Research On Wind Turbine Selection And Comprehensive Evaluation

With the rapid development of wind power industry, the global wind turbines installed scale continues to increase. Grid wind power began to take off from the 1980 s, after more than 30 years of efforts, China’s wind power business has achieved high grades. With the growing number of wind turbine machine manufacturing enterprises, the technical route and the performance parameters of different wind turbines also emerge in endlessly. How to reasonably select the wind power turbine in order to make the units and wind resources of the wind farm area in the most optimal matching, optimal the wind turbine operation performance, require the highest return on investment, get timely maintenance and run for a long time, has became the core subject for researchers.In this paper, synthetic evaluation on wind turbine type selection are regard as the goal. Comprehensively considering the indicators related to wind turbines during the construction and operation of the wind farm, these indicators are reasonably classified into 5 categories, including technical performance, wind farm adaptability, economic performance, operation performance and product technical service, which also contains more subordinate indicators, and the basis and significance of each indicator selection have been detailed analyzed. Then, the selection and comprehensive evaluation index system of wind turbine is established. The BP neural network based on particle swarm optimization method are introduced in constructing the wind turbine selection and comprehensive evaluation model. The analytic hierarchy process and expert scoring are used respectively to determine the weights of the indexes and the indexes’ score. Then, the particle swarm algorithm are applied to optimize connection weights and threshold of BP neural network. The optimized connection weights and threshold, the network of all kinds of initialization parameters and wind turbine samples, are all plugged in to the BP neural network for training and testing. The empirical analysis verifies that the BP neural network model based on particle swarm optimization to compare accurate evaluation of comprehensive performance level of the wind turbine, is superior to the traditional single BP neural network model, which can be applied in the actual selection comprehensive evaluation problem of wind turbines.