Dynamic Restoration For Distribution Network Based On Multi-agent Theory

Distribution service restoration is a complicated multi-objective, multi-period,scattered, nonlinear combination optimization problem. the essence of it is a set ofswitches action combination which is obtained under the restriction of satisfying thedistribution network connectivity, radiation and the load ability of equipments, achievingtargets such as rapid recovery of load power supply, minimizing switch operations,obtaining the minimum network loss and so on. On the basis of extensive reading ofrelevant domestic & international literature, this paper does the research on dynamicrestoration during multi-fault rush repair and smart distribution restoration by using theimproved ant colony algorithm based on the theory of spanning tree and the multi-Agenttheory. The main contents include:First, based on the further researches of network topology and power flow calculation,distribution network is equivalent to an undirected graph by using graph theory; it hasanalyzed the principle of the Depth-First Search method and the Breadth-First Searchmethod, combines both Breadth-First Search method and back/forward sweep algorithm,then forms the improved back/forward sweep algorithm, meanwhile, its process is given.Then, studying the dynamic restoration for the multi-periods during the multi-faultsrush repair in distribution network, a bi-level optimization model aimed at maximumrecovery loss of electrical power and minimum network loss is put forward. Restoration ofvarious periods using different agents is responsible for the implementation of distributedparallel computing independently; In the course of coordination, the use of interruptibleload ensure the priority of critical load and reduce the number of switching operations; Inaddition, the simulation that carried on by the improved ant colony algorithm based on thetheory of spanning tree in the static restoration verifies the validity of the proposedmethod.Finally, further research on smart distribution service restoration are taken,multi-agents system for the smart distribution service restoration is built as well, then aintelligent restoration framework and method based on multi-agents theory is proposed, and the strategy of self-diagnosis, self-healing and self-optimization is given. Meanwhile,the improved ant colony algorithm based on the theory of spanning tree is proposed asoptimum strategy of network optimization agent in.