Design and Calibration Method for Built-In Temperature Sensor in Rocker Arm of the Shearer

The intelligent coal shearer has introduced unmanned systems and fault diagnosis technology, achieving automation in many tasks, reducing reliance on manpower, and improving work safety. By monitoring and analyzing the operational data of the equipment, the fault diagnosis system can timely detect potential issues and provide corresponding solutions. Accurate and timely preventative maintenance can reduce equipment failures and downtime, thereby improving production efficiency and reliability.This study proposes a built-in sensor system, including design, installation, and communication scheme, for rocker arm of the shearer. The system aims to achieve high-quality data and conduct real-time monitoring for fault diagnosis of rocker arm in a shearer. Unlike traditional vibration sensor which contains intense noise from coal cutting, the designed built-in sensor system enables real-time acquisition of infrared images on the gear tooth surface, thus providing more intuitive information for fault diagnosis. Additionally, a temperature calibration method named Poly-Forest Bagging (PFB) is proposed to solve the nonlinearity and drift problems in temperature measurement using infrared sensors. This algorithm integrates polynomial regression and random forest under the framework of bagging. Through calibration experiments and comparisons with individual models, it demonstrates that the PFB algorithm exhibits superior accuracy, overfitting resistance, and robustness compared to individual models, significantly enhancing sensor precision.