Machine Controller Configuration Using Reduced State-Space Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Model predictive control systems for complex machines require significant computing resources and manual effort, and data-driven machine learning models provide uninterpretable control characteristics, making certification challenging.
Innovation Solution
A machine controller is configured using a first signal converter to reduce state signals into a lower-dimensional space and a second signal converter to generate optimized control signals through a conversion rule, allowing for efficient and interpretable control of machines like robots and turbines.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If model predictive control is used to control complex machines, then control optimization is improved, but computing resource requirements increase significantly
Solution Approach 1:
The patent performs simulations for various operating conditions in advance to generate an operating condition model before actual control is needed. This pre-computation stores control strategies for different scenarios, allowing the controller to quickly retrieve and apply pre-determined control actions during real-time operation without performing complex simulations, thus reducing computing resource requirements while maintaining control optimization.
2Measurement precision
If manual derivation of control rules from simulations is performed, then control accuracy is improved, but manual effort increases greatly
Solution Approach 1:
The system automatically generates the operating condition model by performing simulations and deriving control rules without requiring manual intervention. The controller autonomously processes simulation data, identifies operating conditions, and formulates control strategies, eliminating the need for manual rule derivation while maintaining high control accuracy.
3Adaptability or versatility
If data-driven machine learning models are used for control, then control flexibility is improved, but interpretability and certification difficulty increase
Solution Approach 1:
The patent introduces an operating condition model as an intermediary between the machine learning model and the control system. This model explicitly represents operating conditions and control strategies in a structured format that is both flexible for adapting to different scenarios and interpretable for certification purposes, bridging the gap between data-driven flexibility and human-understandable control logic.
Data Source
AI summary
To configure a machine controller for a machine, a plurality of state signals of a first state space is read in, each state signal being assigned an optimized control signal. Using the state signals, a first signal converter is trained to convert state signals from the first state space into a second state space which is dimension-reduced in comparison with the first state space. A second signal converter is trained to reproduce corresponding optimized control signals by converting reduced state signals by means of the conversion rule. Thus, the machine controller is designed to convert a state signal of the machine into a reduced state signal by means of the trained first signal converter and to convert the reduced state signal into an optimized control signal by means of the trained second signal converter, the optimized control signal being used to control the machine.