City Power Grid Optimization Planning Problem

The electrical power system planning research usually includes the power source planning and the electrical network planning. The urban electrical network planning may further be divided into the transmission network planning (principal network planning) and the distribution network planning. According to the actual planning work, the power source (mostly transformer substation) locating question and the transmission network planning question both are important and difficult. Carrying on the optimization planning to these two questions will greatly enhance the economy, reliability, and security of the whole electrical network.Aiming at the special question of transformer substation locating, this thesis proposes a two-phase optimization planning approach to multi-goal substation locating. This approach is practical and its algorithms are simple. In the first phase, a practical and scientific mathematical model is established, and it carries on the optimization planning using the computer program. We can achieve the most economic solution as the initial solution for the further planning. In the second phase, the experts will give several candidate solutions for the initial solution. Then valuing their geographical fitness based on the analytic hierarchy process(AHP) respectively. Finally, find the resultant solution with highest comprehensive fitness degree, optimize the whole network.Meanwhile, a model for transmission network planning in power system with GA is developed in this thesis. The model uses an evolutionary computation technique to find out the most economical scheme for network expansion while satisfaction with the stability and security constrains. By comparison of popular evolutionary techniques, it is concluded that genetic algorithm is one of the best choices until now. To overcome the disadvantages of conventional genetic algorithm, accelerated genetic algorithm is suggested. The need for Advanced Genetic Algorithm leads us to the virtual population for the formation of high quality chromosomes in a population. This accelerated genetic algorithm searches the solution space in the neighborhood of the best candidate solution to produce solutions closer to the global optima.