The Theory Of Stochastic Programming In Applied Research, Wind Power System Analysis

As one of the most matured renewable energy technologies, wind power generation is confronted with an unheard of developing opportunity when the global environment concern and fossil energy shortage change the traditional electric power industry from growth to quality. Different from conventional generators, wind power is an intermittent energy source, and its stochastic characteristic has a great impact on the traditional power system's operation. Electric power utilities' concern about the detrimental effects due to large-scale wind power integration has impeded wind power's development. As a result, It become more and more necessitous to study the wind power penetration level allowed by main system and the economic load dispatch in wind power integrated system.In past studies, dynamic simulation is often used to analyze the influences of wind power on the power quality, security and stability of system. Among a set of optional wind powers, the largest one without detrimental effects is usually regarded as wind power penetration limit. There're also reports on the adoption of probability methods to evaluate the penetration limit. However, these methods are not systemic, and the calculation efficiency mainly depends on researcher's experience and knowledge about the studied system. In essence, wind power penetration limit is the maximum wind power allowable by system subject to various technical constraints such as bus voltage magnitude, transmission capability, system frequency deviation, spinning reserve, etc. Naturally, an optimization method is proposed to calculate the penetration limit in this paper, and random variable is introduced to describe the wind power's stochastic characteristic in details. According to stochastic programming theory, a Chance Constrained Programming (CCP) model is used to tackle such optimization problem containing random variables.The well-balanced operation of integrated wind farm needs conventional units' match accompanying by the change of generation costs. Since wind farm's outputs are stochastic, the costs are also random variables. Obviously, economic load dispatch in wind power integrated system goes beyond the traditional category, a new algorithm isneeded. In order to minimize generation costs by dispatching loads among the conventional units and wind farms, an optimal real power flow algorithm based on CCP is presented out in this paper.To overcome the difficulty in solving CCP problem, a genetic algorithm based on stochastic simulation technique is applied to calculate the wind power penetration limit and optimal real power flow. Although more time-consuming than traditional optimization method, it's a feasible solution for complicated CCP problem. The test results on IEEE30 system demonstrate the effectiveness of the algorithm.Xinjiang power system is a medium-sized network connected with Dabancheng, the largest wind farm in china. It is representative to analyze wind power integration problem in Xinjiang. To verify the practicability of proposals in this paper, some studies about wind power penetration limit and optimal real power flow in Xinjiang system are conducted.