Predictive Vehicle Speed Control for Low-Compute Hardware

Resolve Bottlenecks,
Find Innovative Solutions
Generate 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

VSEngineering 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

Engineering Contradiction:
Improvefuel consumptionVSAvoidcomputational resources and hardware
Core Design Contradiction:
Loss of energyVSDevice complexity

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.

Inventive Principle:
Principle #1Segmentation

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.

Inventive Principle:
Principle #35Parameter changes

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

Engineering Contradiction:
Improveoptimality of operating strategyVSAvoidcalculation speed
Core Design Contradiction:
ReliabilityVSProductivity

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.

Inventive Principle:
Principle #2Taking out (Extraction)

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.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP2881297B1Method for predictive influence of a vehicle speed
Publication Date: 2020.03.18 ROBERT BOSCH GMBH
  • EP2881297B1 patent drawingFigure 1
  • EP2881297B1 patent drawingFigure 2
  • EP2881297B1 patent drawingFigure 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.