Two-Stage Economic Dispatch For Output Adjustment Of Generstion Units And Frequency Constraints Considering The Randomness Of Wind Power

With energy and environmental problems become increasingly serious, new energy has been considerable development. As a kind of renewable and clean energy, wind power has become an important energy that can replace fossil energy. However, the random characteristic of wind power is a challenge for power system dispatch. Moreover, with the development of smart grid, forecasting information of radom variables can be obtained from different channels and the closer the operation-time, the more accurater information can be obtained. Therefore, in order to gain greater social benefits, it is particularly important to make many times decisions before operation-time and eatablish an integration model. In this paper, on the basis of multi-stage decisions, the two-stage economic dispatch integration model considering wind randomness and its calculation method are proposed.Due to the randomness of wind power and operation risk of power system, Conditional Value-at-Risk(CVa R) method is introduced in this paper, and then set up a two-stage economic dispatch integration model considering CVaR constraints. For solving the model, smoothing method is used to make the model into a continuous optimization problem, and then dual dynamic programming algorithm is adopted to solve the smoothing model. Numerical experiments using IEEE-30 nodes system for example, by setting different confience levels, different risk values, whether consider wind power and valve-point effect, we analyze the vary of total cost of system power generation. Furthermore, in order to ensure the system frequency can return normal in a short time after contingency and not lead to frequency crash, a two-stage economic dispatch integration model considering frequency constraints of post-contingency is established. In this model, the penalty costs of overvalued or undervalued the available output of wind turbines are added and join the frequency constraints in the second-stage. Due to the non-smooth characteristic of model, the smoothing function method is also used to make the model smooth, and IEEE-30 node system is used to validate the proposed model. We analyze the effect of frequency constraint and wind power punishment on operating costs, the effect of available output of wind turbines on the output of conventional units in stage one, etc. The results verify the efficiency of the new model.The study of problems in this paper integrated current power system operation environment, fully considered the much information conditions in power system smart run environment, applied the ideology of multi-stage decisions in system economic operation, proposed new models and algorithm are using effective ways in mathematical and economic area. The methods in this paper can effectively solve the economic dispatch problem of post-contingency and provide a scientific method for the optimal operation of the system optimal decisions.