Optimization of Thermal Network Parameters for Generator-Motor Based on Genetic Algorithm

The working conditions of generator-motors in pumped storage units are diverse. Therefore, online temperature prediction of generator-motors is crucial for the safe operation of pumped storage units. This article proposes an improved thermal network model based on genetic algorithm to optimize thermal resistance parameters for online temperature calculation of generator motors. Firstly, an equivalent thermal network model for stator-rotor separation was established based on the structure of the generator motor; Then, to address the issue of poor computational accuracy in general thermal network models, genetic algorithms are used to optimize the parameters of the thermal network; Next, in response to the difficulty of searching for the optimal solution using genetic algorithms, a segmented parameter optimization method for thermal network thermal resistance is proposed; Finally, the effectiveness of optimizing the thermal network parameters was verified by comparing the improved thermal network under different operating conditions with the finite element simulation results.