On-Board Power Supply Predictive Control Using Third-Vehicle Probabilities
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing on-board network control systems in hybrid vehicles struggle to efficiently manage charge flows due to the complexity of operating parameters, requiring excessive data exchange and uncertainty about available models for vehicle control, especially with the integration of navigation and vehicle-to-vehicle communication.
Innovation Solution
A control device and method that utilizes a central database to predict charge flows by determining probabilities of third-party vehicle operating actions based on georeferenced data, allowing for optimized energy storage management through adaptive and reflex-augmented reinforcement learning algorithms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If operating actions are determined using only vehicle operating parameters, then the control system is simple, but the prediction accuracy of charge flows is insufficient
Solution Approach 1:
The system performs preliminary actions by determining probabilities of third-party vehicle operating actions in advance using reinforcement learning algorithms. These probabilistic predictions are stored and later used to optimize charge flow management, allowing the system to prepare energy storage conditioning before actual events occur, thereby improving prediction accuracy without requiring complex real-time analysis of all operating parameters
Solution Approach 2:
The patent introduces probability values as an intermediary element between raw operating parameters and control decisions. Instead of directly processing complex multi-vehicle operating data, the system converts third-party vehicle operating actions into probabilistic representations that serve as simplified inputs for charge flow prediction, reducing system complexity while maintaining predictive accuracy
2Measurement precision
If navigation systems and terrain parameters are integrated, then the prediction of charge flows is improved, but the data exchange volume increases
Solution Approach 1:
The system extracts only the essential probabilistic information needed for charge flow prediction from complex navigation and terrain data. By using reinforcement learning to determine probabilities of specific operating actions (such as regeneration likelihood) rather than processing all raw navigation parameters, the system maintains accurate charge flow predictions while significantly reducing data exchange volume
Solution Approach 2:
The patent transforms detailed navigation and terrain parameters into simplified probability values that represent the likelihood of specific operating actions. This parameter transformation allows the system to retain the predictive value of navigation integration while reducing the complexity and volume of data that must be exchanged between systems
3Measurement precision
If operating parameters of third-party vehicles are incorporated, then the prediction accuracy increases, but the uncertainty about suitable models increases
Solution Approach 1:
The system performs self-service by using its own reinforcement learning algorithms to determine probabilities of third-party vehicle operating actions based on available data. Rather than relying on external systems to provide pre-processed model data, the vehicle independently processes operating parameters and generates probabilistic predictions, ensuring reliable operation even when external data sources are unavailable or uncertain
Solution Approach 2:
The patent implements feedback mechanisms where the system continuously refines its probability determinations based on actual operating outcomes. By comparing predicted charge flows with actual results and adjusting the reinforcement learning models accordingly, the system improves prediction accuracy over time while reducing uncertainty about model suitability through continuous validation and adaptation
Data Source
AI summary
A control device for operating an on-board power supply system of a motor vehicle includes an input unit which is configured to determine operating parameters of the on-board power supply system of the motor vehicle and/or one or more environment parameters of the motor vehicle and to forward them to a processing unit of the control device. At least one environment parameter is a probability of an operating action of a third-party vehicle.