Improving of Permanent Magnet Synchronous Machine Control Efficiency

The paper describes the problem of permanent magnet synchronous machine control efficiency improving. The main goal is to evaluate rotor temperature, stator yoke, tooth and winding temperature. To solve the problem, mathematical modeling method was chosen. The method has number of economical and operation advantages in comparison with injection method and flow tracking method. Machine’s heat temperature forecast was based on solution trees gradient boosting model. Nowadays this model is popular in forecasting models because of its high generalization ability. In order to get mathematical model, data about environmental temperature, cooling agent temperature, machine’s speed, current rate and voltage power is taken into account. Experiment result is the comparison between input and forecast values of rotor and stator temperature on different sections. Determination coefficient lies in range from 0,953 to 0,985. The result is satisfactory for rotor and stator temperature evaluation. Suggested approach for machine’s temperature evaluation can be a part of the control system. Such system will be useful during operator’s work with electric drive, because the forecast of drive’s parts overheat allows operator to make relevant actions to eliminate emergency situation. Reduction, and in perspective exception of situations, when the temperature of the machine is about to high permissible limit, can safe winding quality. In general, suggested system should increase operational life of the electric drive.