Auto-tuning Controller Parameters for Vacuum Deposition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing control systems, particularly in vacuum deposition and other complex processes, face challenges in automatically adjusting controller parameters to meet performance specifications due to non-linear relationships between actuator inputs and sensor outputs, requiring manual adjustments that are impractical in real-world scenarios.
Innovation Solution
The method involves determining an average value for actuator offsets and calculating optimum controller parameters using a function that considers the time to setpoint, phase difference, and control algorithm, allowing for automatic adjustment of controller parameters to achieve specified performance characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual adjustment of controller parameters is used, then performance specifications can be met, but the system requires human intervention and is impractical for real-world scenarios
Solution Approach 1:
The system performs self-tuning by automatically determining system parameters (A and B) through test routines and calculating optimum controller parameters without human intervention. The controller executes the tuning method autonomously, making the system self-sufficient for parameter adjustment while maintaining performance specifications.
Solution Approach 2:
The invention changes the parameters of the controller (Kp, Ki, Kd) based on determined system parameters A and B. By dynamically adjusting these parameters according to the calculated values and desired time to setpoint, the system automatically adapts to meet performance specifications without manual intervention.
2Ease of operation
If automatic parameter adjustment is implemented, then ease of operation improves, but the non-linear relationship between actuator inputs and sensor outputs complicates the adjustment process
Solution Approach 1:
The system determines system parameters A and B through controlled test routines that characterize the non-linear relationship between actuator and sensor. These parameters are then used in calculations to derive controller parameters, transforming the complex non-linear problem into a systematic parameter determination and calculation process.
Solution Approach 2:
The method uses feedback from sensor measurements during test routines to determine system parameters. By monitoring sensor outputs in response to known actuator inputs, the system characterizes its own behavior and uses this feedback information to calculate appropriate controller parameters that compensate for non-linearities.
3Extent of automation
If comprehensive modelling of all process parameters is performed, then self-contained control is achieved, but the complexity and difficulty of modelling increases significantly
Solution Approach 1:
Instead of comprehensive modelling of all process parameters, the invention determines two key system parameters (A and B) through simplified test routines. This parameter reduction approach achieves practical self-contained control by focusing on the essential dynamics without requiring exhaustive system modelling.
Solution Approach 2:
The method extracts the essential dynamic characteristics of the system by determining parameters A and B through targeted test routines. Rather than modelling the entire complex system, the invention extracts the critical information needed for control parameter calculation, simplifying the overall approach while maintaining effectiveness.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
This invention relates to the automated tuning and calibration of feedback control of systems and processes. According to the present invention a series of actuator pulses are automatically performed and, based on the gradient of the sensor response, information is determined on the dynamics of the system to be controlled (this will be known as the system identification procedure). This is preceded by an automatic sensor calibration procedure in order to determine the controller's window of operation. Based on the information collected during the system identification procedure controller parameters are automatically calculated for a specified time for the sensor to reach the setpoint. The present invention relates to any system that is managed and/or controlled by a controller and/or control algorithm. The present invention relates to the use of the present method for parameterisation of any control algorithm, for example, PID, PI, P, PDF.