SIMULATION OF THE WIND GENERATOR. STUDY WIND SPEED CONTOUR

Today, there are several ways to find and maintain the maximum power value. This may be as already built-in controllers on the search engine power value (or MPPT controllers), as well as separate intelligent systems (with phase-logic or neural networks). An example of using a neural network for finding a point of maximum power in a power supply system with a renewable energy source (wind generator) is considered in this paper. Neuronal network is two-layer (the first layer is a sigmoid (in which four neurons), and the second is a linear function (two neurons). In order to find out the weights and coefficients of the function, the neural network was investigated. It was trained on models whose results were known in advance. A further trained neuron network is small in dealing with different tasks, during which time she constantly retrained. This happened as long as the results did not fall within the tolerable limits of error deviation. The construction of the neural network was performed in the Matlab program.