Converter Control Loop Tuning With ML-Based PID Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional PID controller tuning methods, especially in digital control systems for electrical drives, often result in reduced stability margins due to neglecting sampling effects, and require experienced personnel for manual tuning, which is time-consuming and inefficient for systems with long loop times.
Innovation Solution
A method utilizing machine learning models to acquire actual values from drive systems and infer adjustments for control parameters, such as proportional gain and integral time, to improve control accuracy and stability, automatically optimizing control loops based on actual rotation speed, torque, and control differences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional PID controller tuning methods are used in digital control systems, then the control loop can be implemented, but stability margins are reduced due to neglecting sampling effects
Solution Approach 1:
The patent transforms continuous-time PID parameters (Kp, Ki, Kd) into discrete-time parameters (Kp_s, Ki_s, Kd_s) by applying sampling period compensation formulas. This parameter transformation accounts for sampling effects while maintaining control accuracy, resolving the contradiction between stability and precision in digital implementations.
2Manufacturing precision
If manual tuning methods are used for control parameters, then experienced personnel can optimize the control loop, but the process is time-consuming particularly for systems with long loop times
Solution Approach 1:
The patent enables the control system to automatically determine optimal PID parameters based on measured process characteristics (rise time, settling time, overshoot) without requiring external expert intervention. The system self-adjusts by calculating parameters from its own operational data, eliminating time-consuming manual tuning while maintaining optimal control accuracy.
3Ease of operation
If control parameters are fixed during operation, then the control loop is simple to operate, but the system cannot adapt to wear or environmental changes
Solution Approach 1:
The patent transforms static PID parameters into dynamic parameters that automatically adjust based on real-time process characteristics. The system continuously monitors performance metrics and recalculates optimal parameters, enabling adaptation to wear and environmental changes while maintaining ease of operation through automated adjustment.
4Loss of time
If automated tuning methods are implemented, then tuning time is reduced, but device complexity increases
Solution Approach 1:
The patent introduces measurement intermediaries (timers and detectors) that automatically capture process characteristics (rise time, settling time, overshoot) and feed them to parameter calculation units. This intermediary measurement system enables automated tuning without requiring complex artificial intelligence, balancing reduced tuning time with manageable system complexity.
Data Source
AI summary
A method of optimizing a control loop of a converter, such as a control system of the converter, includes acquiring actual values of a drive system powered by the converter, inferring, based on at least one machine learning model and the actual values, one or more adjustments of control parameters of the control loop for improving the control accuracy, and outputting the one or more adjustments for adapting control parameter values.