Error compensation for instantaneous angular speed under industrial robot operating conditions

Abstract

The instantaneous angular speed (IAS) signal is currently being increasingly applied in the fault detection of rotating machinery, in the actual operating conditions of industrial robot joints, encoders experience complex variable rotation speeds. This presents significant challenges for fault detection in industrial robots based on the IAS signal collected by the encoder. This paper proposes an improved error compensation method specifically designed for non-stationary conditions. A coefficient of variation-guided local polynomial fitting method (VLPF) is introduced to mitigate engraving and subdivision errors in incremental grating encoders operating at varying rotation speeds. First, an error compensation model is developed by analyzing these errors. Next, the relationship between the coefficient of variation (Cv), based on power spectral density (PSD), and the optimal fitting window length at different speeds is established through simulation analysis. The IAS signal of the encoder is then collected from an industrial robot RV test bench and corrected using the proposed method. Finally, simulation and experimental results show that the VLPF method effectively suppresses IAS signal errors during rapid changes in rotation speed. The proposed error compensation method under operating conditions reduces encoder errors by approximately 63.37%, which will be more advantageous to use IAS signal for fault detection.