Forecasting the magnitude of the electrical load in electrical complexes of aircraft

BACKGROUND: The issue of forecasting, analysis and control of electrical load becomes more significant both within the framework of the electrical complex of the summer apparatus as a whole, and for certain groups of electric energy consumers. Forecasting the electrical load is necessary to solve the problem of optimizing the operational state of an electrical complex or system, under constantly changing conditions and changing environment, which entails a change in power. Recently, a relatively new method has often been used, which is based on fuzzy logic. This method is a symbiosis of fuzzy logic and neural networks, which includes the main properties characteristic of these areas. Thanks to the use of a well-established fuzzy neural technology implemented in a correctly designed and trained fuzzy neural network for predicting electrical loads, it became possible to ensure sufficiently high accuracy and speed of load prediction.

AIMS: The purpose of the work is to analyze methods for predicting the electrical load of the aircraft’s electrical and technical complex, as well as to determine the most optimal methodology for predicting the electrical load of an autonomous aircraft used in the armed forces.

METHODS: Modeling of working conditions is performed in the Matlab program and its applications – Simulink. At the same time, with the help of the fundamental blocks of this application, models of the physical components of the electrical complex of the aircraft are created.

RESULTS: For a substantive assessment of the prediction of the magnitude of the electrical load of the electrical complex of the aircraft, an algorithm implemented on a computer has been developed. It provides for the implementation of retrospective calculations taking into account the amount of power generated, the duration of operation and the flow of electricity in the electrical complex as a whole.

CONCLUSIONS: Due to the use of a well-established fuzzy neural technology for predicting electrical loads, implemented in a correctly designed and equipped fuzzy neural network, it became possible to provide the necessary accuracy and speed of predicting electrical loads.