Research On Network Reconfiguration Model And Algorithm Of Ladder-shaped Shipboard Power System

Shipboard power system network fault reconfiguration is a significant part in shipboard energy management, as well as an important approach to restore power supply and improve shipboard survivability. In SPS, the aim of network reconfiguration is to isolate faults and restore power supply through changing the statuses of switches in network. In this process, load flow calculation is a basic sub-problem of network reconfiguration, which needs to conduct power flow calculation according to the status of branch switches. The task of load flow calculation is to calculate the power flow distribution of the running system. As the topological structure of SPS getting more complicated and its capacity continuous growing, ladder-shaped network is applied more general. This type of network is an inevitable choice through developing of SPS in spite of its complex scheduling algorithms. Therefore, it’s a research hotspot as well as a research challenge to design load flow calculation and network reconfiguration model and optimization algorithms suitable for ladder-shaped SPS.This paper first gives a detail description of network reconfiguration problem in SPS and introduces ladder-shaped shipboard power system structure. Then mathematic models of devices in SPS are given. After that, this paper established integrated mathematic model of ladder-shaped SPS.Then, this paper introduces several classical methods and their application in ladder-shaped SPS. Next considering the structure of ladder-shaped SPS, this paper proposed a parallel load flow algorithm. This algorithm divides the entire system into concentrated supply network and distribution subnets, and allocate different and independent computing element to every division. Node potential method and forward-backward method are respectively applied in concentrated supply network and distribution subnets.Then, aiming at network fault reconfiguration in ladder-shaped SPS, this paper tries to solve it using 0-1 knapsack problem with packing constraints, which is a typical algorithm in dynamic programming algorithms. In particular, this paper takes supply capacity of SPS as knapsack, loads as goods, and power consumed by loads is regarded both as weight and value of goods. What’s more, this paper improves the application of PSO algorithm in ladder-shaped SPS. In order to express loads which are supplied with multiple paths, this paper applies multistage coding scheme and lowers dimension of particles. At the same time, variables in PSO algorithm are also optimized here. This paper conducts simulation experiments to evaluate proposed algorithms based on typical multi-station cases. And results show that proposed algorithms has strong searching ability and faster convergence speed.Finally, this paper also introduces the system design architecture and main technologies based on proposed parallel load flow algorithm and discrete PSO reconfiguration algorithm.