Operation Machine Control Switching for Data Deviation Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing control systems face reliability issues when new input data and output data deviate from previously obtained data, leading to potential failures and malfunctions in operation machine control.
Innovation Solution
A control device and system that acquires attribute information including statistics of previous input and output data, evaluates the deviation of new data from these statistics, and adjusts or switches controllers to maintain reliability, using an evaluation part to compare new data with established thresholds and measurement information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a controller generated based on previously obtained input data and output data is used to control an operation machine, then control reliability is high when new data is similar to previous data, but control reliability deteriorates when new data deviates from previous data
Solution Approach 1:
The patent implements a feedback mechanism where the evaluation part continuously monitors new input data and new output data against attribute information from previous data. When deviation exceeds a threshold, the system provides feedback to switch from the learned controller to a safety-oriented controller, ensuring control reliability adapts to changing data conditions
Solution Approach 2:
The system dynamically switches between different control modes based on data similarity. The control part selects between a learned controller (for similar data) and a safety-oriented controller (for deviant data), making the control system adaptive rather than static, thereby maintaining reliability across varying data conditions
2Productivity
If the controller uses learned relationships from previous data, then control performance is optimized for known patterns, but control safety deteriorates when encountering unknown or deviant patterns
Solution Approach 1:
The evaluation part acts as an intermediary between the learned controller and the operation machine. It assesses whether new data deviates from learned patterns and mediates controller selection, preventing direct use of potentially unsafe learned control for deviant data while maintaining efficient learned control for normal data
Solution Approach 2:
The system prepares multiple controllers in advance: a learned controller for optimal performance and a safety-oriented controller for unknown conditions. This prior preparation cushions against the risk of using learned control on deviant data, ensuring safety is already in place before deviation occurs
Data Source
AI summary
An objective of the present invention is to provide a control device, control system, control method, and computer-readable storage medium, for enabling verification of the reliability of operation machine control. Provided is a control device comprising: a control part comprising a controller for outputting output data with regard to input data, said control part serving to control an operation machine using the controller; an acquisition part for acquiring attribute information including statistics of previously obtained input data and output data; and an evaluation part for, on the basis of a comparison of the attribute information with new input data being newly inputted into the controller and/or new output data being newly outputted from the controller with regard to the new input data, evaluating to what extent the new input data and/or the new output data deviate from the statistics.