Algorithm Research On Network Reconfiguration Of Ship Power System Under Fault Condition

The ship power system with fault or abnormal operation will result in equipment damage and even power interruption,which endangers the safety and reliability of ship.How to quickly restore the ship power network from electrical fault to a stable and reliable state,realize the self recovery of ship power system from network faults,has become an important task in modern ship power station.The essence of ship power network fault recovery is to build an automatic management system for power network to regulate,dispatch and control the power network under fault condition,minimize the impact of various faults on ships,and enhance the vitality of ships.In order to achieve self recovery of ship power system from network faults,topology analysis technology,network reconstruction optimization technology and system stability analysis technology are deeply researched in this paper.A topology analysis method for large ship power system is proposed.A data information model based on graph theory is established.The power and load are regarded as electrical nodes,the transformers,switches,cutters,grounding switches and other two terminal elements are regarded as the branches connecting to electrical nodes.A compact data table is used to store topological information.The topology analysis of ship power network is divided into two parts:the static analysis and the dynamic tracking.Based on the tree search method to traverse the whole network,and a set of topology information reflecting the inherent connection of the grid structure are generated.A node marking algorithm is used to calculate the connectivity of the affected bus,to realize the dynamic updating of the network topology information.A novel multi-swarm competitive particle swarm optimization algorithm based on single objective optimization technology is proposed in this paper to solve the problem of ship power network fault reconfiguration.We use the competitive relationship between sub-swarms,reinitialize the inferior sub-swarms which in stagnation with the help of chaotic mapping to improve the global searching ability.In order to cooperate with the multi-swarm competition mechanism,an adaptive inertia weight strategy is proposed based on the distribution range of the member history optimal location.the historical population location information is used to divided the population members into "exploration" and "Exploitation",which dynamically adjust the task of members to prevent the premature convergence and stagnation.The elitist learning strategy is adopted to help the members to improve the search accuracy.In the single objective reconfiguration of ship power network,weighted method is used to normalize each sub-objective to integrate them into a total objective function.The new algorithm can effectively provide a reliable reconstruction scheme for the network reconfiguration of the ship power system through an example test.Based on multi-objective optimization strategy,a multi-objective particle swarm optimization algorithm with variable neighborhood is proposed,and based on this algorithm,a multi-objective discrete particle swarm optimization algorithm is generated to solve the multi-objective reconfiguration for ship power network.The variable neighborhood strategy is proposed to help the algorithm to maintain a good diversity,and external archive,the variable neighborhood strategy is used to reduce the computational cost of the algorithm.In the discrete improvement of this Multi-objective particle swarm optimization,the variables are expressed in discrete form,the local optimal of each particle is selected from the external archive in a random manner.A multidimensional perturbation strategy is proposed to overcome the shortcoming that the single dimensional disturbance can not jump out of the local optima at some time by only one step transformation.Through a series of experimental tests,the remarkable advantages of the new algorithm are verified in terms of the convergence and diversity of the optimal solution set.The network topology of the ship power system is complex,and the transient stability problem is serious.A hybrid transient stability analysis method for ship power system is presented.Based on the coherency principle,the ship power system is equivalent to a two-machine system.Combining the time-domain simulation and the direct method,the energy function is constructed for ship power system,the quantitative analysis of the transient stability of the system is carried out.The validity of the proposed transient stability analysis method is verified by the simulation experiment.