Unit Commitment For Wind-Thermal Hybrid Power Systems

With the development of wind power, the proportion of wind power in the power grid is increased constantly. Because of the dynamic and stochastic characteristics of the wind power generation, the influence of wind power in economic dispatching of generating units should be taken into consideration, and the conventional methods are no longer suitable for power systems with integration of large-scale wind farms. The wind power presents a new challenge for the optimal scheduling problem.In this thesis, the development and current research status of the optimal dispatching of generating units are reviewed, and the impact of wind power is analyzed. Due to dynamics and stochastic of the wind farms, a stochastic model for the economic dispatch in unit commitment is presented based on the chance constrained programming algorithm by considering the technical conditions of the system operation characteristics. The algorithm is modified based on equivalence principle. It describes the related constrained conditions by the form of probability, and stochastic simulation technique is introduced to evaluate the constraints.Two algorithms are presented to solve the unit commitment problem. Dynamic programming is first adopted.Based on the Bellman optimization principle, the algorithm can solve the optimization problem containing multi-stage decision process.A complicated problem can thus be divided into several stages and solved in turn. It can find the global optimal solution and guarantee the convergence. The second algorithm is a hybrid intelligent algorithm.The outer layer is unit commitment solved with the quantum-inspired binary particle swarm optimization, and the inner layer is load dispatching solved with quadratic programming. The quantum-inspired binary particle swarm optimization is improved based on greedy mutation strategy, parts of the individuals are mutated through quantum bits to a certain algebra. It is easier to get the global optimal solution compared with the traditional quantum-inspired binary particle swarm optimization algorithm.The algorithms are tested for a system consisting of a wind farm with10units. Scheduling schemes are solved for two situations, with and without considering the generation rate constraints, under different degree of confidence. In the case of stochastic of wind power generation, the computation results verify the feasibility and effectiveness of the proposed model and algorithms, taking into account both the economy and reliability of the wind-thermal hybrid power system.