Vehicle Control System Reinforcement Learning for Engine Adaptation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Experts spend significant man-hours to adapt the operation amounts of internal combustion engine drive systems to match vehicle states, making it inefficient and time-consuming to set appropriate operation amounts for optimal fuel consumption, exhaust characteristics, and drivability.
Innovation Solution
A method involving the storage of relationship prescription data that prescribes the relationship between vehicle states and operation amounts, using reinforcement learning to update these data based on detection values from sensors, thereby optimizing operation amounts for improved fuel efficiency, exhaust characteristics, and drivability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If filtering methods are used to set throttle valve operation amounts, then appropriate operation amounts can be achieved, but significant man-hours are required for expert adaptation
Solution Approach 1:
The system performs self-learning through reinforcement learning, automatically adapting the relationship prescription data based on detected vehicle states and calculated rewards, eliminating the need for expert adaptation while maintaining high precision in operation amount setting
Solution Approach 2:
The system implements feedback loops where detection values from sensors are used to calculate rewards, which then update the relationship prescription data through reinforcement learning, enabling continuous automatic optimization without manual intervention
2Productivity
If expert adaptation is performed to optimize drive system operation, then fuel consumption and exhaust characteristics are improved, but the process is time-consuming and inefficient
Solution Approach 1:
The patent replaces manual expert adaptation (mechanical/process-based) with automated reinforcement learning algorithms, substituting human expertise with computational intelligence that rapidly optimizes operation amounts without time-consuming manual processes
Solution Approach 2:
The system dynamically changes operation parameters through automated learning, adjusting the relationship between vehicle states and operation amounts based on detected performance metrics, enabling rapid optimization without fixed expert-defined parameters
Data Source
AI summary
A method of generating vehicle control data includes: storing, with a storage device, relationship prescription data; operating, with an execution device, an operable portion of an internal combustion engine; acquiring, with the execution device, a detection value from a sensor that detects the state of the vehicle; calculating, with the execution device, a reward; and updating, with the execution device, the relationship prescription data using update mapping determined in advance, the update mapping using the state of the vehicle based on the detection value, an operation amount used to operate the operable portion, and the reward corresponding to the operation as arguments, and returning the relationship prescription data which have been updated such that an expected profit for the reward calculated when the operable portion is operated in accordance with the relationship prescription data increases.