Multi-Objective Allocation Of Distributed Generation In Distribution Network

In the requirement of energy saving and environmental protection policy and the strategy of sustainable development, distributed generation has been widespread concern with the characteristics of environmental friendliness, flexible installation and economics of power supply. It brought a lot of good effects when the distributed generation was integrated into power distribution reasonably, such as reducing the load peak and off-peak and the losses of power grid, relieving the congestion of power transmission and distribution, increasing the reliability of power supply, improving the voltage stability of system. In this paper, on basis of analyzing the characteristics different types of distributed generation and researching the impacts of the distributed generation integrated into power distribute network, to comprehensive costs, network power loss, node voltage deviation as objective function, an optimal allocation model of multi-objective power distributed generation was established.Aiming at solving the problem that the global optimal particle is difficult to determine w hen the particle swarm algorithm is applied to multi-objective optimization, an improved mult i-objective particle swarm algorithm was proposed. In order to solve the problem of selecting the global optimal particle, the optimal guidance strategy was applied, and the fast non-dominated sorting strategy, external elite archiving strategy and small disturbance variation mechanism were introduced into the particle swarm algorithm. Consequently, the global convergence of the algorithm and the convergence speed are improved as well as the diversity and integrity of the Pareto optimal solution set are maintained while ensuring the population to move in the Pareto optimal front.Taking power distribution network of IEEE33nodes and the test system of69nodes as example, the model is solved by the improved multi-objective particle swarm algorithm and non-dominated sorting genetic algorithm Ⅱ. The IMOPSO is better than NSGA-II in the convergence speed and global convergence, and it’s effective and feasible for IMOPSO to be applied to solve the model.