Cyber-Physical Control Process Generation Using Potential Fields
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for automated control process generation in cyber-physical systems lack a general approach to describe target constellations and transitions from arbitrary start constellations, hindering the automatic generation of control processes.
Innovation Solution
A method specifying control processes for cyber-physical systems involves modeling cyber-physical objects with parametrized functions, assigning parameters representing degrees of freedom, and using artificial potential fields to evaluate progress and generate control instructions through experimentation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If global path planning algorithms are used to evaluate all possible configurations, then a collision-free path can be found, but the method requires prior knowledge of the complete operative environment and high computational resources
Solution Approach 1:
The patent segments the configuration space into discrete configurations and uses systematic exploration methods (RRT, PRM) to divide the path planning problem into manageable phases: preprocessing to generate collision-free configurations and query phase to find paths between them
Solution Approach 2:
The patent performs preliminary actions by pre-computing and storing collision-free configurations in a roadmap structure before actual path planning is needed, allowing faster query-time path finding without re-evaluating all configurations
2Ease of operation
If local path planning algorithms are used to find trajectories from actual configuration, then situational knowledge is utilized, but complete path planning from start to target configuration is not achieved
Solution Approach 1:
The patent implements feedback mechanisms where the path planner continuously evaluates the current configuration against the roadmap of pre-computed collision-free configurations, using shortest path algorithms (Dijkstra) to dynamically determine optimal paths based on current situational knowledge while guaranteeing complete path planning from start to target
3Ease of operation
If evolutionary artificial potential field method is used for configuration planning, then attractive and repulsive forces guide movement, but the method may get trapped in local minima
Solution Approach 1:
The patent employs dynamic path planning algorithms (RRT, PRM) that adaptively explore the configuration space rather than following static potential field gradients, allowing the system to dynamically adjust exploration strategies and escape local minima by randomly sampling new configurations or revisiting previously explored areas with different parameters
4Productivity
If automated control process generation is implemented, then control efficiency is improved, but lack of general approach to describe target constellation and transitions prevents viability
Solution Approach 1:
The patent creates a universal control process generation framework that can handle diverse cyber-physical systems by defining abstract concepts (control process targets, observation targets, constellations) that apply generally across different domains, allowing the same experimental approach to generate control processes for various systems with different parameters and objectives
Solution Approach 2:
The patent uses parameter-based representations where control processes are defined by configurable parameters (target constellations, observation selections, experimental settings) that can be adjusted to suit different applications, enabling automated generation across diverse systems through parameter variation rather than system-specific customization
Data Source
AI summary
The present invention relates to the field of automated control process generation for control of a cyber-physical system. In more detail the present invention relates to a method of specifying a control process for a cyber-physical system and a related control process specifying engine. Control process specification is achieved by specifying at least one control process observation target to be reached by the control process. Then, according to the present invention a method of generating at least one control process instruction for the specified control process and a related control process experiment execution engine are used for automated control process generation through execution of control process experiments.