T-S Fuzzy Modeling Of FAST Cable Node Displacements

Astronomical telescope is also known as radio telescope. Because of its features, like strong perspective, remote observing distance wide observing range and so on. Astronomical telescope will become an important observation equipment at present of the astronomical observation. Astronomical telescope is able to explore the mysteries of the universe with high resolution and high sensitivity. The Active reflector of 500m aperture spherical radio telescope (FAST) is composed of more than 2400 cable nodes. Only if we real-time dynamic positioning control the nodes, FAST active reflector can successfully complete observing mission. Therefore establishing the model of node path has the vital significance.The moving process of FAST has many features, such as multiple input, nonlinear, time-varying, large inertia etc. Meanwhile the number of nodes and interference factors is so large that it is difficult to measure the node moving process directly. According to the 50m model of FAST in this thesis, T-S model is applied to the forecast of cable node moving model. This paper use the method that antecedent and consequent are identified apart. Firstly, FCM algorithm is used to identify the structure and parameters of fuzzy antecedent, then least square method is used to identify parameters of the fuzzy consequent model. In order to improve the precision and speed of identifying model, the fourth chapter improve the identification method of the third chapter. Aiming at the problems of FCM algorithm, such as the initial value sensitivity、Slow convergence speed、Easy to fall into local minima etc, the fourth chapter use subtractive clustering algorithm to identify the structure and parameters. Then the PSO-BP hybrid algorithm, instead of the least squares method, is used to identify the consequent parameters. finally, the model is simulated by MATLAB using the data gathering in the field.The simulation results show that the improved T-S fuzzy model can predict the model of node displacement well. If the number of rules and parameters is certain, T-S fuzzy model, based on FCM and least square method, can only identify the mean square error of only accuracy. If we want to improve the model precision, we must increase fuzzy rules. But this makes the model more complex. The improved T-S fuzzy model, based on PSO-BP algorithm, essentially solves the problem above. Obviously, the model that do not increase the number of fuzzy rules and improve the identification accuracy, is more suitable for practical engineering field. According to the forecast model of node path, Workers can timely control and regulate the model, in order to ensure the normal operation of the work. Therefore, the improved T-S fuzzy model not only has theoretical significance, but also has practical application value.