Novel Energy Perspective on Error Filtering Analysis in Precision Control Circumstances

Bingyang Hou,Ze Wang,Chuxiong Hu,Yu Zhu

DOI: https://doi.org/10.1109/tmech.2024.3480931

2024-01-01

Abstract:Error filtering analysis is a crucial process in precision control scenarios, as it indicates whether any residual errors remain to be eliminated, thereby potentially improving control methods. Strategies for error filtering commonly rely on frequency domain analysis, with adjustments to the bandpass bandwidth enabling the acquisition of effective components. However, the classic filtering analysis has critical drawbacks, as it necessitates different bandpass bandwidths for varying control scenarios, and obtaining effective components is often empirical. This article introduces a novel error filtering method based on the energy perspective, which offers enhanced potency in separating primarily eliminable components in errors. To some extent, the proposed energy-based filter demonstrates adaptive bandpass properties by automatically adjusting its bandwidth to suit various circumstances and systems. This approach has led to an novel design of the iterative learning control feedforward controller aimed at achieving high-precision motion control. The primary contribution is the innovative energy filtering perspective, improving upon the prevailing frequency domain-based analysis view. Comparative experiments have demonstrated the effectiveness of the proposed method in selecting and compensating for eliminale components in errors.