Electric Vehicle Range Prediction Using Segmented Energy Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for predicting the range of electric vehicles are inaccurate due to failure to consider variable speed profiles, regenerative braking, and energy source degradation, leading to range anxiety and unexpected charge loss.
Innovation Solution
A method and system that uses a range predicting system to receive input data, predict energy consumption using data-based and physics-based models, and account for energy degradation, allowing for accurate determination of remaining energy and vehicle range.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If existing physics-based models use constant speed assumption for link traversal, then calculation complexity is reduced, but measurement precision of energy consumption deteriorates
Solution Approach 1:
The patent segments the vehicle journey into multiple links with nodes at intersections, and further segments each link into phases (constant speed, acceleration, deceleration). This allows the system to apply simplified constant speed calculations for most of the link while adding acceleration/deceleration energy calculations only where necessary, balancing computational complexity with prediction accuracy.
Solution Approach 2:
The patent applies different calculation methods to different portions of the journey: constant speed assumption for link traversal, but variable speed profiles for acceleration and deceleration phases. This local differentiation allows the system to maintain low computational complexity for the majority of the calculation while improving accuracy at critical phases where speed changes occur.
2Device complexity
If existing techniques determine energy consumption only based on real-time values, then computational simplicity is maintained, but reliability of range prediction deteriorates due to ignoring aging effects
Solution Approach 1:
The patent performs preliminary determination of the energy degradation factor based on accumulated usage data (distance traveled, charge-discharge cycles, temperature exposure) before calculating the final energy consumption. This allows the system to account for battery aging effects in advance, improving prediction reliability without significantly increasing computational complexity during real-time operation.
Solution Approach 2:
The patent incorporates feedback from historical usage data to continuously update the energy degradation factor. The system monitors accumulated distance, charge-discharge cycles, and temperature exposure, then uses this feedback to adjust the degradation factor, ensuring that range predictions remain reliable as the battery ages.
3Ease of operation
If existing models ignore variable acceleration and deceleration during navigation, then ease of operation is improved, but measurement precision of energy consumption deteriorates
Solution Approach 1:
The patent segments the motion profile into distinct phases: constant speed traversal of links, acceleration phases at nodes, and deceleration phases. This segmentation allows the model to maintain simplicity for the majority of the journey (constant speed links) while incorporating variable acceleration calculations only where necessary, improving accuracy without significantly complicating the overall model.
Solution Approach 2:
The patent applies variable acceleration and deceleration calculations locally at specific nodes and phases rather than throughout the entire journey. This local application maintains the ease of operation for the overall model while improving measurement precision at critical points where speed changes occur, such as during regenerative braking events.
Data Source
Figure 1A~1B
Figure 2A
Figure 2B
AI summary
Disclosed subject matter is related to field of telematics that provides method and a range predicting system for predicting range of an electric vehicle. The method includes receiving, by the system, input data from data sources. Thereafter, the system predicts energy consumption along one or more routes based on first subset of input data using data based and physics based models. The system then determines available energy based on second subset of input data and an energy degradation factor, and consumed energy based on third subset of input data. Thereafter, the system determines remaining energy level based on the available energy and the consumed energy. Finally, the system determines a range of the electric vehicle based on the remaining energy level and the predicted energy consumption along the one or more routes. The present disclosure enhances accuracy in determining range, and avoids unexpected loss of charge or breakdowns.