Aiming at the characteristics of randomness and volatility of the power output of multivariate decentralized resources affected by natural conditions, this paper proposes a new type of intelligent cluster regulation and control strategy. A multivariate equivalent modeling method is designed to establish a cluster voltage control model to enhance the system’s ability to dissipate decentralized resources. The krill swarm algorithm is improved, combining inverse learning and Powell’s local search strategy to solve the optimization problem in dynamic reconfiguration. Simulation results show that the improved krill swarm algorithm improves the voltage of each node more than the standard krill swarm solution, and most of the node voltages can reach more than 0.95p.u. The simulation results show that the improved krill swarm algorithm can improve the voltage of each node more than the standard krill swarm solution. Comparing with the regulation strategy that does not consider the improved krill swarm algorithm for dynamic reconfiguration optimization of the distribution network and does not use the cluster voltage control model, the average wind abandonment rate, light abandonment rate, and total operating cost of this paper’s regulation strategy are 4.135%, 4.315%, and 620,000 yuan, respectively, which are the best performance among the three regulation strategies.
