The Outside Trajectory Measuring Dataprocessing Based On Neural Network

Microwave united measuring and control system (MUMCS) is one of the most popular measuring and control equipments and plays an important role in the measuring and control equipment of aircraft in China. As the main data source of MUMCS, remote measuring data (RMD) is the primary basis of aircraft orbit measuring and control. Moreover, It is very important to improve the processing precision and capacity to singular data of MUMCS remote measuring data. Treated as a crucial part of aircraft measuring and control, the processing of RMD is also the primary basis of evaluating condition of measuring and control equipments (MCE) and aircraft spacing pose. Data processing has developed from the simply afterwards processing to the sub real-time, real-time and intelligent process. After the explanation of basic theory of neural network (NN) and thorough consideration in character of real-time data processing and intelligent processing, depending on the many years working experience in afterwards processing of outside measuring data, the author applied wavelet multi-resolution analysis and NN to the data processing, mainly aiming at the existent questions in the data processing and developmental orientation of outside measuring data processing. For its proper capacity of non-linear adaptation in information processing overcomes the limitation brought by the traditional processing methods, NN applied successfully in the expert system and forecasting fields. Nowadays, combined with other subjects and became a leading developmental orientation of artificial intelligence, artificial neural network is developing in the way of simulating human cognition. Based on the Fourier analysis, wavelet analysis, a time-frequency analysis method, has essential progress in contrast with Fourier analysis. It provided a self-adaptive analysis method that can localize simultaneously in both time field and frequency field. In addition, wavelet analysis has strong agility in the local time-frequency analysis, which is capable of focusing on the arbitrary details whether in time field or frequency field. According to the data singularity often happened in data processing, this article establishes a data processing model and provides the solution. The results obtained by experiments are satisfactory.