Adaptive Fuel Consumption Forecasting Model
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for forecasting fuel consumption and optimizing vehicle operations, such as choosing roads and gear usage, are not accurate enough, leading to potential increased fuel consumption and environmental impact due to inefficiencies in onboard electronic control systems.
Innovation Solution
A method using a mathematical model to forecast the evolution of journey data, such as fuel consumption, by defining a function based on input parameters, running reference trips, measuring and computing values, comparing results, and adjusting the model to minimize differences, allowing for improved selection of vehicle running conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a mathematical model with fixed parameters is used for forecasting fuel consumption, then the computation is simple and fast, but the forecasting accuracy is insufficient
Solution Approach 1:
The patent transforms a static mathematical model into a dynamic adaptive model by introducing real-time parameter adjustment mechanisms. The model parameters are continuously updated based on actual measured data from the vehicle journey, allowing the forecasting system to adapt to changing driving conditions, vehicle states, and environmental factors, thereby improving accuracy without requiring a completely complex system architecture
Solution Approach 2:
The forecasting system performs self-calibration by automatically adjusting its own parameters based on the difference between predicted and actual fuel consumption measurements. This self-learning capability allows the model to improve its accuracy over time without external intervention, effectively resolving the contradiction between model simplicity and forecasting precision
2Measurement precision
If real-time data collection and model adjustment are implemented, then the forecasting accuracy improves, but the computational load and processing time increase
Solution Approach 1:
The patent implements partial model adjustment by updating only the critical parameters that have the most significant impact on forecasting accuracy, rather than recalculating the entire model. This selective parameter adjustment reduces computational overhead while maintaining improved accuracy, effectively balancing processing time and forecasting precision in real-time operation
3Measurement precision
If the model is adjusted frequently based on measured data, then the forecasting accuracy improves, but the system complexity and adjustment difficulty increase
Solution Approach 1:
The patent establishes a closed-loop feedback system where the difference between measured and forecasted fuel consumption is continuously fed back to adjust model parameters. This systematic feedback mechanism automates the model adjustment process, reducing the complexity of manual tuning while maintaining high forecasting accuracy through continuous optimization
Data Source
Figure 1
Figure 2
AI summary
The magnitude (Y) of a data associated to a journey of an automotive vehicle is expressed by a function (f) of at least one input parameter (x). This method includes at least the steps of: a) defining (101) a first model (ft=0) of the function; b) running (102) the vehicle on a reference trip, the input parameter (x, m, p) and the magnitude (Y) being measured (YM, XM) during or at the end of the reference trip; c) computing (103) a value (Yc) of the magnitude by using the first model (ft=o) of the function (f) and the value of the parameter (XM) measured at step b); d) comparing (104) the values (YM, Yc) of the magnitude at said time; and e) adjusting (105) the function (ft=1) in a way corresponding to the reduction of the difference between the measured value (YM) and the computed value (Yc).