Network Reconfiguration Of Distribution System With Distributed Generations

In recent years,distributed generation(DG)technology develops rapidly due to its’ flexible and environment-friendly nature.However,when a large number of DGs are connected to the system,their influence on the reliability and economy of distribution network cannot be ignored.The management of the DGs that aims to enhance the economics and environmental effect of the DGs has been a research focuses in smart distribution network.As an effective optimal operation measure in intelligent distribution network,distribution system reconfiguration is realized by changing the combination of states of section switches and tie switches to improve the economy,safety and distributed power accommodation capacity of the distribution system.The intelligent algorithm and heuristic algorithm of static network reconfiguration,and the dynamic reconfiguration algorithm are studied in the thesis.In the survey of intelligent algorithm of static network reconfiguration,the genetic algorithm is applied to solve the reconfiguration problem of distribution system with wind power generators.The strategy of searching random spanning trees is used to ensure the radial structure of network in distribution network reconfiguration.In addition,the forward and backward sweep method is applied to calculate the power flow of the distribution system with wind power generators.The typical IEEE33 test system is used to illustrate the validity of the proposed approach,and the reconfiguration results are compared among different power output of wind turbines.Simulation results show that the proposed reconfiguration method can reduce the power loss and improve the voltage quality of distribution system.In the survey of heuristic algorithm of static network reconfiguration,a quick reconfiguration algorithm based on vector shifting is proposed.According to the simplified flow model,the power vectors and resistance vectors which are utilized to calculate the variations of power loss in branch-exchange are formed.The new vectors of power and resistance after each branch-exchange are obtained by simple shift operation.In the process of network reconfiguration,vector operation rather than load flow calculation is used to get the increment value of active power loss after each branch-exchange.The variation rules of power loss during branch exchanging are obtained by analyzing the expression of power loss increment.The simulation results of the IEEE33 and IEEE69 test systems show that the vector shifting based distribution network reconfiguration algorithm is effective.In the survey of the dynamic network reconfiguration algorithm,a co-evolution algorithm based dynamic network reconfiguration algorithm is proposed.First of all,the natural time periods are merged into the reconfiguration interval based on the constrain of action number of switches and the variation of loads.Then the hash table based genetic algorithm is used for static reconfiguration in each interval.Finally,according to the solution set obtained from the static reconfiguration,co-evolution algorithm based dynamic reconfiguration is operated to further reduce the operation cost of distribution system.The simulative results of the IEEE33 system proved that the proposed dynamic reconfiguration algorithm can obtain effective reconfiguration schedules with less computation.