Predictive Vehicle Speed Control for Low-Compute Hardware
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for calculating optimized vehicle operating strategies require significant computational resources and hardware, making them unsuitable for use with limited computing power or low hardware capabilities.
Innovation Solution
A model-predictive method that restricts calculations to physically possible speed changes and early detection of non-optimal states, reducing the number of calculation steps and hardware requirements by creating a limited speed corridor and excluding non-optimal states from the calculation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If a model-predictive method is used to calculate optimized vehicle operating strategies, then fuel consumption optimization is achieved, but significant computational resources and hardware are required
Solution Approach 1:
The prediction horizon is divided into multiple time steps, and the continuous speed profile optimization is segmented into discrete speed transitions at specific locations. This segmentation allows the complex continuous optimization problem to be broken down into manageable discrete decisions, reducing computational complexity while maintaining optimization effectiveness.
Solution Approach 2:
The invention changes the parameter representation from continuous speed profiles to discrete speed transitions at specific locations. By parameterizing the speed profile as a sequence of discrete speed changes at predetermined locations rather than continuous functions, the computational burden is significantly reduced while still achieving fuel consumption optimization.
2Reliability
If all possible speed transitions are calculated to find the optimal speed profile, then the most favorable operating strategy is determined, but the number of calculation steps increases significantly
Solution Approach 1:
The invention extracts and focuses only on the critical decision points - the speed transitions at predetermined locations - rather than calculating all possible continuous speed variations. By taking out only the essential discrete decisions that fundamentally affect fuel consumption, the calculation complexity is reduced while maintaining the ability to find the optimal strategy.
Solution Approach 2:
Instead of calculating all possible continuous speed profiles (excessive action), the invention calculates only the discrete speed transitions at predetermined locations (partial action). This partial approach is sufficient to achieve the optimization goal without the computational burden of exhaustive continuous optimization.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method for predictively influencing a vehicle speed, wherein the vehicle speed on a route is influenced by a changing parameter within a predetermined speed corridor, using segments with a defined segment length and speed steps, is to be improved in such a way that a computer with limited computing power can be used or the requirement for usable hardware is low.This is achieved by dividing the ahead route into segments (S), to which different speed steps (GA) are assigned, where intersecting segment values and speed values are defined as nodes in a two-dimensional diagram, and where each node is assigned a discrete position on the ahead route of the vehicle in conjunction with a discrete speed, that at least one entry can be stored in a storage medium for each node, where a cost value can be stored for the respective node, and that cost values stored in the storage medium for several nodes are compared with new cost values in order to determine a cost-optimal vehicle speed on a route.