Optimal Methods Study On Transmission Network Expansion Planning

Electric power is the important motive force of modern society. By reasonable planning great social and economic benefit can be attained. There is immediate significance to carry out the research of power system planning.With the fast development of modern power system, more and more goals are becoming planning objectives. Traditional single objective planning cannot meet practical planning demands. Based on the DC power flow model, a practical single-stage power planning model with multiple objectives is set up in this thesis. The "N-1" check is also considered in the model to fulfill the basic secure operation requirement. Conventional optimization techniques have difficulty in dealing with complex constraints and often result in "dimension disaster". So, genetic algorithm is used in transmission network planning. Furthermore, a new multi-population genetic algorithm based on standard genetic algorithm is adopted to solve the problem of enclosed competition. Multi-objective problems convergent speed is fastened by the method. Satisfied results and fine flexibility are seen in numerical examples by compound use of the model and multi-population genetic algorithm.Transmission long-term planning is a dynamic, multi-stage and nonlinear combinative optimization problem. The thoughts of how to carry out multi-stage transmission network planning are analysed. Based on a mathematical model which conforms to constraints, and has minimum discounted value of the sum investment and loss-cost in whole planning period, a new binary coding genetic algorithm is presented to solve multi-stage planning problem in this thesis. The method is proved to be efficient by test example.Many uncertainties such as power supply and fluctuating load still exist even if a certain transmission network structure has set up in planning period. For the purpose of getting more system information, fuzzy theory is adopted in power flow analysis. By DC fuzzy load flow methods, a numerical example is tested. All the works above lay the foundation of power network flexible planning.