EV Thermal Management Control for Range and Cabin Comfort
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current thermal management systems in electric vehicles face inefficiencies in achieving optimal balance between user comfort and driving range due to complex calibration requirements and suboptimal control strategies, leading to increased energy consumption and reduced performance across varying conditions.
Innovation Solution
A data-driven supervised learning model, integrated with a control optimization unit, generates optimal control setpoints for the thermal management system, utilizing AI-based control to minimize energy consumption and maximize the coefficient of performance (COP) by considering user requests, cabin temperature, and driving conditions, thereby reducing calibration efforts and ensuring continuous high efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If traditional control strategies are used for thermal management, then the system structure is simple, but energy consumption increases and coefficient of performance decreases
Solution Approach 1:
The patent replaces traditional mechanical control strategies with an AI-based supervised learning model. The control device uses machine learning algorithms to predict optimal control outputs based on input parameters, substituting conventional control mechanisms with intelligent computational systems that adapt to varying operating conditions, thereby reducing energy consumption while managing complexity through software-based solutions.
Solution Approach 2:
The patent dynamically changes control parameters based on real-time operating conditions. The supervised learning model processes multiple input parameters (ambient temperature, cabin temperature, vehicle speed, etc.) and generates optimized control outputs that adjust thermal management system parameters continuously, enabling the system to adapt to different scenarios and minimize energy consumption across varying operating conditions.
2Manufacturing precision
If complex calibration is performed to optimize thermal management, then control precision improves, but calibration time and effort increase
Solution Approach 1:
The patent performs preliminary training of the supervised learning model using extensive datasets before actual operation. The model is pre-calibrated with various operating scenarios and conditions during the development phase, storing optimized control strategies in its neural network structure. This preliminary action eliminates the need for time-consuming calibration during vehicle operation or deployment, as the model already contains pre-learned optimal control patterns for diverse conditions.
Solution Approach 2:
The supervised learning model enables the thermal management system to self-optimize without requiring manual calibration. The model automatically adapts to new operating conditions by processing real-time sensor data and generating appropriate control outputs, eliminating the need for external calibration interventions. The system serves itself by continuously learning from operational data and adjusting control strategies autonomously.
3Reliability
If thermal management system is optimized for various conditions, then coefficient of performance improves, but system complexity increases
Solution Approach 1:
The patent implements a universal supervised learning model that handles multiple thermal management functions through a single control device. The model processes various input parameters and generates control outputs for different system components (compressor, condenser, evaporator, etc.), enabling one system to perform multiple optimization functions across diverse operating conditions, thereby improving coefficient of performance without proportionally increasing system complexity.
Solution Approach 2:
The patent introduces an intermediary layer between sensors and actuators in the form of the supervised learning model. This intermediary processes raw sensor data, predicts optimal control strategies, and generates control signals for system components. The model acts as a intelligent mediator that simplifies the control architecture by replacing multiple specialized controllers with a single unified learning-based system that handles all optimization tasks.
Data Source
AI summary
A thermal management system for an electric vehicle includes a thermal system with a control device. The control device generates lower level control outputs for operating the thermal management system based on inputs including user requests and parameters detected by a sensor device. The control device computes a cost function from the inputs to generate intermediate outputs, computes optimal control setpoints based on the intermediate outputs, and computes the lower level control outputs based on the optimal control setpoints for operating the thermal management system. The inputs are selected from a group of parameters defining target air conditions in a cabin, ambient conditions, thermal system conditions and vehicle states. The intermediate outputs comprise coefficient of performance of the thermal system and/or power consumption parameters of electric components. The optimal control setpoints comprise operating parameters of the thermal systems and/or temperature conditions.