Power Management Optimization In All-Electric Ship Micro-Grid

As shipping industry consumes a great of fuels,it will make great sense to improve operation efficiency while decrease carbon emission and energy consumption by adopting new technology.With the rapid development of renewable energy,artificial intelligence,power electric and so on,all-electric driven combining renewable energy ship becomes the important way to decrease carbon emission and energy consumption and improve the performance of operation in ship.For ocean navigation,the load fluctuates with the operation changes and the output of renewable source can make impact on power grid.So this paper does lots of research about the power management for a real all-electric solar-diesel-ESS(energy storage system)ship.This paper can be summarized as follows:(1)This paper deeply analyzes the load,the route and operating conditions of a real all-electric solar-diesel-ESS ship.Based on the load data and statistical analysis,the load probability distribution function is obtained by Fourier series fitting.Then,the hourly solar radiation intensity probability distribution function which is lognormal distribution along the route can be obtained using data from Geo-Model Solar company weather station.(2)This paper builds the ship micro-grid model including diesel generator,photovoltaic system and ESS.Then this paper builds the optimization model for ESS sizing by taking the cost and CO2 emission as the objectives for every random scene and gets the final optimal size by adding the results of each random scene with its weighting to fully consider the influence of power fluctuation.In addition,based on the relationship between propulsion power and ship speed in regular cruising,this paper establish the model of ship power management to optimize the speed arrangement and the dispatch of generator unit.(3)This paper uses the probability distribution function to generate enough random scenes by slicing sample method and proposes a random optimization method with clustering: The random scenes could be clustered into several clustering centers which can represent the characteristic of all scenes to decrease computing time.As the existing clustering algorithm performs badly for high dimensional data,this paper proposed a multi stage adaptive clustering algorithm(Adaptive Multi-Clustering Algorithm,AMCA)to solve the problem in this paper.It proves that the AMCA algorithm has good clustering performance by test,and can reduce the computing time.(4)This paper calculates the model of optimal sizing for ESS using multi-objective particle swarm algorithm(Multi-Objective Particle Swarm Optimization,MOPSO)with the data of clustering centers and gets the final optimal size by adding the results of each clustering center with its weighting.Compared with the situation without ESS,the optimal ESS can effectively reduce the cost and CO2 emission.Then the model of power management also is solved using MOPSO and the comparison between the optimal result and the result in fixed speed proves that the proposed method can improve the operation efficiency.