Research On Group Search Optimization Algorithm In The Application Of Mid-voltage Distribution Network Planning

Today, it is lower level for distribution system development in our country. With the speeding up of urbanization process and the continuous expansion of new district construction in many cities, the existing distribution grids has been lagging behind the development of urban economy. Reasonable distribution network planning can effectively solve many currently existing problems in urban distribution grids such as weak network configuration, high line loss, and unreasonable planning, and etc.On the analysis of the problems of some commonly-used algorithms, this dissertation establishes the mathematical model of distribution network planning taking economy as the aim with various constraints considering, and proposes a new random search algorithm, named as group search optimization algorithm, to be used to solve the problem. In the study and analysis of group search optimization algorithm, to aim at some flaws such as single optimization way of the followers, and easy to be trapped in local optimum, and etc, this dissertation proposes an improved group search optimization algorithm where strategy of fish swarm optimization mechanism is introduced into, and then applied to solve the distribution network planning problems. The simulation results show the superiority of the proposed algorithm. The main content is below in this dissertation.(1) To investigate and analyze some commonly-used algorithms currently existed on distribution network planning, and give more attention to the shortcomings and deficiencies of these algorithms, and point out necessity and possibility to propose a new random optimization algorithm for solving the distribution network planning problem.(2) To propose an improved power flow calculation method based on comparison and analysis on diverse power flow calculation methods, and apply it to solve the distribution network planning problems. The results verified by examples show that the proposed method is efficiency and precise.(3) To conduct deeper investigations and improvements on group search optimization algorithm as follows. To aim at a fact that the group search optimization algorithm is easy to fall into the local minimum, the artificial fish swarm algorithm is introduced into by setting of the scrounger searching to enhance the global searching ability. Six benchmark functions are used to evaluate the performance of the proposed method. Experimental results show that convergence speed of the improved group search optimization algorithm is faster and more efficient than traditional group search optimization algorithm.(4) To establish the mathematical model of distribution network planning with aim function including grid loss and construction investment, and apply the improved group search optimization algorithm to solve the distribution network planning problem. The simulation resultsshow that the improved group search optimization algorithm is much faster and more efficient than group search optimization algorithm and artificial fish swarm algorithm through the simulations and comparisons on two examples with artificial fish swarm algorithm in distribution network planning.