Research On The Whole Network Control Strategy Of FAST Based On PSO

500m Radio Telescope(Five-hundred-meter Aperture Spherical Radio Telescope,FAST),for the major national scientific projects.FAST whole network control is one of the basic work of the entire network and core technology deformation control strategies and adaptive modeling research program FAST initiative reflecting surface,so this initiative reflecting surface for FAST explore strategies to control the entire network has more important theoretical value and practical significance.According to astronomical observations needed,FAST initiative reflecting surfaces need to be tracked in real time to adjust.But,FAST reflective surface is a large scale,and the error of node action exists in the node itself features,coupled with the impact of external factors,so that truly parabolic form is not very smooth,there is even a partial error can not be ignored.Therefore,this thesis propose PSO(Particle Swarm Optimization,PSO)as FAST entire network node control strategy studies to enable instantaneous paraboloid when observing observation precision can meet the greatest degree.First,focusing on learning the theory of PSO’s in-depth study to understand the operational processes and design parameters,the ultimate combination of theoretical analysis summarizes the advantages and disadvantages of PSO;secondly,analyzes the overall system architecture and FAST initiative reflecting surface deformation of the entire network strategy and to develop according to the specific characteristics of the reflecting surface of the FAST initiative corresponding control strategy,focus on mining can be characterized FAST initiative reflecting surface deformation precision of indicators to determine the input and output of PSO algorithm;again,slow convergence and easy for PSO "premature" limitation proposed sub-dimension evolution PSO(Particle Swarm Optimization based on the Evolution of Sub-dimension,sdPSO).Then on the integrated sdPSO algorithm principle,elaborated algorithm implementation steps and processes.And needle PSO and sdPSO standard function simulation comparison test results show the effectiveness in the fight against PSO sdPSO shortcomings;Finally,the FAST control the entire network modeling to determine the model parameters and simulation test.Simulation results show the effectiveness of the improved algorithm and the model proved the feasibility of improved particle swarm optimization algorithm as the FAST whole network control strategy.The innovation of this thesis is the first application of the PSO and sdPSO optimal control problem in the FAST initiative reflecting surface,and a new expression and a node parabolic precision large errors priority adjustment strategy;verify the particle swarm optimization algorithm based on the entire network FAST control strategy is feasible and effective for the entire network FAST control strategy laid the foundation work has a certain practical significance.