Rotor Circle Measurement Analysis of Hydro-Generator Based on Improved Parrot Optimization Algorithm

With the rapid development of China’s hydropower industry, hydro-generator units are becoming increasingly complex, large-scale, and highly integrated. Rotor roundness analysis of hydro-generators is crucial for ensuring the safe operation of units and improving the economic and social benefits of hydropower stations. Aiming at the problems of traditional data processing algorithms for rotor roundness measurement, such as insufficient accuracy and vulnerability to abnormal data interference, this paper introduces and improves the Parrot Optimization Algorithm (PO), proposing the Improved Parrot Optimization Algorithm (IPO). The algorithm capabilities are enhanced by initializing the population with Halton sequences, introducing t-distribution perturbation, and constructing adaptive parameters. A hybrid dynamic loss function including Huber loss and least squares loss is constructed as the objective function for circle fitting. In an engineering case study of a power station, IPO is compared with the traditional least squares method, PO, and PSO. The results show that IPO has a fitted radius closer to the ideal value, smaller eccentricity, lower fitness value, faster convergence speed, and higher accuracy. The adjustment suggestions provided by it meet the specification requirements in practical operations, proving the superiority and accuracy of the rotor roundness analysis model based on IPO.