Monitoring and predicting the states of electric drives of a wheeled robot

This study proposes an approach for monitoring and predicting the state of a wheeled robot's drives based on the analysis of empirical data and the dynamics of changes in the reductor's temperature and the motor's current. A test bench was developed to simulate various robot drive operational modes—from light load to extreme operating conditions. During the experiments, time series data for four load modes were obtained. The least squares method was applied for predicting temperature regimes by constructing polynomial regression models. Based on the data analysis, a state matrix with drive operation recommendations was formed, and a decision-making algorithm was formalized. The results demonstrate the potential for effective forecasting of abnormal operating modes in electric drives.