Vehicle Energy Prediction Using Historical Journey Error Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle range prediction systems are inaccurate as they do not consider the full nature of the journey, leading to incorrect energy requirement estimates, particularly for vehicles with long refueling/recharging intervals or sparse refueling/recharging stations.
Innovation Solution
The system adjusts the energy prediction algorithm to minimize aggregate error for historical journeys, using both variant and invariant vehicle data, as well as journey data, to provide more accurate energy requirement predictions for future journeys.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing vehicle range prediction systems use recent consumption data extrapolation, then the system complexity is low, but the measurement precision of energy requirement prediction deteriorates
Solution Approach 1:
The system performs preliminary actions by collecting and storing journey data, vehicle data, and energy consumption data before making predictions. Historical journey information is gathered in advance and used to train machine learning models, enabling more accurate predictions without requiring complex real-time calculations during actual journey planning
Solution Approach 2:
A machine learning model serves as an intermediary between raw data inputs and energy requirement predictions. The model processes historical journey data, vehicle specifications, and consumption patterns to generate accurate predictions, bridging the gap between simple data collection and precise energy estimation without requiring direct complex calculations
2Measurement precision
If the system collects and processes comprehensive journey data and historical information, then the measurement precision of energy prediction improves, but the loss of time for data processing increases
Solution Approach 1:
The system performs preliminary data processing by collecting and organizing journey data, vehicle data, and energy consumption data in advance. Historical information is pre-processed and stored in a structured format, allowing the machine learning model to make predictions without requiring time-consuming real-time data collection and processing during actual journey planning
Solution Approach 2:
The system uses feedback from actual journey outcomes to continuously improve predictions. After each journey, actual energy consumption data is fed back into the system to retrain and refine the machine learning model, progressively improving prediction accuracy without requiring increased processing time for future predictions
Data Source
Figure 1
Figure 2
Figure 3
AI summary
There is described a system for determining an energy requirement of a vehicle for a journey. The system may comprise a predictor mechanism to predict, using an energy prediction algorithm, a vehicle energy requirement for the journey. The system comprises an updater mechanism configured to refine the energy prediction algorithm for the vehicle by determining for each of a number of historical journeys undertaken by the vehicle, an error between an actual vehicle energy usage for the historical journey and a predicted energy usage derived using the energy prediction algorithm for the historical journey. An aggregate error is calculated from the errors of the number of historical journeys. The updater is arranged to adjust the energy prediction algorithm to reduce the aggregate error.