Based On Multi-Resource Constrained Reactive Scheduling Distribution Network Fault Repair

Project scheduling theory is a new project schedule management tools and resource optimization methods. Compared to traditional project schedule management method, project scheduling theory takes into account resource constraints and task execution mode, the model is more close to reality, scheduling is relatively reasonable. In the research of project scheduling, this paper discusses the reactive scheduling of the distribution network fault repair model and algorithm, and improved the genetic algorithm. In the conventional multi-mode resource-constrained project scheduling (MRCPSP) study, a task can be performed by multiple mode, resource requirements is different under different execution modes, uninterrupted completion of task scheduling, resource and its duration are fixed. Due to the presence of lead in the construction process, the risk of interruption of the task, resource use and the duration of the plan does not match in distribution detwork repair process, so the project should have a dynamic scheduling scheme of control methods. Reaction scheduling can dynamically adjust the project of distribution network fault repair plan based on resource consumption and schedule changes.The problem of distribution network repair is an important part of distribution maintenance. When large-scale distribution network broken, effective and timely repairation to the point of failure is essential. Therefore, this paper introduces project scheduling theory which originates from project management theory, uses a phased and repeatedly repair and restoration method to repair distribution network, aims at shortest repair time and minimal loss of the loss of power, finally, finds a reasonable way to solve the model of fault repair scheduling problem of distribution network. Because of between each point of distribution network fault repair tasks with no logical relationship, Firstly, each fault repair tasks of distribution network is setted a repair priority. And according to priority of fault repair task, each phase selects some avalible task, setting a logical relationship between the fault repair tasks. Secondly, this paper considers the problem of repair accompany with recovery. At the end of each stage of repair, switching operation makes the outage region to restore power. As all the the fault repair tasks are assigned to different stages, established Node Network of fault repair work of distribution network, So that the model is avalible to project scheduling theory. Therefore, this paper put forward the reactive scheduling problem of distribution network fault repair. When the initial plan does not comply with the repair time, adjustment the successor fault repair tasks, and making work of distribution network repair is still the optimal. In the choice of algorithm, this paper use a more mature and improvement GA, generating the initial population by using serial scheduling mechanism. Cross-way between individuals is multi-point crossover. Variation consider the logical relationship between tasks. Finally, the use of numerical example shows the rationality of the model and algorithm.