Industrial Control under Non-Ideal Measurements: Data-Based Signal Processing as an Alternative to Controller...

Area of Science:

• Industrial Process Control
• Data-Driven Modeling
• Signal Processing

Background:

• Industrial processes are complex and non-linear, requiring robust control strategies.
• Existing control strategies are vulnerable to noise and delays in measurement data.
• Current denoising and delay correction methods often involve complex, scenario-specific designs.

Purpose of the Study:

• To propose a data-based approach for denoising and correcting measurement delays.
• To simplify the design process for these techniques.
• To decouple the solution from specific control strategies and industrial scenarios.

Main Methods:

• A complete data-based methodology was developed using only input-output data pairs.
• The approach focuses on denoising and delay correction.
• The method was applied to a Wastewater Treatment Plant (WWTP) but is generalizable.

Main Results:

• A minimum Root Mean Squared Error (RMSE) improvement of 63.87% was achieved through the proposed denoising approach.
• The overall system performance demonstrated comparable or superior results to scenario-optimized methods.
• The data-based approach proved effective without requiring scenario-specific optimization.

Conclusions:

• The proposed data-based approach offers a simplified and effective solution for measurement denoising and delay correction in industrial environments.
• The method's decoupling from specific scenarios and control strategies enhances its versatility.
• This approach provides significant improvements in measurement accuracy and overall system performance.