Control Loop Interaction Detection via Cross-Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current diagnostic methods for automation systems do not effectively monitor and evaluate interactions between control loops, which can lead to suboptimal system performance and energy inefficiency due to unnoticed cross-influences between interconnected control loops.
Innovation Solution
A diagnostic device and method using a discrete-time cross-correlation function to analyze interactions between control loops by evaluating the cross-correlation of controlled process variables and manipulated variables, allowing for graphic representation and identification of significant interactions without requiring scaling factors or prior knowledge of the system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If control loops are monitored using traditional diagnostic methods (variance analysis, overshoot evaluation), then individual control loop performance can be assessed, but interactions between control loops cannot be detected
Solution Approach 1:
The patent combines multiple diagnostic approaches by integrating cross-correlation analysis with traditional control loop monitoring. The cross-correlation function merges information from multiple control loops to detect interactions, while maintaining the ability to assess individual loop performance through variance and overshoot evaluation.
Solution Approach 2:
The cross-correlation function serves as an intermediary tool that analyzes the relationship between controlled variables of different control loops. It mediates the detection of interactions by quantifying the degree of coupling between loops without requiring modification to the individual loops themselves.
2Productivity
If interactions between control loops are analyzed using cross-correlation functions, then system-wide optimization can be achieved, but the complexity of the diagnostic system increases
Solution Approach 1:
The cross-correlation function is a universal mathematical tool that can be applied to any pair of control loops without requiring loop-specific parameters. This multi-functional approach allows the same diagnostic method to detect interactions across the entire control system, reducing the need for multiple specialized diagnostic tools.
Solution Approach 2:
The patent transforms the complex problem of interaction detection into a simpler parameter analysis by using cross-correlation coefficients. The interaction strength is quantified through parameter changes in the cross-correlation function, making it easier to assess and compare interactions across different loop pairs.
3Ease of manufacture
If traditional diagnostic methods are used, then the diagnostic system remains simple to implement, but scaling factors and background knowledge are required
Solution Approach 1:
The cross-correlation-based diagnostic system is self-sufficient and does not require external scaling factors or process-specific background knowledge for implementation. The method automatically adapts to different control loops by computing correlations directly from the measured signals, eliminating the need for manual calibration or expert knowledge input.
Data Source
AI summary
A diagnostic device and method for monitoring operation of automation system control loops includes an evaluation device and a data memory storing sequences of actual-value data of the control loops. An absolute value maximum of a cross-correlation function, for an excitation resulting from changes in a setpoint setting of one control loop, is determined as a first quantitative measure for positive time lag of the sequence of actual-value data of another control loop, and a numeric parameter evaluating an interaction effect of the one control loop on the other is determined as a function of the first quantitative measure. Strength and direction of the interaction effect are displayed. The calculations are repeated for all pairs of monitored control loops and a matrix of the numeric parameters is displayed.