Vehicle Motor Thermal Control Using Reinforcement Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current thermal control methods for vehicle motors are either too imprecise (rule-based) or economically infeasible (optimization-based), failing to optimize cooling efficiently while limiting electricity consumption, especially in complex systems.
Innovation Solution
A reinforcement learning algorithm, specifically using an actor-critic architecture and thermodynamic reward functions, is implemented to optimize thermal control by iteratively determining control actions, updating the control function based on system performance, and minimizing thermodynamic irreversibilities, thus enabling efficient thermal management without extensive experimental data or complex modeling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If rule-based thermal control is used, then implementation is simple, but control precision is insufficient and cannot achieve optimum control
Solution Approach 1:
The patent transforms the control approach from rule-based to data-driven by changing the fundamental parameter representation. Instead of using predefined rules, the system uses neural networks to learn optimal control parameters from operational data, enabling precise thermal control while maintaining implementation simplicity through automated learning.
Solution Approach 2:
The patent replaces the mechanical rule-based control system with an intelligent software-based neural network system. This substitution allows the system to automatically adapt and optimize thermal control without requiring manual rule definition, achieving both precision and ease of operation.
2Measurement precision
If optimization-based thermal control with theoretical models is used, then control precision is improved, but system complexity increases and design cost becomes too high
Solution Approach 1:
The patent uses neural networks to create a virtual model of the thermal system that learns from operational data rather than requiring complex theoretical modeling. This copied behavioral model achieves precise control predictions without the complexity of detailed theoretical models, reducing system complexity while maintaining control precision.
Solution Approach 2:
The system performs self-learning and self-optimization through neural networks that automatically adapt to the specific thermal characteristics of the motor and cooling system. This eliminates the need for extensive manual modeling and validation, reducing design cost and system complexity while achieving precise control.
3Measurement precision
If optimization-based thermal control with theoretical models is used, then control precision is improved, but validation cost increases significantly
Solution Approach 1:
The patent implements continuous feedback learning where the neural network is trained on actual operational data from the thermal system. This feedback mechanism allows the system to validate and improve control precision using real-world data rather than requiring extensive separate validation experiments, significantly reducing validation costs.
Solution Approach 2:
The system performs preliminary learning and adaptation during normal operation by continuously training the neural network on operational data. This preliminary action embeds validation into the operational process itself, eliminating the need for separate expensive validation phases while maintaining control precision.
4Reliability
If thermal control is applied continuously, then motor temperature is maintained within acceptable range, but electricity consumption increases and reduces vehicle range
Solution Approach 1:
The patent applies partial cooling action by using neural networks to determine the precise cooling capacity needed at each moment rather than continuous full-capacity cooling. This partial action maintains temperature reliability while minimizing electricity consumption by applying cooling only when and to the extent necessary.
Solution Approach 2:
The system dynamically adjusts cooling capacity based on real-time thermal conditions and motor operating parameters. The neural network continuously optimizes the cooling strategy, varying the cooling intensity to match actual thermal needs, thereby maintaining reliability while reducing overall energy consumption compared to static continuous cooling.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach allows for optimized thermal control of vehicle motors, reducing electricity consumption and extending motor longevity, while being adaptable to various systems and cost-effective, without the need for complex modeling or experimental validation.
Implementation Method 1
a cooling system, and at least one actuator suitable for varying a capacity for cooling the motor by means of the cooling device
Implementation Method 2
In the case of electric motors, for example, they generate a lot of heat by the Joule effect
Data Source
AI summary
The disclosed computer-implemented method optimizes thermal control of a vehicle motor, the vehicle including a cooling device including an actuator varying cooling capacity, the method including training a reinforcement learning algorithm including the iterative steps: 1) determining an action to control an actuator by applying a control function to a current state of the thermal system, and implementing the action; 2) determining a modified state of the thermal system after implementing the action; 3) calculating, by implementing a thermodynamic reward function of the motor, a reward value based on the modified state of the thermal system, and the action; 4) updating a function for estimating thermal performance based on the current state of the thermal system, the modified state of the thermal system, the action and the reward; and 5) modifying the control function based on the update of the function for estimating thermal performance.