Research On Distribution Network Reconfiguration With Distributed Generation Based On Harmony Search Algorithm

Renewable distributed generation has the advantages of clean, energy saving, green and environmental protection. In recent years, it has developed rapidly in the power grid. The output power is random which is limited by the natural factors. With the increasing proportion in the distribution network, it will bring serious challenges to the security and economic operation. In order to make safe and economical operation of the distribution network, the model and algorithm of distribution network reconfiguration and day-ahead optimization scheduling with distributed generation were studied in this paper.The correlation of distributed generation output is studied. This paper established the system power loss expectation value, the expected energy not supplied,and the voltage stability margin of the multi-objective distribution network reconfiguration model. Spatial correlation of wind and solar power generation is used nataf transformation to process. Distribution network probabilistic load flow is calculated based on Latin hypercube sampling Decimal harmony ring network coding strategy and the multi-objective harmony search algorithm is used to solve the model. The performance of probabilistic power flow and the effect of distributed power correlation on the distribution network reconfiguration are analyzed based on the PG&E 69-bus system. And the effectiveness of the algorithm and the model is verified by the simulation analysis.Aiming at the shortage of time point reconfiguration, a model of distribution system reconfiguration containing wind power is proposed to estimate the energy loss. Considering the time series variation of wind power output and load, in order to improve the efficiency of the model, K means clustering analysis of wind power output and load is used to estimate the system energy loss at given time interval. The premature judgment mechanism, crossover operator and adaptive dynamic parameters are introduced to improve the harmony search algorithm performance. Numerical simulation is on IEEE 33-bus and PG&E 69-bus system. The results show that wind power output and load can be a good clustering and clustering process can improve the reconfiguration efficiency. Compared to other algorithms, improved harmony search algorithm convergence performance and computation speed is better.Day-ahead optimal scheduling model is established by the minimum distribution network operation cost in power market environment, also considering the influence of the distribution network reconfiguration on the model. The switching cost and interruptible load compensation expense is considered in the objective function. The effect of distribution network reconfiguration, scenery prediction error and the renewable energy penetration are analyzed respectively. The example analysis of the modified IEEE 33-bus distribution network verified that the day-ahead optimization scheduling considering the distribution network reconfiguration can reduce controllable distributed generating capacity and interruptible load reductions.Also it is able to reduce the operation cost, so as to ensure the system security and economic operation.