Development of power flow with distributed generators and reconfiguration for restoration of unbalanced distribution systems.

摘要

With the increasing interest in distribution automation, distribution power flow is important for applications like VAR planning, switching, state estimation and especially optimization. Typically, a distribution system originates at a substation and continues to a lower voltage for delivery to the customers. There are several tools for transmission system analysis. These tools include Newton Raphson, Gauss Seidel and fast decoupled techniques. These techniques however sometimes fail to converge when applied to distribution systems due to their higher resistance/reactance (R/X) ratio of the lines, making them ill conditioned. Distribution systems typically have a radial topological structure where the loads are not always constant power. With the increase in distributed generation (DG) there is a critical need to develop analysis tools to study the effect they will have on the distribution systems. Also, shipboard power systems are different from terrestrial distribution systems, as they are tightly coupled and have multiple generators. This dissertation focuses on developing a software program to perform the power flow analysis of terrestrial as well as shipboard power systems. Components are modeled considering the mutual coupling of cables and the tightly coupled nature of the ship systems. The algorithm is built and tested on IEEE test cases. The distributed generator is modeled as both a PQ (constant power factor) and a PV (constant voltage) node.; This dissertation also focuses on reconfiguration for restoration of unbalanced distribution systems. Reconfiguration is changing the status (OFF/ON) of switches and reconfiguration for restoration is changing the switch status to maximize the supply to loads that are left unsupplied after fault removal. Methods exist for restoration of distribution systems and can be categorized into heuristics, knowledge based, metaheuristics and intelligent techniques. However, the application of these methods have not considered the unbalanced nature of distribution system operation with mutual coupling. The restoration in this dissertation is achieved using optimization with multiple objectives; that of maximizing the load giving priority to vital loads and minimizing the number of switch operations. Also a restoration scheme for shipboard power systems with an IPS and distributed generation has been developed. Restoration with possible islanding is demonstrated.