Bayesian Embedded Tuning for Model-Less Controller Adaptation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Model-less controllers, such as PID controllers, in industrial automation systems face challenges in adapting to changes in operating conditions and disturbances due to their inability to account for future system responses, leading to inefficiencies and reduced robustness.
Innovation Solution
A method using a Bayesian optimization algorithm to generate a model of the system's operational characteristics based on normal operation data, with minimal excitations to enrich information content, allowing for real-time tuning of parameters without disrupting the system's operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If model-less controllers are used to control industrial automation systems, then device complexity is reduced and ease of operation is improved, but adaptability to changes in operating conditions deteriorates
Solution Approach 1:
The system performs preliminary actions by continuously collecting operational data and generating a dynamic model of the controlled object during normal operation. This preliminary modeling enables the model-less controller to adapt to changes in operating conditions without requiring complex predefined models or manual reconfiguration, thus maintaining ease of operation while improving adaptability.
Solution Approach 2:
The system implements feedback by continuously monitoring the controlled object's response to controller commands and using this information to update the operational model. The Bayesian optimization algorithm processes this feedback to refine controller parameters in real-time, enabling the model-less controller to adapt to changing conditions while maintaining simple operation through automated parameter adjustment.
2Manufacturing precision
If traditional tuning methods are used for model-less controllers, then manufacturing precision may be maintained, but loss of time increases due to system shutdowns and manual intervention
Solution Approach 1:
The system implements self-service by automatically performing controller tuning without human intervention. The Bayesian optimization algorithm autonomously analyzes operational data, generates models, and adjusts controller parameters in real-time, eliminating the need for manual tuning operations that would require system shutdowns and technician intervention, thus reducing time loss while maintaining performance.
Solution Approach 2:
The system ensures continuity of useful action by performing tuning operations during normal system operation rather than requiring shutdowns. The Bayesian optimization algorithm continuously processes operational data and adjusts parameters in real-time, allowing the controlled process to maintain continuous production while the controller adapts to changing conditions, thereby eliminating time loss associated with shutdowns.
3Measurement precision
If excitation inputs are applied to enrich information content for model generation, then measurement precision improves, but productivity decreases due to disruption of normal operation
Solution Approach 1:
The system applies partial action by using only the information naturally available during normal operation to generate the operational model. The Bayesian optimization algorithm processes existing operational data without requiring additional excitation inputs that would disrupt production, achieving sufficient measurement precision for effective controller tuning while maintaining continuous productivity through partial utilization of available data.
Data Source
AI summary
A method may include receiving data representative of one or more commands generated by a model-less controller to control operations of devices within a system and output parameters associated with the devices of the system. The method may also include determining whether the data is indicative of a change in operational characteristics of the system and generating a model representative of the operational characteristics of the system as a function of the data based on a Bayesian optimization algorithm in response to the data being indicative of the change. The method may also involve transmitting an excitation input to the devices in response to the data not being indicative of the change, receiving updated output parameters associated with the devices of the system after the excitation input is transmitted, and generating the model based on the updated output parameters and the excitation input.