Font Size: a A A

Research On Load Classification And Hierarchical Management Of Cruise Integrated Power Network Based On Multi-Agent

Posted on:2020-11-29Degree:MasterType:Thesis
Country:ChinaCandidate:C M WangFull Text:PDF
GTID:2392330590950983Subject:Power electronics and electric drive
Abstract/Summary:PDF Full Text Request
The structure of the cruise integrated power grid is complex and the load is various.It is difficult to meet the requirements of the cruise ship grid by the traditional centralized control management.It is urgent to introduce new technologies to control and manage the cruise ship grid.In recent years,the rapid development of multi-agent technology,combined with its characteristics of distribution,adaptability and coordination,will become an important means of cruise ship grid management.During the operation of the cruise ship,the operating conditions are complex and changeable.The traditional load classification and classification method has certain limitations and cannot adapt to the diverse characteristics of the cruise ship operating state.In view of the characteristics of large load randomness during the operation of the cruise ship,this paper simulates the load usage under various working conditions according to the statistical idea,and finds the load probability distribution function under each working condition.According to the probability distribution function under each working condition and using sampling method to simulate the stochastic load scenario of cruise ship's whole voyage,then classify the stochastic load scenarios,and optimize the storage capacity and energy management of integrated power grid according to the classification results.Finally,according to the load importance of different loads under different working conditions,the load is re-classified,then the impact of load grading in energy coordination is considered,and multi-agent technology is introduced at the same time,and it is on the JADE platform for the cruise ship grid.The various Agent functions are designed.Through the establishment of a MAS-based cruise ship grid distributed control structure,and the development of a corresponding hierarchical management strategy,the energy of the entire cruise ship grid is coordinated.The specific research includes the following aspects:First,establish a cruise ship grid model and conduct a detailed analysis of the power load of the cruise grid.Then,the comprehensive grid load is reclassified and classified according to the normal sailing conditions,inbound and outbound,and parking conditions of the cruise ship,and the usage probability of the load under each working condition is counted;Secondly,based on the statistical power and usage probability of each load under various working conditions,the histogram of load frequency distribution under each working condition is obtained based on statistical thought,and then the load probability distribution function is fitted by Fourier series.According to the fitted function,a sampling method is used to generate a large number of random load scenarios that can represent the entire navigation process;Again,the generated multiple sets of load random scenes are classified.Based on the comparison of several fuzzy clustering algorithms,the AM-PCM(Automatic Merging Possibilistic Clustering Method)algorithm is selected for clustering.Then,based on the large dimension of the load scene data,the AM-PCM algorithm is improved.The Adaptive Multi-Clustering Algorithm(AMCA)is used to verify the AMCA algorithm.The results verify the effectiveness of the algorithm.Then,due to the randomness of the load during the operation of the cruise ship,it is necessary to allocate a certain capacity of energy storage capacity.Before the energy optimization management,the storage capacity of the cruise ship is optimized according to the classification result of the load,and the capacity of the energy storage system is established.The model is optimized,the results of clustering are substituted into the model,and the multi-objective particle swarm optimization algorithm is used to solve the optimal energy allocation of the energy storage system.Then,the cruise ship grid energy management optimization model is established to optimize the navigation status of the cruise ship within one day.The optimal energy storage capacity is substituted into the model,and the model is optimized by the multi-objective particle swarm optimization algorithm and compared with the navigation state before optimization.The results show that by adjusting the speed of the cruise ship within one day,not only the efficiency of the cruise ship is improved,but also the emission rate of nitrogen dioxide is reduced.Then,due to the previous energy management of the cruise ship,the impact of load grading was not considered.In this paper,by introducing multi-agent technology,the multi-agent distributed control structure is used to hierarchically manage the cruise load.According to the characteristics of the cruise ship network,the functions of each agent in the cruise ship network and the communication between the agents are designed on the JADE platform.The communication between Simulink and JADE platform is realized by the MACSim JX interface program,and the data interaction between the two platforms is realized.Finally,by introducing the distributed control theory of multi-agents,the impact of load grading on cruise energy management is considered,and corresponding coordinated control strategies are formulated.The distributed control structure of MAS-based cruise ship power grid is constructed.The multi-agent autonomy,distribution,adaptability and coordination are used to dynamically manage and control the load and power supply in the cruise ship network.Through the coordinated control between the power sources in the power grid and the hierarchical control of the load under various working conditions,the supply and demand balance of energy in the power grid is maintained.
Keywords/Search Tags:Cruise integrated power grid, Multi-Agent, Multi-Objective Particle Swarm Optimization, Energy management, JADE development platform
PDF Full Text Request
Related items