figure 6The speed curve in this embodiment describes the speed changes throughout the driving process. Since it is not limited by the cruise speed, the speed change fluctuation range in the planned state is large. That is, it tries to pass the current traffic light within the feasible speed range, rather than in the ACC state, where if the current cruise speed cannot pass the current traffic light, it can only decelerate to meet the requirement of passing the current traffic light in the next green light window.

[0117] Figure 7 and Figure 8 The changes in State of Charge (SoC) and battery degradation throughout the entire cycle are shown separately. Figure 5 The ST graph clearly shows that the fluctuations in the SoC and battery degradation curves under the two driving states reflect the changes in the vehicle's state during driving. For SoC, the curve under the planned driving state is lower than that under the ACC driving state, indicating that the SoC fluctuation under the planned driving state is smaller throughout the entire driving cycle. Regarding battery degradation, the results show that battery degradation caused by ACC driving is greater than that under the planned driving state. Therefore, driving the vehicle under the planned state is beneficial for extending battery life.

[0118] Table 2 Evaluation Indicators for Energy Management Strategies

[0119]

[0120] Table 2 shows the results of various performance indicators under two driving conditions. It can be seen that the performance indicators under the planning state have significant advantages and the total cost is greatly reduced, indicating that the control strategy proposed in this invention has good energy saving and economy.

[0121] It will be readily understood by those skilled in the art that the above embodiments are only for illustrating the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the content of the present invention and implement it accordingly. They are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An intelligent fuel cell vehicle energy management strategy for complex dynamic urban traffic environments, characterized in that: The speed optimization module updates the intelligent fuel cell vehicle's own state according to the time sequence, and combines it with the current moment. k Based on the traffic light status, vehicle speeds, and road gradient information, a feasible target speed for passing through the traffic light is calculated. Simultaneously, based on the target speed, GRAMPC based on the fast projection gradient method is used to obtain the current time step. k safe speed v The energy management control module optimizes the MPC weight coefficients based on a multi-objective performance function, according to the safe speed. v It combines the battery system and traction system to allocate the driving power demand in real time. At the same time, it coordinates with the energy recovery module based on fuzzy control to carry out efficient braking force distribution. Finally, the intelligent fuel cell vehicle outputs driving force in the form of torque, and then feeds back the real-time speed output by the vehicle to the speed optimization module as the input at time k+1 to continuously optimize the current speed trajectory. Includes the following steps: (1) Establish the longitudinal dynamics model of the vehicle wherein , F t is a vehicle driving force, x is a vehicle system state quantity, wherein s is a vehicle travel displacement over time, v is a vehicle travel speed; (2) Establish the target speed of the vehicle According to the current location of the vehicle and the signal light information, a feasible target speed is selected As follows: Make: in This refers to the position between the vehicle and the traffic light. and These are the window times for red and green lights, respectively. This represents the total cycle time of the traffic light. Represents the number of cycles of the traffic light. Function generation Divide by The remainder, This is the lower limit of the road speed constraint. This is the upper limit of the road speed constraint; (3) Establish a vehicle system energy model Braking energy recovery has a significant impact on the energy consumption of vehicle systems. Introducing braking energy recovery into the vehicle speed optimization objective function can, over a period of time… Energy consumption of the vehicle system for: in in, It is the energy that passes through the wheels during propulsion. It is the energy that passes through the wheels during braking, of which It is the energy flowing towards regenerative braking. It is the energy flowing towards friction braking. and All are negative values. It is the instantaneous power of air resistance. It is the instantaneous power of rolling resistance. It is the instantaneous power of the slope resistance. It is motor power loss. It's the motor efficiency. It is the rotational mass conversion factor. It's about vehicle quality; (4) Establish the objective function for optimizing vehicle speed Get the current time k After energy output, i.e. energy consumption, it is necessary to calculate the optimal safe speed. v Matching the vehicle's driving force, based on Pontryagin's minimum principle, transforms the time-based optimization control problem into the following form: in, For the cost of points, For output power, and These are the weighting coefficients. For the desired speed, For output torque, For minimum torque, Maximum torque; (5) Establish a regenerative braking energy recovery strategy During vehicle braking, the distribution of braking force is affected by many nonlinear and complex factors. A multivariable fuzzy controller structure is adopted, with the controller inputs being SoC and vehicle speed. v Braking intensity z, the output variable is the regenerative braking force distribution coefficient. ; (6) Establish energy management and control strategies According to the MPC algorithm, a discrete-time system state-space model is used as the model prediction. k state quantity at time 1 Defined as The control quantity is ,in For fuel cell current, SoC In a charged state, For fuel cell voltage, For the lithium battery current; in this model, all states of the vehicle system are measurable, in There are state measurement values ​​at all times. Based on the fundamental principles of predictive control, the constrained MPC optimization problem can be described as follows: in It controls the increment sequence, which is obtained by solving... As the next moment k The control increment at time +1 is achieved through a DC converter. k Energy distribution between the fuel cell and lithium battery at time +1.

2. The intelligent fuel cell vehicle energy management strategy for complex dynamic urban traffic environment according to claim 1, characterized in that: The vehicle driving force in step (1) We obtain it from the following formula: in For rolling resistance, For slope resistance, For air resistance, To increase resistance.

3. The intelligent fuel cell vehicle energy management strategy for complex dynamic urban traffic environment as claimed in claim 1 wherein: In step (2), when a vehicle passes through the traffic light during the current green light window, the speed constraint range that allows the vehicle to avoid stopping during the red light window is: ,in: is a lower speed limit, is an upper speed limit.

4. The intelligent fuel cell vehicle energy management strategy for complex dynamic urban traffic environments as described in claim 1. characterized in that In step (4), for time-based optimization control problems, the projection gradient method based on the respective first-order optimality conditions is used, which requires defining the Hamiltonian: wherein is a concomitant state, is a sampling instance.

5. The intelligent fuel cell vehicle energy management strategy for complex dynamic urban traffic environments according to claim 1, characterized in that: To ensure driving safety, a minimum safe distance model is incorporated into the optimization control problem in step (4): If, after the vehicles in front and behind have braked to a stop, to ensure that a rear-end collision does not occur, then and Must meet , Then we have: in, The stopping distance between the vehicle in front and the target vehicle. The parking distance between the following vehicle and the target vehicle. The minimum safe driving distance between the target vehicle and the vehicle in front. The minimum safe driving distance between the target vehicle and the vehicle behind it. This is the braking distance of the vehicle in front. The braking distance of the target vehicle. This is the braking distance of the vehicle behind.

6. The intelligent fuel cell vehicle energy management strategy for complex dynamic urban traffic environments according to claim 1, characterized in that: In step (5), based on the constructed fuzzy controller structure, five linguistic values ​​are used to describe the input and output variables, namely minimum, small, medium, large, and maximum. The physical domains of the input and output variables are respectively: SoC: [0, 1], vehicle speed: [0, 30], braking intensity z: [0, 1], and regenerative braking distribution coefficient. [0.5, 1], and based on the above rules, 125 fuzzy rules are formulated.

7. The intelligent fuel cell vehicle energy management strategy for complex dynamic urban traffic environments according to claim 1, characterized in that In the energy management control model, the state of charge of the battery is: in for SoC Initial value, For battery capacity, This represents the battery current.

8. The intelligent fuel cell vehicle energy management strategy for complex dynamic urban traffic environment as claimed in claim 1, wherein: The energy management and control model transforms the constrained MPC optimization problem into the following QP problem for solution: satisfy: in, Represents the Hessian matrix. , This represents a sequence that controls the increment. G For intermediate matrix variables, and , This is the constraint matrix.

Images