A method and system for thermal management of a traction battery for a vehicle

Abstract

This invention discloses a thermal management method for automotive power batteries, comprising the following steps: S1, constructing a system model including a battery operating condition parameter dataset and a real-time parameter dataset for the charging / discharging scenario of the automotive power battery; S2, obtaining optimization objectives based on the real-time collected battery operating condition parameters, environmental parameters, temperature control model, and charging / discharging control model, the optimization objectives including an initial temperature control strategy and an initial charging / discharging strategy; S3, learning the optimal temperature control strategy and charging / discharging strategy based on the initial temperature control strategy and the initial charging / discharging strategy using a dual-delay deep deterministic strategy gradient algorithm based on preference prediction and synchronous control; S4, determining the optimal temperature control strategy and charging / discharging strategy for real-time control of the charging / discharging of the automotive power battery. This invention can synchronously control the temperature control and charging / discharging of automotive power batteries, improving charging and discharging efficiency while ensuring battery life.

Description

Technical Field

[0001] This invention relates to the field of new energy technology, and in particular to a thermal management method for automotive power batteries. Background Technology

[0002] As the primary energy storage component of electric vehicles, the performance of the power battery directly impacts the overall performance of the vehicle. However, batteries generate a significant amount of heat during charging and discharging. If this heat cannot be dissipated effectively and promptly, it can lead to excessively high or unevenly distributed battery temperatures, affecting charge-discharge cycle efficiency, power output, and energy utilization. In severe cases, it can even cause thermal runaway, threatening system safety and reliability. Therefore, thermal management of power batteries is of paramount importance. Current thermal management strategies automatically trigger the start and stop of active heating and cooling of the battery by setting fixed temperature thresholds. This process consumes a large amount of electrical energy, impacting the vehicle's range.

[0003] In existing technologies, thermal management is based on the user's operating conditions, providing cooling or heating based on the demand from the power battery. However, different customers have different operating conditions. Currently, intelligent computing strategies can make adaptive calculations for each vehicle under different operating conditions, minimizing energy consumption for vehicles with mild operating conditions and ensuring battery life for vehicles with severe operating conditions. However, they cannot know the impact of the current operating conditions on the power battery's lifespan, which can easily lead to a reduction in the power battery's lifespan. Furthermore, the inability to allocate demand according to the actual degradation of the vehicle results in increased energy consumption for vehicle thermal management. Summary of the Invention

[0004] To overcome the above-mentioned defects, the present invention aims to provide a thermal management method and system for automotive power batteries, which can synchronously control the temperature and charging / discharging of automotive power batteries, thereby improving charging / discharging efficiency while ensuring battery life.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A thermal management method for a vehicle power battery includes the following steps:

[0007] S1. For the charging / discharging scenario of automotive power batteries, construct a system model that includes a battery operating condition parameter dataset and a real-time parameter dataset, and establish a battery life prediction model, a temperature control model, a charging / discharging control model, and a calculation model.

[0008] S2. Based on the real-time collected battery operating parameters and environmental parameters, as well as the temperature control model and charge / discharge control model, the optimization objective is obtained, which includes the initial temperature control strategy and the initial charge / discharge strategy.

[0009] S3. Based on the initial temperature control strategy and the initial charge / discharge strategy, a dual-delay deep deterministic strategy gradient algorithm based on preference prediction and synchronous control is used to learn the optimal temperature control strategy and charge / discharge strategy.

[0010] S4. Determine the optimal temperature control strategy and charging / discharging strategy for real-time control of vehicle power battery charging / discharging.

[0011] Preferably, step S1 includes the following sub-steps:

[0012] S1.1 Incorporate the temperature control equipment, charging equipment, and discharge load of the vehicle power battery into the modeling to construct a system model;

[0013] S1.2. Construct computing models based on different computing service carriers;

[0014] S1.3. Establish a battery life prediction model based on the battery operating condition parameter dataset and real-time parameter dataset, and predict the battery life based on the battery charging / discharging current and cycle number.

[0015] S1.4. Based on the battery's charging parameters and / or discharging load parameters, construct a charging / discharging control model;

[0016] S1.5. Construct a temperature control model based on the parameters of the temperature control equipment and the battery temperature constraints.

[0017] Preferably, step S2 includes: based on the multiple models established in step 1, defining the optimization objective, namely, using the maximum charging current under the constraints of ensuring battery life and battery temperature; or, using the maximum discharging current under the condition that the vehicle's operation meets the regulations.

[0018] Preferably, the part of learning the optimal temperature control strategy and charging / discharging strategy in step S3 is modeled as a multi-agent partially observable Markov decision process, in which the temperature control device, charging device and discharging load are all regarded as agents with independent decision-making capabilities; the decision-making process is solved by a dual-delay deep deterministic policy gradient algorithm based on preference prediction and synchronous control, global state information is introduced in the centralized training phase, and each agent makes decisions based only on local observations in the distributed execution phase.

[0019] Furthermore, the dual-delay deep deterministic strategy gradient algorithm based on preference prediction and synchronous control introduces an attention mechanism to characterize the collaborative relationship between different batteries and edge computing centers. Based on the obtained temperature control strategy and charging / discharging strategy, a synchronous control mechanism is introduced according to the state of the temperature control device, charging device, and discharge load to maximize charging / discharging efficiency and improve battery life.

[0020] Preferably, in step S3, a pre-processed preference network is introduced as a pre-network for deep reinforcement learning to select a suitable vehicle power battery for synchronous control.

[0021] Preferably, in step S4, the agent makes a decision based on its current observation state output, temperature control strategy, and charging / discharging strategy. The decision includes starting and stopping the temperature control device and adjusting the charging / discharging current.

[0022] Preferably, the battery operating parameters include the battery's nominal parameters, and the real-time parameters include charging and discharging current, remaining capacity, surface temperature, and ambient temperature and humidity.

[0023] The present invention also discloses a thermal management system for a vehicle power battery, which adopts the above-mentioned thermal management method for a vehicle power battery.

[0024] Preferably, the thermal management system includes:

[0025] The sensing device includes a current transformer, a temperature and humidity sensor, and several temperature sensors. The current transformer is used to monitor the charging and discharging current of the vehicle power battery in real time. The temperature and humidity sensor is used to monitor the temperature and humidity of the space where the vehicle power battery is located. The several temperature sensors are used to monitor the surface temperature of each battery pack respectively.

[0026] An edge computing center, which is used to execute steps S1-S4;

[0027] The storage unit is used to store battery operating condition parameter datasets, real-time parameter datasets, and data related to the edge computing center;

[0028] The sensing device is electrically connected to the edge computing center.

[0029] The beneficial effects of this invention are as follows:

[0030] 1. This invention can synchronously control the temperature and charging / discharging of automotive power batteries, thereby improving charging / discharging efficiency while ensuring battery life.

[0031] 2. This invention employs a dual-delay deep deterministic strategy gradient algorithm based on preference prediction and synchronous unloading to learn and determine the temperature control strategy and the charging / discharging strategy, resulting in precise control and a high degree of intelligence.

[0032] 3. This invention fully considers battery operating parameters and real-time parameters, while also taking environmental parameters into account, making the temperature control strategy and charge / discharge strategy more accurate. Attached Figure Description

[0033] Figure 1This is a flowchart of the thermal management method for automotive power batteries disclosed in this invention.

[0034] Figure 2 This is a schematic diagram of the thermal management system for automotive power batteries disclosed in this invention. Detailed Implementation

[0035] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings.

[0036] Example 1

[0037] See Figure 1 This embodiment discloses a thermal management method for a vehicle power battery, specifically including the following steps:

[0038] S1. For the charging / discharging scenario of automotive power batteries, construct a system model that includes a battery operating condition parameter dataset and a real-time parameter dataset, and establish a battery life prediction model, a temperature control model, a charging / discharging control model, and a calculation model.

[0039] S2. Based on the real-time collected battery operating parameters and environmental parameters, as well as the temperature control model and charge / discharge control model, the optimization objective is obtained, which includes the initial temperature control strategy and the initial charge / discharge strategy.

[0040] S3. Based on the initial temperature control strategy and the initial charge / discharge strategy, a dual-delay deep deterministic strategy gradient algorithm based on preference prediction and synchronous control is used to learn the optimal temperature control strategy and charge / discharge strategy.

[0041] S4. Determine the optimal temperature control strategy and charging / discharging strategy for real-time control of vehicle power battery charging / discharging.

[0042] The following details the specific implementation steps:

[0043] Steps S1 and S2 involve model building and optimization target determination. Specifically, during vehicle charging or driving, the temperature of the power battery needs to be maintained within a certain range. The optimal operating temperature for commonly used lithium batteries is between 15°C and 30°C, with a single-cell temperature difference of <5°C. When the onboard controller's computing power cannot complete the task within the specified time or a better temperature control strategy is available, the task will be downloaded to the edge computing center. By establishing a power battery life prediction model, a temperature control model, a charge / discharge control model, and a computational model, the temperature control strategy and the charge / discharge strategy are optimized to extend the power battery's lifespan while ensuring charging / discharging efficiency.

[0044] Step S3 employs a dual-delay deep deterministic policy gradient algorithm based on preference prediction and synchronous unloading to learn and determine the temperature control strategy and charge / discharge strategy. Based on this, the preference prediction network of agents such as the temperature control device, charging / discharging device, and discharging load outputs the prior probability distribution of temperature control and charge / discharge based on the operating conditions, used to characterize the relative temperature control strategy and the priority of the charge / discharge strategy of the power battery under the current operating conditions. Simultaneously, the actor network generates temperature control actions and charge / discharge actions based on the current operating conditions. Subsequently, based on the current temperature control actions and charge / discharge actions, the computational load under the current environment is predicted, and it is determined whether the computation should be executed in the edge computing center. After executing the action, a system-level reward value and the system state at the next moment are returned. After obtaining the interaction results, the interaction results are placed in a buffer pool. A batch of historical experiences is further randomly sampled from the experience pool for network updates. For each sampled data, the target action for the next moment is first generated through the target actor network, and noise is pruned to the target action to achieve policy smoothing. Subsequently, the target Q-value is calculated using the target critic network, and the smaller value between the two critic outputs is taken as the temporal difference objective. The parameters of the two critic networks are then updated by minimizing the mean squared error between the current critic output and the target Q-value. After the critic network is updated, the actor network is not updated immediately. Instead, according to a delayed update mechanism, the actor network is updated only based on the deterministic policy gradient when a preset update interval condition is met. The above-described time-slot-based interaction, storage, and update process continues within one round until the system time ends or the task termination condition is met. After training, the optimal temperature control policy and charging / discharging policy are obtained to maximize the overall system benefit. This multi-agent reinforcement learning framework includes... Each agent continuously updates its strategy to approximate the optimal decision through constant interaction with the dynamic edge computing environment. For any agent... It maintains the following three core network structures internally:

[0045] 1) Action Network It is responsible for generating decision-making actions.

[0046] 2) Dual Critics Network and Together, we assess the value of the action.

[0047] 3) Preference prediction network Based on the current state, the prior preference information for task unloading is output to guide the exploration direction in the high-dimensional hybrid action space. The specific update process is as follows:

[0048] The sample batch size in the experience pool is denoted as B. The loss function of the agent's dual-critic network is shown in equation (1):

[0049]

[0050] in This represents the action of each agent. The action network is updated via deterministic policy gradients as shown in equation (2):

[0051]

[0052] The target network soft update is shown in equation (3):

[0053]

[0054]

[0055] in To update the coefficients, Represents an actor network. Indicates the critic network. This indicates a preference for certain networks.

[0056] Example 2

[0057] Based on Example 1, this example discloses a thermal management system for a vehicle power battery. This thermal management system employs the thermal management method for a vehicle power battery disclosed in Example 1, specifically as follows: Figure 2 As shown:

[0058] The sensing device in this embodiment includes a current transformer, a temperature and humidity sensor, and several temperature sensors. The current transformer is used to monitor the charging and discharging current of the vehicle power battery in real time. The temperature and humidity sensor is used to monitor the temperature and humidity of the space where the vehicle power battery is located. The several temperature sensors are used to monitor the surface temperature of each battery pack respectively.

[0059] Edge computing center, used to execute steps S1-S4;

[0060] The storage unit is used to store battery operating condition parameter datasets, real-time parameter datasets, and data related to the edge computing center;

[0061] The sensing device is electrically connected to the edge computing center.

[0062] Of course, the present invention may have other various embodiments. Without departing from the spirit and essence of the present invention, those skilled in the art can make various corresponding changes and modifications according to the present invention, but these corresponding changes and modifications should all fall within the protection scope of the appended claims.

Claims

1. A method for thermal management of a traction battery for a vehicle, characterized in that Includes the following steps: S1. For the charging / discharging scenario of automotive power batteries, construct a system model that includes a battery operating condition parameter dataset and a real-time parameter dataset, and establish a battery life prediction model, a temperature control model, a charging / discharging control model, and a calculation model. S2. Based on the real-time collected battery operating parameters and environmental parameters, as well as the temperature control model and charge / discharge control model, the optimization objective is obtained, which includes the initial temperature control strategy and the initial charge / discharge strategy. S3. Based on the initial temperature control strategy and the initial charge / discharge strategy, a dual-delay deep deterministic strategy gradient algorithm based on preference prediction and synchronous control is used to learn the optimal temperature control strategy and charge / discharge strategy. S4. Determine the optimal temperature control strategy and charging / discharging strategy for real-time control of vehicle power battery charging / discharging.

2. The method of thermal management of a traction battery of a vehicle according to claim 1, characterized in that, Step S1 includes the following sub-steps: S1.1 Incorporate the temperature control equipment, charging equipment, and discharge load of the vehicle power battery into the modeling to construct a system model; S1.

2. Construct computing models based on different computing service carriers; S1.

3. Establish a battery life prediction model based on the battery operating condition parameter dataset and real-time parameter dataset, and predict the battery life based on the battery charging / discharging current and cycle number. S1.

4. Based on the battery's charging parameters and / or discharging load parameters, construct a charging / discharging control model; S1.

5. Construct a temperature control model based on the parameters of the temperature control equipment and the battery temperature constraints.

3. The method of thermal management of a traction battery for a vehicle according to claim 2, characterized in that, Step S2 includes: based on the multiple models established in step 1, clarifying the optimization objective, namely, using the maximum charging current under the constraints of ensuring battery life and battery temperature; or, using the maximum discharging current under the condition that the vehicle's operation meets the regulations.

4. The method of thermal management of a traction battery of a vehicle according to claim 3, characterized in that, In step S3, the part that learns the optimal temperature control strategy and charging / discharging strategy is modeled as a multi-agent partially observable Markov decision process, in which the temperature control device, charging device, and discharging load are all regarded as agents with independent decision-making capabilities. The decision-making process is solved using a dual-delay deep deterministic policy gradient algorithm based on preference prediction and synchronous control. Global state information is introduced in the centralized training phase, and each agent makes decisions based only on local observations in the distributed execution phase.

5. The method of thermal management of a traction battery of a vehicle according to claim 4, characterized in that, The dual-delay deep deterministic strategy gradient algorithm based on preference prediction and synchronous control introduces an attention mechanism to characterize the collaborative relationship between different batteries and edge computing centers. Based on the obtained temperature control strategy and charging / discharging strategy, a synchronous control mechanism is introduced according to the state of the temperature control device, charging device, and discharge load to maximize charging / discharging efficiency and improve battery life.

6. The method of thermal management of a traction battery of a vehicle according to claim 4, characterized in that, In step S3, a pre-processed preference network is introduced as a pre-network for deep reinforcement learning to select a suitable vehicle power battery for synchronous control.

7. The thermal management method for a vehicle power battery according to claim 4, characterized in that, In step S4, the agent makes a decision based on its current observation state output, temperature control strategy, and charging / discharging strategy. The decision includes starting and stopping the temperature control device and adjusting the charging / discharging current.

8. The thermal management method for a vehicle power battery according to claim 1, characterized in that, The battery operating parameters include the battery's nominal parameters, and the real-time parameters include charging and discharging current, remaining capacity, surface temperature, and ambient temperature and humidity.

9. A thermal management system for a vehicle power battery, characterized in that, The thermal management system adopts the thermal management method for vehicle power batteries as described in any one of claims 1-8.

10. The thermal management system according to claim 9, characterized in that, include: The sensing device includes a current transformer, a temperature and humidity sensor, and several temperature sensors. The current transformer is used to monitor the charging and discharging current of the vehicle power battery in real time. The temperature and humidity sensor is used to monitor the temperature and humidity of the space where the vehicle power battery is located. The several temperature sensors are used to monitor the surface temperature of each battery pack respectively. An edge computing center, which is used to execute steps S1-S4; The storage unit is used to store battery operating condition parameter datasets, real-time parameter datasets, and data related to the edge computing center; The sensing device is electrically connected to the edge computing center.

Images