Multi-objective Optimized Reconfiguration For Active Distribution Network Based Stochastic Load Flow

As an advanced stage of smart grid technologies, active distribution network can actively control some of the distributed generators and other end-use energy systems in network in order to achieve the purpose of optimizing the distribution network operation. The impacts of distributed generation penetration on distribution network operating are not only in terms of changing loss, but also in other aspects such as power quality, voltage level and planning of distribution network. To ensure the safety and reliability of the grid while maximizing the benefits of distributed power access, apart from actively controlling distributed generators, adjusting the status of the switch in the network using method of distribution network reconstruction is also needed to combine with. Under the concept of active distribution network, make further optimizing on the network topology will better improve the economics of grid operation and efficiency of distributed generation in condition of distributed generation penetration.Common methods of building distribution network reconfiguration model, commonly used basic algorithm, the concept of active distribution network and its relationship with distributed generation are introduced in this paper. The paper also introduces several common forms of distributed generation and the impacts of distributed generation penetration. This paper gives output expressions of wind power and photovoltaic power generation and lays special stress on inaccuracy and randomness of output forecast wind power and photovoltaic power generation.The method of power flow calculation with distributed generation is discussed and the distributed generation can be of two types:output-controllable DG and output-uncontrollable DG. Due to actual forecasts of loads and output of uncontrollable DG are not very accurate, using method of 2n+l point estimate stochastic load flow in the model enables the results of optimizing to adapt better to uncertain conditions.Combined with multi-objective optimization theory this paper proposes an active distribution network reconfiguration model, and expected value minimization of total loss and loss minimization of distributed generation are optimization objectives. The multi-objective optimization problem is solved by solving Pareto solution set. At last, simulation with IEEE-33 bus system are performed which using multi-objective differential evolution algorithm. The test results show improvement of power supply capability and voltage level of each node in condition of distributed generation penetration, and verify the correctness and effectiveness of the reconfiguration model.