Kolotov M. The semiconductor converters control in the electronic platform orientation system

The dissertation is devoted to the development of theoretical and practical principles of prediction control of semiconductors convertors for orientation systems of platforms. As a result of comparison and modeling of switching 4-shoulder inverter with zero point is chosen as the most effective device for orientation system of electronic platforms. Predicting the values of magnetic induction to eliminate distortions, such as magnetic disturbances and anomalies give more accurate targeting platform in orbit. Forecasting energy consumption allows saving energy on the platform in case of forecasting its lack by banning a maneuver. It is proposed the mathematic instrument for prediction such as neural networks with gamma memory and data preprocessing. Data which enter to the network is appropriated to filter out unwanted fluctuations using wavelet transformation in oriented basis, as the most effective compared with other types of wavelet transformations. The software for control transformers can predict the value of magnetic induction and power consumption in real time.