EV Range Estimation Using Location- and Driver-Specific Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for estimating the energy range of electric or hybrid electric vehicles are limited in accuracy due to factors like changes in location-specific and time-specific attributes, and they often require large databases for comparison, raising privacy concerns.
Innovation Solution
A system that uses a hierarchical learning framework to update an energy consumption estimation model based on the vehicle's location and time, leveraging data from edge/cloud servers and connected vehicles to provide more accurate energy range estimates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a generic energy range estimation model is used for all locations and drivers, then the system complexity is reduced, but the measurement precision of energy consumption estimation deteriorates
Solution Approach 1:
The estimation system is segmented into a generic base model and location-specific/driver-specific customized models. The base model provides general energy consumption estimation, while customized models are applied selectively based on location and driver characteristics, thereby maintaining system simplicity while improving accuracy where needed.
Solution Approach 2:
The system applies local quality by using location-specific models for particular geographical areas and driver-specific models for individual driving patterns. These customized models are deployed only in their respective local contexts rather than universally, improving estimation accuracy for specific conditions without increasing overall system complexity.
2Measurement precision
If location-specific and driver-specific customized models are deployed, then the measurement precision of energy consumption estimation is improved, but the device complexity increases
Solution Approach 1:
The system dynamically selects and switches between generic and customized models based on real-time conditions such as current location and driver identity. This dynamic adaptation allows the system to use complex customized models only when necessary, thereby improving accuracy without permanently increasing system complexity.
Solution Approach 2:
Instead of maintaining entirely separate complex systems for each location and driver, the solution uses copied and adapted versions of the base model. Location-specific and driver-specific models are created by training copies of the generic model with local data, reducing the complexity burden while maintaining precision benefits.
3Measurement precision
If large databases with historical data from multiple vehicles are used for model training, then the measurement precision is improved, but privacy concerns are raised
Solution Approach 1:
The system processes and stores data locally at edge servers near specific locations rather than centralizing all data in a large database. This localized data handling improves model training accuracy for specific regions while reducing privacy risks by keeping sensitive data distributed and accessible only to authorized local systems.
Solution Approach 2:
Edge servers act as intermediaries between vehicles and the cloud, performing local data processing and model training. This intermediary layer enables the system to utilize historical data for improved precision while maintaining privacy by preventing direct access to raw sensitive data and implementing data anonymization and secure processing protocols.
Data Source
AI summary
Systems and methods are provided for estimating the energy range of an electric vehicle, including updating an energy consumption estimation model based on the location of a vehicle or based on the current time. One embodiment comprises obtaining a route of a vehicle and deploying an energy consumption estimation model to provide an energy consumption estimate along the route from a first location to a destination. In response to determining that the vehicle enters a subsequent location, the energy consumption estimation model is updated with a location model received from an edge/cloud server, to estimate an energy range along the route through the subsequent location.