Optimization of Distributed Generation Based on Generative Adversarial Network and Improved Differential Evolution Algorithm

The uncertainty of distribution generation(DG) output is generated by convolutional generative adversarial network, and the upper and lower limits of DG output are described. A multi-objective DG optimal configuration model is established, and the weight of each objective function is determined by analytic hierarchy method, so as to transform the multi-objective problem solution into a single-objective problem solution. Aiming at the shortcomings of weak global search ability of differential evolution algorithm, an improved differential evolution algorithm is proposed, which introduces dynamic scaling factor and cross probability, and adds local search strategy to improve the convergence speed and search accuracy of the algorithm. Finally, based on the simulation analysis of IEEE 33 node distribution network, it is verified that the improved differential evolution algorithm can effectively improve the voltage distribution and branch loss of the distribution network.