Research On Distribution Network Multifault Repair Strategy And Emergency Management System Architecture

The power distribution network provides power to the energy consumers directly,which also plays an important role between power producers and users.Natural disasters,such as earthquakes,ice storms and typhoons,can cause multiple faults occurence and blackouts in the distribution network.If the prevention is improper or repair work is not timely,it will affect the normal life of people,the normal running of many important institutions like government departments and hospitals,the normal developing rescue work and cause huge loss of people's life and property.Hence,it is greatly significant theoretical value and strategic significance to propose an efficient multi-fault rush repairing strategy and propose a distribution network multi-faults emergency management system to reduce the power losses and the impact of blackouts and restore power supply rapidly.Based on this,this paper carries out researches on the multi-fault repair strategy and emergency management system of distribution network.The research content is as follows:First of all,considering the limitation repairing personnel and resources after multi-faults occur in distribution network,the paper establishes a deterministic multi-fault repair model to minimize outage losses and total repairing time.In order to improve the computational speed,storing external archives by improved adaptive grid,oriented mutation based on grid and adaptive external archive are proposed to improve the algorithm convergence performance and calculation speed to improve bacterial colony chemotaxis(BCC)algorithm,and BCC based on grid(GBCC)algorithm are obtained.Convergence of the algorithm is proved in theory.GBCC is tested by the standard test functions,and results are compared with other algorithms to verify the effectiveness of the improved GBCC algorithm.Finally,the discrete GBCC algorithm is used to solve the multi-fault repair strategy model of distribution network and simulation results show the effectiveness of the repair strategy and the fast convergence of the GBCC algorithm.Secondly,taking the uncertainty of fault repairing time into consideration,the paper proposes a distribution network multi-fault robust repair strategy model based on Petri net.Timed Petri net model with inhibitor arcs is used to describe the parallel discrete and dynamic repair process of distribution network.The translation from distribution network faults to Petri net model is defined and the faults' impact factor is obtained.Sensitive fault is defined to coordinate optimality and robustness of the strategy.Latin hypercube sampling method and backward scenario reduction technology are utilized to generate the scene samples considering fluctuation of sensitive faults repairing time.Discrete GBCC algorithm and the improved probabilistic minimax regret criterion are used to solve the problem.The robust repair strategy solving process is given.Examples are used to analyze the superiority of the robust repair strategy and simulation results show the effectiveness and robustness of the proposed method.Then,considering the uncertainty of fault repairing time and in order to improve the computational efficiency and flexibility,the paper proposes a two-stage distribution network multi-fault robust repair strategy model.The uncertainty budget is introduced to control the optimality and robustness of the solution.Linear technique is used to linearize the objective function.In the first stage,the optimal repairing strategy is obtained.In the second stage,the worst-case repairing time contingency scenario is realized in the predicted uncertainty set.The two-stage robust model is decomposed into a master problem and a sub problem,duality theory is used to transform the model into a mixed-integer programming linear model.The solving process of columa and constraint generation(C&CG)algorithm is given.The influence of the uncertainty on the repair strategy is analyzed and simulation results verify the effectiveness and robustness of the proposed method.Finally,to deal with various emergency events which will cause faults in distribution network,according to the functions and features of emergency management,a three-level general emergency management system architecture based on mulitagent is designed.The upper agent is responsible for leading and organizing the whole work in the lifecycle of the emergency management,four middle level agents provide solving method and host the work in different period of the emergency management,and the lower level agents implement the jobs belonging to the four middle level agents.Combined with the characteristics of distribution network multi-faults,distribution network multi-faults emergency management system architecture based on multi-agent fault is designed.A detailed analysis of the work process is presented and prevention,response and self-learning circulation process can be formed to deal with various emergency events in the distribution network.