Vehicle Control Method for Energy Minimization via Dimensionality Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle control methods face challenges in efficiently minimizing vehicle operating costs due to high computational requirements for optimizing vehicle control parameters across multiple dimensions, which can be time-consuming and resource-intensive, especially in real-time applications.
Innovation Solution
The method focuses on finding local minima in two-dimensional subspaces instead of global minima in high-dimensional spaces, allowing for reduced computational effort by keeping other variables constant and optimizing only the most critical manipulated variables, such as vehicle speed and gear, while using a cost map to determine energy consumption based on plausible physical vehicle models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If global minimum search is performed in high-dimensional control parameter space, then optimal vehicle operating cost minimization is achieved, but computational time and processing resources increase significantly
Solution Approach 1:
The patent segments the high-dimensional optimization problem into multiple two-dimensional subproblems by fixing n-1 control variables and optimizing one variable at a time. This divides the complex n-dimensional search space into manageable 2D slices, each requiring separate curve calculations. The segmentation principle directly addresses the contradiction by reducing computational complexity while maintaining sufficient optimization effectiveness for real-time vehicle control applications.
Solution Approach 2:
The patent applies partial action by performing optimization only in the most critical dimensions (two-dimensional subspaces) rather than exhaustively searching all n dimensions. By selecting the two most influential control variables for simultaneous optimization while holding others constant, the system achieves acceptable near-optimal solutions with significantly reduced computational effort, trading complete optimality for real-time feasibility.
2Loss of energy
If iterative optimization is performed across multiple control variables simultaneously, then comprehensive cost minimization is achieved, but processing capacity requirements increase
Solution Approach 1:
The patent segments the multidimensional optimization into sequential two-dimensional problems. Instead of simultaneously optimizing all n control variables which would require complex high-dimensional iterative algorithms, the system performs separate curve minimizations for pairs of variables while holding others constant. This segmentation dramatically reduces the computing capacity required for each optimization step while still addressing the energy minimization objective.
Solution Approach 2:
The patent transforms the n-dimensional optimization problem into a series of two-dimensional problems by changing the dimensionality of the search space. By fixing n-1 variables and optimizing only two variables at a time, the system reduces the computational dimensionality from n to 2, making the optimization tractable for real-time embedded vehicle control systems with limited processing capacity.
3Productivity
If real-time vehicle control optimization is implemented, then energy efficiency is improved, but computational resources and processing time become insufficient
Solution Approach 1:
The patent segments the optimization computation into multiple simple two-dimensional curve minimization tasks that can be executed sequentially or in parallel with minimal computational resources. Each segment optimizes a pair of control variables independently, avoiding the need for complex real-time high-dimensional optimization. This segmentation enables energy efficiency improvements to be achieved within real-time processing constraints of vehicle control systems.
Solution Approach 2:
The patent performs partial optimization by focusing computational effort only on the two most critical control variable dimensions at any given time, rather than simultaneously optimizing all n variables in real-time. This partial action approach achieves sufficient energy efficiency improvements while keeping processing time within acceptable real-time limits, accepting near-optimal rather than perfectly optimal solutions.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
The invention relates to a method for operating a vehicle, comprising the following steps: providing a multidimensional analytical vehicle operating cost function that defines the dependence of a vehicle propulsion energy consumption on a preceding distance traveled or a preceding driving time, as well as on several control variables; calculating a curve of one of the several control variables, wherein the curve minimizes the vehicle operating cost function for the preceding distance traveled or for the preceding driving time, wherein the remaining control variables are kept constant for the calculation of the curve; repeatedly calculating further respective curves of the several control variables, wherein the respective curve of one of the several control variables minimizes the vehicle operating cost function for the preceding distance traveled or for the preceding driving time.wherein, for the calculation of the respective curve, the control variables are specified according to the previously calculated curves and the remaining control variables are kept constant, controlling the vehicle depending on the calculated respective curves of the several control variables while the vehicle travels the route or while the driving time elapses. The invention further relates to a computer program product and a driver assistance system.