Electric vehicle predictive thermal management system based on traffic flow information

Abstract

This invention relates to a predictive thermal management system for electric vehicles based on traffic flow information, comprising: a traffic information acquisition module for acquiring real-time network traffic flow information ahead of the vehicle; an upper-level long-time-domain temperature planner, based on a dynamic programming algorithm, using the traffic flow information to predict future vehicle speed and battery heat generation, and aiming to minimize total system energy consumption, solving for the optimal temperature reference trajectory of the cabin and battery in the long-time domain that satisfies the constraints of the cabin comfort temperature range and the battery safety temperature range; and a lower-level short-time-domain temperature tracker, employing a model predictive control algorithm, guided by the optimal temperature reference trajectory, coordinating and controlling the actuators of the thermal management system in real-time in the short-time domain, dynamically allocating the cooling capacity of the cabin and battery, and achieving accurate temperature tracking and energy consumption optimization. Compared with existing technologies, this invention has advantages such as balancing control accuracy and computational burden, and effectively reducing the energy consumption of the thermal management system.

Description

Technical Field

[0001] This invention relates to the field of electric vehicle thermal management technology, and in particular to a predictive thermal management system for electric vehicles based on traffic flow information. Background Technology

[0002] The global transition to transportation electrification is a key initiative to address climate change and reduce reliance on fossil fuels. Electric vehicles (EVs), as the core of this transition, rely heavily on the effectiveness of their thermal management systems for range and performance. Lithium-ion batteries, a core component of EVs, are extremely sensitive to operating temperatures. Operating in high-temperature environments, especially in regions with frequent heat waves, battery performance deteriorates rapidly. Exceeding the optimal temperature range leads to accelerated capacity decay and increased self-discharge rates, potentially causing battery fires or even explosions, posing a serious safety threat to occupants. Meanwhile, passenger cabin comfort is another important thermal management objective. Unlike internal combustion engine (ICE) vehicles, which have a significant amount of waste heat available for utilization, EV air conditioning systems must draw power directly from the battery. This creates a direct, trade-off between cabin thermal comfort and vehicle range. Therefore, developing an advanced thermal management strategy that can collaboratively manage battery and cabin thermal demands is crucial for promoting reliable and efficient EVs in diverse climate markets worldwide.

[0003] Traditional thermal management strategies often employ rule-based or PID control methods, which are reactive controls. They cannot utilize future traffic information for proactive adjustments, resulting in problems such as high energy consumption, temperature control lag, or overshoot.

[0004] With the widespread adoption of Vehicle-to-Everything (V2X) and intelligent transportation systems, optimizing vehicle energy and thermal management using predictive information such as traffic flow ahead and road gradient has become a research hotspot. However, existing predictive thermal management technologies still face significant challenges in practical applications. On the one hand, existing strategies often employ decoupled control architectures, independently managing the thermal of the cabin or battery system, lacking coordinated scheduling of the overall vehicle thermal load, resulting in limited energy utilization efficiency. On the other hand, the cabin and the power battery exhibit significant differences in thermodynamic characteristics: the battery system has considerable thermal inertia, with a noticeable lag in temperature changes, while the cabin thermal environment responds relatively quickly to adjustments. At the control strategy level, while Model Predictive Control (MPC) is often used to handle such multi-constraint optimization problems, traditional fixed-time-domain MPC struggles to consider the dynamic response characteristics of different subsystems. Relying solely on short-time-domain predictions cannot anticipate long-term battery thermal accumulation, easily leading to adjustment lag and increased energy consumption; while long-time-domain predictions impose a huge computational burden. Therefore, there is an urgent need for a predictive thermal management method that can integrate traffic forecast information, collaboratively manage cabin and battery thermal loads with different time scales, and effectively balance global planning capabilities with real-time control requirements.

[0005] A search revealed that Chinese invention patent application publication number CN120287799A discloses an automatic control method and device for an electric vehicle thermal management system. The method includes: acquiring real-time vehicle state data and external prediction information, wherein the external prediction information includes navigation path information and weather forecast information for the next N time steps; constructing a current system state vector based on the real-time vehicle state data; converting the external prediction information into a predictable disturbance sequence; constructing an MPC optimization problem instance based on the current system state vector and the predictable disturbance sequence; inputting the MPC optimization problem instance into an online optimization solution module to obtain the current optimal control command; and sending the current optimal control command to the drive actuator of the electric vehicle thermal management system to obtain physical drive signals and updated real-time vehicle state data. The process of transforming the external prediction information into a predictable disturbance sequence includes: predicting the battery heat generation power and motor / electronic control heat generation power for the next P steps based on the navigation path information for the next N time steps to obtain battery heat generation power sequences and motor / electronic control heat generation power sequences; predicting the environmental heat load on the cabin and the heat exchange volume on the battery compartment for the next P steps based on the weather forecast information for the next N time steps to obtain heat load sequences and heat exchange volume sequences; performing multi-source disturbance dynamic structured correlation equilibrium on the battery heat generation power sequence, the motor / electronic control heat generation power sequence, the heat load sequence, and the heat exchange volume sequence respectively to obtain battery heat generation power equilibrium sequences, motor / electronic control heat generation power equilibrium sequences, heat load equilibrium sequences, and heat exchange volume equilibrium sequences; and integrating the battery heat generation power equilibrium sequences, motor / electronic control heat generation power equilibrium sequences, heat load equilibrium sequences, and heat exchange volume equilibrium sequences to obtain the predictable disturbance sequence. The existing patent application uses a single control architecture, which has a time-domain mismatch problem when dealing with the slow time-varying thermal inertia of the electric vehicle thermal management system. It is difficult to balance the long-term global planning and short-term dynamic adjustment of energy consumption and temperature control.

[0006] Balancing the forward-looking optimization performance of long-term global planning with the real-time robustness of short-term dynamic adjustment is a core challenge in the field of predictive thermal management for electric vehicles and a technical problem that needs to be solved. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of the prior art and provide a predictive thermal management system for electric vehicles based on traffic flow information.

[0008] The objective of this invention can be achieved through the following technical solutions: According to one aspect of the present invention, an integrated predictive thermal management system for electric vehicles based on connected traffic flow information is provided, the system comprising: The traffic information acquisition module is used to acquire real-time network traffic flow information in front of the vehicle; The upper-level long-time-domain temperature planner, based on the dynamic programming algorithm, uses the traffic flow information to predict future vehicle speed and battery heat generation, and aims to minimize the total energy consumption of the system. In the long time domain, it solves the optimal temperature reference trajectory of the cabin and battery that satisfies the constraints of the cabin comfort temperature range and the battery safety temperature range. The lower-level short-time-domain temperature tracker employs a model predictive control algorithm, guided by the optimal temperature reference trajectory, to coordinate and control the actuators of the thermal management system in real time within the short-time domain, dynamically allocating the cooling capacity of the cabin and battery to achieve accurate temperature tracking and energy consumption optimization.

[0009] As a preferred technical solution, the objective function of the upper-level long-term temperature planner includes terminal cost and stage cost. The stage cost is calculated at each time step and is composed of energy cost, cabin comfort cost and battery temperature range cost weighted together. The terminal cost is calculated in the last time step and is used to penalize the deviation of the terminal's final state from the target.

[0010] As a preferred technical solution, the energy cost is used to penalize the total energy consumption of the thermal management system within the current time step; the cabin comfort cost is used to penalize the degree to which the cabin temperature deviates from the preset optimal temperature reference trajectory at the next moment; and the battery temperature range cost is used to penalize the degree to which the battery temperature exceeds the preset battery safe operating range at the next moment.

[0011] As a preferred technical solution, the state variables in the dynamic programming algorithm include the cabin interior temperature and the average battery temperature, and the control output variables include the compressor power and the allocation coefficient for distributing the cooling capacity of the cabin and the battery.

[0012] As a preferred technical solution, the objective function of the lower-level short-time-domain temperature tracker includes state tracking cost, control magnitude cost, control increment cost, and slack variable cost, wherein the state tracking cost is used to penalize the degree to which the system state deviates from the optimal temperature reference trajectory.

[0013] As a preferred technical solution, the model predictive control algorithm is a nonlinear model predictive control. Its optimization problem is to find the optimal state trajectory and control sequence in the prediction time domain under the premise of satisfying system constraints, so as to minimize the objective function.

[0014] As a preferred technical solution, the system constraints include system dynamic constraints, actuator operation constraints, and temperature boundary constraints described by the cockpit-battery coupled thermal model.

[0015] As a preferred technical solution, the dynamic distribution of cooling capacity is achieved by adjusting the opening of the electronic expansion valve to control the distribution ratio β of refrigerant between the cabin evaporator and the battery cooler, where β represents the ratio of the cooling capacity allocated to the battery to the total cooling capacity.

[0016] As a preferred technical solution, the system predicts future high heat load conditions through the upper-level long-term temperature planner and generates control commands to enable the system to start the active pre-cooling operation of the battery in advance.

[0017] As a preferred technical solution, the connected traffic flow information includes average traffic flow speed, road speed limit information, and real-time congestion status, which is used to predict future driving conditions and their corresponding heat loads.

[0018] Compared with the prior art, the present invention has the following beneficial effects: 1) This invention employs a hierarchical control architecture for predictive thermal management of electric vehicles. The upper layer, based on dynamic programming (DP) algorithm, utilizes network traffic flow information, considering both global energy consumption and dual objectives (cabin comfort and battery safety), to generate long-term optimal cabin and battery temperature trajectories. The lower layer, based on model predictive control (MPC) algorithm, tracks the reference temperature in the short-term domain and coordinates the allocation of cooling capacity between the cabin and battery. Through this hierarchical design, the system can proactively adjust the thermal load, achieving a leap from "passive response" to "proactive prediction" in electric vehicle thermal management, effectively reducing the energy consumption of the thermal management system while ensuring cabin comfort and battery safety. Simulation results show that this method reduces the energy consumption of the thermal management system by 12.69% and 3.54% compared to traditional PID and single-layer MPC strategies, respectively.

[0019] 2) The upper-level long-term temperature planner combines the terminal cost (penalizing final state deviation) with the stage cost (weighted average of energy cost, cabin comfort cost, and battery temperature range cost). This allows the upper-level planner to ensure that the terminal state approaches the target from a global perspective during long-term optimization, while also balancing the trade-offs between "minimizing energy consumption," "cabin temperature comfort," and "battery temperature safety" at local time steps. This refined objective function design makes the "optimal temperature reference trajectory" planned by the upper level more practical for engineering applications, providing a scientific and reasonable guiding benchmark for lower-level real-time control.

[0020] 3) The objective function of the lower-level short-time-domain temperature tracker covers state tracking cost, energy cost, control increment cost, and slack variable cost. This design allows the lower-level controller to accurately track the upper-level long-time-domain planning trajectory in short-time-domain real-time optimization (ensuring the continuity of the global optimization objective), reduce temperature tracking energy consumption and actuator losses through energy cost and increment penalties, and handle constraint conflicts through slack variables (enhancing system robustness). This achieves multiple optimization objectives of accurate tracking, low energy consumption, and strong robustness, ensuring the real-time control performance of the thermal management system.

[0021] 4) Since the cabin-battery coupling model and the refrigeration system model of the electric vehicle thermal management system have nonlinear characteristics, NMPC is more adaptable to nonlinear characteristics than linear MPC. It can maintain high control accuracy under complex thermal management scenarios (such as high-load cooling and drastic temperature fluctuations) and ensure that the system state stably approaches the reference trajectory.

[0022] 5) This invention uses an upper-level long-time domain planner to generate control commands based on future high-heat-load conditions (such as high-speed driving or pre-heating in traffic congestion) to initiate proactive battery pre-cooling in advance. This predictive pre-cooling strategy can lower the battery temperature to a lower, safer range before the arrival of high heat loads, avoiding a sharp rise in battery temperature under high heat loads. On the one hand, it reduces cooling energy consumption during high heat load phases (due to the lower starting temperature, the cooling capacity required to reach the target temperature is reduced), and on the other hand, it extends battery life (avoiding accelerated battery aging due to high temperatures), achieving the dual benefits of "energy saving + battery protection". Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the structure and process of the integrated predictive thermal management system for electric vehicles in this invention; Figure 2 This is a schematic diagram of the refrigeration circuit architecture of the thermal management system in an embodiment of the present invention; Figure 3 This is a schematic diagram of the summer test conditions in an embodiment of the present invention; Figure 4 This is a schematic diagram comparing the simulation results of different control strategies under operating condition 1 in this embodiment of the invention; Figure 5 This is a schematic diagram comparing the simulation results of different control strategies under operating condition 2 in this embodiment of the invention; Figure 6 This is a schematic diagram showing the energy consumption comparison results of different components under different control strategies in working condition 1 of the present invention. Figure 7 This is a schematic diagram showing the energy consumption comparison results of different components under different control strategies in working condition 2 of this embodiment of the invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0025] Example 1 This embodiment relates to an integrated predictive thermal management system for electric vehicles based on connected traffic flow information. The system utilizes traffic flow information provided by the Internet of Vehicles and combines it with an established cabin-battery coupled thermal model to construct a hierarchical control architecture, thereby predictively adjusting the cabin and battery temperatures and optimizing the energy consumption of the vehicle's thermal management system.

[0026] The system adopts a hierarchical control architecture, such as Figure 1 As shown, the upper layer uses a dynamic programming (DP) algorithm to generate the optimal temperature reference trajectory for the cabin and battery over a long time domain using traffic forecast information; the lower layer uses a model predictive control (MPC) algorithm to track the reference temperature and coordinate the allocation of cooling capacity between the cabin and battery in a short time domain. Through this layered design, the system can proactively adjust the thermal load, significantly reducing the energy consumption of the thermal management system while ensuring occupant comfort and battery safety. Simulation results show that this method reduces the energy consumption of the thermal management system by 12.69% and 3.54% respectively compared to traditional PID and single-layer MPC strategies.

[0027] The system includes: Traffic Information Acquisition Module: Real-time traffic flow information ahead is obtained through the connected traffic map, including the average speed and speed limit of the traffic flow ahead of vehicles, congestion status, and other real-time traffic dynamics.

[0028] Upper-level long-term temperature planner: Based on dynamic programming (DP) algorithm, it uses the acquired traffic flow information ahead to predict future vehicle speed and battery heat generation, and solves the optimal temperature reference trajectory for the cabin and battery in the long-term domain. The upper-level long-term temperature planner aims to minimize the total system energy consumption, while satisfying the constraints of the cabin comfort temperature range and the battery safety temperature range.

[0029] Lower-level short-time-domain temperature tracker: Employing a nonlinear model predictive control (NMPC) algorithm, guided by the optimal temperature reference trajectory provided by the upper level, it coordinates and controls actuators such as compressor power, expansion valve opening, coolant pump, and cabin blower power in real time within the short-time domain, dynamically allocating the cooling capacity of the cabin and battery to achieve accurate temperature tracking and energy consumption optimization.

[0030] This system effectively solves the time-domain mismatch problem of traditional single controllers when dealing with slow time-varying thermal inertial systems through a hierarchical structure, and at the same time realizes the forward-looking management of heat load by utilizing traffic forecast information.

[0031] System Modeling: To achieve coordinated thermal management of the cabin and battery, this invention employs an advanced integrated predictive thermal management system architecture, such as... Figure 2 As shown, since this invention focuses on the control of the thermal management system under high summer temperatures, only the refrigeration circuit is shown. The core of this architecture is a heat pump system that includes a refrigerant circuit (blue) and a coolant circuit (green). Correspondingly, the cabin is an air-cooled circuit, while the battery is a liquid-cooled circuit.

[0032] The key operating condition of this invention focuses on the vehicle's cooling needs during hot summer weather: the refrigerant circuit is activated to cool the battery coolant via a heat exchanger, while simultaneously cooling the cabin air via an evaporator. This highly integrated architecture provides the physical basis for achieving global energy optimization. The task of the lower-level short-time domain controller is to intelligently select and regulate the optimal operating mode and component operating status based on current and future needs.

[0033] Constructing a cockpit-battery coupled thermal model includes: 1) Vehicle longitudinal dynamics model The total battery current is calculated using the vehicle's traction force and mechanical power. This invention assumes the road is flat and the total traction force... Calculated using the following formula: , in , and These represent rolling resistance, aerodynamic drag, and inertial force, respectively; variables m, v α and α represent the vehicle's mass, velocity, and acceleration, respectively; parameters g, , , and These correspond to gravitational acceleration, rolling resistance coefficient, air density, drag coefficient, and frontal area, respectively.

[0034] 2) Cooling capacity calculation model COP is commonly used to describe the energy efficiency of air conditioning (AC) systems, and it is defined as follows: ,in( This represents the total cooling capacity of AC. (Work done by AC).

[0035] .

[0036] As a key component of the AC system, the condenser's heat dissipation efficiency is closely related to and positively correlated with air convection velocity; that is, the faster the air convection velocity, the higher the heat dissipation efficiency. Furthermore, ambient temperature is also a key factor affecting COP (Coefficient of Performance). Therefore, this invention characterizes it as vehicle speed (…). ) and ambient temperature ( The multivariable polynomial function is shown below: , The main working components of an AC system include the compressor, blower, and water pump. Using the COP formula, the total cooling capacity can be calculated as follows: , , in, This is the total cooling capacity. Indicates the compressor power consumption. Indicates the power consumption of the blower. This indicates the power consumption of the coolant pump.

[0037] AC primarily cools two parts: the cabin and the battery. Figure 2 As shown in the diagram, the electronic expansion valve EXV is responsible for switching the refrigerant circuit of the battery heat exchanger and also for distributing heat between the two subsystems (battery and cabin). Here, the distribution coefficient is defined. It represents the ratio of the cooling capacity allocated to the battery to the total cooling capacity, as shown below.

[0038] , , in, The opening degree of the electronic expansion valve control, Indicates the battery's cooling capacity. Refers to cabin cooling capacity.

[0039] 3) Battery thermal model This invention treats the battery pack as an equivalent lumped heat capacity, described by a first-order ordinary differential equation: , in, Indicates battery pack quality, superscript To represent the first-order differential, It's the battery temperature. It is the battery's thermal capacity. The heat is removed by the cooling system. It is through natural convection and radiation heat exchange with the environment. The battery's heat output power can be calculated using the following formula: , , , , , , in, This represents the total traction force, i.e., the driving force of the wheels; Indicates mechanical power. Indicates motor power. Indicates traction efficiency. Indicates regeneration efficiency; This indicates the total power consumption of the refrigeration equipment. Indicates the compressor power consumption. Indicates the power consumption of the blower. Indicates the coolant pump's power; Indicates the total battery current. Indicates the nominal voltage; Indicates the internal heat generation rate. Indicates the internal resistance of a single battery cell. The value represents the battery current, and ns and np represent the number of batteries connected in series and in parallel, respectively.

[0040] 4) Cockpit thermal model Similarly, this invention also models the cockpit as a lumped heat capacity: , in, and These are the cabin temperature and the cabin equivalent heat capacity, respectively. It refers to the cooling / heating capacity of the air conditioning system. It is the solar radiation heat load. It is convective heat exchange with the external environment. It is the heat generated by passengers, superscript It represents the first-order differential.

[0041] A hierarchical integrated predictive thermal management strategy: This invention proposes a hierarchical framework integrating a long-term temperature planner and a short-term temperature tracker. First, the system integrates multi-source predictive data, including traffic flow velocity, V2V information, and environmental information, to estimate future cabin and battery thermal loads. The upper layer (DP layer) generates a long-term optimal temperature reference trajectory, guiding the short-term temperature tracker towards global energy efficiency and system coordination. The lower layer (MPC layer) regulates actuators (compressors, pumps, fans, valves, etc.) to minimize energy consumption while meeting safety and comfort constraints.

[0042] Long-term dynamic programming layer (DP): The role of the upper layer (DP layer) is to plan the optimal temperature change curve for the cabin and battery within a long-term prediction period, and to send the optimal temperature reference trajectory to the MPC layer at a fixed frequency. The following introduces the design of the dynamic programming algorithm, the core of which is to find an optimal temperature reference trajectory that takes into account energy consumption, cabin comfort and battery safety under the premise of satisfying system constraints.

[0043] At any discrete time step k The system's state variables From the cabin interior temperature and average battery temperature Composition, defined as: Control output Including compressor power And the refrigerant distribution ratio between the cabin evaporator and the battery cooler. To accurately track the temperature reference trajectory provided by the upper layer, the control output is defined as follows: ,in U It is the feasible region of the control quantity.

[0044] The objective function of dynamic programming is given the planning time domain. N t Within, minimize the cumulative total cost. J The total cost consists of the terminal price. and accumulated stage costs L k composition.

[0045] , Terminal cost in the last time step (i.e. ) is calculated at the time of calculation to penalize deviations between the final state and the target: ,in L comfort,terminal For cabin comfort, end cost, L battery,terminal The specific calculation for the battery terminal cost is as follows: , , in, T i,N This represents the final cabin temperature (indicating the actual temperature). T i,ref,N For the target cabin temperature, T bat,N This represents the final battery temperature (indicating the actual state). T bat,ref,N For the target battery temperature, αcomfort and α battery These represent the comfort weighting coefficient and the battery temperature weighting coefficient, respectively.

[0046] Phase Cost exist Each step of the calculation consists of three parts: energy cost, cabin comfort cost, and battery temperature range cost. , Energy cost Used to penalize the total energy consumption of the thermal management system within the current time step k. , in, P cp,k For compressor power, P blower For fan power, P pump For water pump power, ω e For energy cost weighting coefficient, △ t For time step.

[0047] Cabin comfort cost The cabin temperature is used to punish the next moment. Deviation from the preset optimal temperature reference trajectory To what extent, ω comfort Weighting factor for cabin comfort cost: , Battery temperature range cost Punish the battery temperature at the next moment by using a soft contract. Exceeding the preset battery safe operating range To what extent, , in and These represent the lower and upper temperature limits of the battery's safe operating range, respectively. ω battery This is the cost weighting coefficient for battery temperature range. When the temperature is within the battery's safe operating range, this cost is 0; when the temperature exceeds the battery's safe operating range, the cost increases with the square of the deviation.

[0048] In summary, this dynamic programming strategy aims to obtain the optimal control sequence under the conditions of system dynamics and various constraints by solving the above objective function minimization problem. This allows for the acquisition of the optimal temperature change sequence for the cabin and battery, which serves as a reference for the model predictive controller. This design achieves an optimal trade-off between energy consumption, comfort, and battery safety.

[0049] Short-term trajectory tracking layer: To track the optimal temperature reference trajectory generated by the long-term dynamic programming (DP) layer while improving the energy efficiency of the thermal management system, this invention designs a nonlinear model predictive controller (NMPC). The core idea of ​​MPC is to solve a finite-time optimal control problem online in each control cycle based on the current system state and the prediction of future disturbances, and apply the first element of the obtained control sequence to the system.

[0050] At each discrete control time k, the MPC controller needs to solve the following optimization problem: , in, For the current moment k Measured system status (cabin temperature) Battery temperature ), as the initial condition for optimization.

[0051] Objective: Under the premise of satisfying system constraints, find the optimal state trajectory X and control sequence U in the prediction time domain, so that the objective function... J MPC Minimum.

[0052] To predict the trajectory of state changes in the time domain starting from the current time k, This represents the system state predicted at the current time k, in the future at time i.

[0053] For future predictions within the time domain ( N p-1 The control sequence within the step, To predict the control input at time i in the future (in order of compressor power) Expansion valve opening Battery water pump power P bp Cabin blower power P bl NMPC obtains this optimal control sequence by solving an optimization problem, and then only executes the first step u. k And then optimize it again at the next moment.

[0054] This is a sequence of slack variables used to handle soft state constraints.

[0055] also, and These represent the optimal temperature reference trajectory and the predicted perturbation sequence, respectively. External perturbations include ambient temperature. Battery heat generation Solar radiation Passenger heat load and vehicle speed , forming the perturbation vector .

[0056] objective function Designed to balance multiple control objectives, it consists of a weighted average of four components: state tracking cost, control magnitude cost, control increment cost, and slack variable cost. , Among them, state tracking cost This term is used to penalize the degree to which the predicted state trajectory deviates from the optimal temperature reference trajectory issued by the upper-level DP: , in, y j∣k The predicted system state at time j from time k; y ref,j : The temperature reference value at time j in the future (output of DP layer); W track : State tracking weight matrix, which allows setting the tracking priority for cabin and battery temperatures (e.g., when battery safety has a higher priority). T bat (The corresponding weight is larger); N is the MPC prediction time domain.

[0057] Controlling the size and cost This penalty controls the size of the input to limit the overuse of the actuator and indirectly save energy. , in, The control input for predicting the future time j at time k ( ); W effort The control quantity weight matrix allows for setting greater weights for components with high energy consumption (such as compressors) to enhance energy-saving effects.

[0058] Controlling incremental costs This penalty is applied to the rate of change of the control quantity between consecutive control steps to ensure the smoothness of the control action and reduce actuator wear and system oscillation. , in, The increment of control quantity between continuous control steps; W ΔuIncremental weight matrix: assign greater weights to actuators that are slow to respond or sensitive to wear (such as expansion valves).

[0059] Slack Variable Costs This penalty is applied to the magnitude of the slack variables of the soft constraints, ensuring that the constraints are satisfied as much as possible: , in, These are slack variables (non-negative) used to relax the hard temperature constraint; The larger the value of the relaxation weight coefficient, the heavier the penalty for constraint deviation, and the closer the system is to "hard constraint" control.

[0060] To ensure physical feasibility and safety, the optimization problem must satisfy discretized system dynamics constraints, actuator operation constraints, and thermal boundary constraints.

[0061] System dynamic constraints: , Actuator operating constraints: , Thermal boundary constraints: , in, The system state at time j is predicted from time k. , , These represent the system state, control input, and disturbance at time k, respectively. P cp,j∣k , , , These are the compressor power, battery water pump power, blower power, and expansion valve opening, predicted at the current time k for the future time j. , These are the minimum and maximum values ​​of the compressor power, respectively. and These are the minimum and maximum power values ​​of the battery-powered water pump, respectively. and These are the minimum and maximum values ​​of the blower power, respectively; and These are the minimum and maximum values ​​of the expansion valve opening, respectively. and These represent the battery temperature and cabin temperature predicted at time j, respectively, at time k. , These are the lower and upper limits of the battery's safe operating temperature range, respectively. and These are the basic comfort lower and upper limits for cabin temperature, respectively. and These are the relaxation variables for battery temperature constraints and cabin temperature constraints, respectively, and the ranges within which offsets are allowed under extreme conditions. This refers to all slack variables and has a non-negative value.

[0062] Example 2 This embodiment relates to the verification of an integrated predictive thermal management system for electric vehicles based on connected traffic flow information. The effectiveness of the system of the present invention is verified through simulation.

[0063] Build a summer test scenario with a total duration of 6500 seconds, such as... Figure 3 As shown. This operating cycle cascades two standard China Light Vehicle Driving Cycle (CLTC) sequences, including a high-speed driving phase (vehicle speed 100 km / h, lasting 1200s) and ending with a section of urban congestion, with the ambient temperature set at 43°C and the target cabin temperature set at 27°C. To objectively evaluate the management system (Hierarchical Integrated Predictive Thermal Management) (HIPTM strategy) proposed in this invention, its performance was compared with two benchmarks: a standard PID controller (Benchmark 1) and a simplified IPTM (Integrated Predictive Thermal Management) framework (Benchmark 2). Unlike the system proposed in this paper, Benchmark 2 uses fixed temperature setpoints for the cabin and battery and does not include long-term optimized trajectories.

[0064] To verify the effectiveness of the proposed strategy, this invention designed two comparative simulation scenarios: Scenario 1 (strong cooling demand) simulates the high concurrent cooling load faced by the system by initializing the cabin and battery to 45°C and 38°C respectively; Scenario 2 (differentiated thermal demand) sets the initial temperatures of the cabin and battery to 43.5°C (cabin) and 34.5°C (battery) respectively, aiming to evaluate the system performance under the conflicting objectives of requiring rapid cabin cooling while maintaining stable battery temperature. The simulation results under different scenarios are summarized in Table 1, and a comprehensive evaluation is conducted from three dimensions: energy efficiency, passenger comfort, and battery protection. Specific indicators include TMS energy consumption for quantifying the impact on range, cabin cooling time (to the target ±1°C range) reflecting system responsiveness, and average cabin temperature and average battery temperature.

[0065] Table 1

[0066] Figure 4 and Figure 5Simulation results for scenarios 1 and 2 are presented respectively. It can be seen that during the initial cooling phase (0–3000 seconds) in both scenarios 1 and 2, the PID and IPTM strategies rapidly reduced the cabin temperature to approximately 28°C to maintain occupant comfort. However, IPTM achieved energy savings compared to the PID baseline, primarily due to its smoother battery cooling trajectory, effectively mitigating the significant overshoot observed in the PID strategy at 50 seconds and 1500 seconds in scenario 1.

[0067] In contrast, the HIP™ strategy utilizes a long prediction time of 1800 seconds to optimize the thermal trajectory. Regarding the battery, HIP™ anticipates the temperature rise caused by future high-speed driving and initiates proactive pre-cooling to prevent overheating. In the cabin, this strategy reduces compressor load by moderately slowing the cooling rate during the rapid cooling phase. This strategic reduction is a key difference between HIP™ and IP™ in terms of energy efficiency. Subsequently, the cooling rate accelerates, reaching the target 28°C at 3000 seconds, thus ensuring that passenger comfort is not compromised.

[0068] During the temperature maintenance phase (3000 s – 6500 s), IPTM and HIPTM maintain the cabin temperature at approximately 0.5°C above the target, relative to the PID baseline in Condition 1. This slight relaxation reduces energy consumption without sacrificing occupant comfort. Simultaneously, these strategies prioritize cooling the battery, ensuring strict adherence to thermal safety thresholds.

[0069] Furthermore, the HIPTM strategy of this invention utilizes its predictive capabilities to anticipate significant battery heat generation during the high-power range (3500s–5000s) in operating conditions 1 and 2. Therefore, it initiates active pre-cooling at 3100s, achieving a deeper cooling effect than IPTM. This predictive operation alleviates compressor load during peak demand periods (4000s–5000s), thereby unlocking additional energy-saving potential.

[0070] Figure 6 and Figure 7The component-level energy consumption under operating conditions 1 and 2 is presented separately, illustrating a strategic trade-off shared by IPTM and HIPTM. Specifically, these strategies result in minimal additional energy consumption in low-power auxiliary equipment (i.e., coolant pumps and blowers) to alleviate the load on high-power compressors. Comparing compressor energy consumption, for example, in operating condition 1, HIPTM (11570 kJ) < IPTM (12384 kJ) < PID (14521 kJ), with HIPTM achieving 20.32% energy savings compared to IPTM and 14.72% compared to PID. This is because HIPTM utilizes its predictive capabilities to initiate active pre-cooling 3100 seconds before the arrival of the high-power range (3500-5000 seconds), effectively mitigating the compressor load during peak demand periods (4000-5000 seconds). Total energy consumption: HIPTM (14661 kJ) < IPTM (15255 kJ) < PID (16792 kJ).

[0071] In terms of total energy consumption of the thermal management system, HIPTM saves 12.69% more energy than IPTM and 9.15% more than PID. This clearly demonstrates that despite the minor additional energy consumption of pumps and blowers, the HIPTM strategy achieves a net reduction in total energy consumption and improves overall system efficiency by significantly reducing the load on high-power compressors. This synergistic approach achieves a net reduction in total energy consumption, thereby improving overall system efficiency.

[0072] Therefore, the layered integrated predictive thermal management strategy (HIPTM) of the present invention has the advantages of predictive precooling, strategic trade-offs, and a balance between comfort and safety, specifically: Predictive pre-cooling: By predicting high-load operating conditions, cooling is initiated in advance to lower the battery temperature to a lower level, thereby delaying or reducing the cooling requirements in the high-power area.

[0073] Strategic trade-off: generating minimal additional energy consumption in low-power auxiliary equipment (water pumps, blowers) to reduce the load on high-power compressors, achieving an energy-saving effect of "trading small for large".

[0074] Balancing comfort and safety: By slightly relaxing the cabin temperature (by about 0.5°C), energy consumption is reduced without sacrificing comfort, and cooling is prioritized for the battery to ensure it strictly adheres to thermal safety thresholds.

[0075] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A predictive thermal management system for electric vehicles based on traffic flow information, characterized in that, The system includes: The traffic information acquisition module is used to acquire real-time network traffic flow information in front of the vehicle; The upper-level long-time-domain temperature planner, based on the dynamic programming algorithm, uses the traffic flow information to predict future vehicle speed and battery heat generation, and aims to minimize the total energy consumption of the system. In the long time domain, it solves the optimal temperature reference trajectory of the cabin and battery that satisfies the constraints of the cabin comfort temperature range and the battery safety temperature range. The lower-level short-time-domain temperature tracker employs a model predictive control algorithm, guided by the optimal temperature reference trajectory, to coordinate and control the actuators of the thermal management system in real time within the short-time domain, dynamically allocating the cooling capacity of the cabin and battery to achieve accurate temperature tracking and energy consumption optimization.

2. The predictive thermal management system for electric vehicles based on traffic flow information according to claim 1, characterized in that, The objective function of the upper-level long-time-domain temperature planner includes terminal cost and stage cost. The stage cost is calculated at each time step and is composed of energy cost, cabin comfort cost and battery temperature range cost weighted together. The terminal cost is calculated in the last time step and is used to penalize the deviation of the terminal's final state from the target.

3. The predictive thermal management system for electric vehicles based on traffic flow information according to claim 2, characterized in that, The energy cost penalizes the total energy consumption of the thermal management system within the current time step; the cabin comfort cost penalizes the degree to which the cabin temperature deviates from the preset optimal temperature reference trajectory at the next moment; and the battery temperature range cost penalizes the degree to which the battery temperature exceeds the preset safe operating range at the next moment.

4. The predictive thermal management system for electric vehicles based on traffic flow information according to claim 1, characterized in that, The state variables in the dynamic programming algorithm include the cabin interior temperature and the average battery temperature, while the control output variables include the compressor power and the allocation coefficient used to distribute the cooling capacity of the cabin and the battery.

5. The predictive thermal management system for electric vehicles based on traffic flow information according to claim 1, characterized in that, The objective function of the lower-level short-time-domain temperature tracker includes state tracking cost, control magnitude cost, control increment cost, and slack variable cost, wherein the state tracking cost is used to penalize the degree to which the system state deviates from the optimal temperature reference trajectory.

6. A predictive thermal management system for electric vehicles based on traffic flow information according to claim 1 or 5, characterized in that, The model predictive control algorithm is a nonlinear model predictive control. Its optimization problem is to find the optimal state trajectory and control sequence in the prediction time domain under the premise of satisfying system constraints, so as to minimize the objective function.

7. A predictive thermal management system for electric vehicles based on traffic flow information according to claim 6, characterized in that, The system constraints include system dynamic constraints, actuator operation constraints, and temperature boundary constraints described by the cockpit-battery coupled thermal model.

8. A predictive thermal management system for electric vehicles based on traffic flow information according to claim 1, characterized in that, The dynamic distribution of cooling capacity is achieved by adjusting the opening of the electronic expansion valve to control the distribution ratio β of refrigerant between the cabin evaporator and the battery cooler, where β represents the ratio of the cooling capacity allocated to the battery to the total cooling capacity.

9. A predictive thermal management system for electric vehicles based on traffic flow information according to claim 1, characterized in that, The system predicts future high heat load conditions through the upper-level long-time-domain temperature planner and generates control commands to enable the system to initiate active pre-cooling of the battery in advance.

10. A predictive thermal management system for electric vehicles based on traffic flow information according to claim 1, characterized in that, The networked traffic flow information includes average traffic flow speed, road speed limit information, and real-time congestion status, which is used to predict future driving conditions and their corresponding heat loads.

Images