Optimal Allocation Of Distributed Generation With Multi-Objective Considering Uncertainties

Distributed generation (DG) has the characteristics of flexibility, economy and environmental protection. Reasonable placement of DGs can bring many advantages to the network, such as network loss reduction and voltage quality improvement.The development of DGs is one of the best practices to balance the conflict between energy supply and environmental protection. However, as large-scale renewable DGs and plug-in electric vehicles (PEV) integrated into the distribution network, uncertainties such as PEV charging power, stochastic wind speed and illumination intensity could bring potential risk to the safe operation of distribution system. Thus, it is of vital importance to consider the uncertainties above in DG optimal allocation problem to adapt to the development of active distribution network.Firstly, this paper expounds the influence of uncertainties of renewable energy generation and PEV charging power on distribution network, including the impact on voltage level, network loss, system reliability and distribution network planning. Meanwhile, to ensure the planning results be consistent with the actual situation, the probabilistic models of wind power generation, photovoltaic generation and PEV charging power are established according to the random characteristics of renewable energy DGs and PEV charging behavior.Secondly, after introducing the basic conception and characteristic of Chance Constrained Programming (CCP), a multi-objective optimal DG allocation (ODGA) model considering environmental benefit, total DG cost and power loss is established under the CCP framework, wherein three types of DGs including wind power, photovoltaic and micro-turbine are taken into account, and voltage amplitude and line transmission capacity are set as chance constrains. In the proposed mathematical model, the comprehensive objective function is composed of three sub-objective functions by means of weighted method. The weighting coefficients for sub objectives are determined by a multi-objective comprehensive evaluation method called Analytic Hierarchy Process (AHP).Thirdly, this paper proposed a new approach named correlation Latin hypercube sampling Monte Carlo simulation embedded modified crisscross optimization algorithm (MCSO-CLMCS) for solving the optimization model. The method firstly used CLMCS to calculate the probabilistic load flow (PLF) according to the probability model of renewable energy generation and PEV charging power. Then the constraint conditions were checked and the objective function value was obtained using the result of PLF. Finally, the optimal planning scheme was obtained by MCSO algorithm with self-adaptive mutation mechanism.Finally, in order to validate the feasibility and effectiveness of the proposed model and method, simulation tests are conducted on the typical IEEE 33-bus distribution system and PG&E 69-bus distribution system. The results show that the proposed model and method can obtain reasonable ODGA schemes, which can improve the security, economy and environmental benefits of distribution system operation. Simulation results also show that the proposed MCSO-CLMCS algorithm has high convergence precision and computing speed, and is suitable for solving complicated multi-objective DG optimization problem.