Power network planning is a multi-stage dynamic optimization process,in essence,is to determine a dynamic investment program of the electrical equipment required to be elected in the various planning stages.Therefore,Dynamic power network planning is a complex,large-scale discrete combination optimization problem,and it has a non-linear,discrete,multi-modal,multi-objective and the objective function cannot be differential etc.characteristics.For this,the artic discusses the basic principles,theories and methods of dynamic network planning,introduces its the optimality theorem and the basic thought.And,on this basis,the artic presents the basic operation process and form design of dynamic planning.Meanwhile,it applies the electromagnetism-like mechanism(ELM)to Multi-objective Dynamic Power Network Planning.According to the dynamic of power network plannig,the artic establish a dynamic planning model which can describe the power grid economy,reliability,N-1 static of safety objectives.But these goals is contradictory,so the paper uses the weighted method and the penalty function method to transform these target functions into the minimum cost single target model in power grid planning,and then unified measure to find the merits of Power Network Planning Scheme.this mutil-objective linear weighted coordination method can improve the overall satisfaction of the planning scheme.In the trial,a 19 nodes system and a 46 nodes system will be as the research object and test,and they will be coded with decimal system.ELM algorithm has obvious advantage in searching efficiency and fault-tolerant aspects compared with that of GA and AA.The results show that ELM algorithm can effectively solve the large-scale combinatorial optimization problems such as network planning with discrete variables.This is not only for the solving of the problem open up a new way,but also promoted the development of network planning method based on artificial intelligence methods. |