Acquisition, Tracking And Controlling Of ATP System On Mobile Platform

Despite the fact that free space optical communication is wildly applied in military and civilian areas. still there are many technical problems needed to be resolved. among which ATP is the most important one. ATP is the premise for realizing space optical communications based on mobile platform. But the mobility of the platform leads to larger moving range of beacon light and further results in bigger errors in tracking and pointing. Therefore, more advanced tracking algorithm is needed to accurately acquire real-time beacon moving trajectory. Also more advanced control algorithm is required to control ATP so that the beacon is maintained to be around the center filed of view.The concept. characteristics and key technologies of space optical communication are first introduced. followed by development status and trends in and abroad China. Major products and experiments in optical comminication in the world are also introduced in details.And then ATP composition and working principle are discussed. Compound axis systems constructed by the coase tracking and fine tracking system are analysed in control model. control flow and control theory. Also the error sources in tracking and pointing are analysed.Besides, based on antenna scanning method, a new method with GPS is applied to initailly acquire light beacon. The formula of azimuth angle. pitch angle and coordinate transformation calculation with GPS coordinate is given. The algorithms are implied in fixed point DSP through floating point function library. By comparing with simulated results, it proves that DSP is much better in precision. The results of ground GPS acquring experiments demonstrate that GPS is very effective in initial beacon acquisitions and can reduce time in beacon acquiring.The overall process of beacon tracking on mobile platform is simulated. Wiener filtering in Wavelet Shrinkage is proposed on the basis of usual image denosing methods. Simulaton experimets results demonstrate that this method is much better than others in denoising and at the same time preserves the spot edge information. To acquire spot loction effctively. histogram thresholding methods and centroid location method are proposed. Based on usual maneuvering targets models, a beacon motion model under polar coordinates is established with "current" statistical model and the beacon location information is used as obsevation model. With the above mentioned models, an improved real-time particle filter algorithm based on resample threshold decision is applied in predictive beacon tracking, so that the degeneracy phenomenon is relieved and the real-time is improved. Simulations and indoor tracking experimets present that this algorithm can realize beacon predictive tracking and the precision is much better compared with extended Kalman filtering algorithm.Generized predictive explicit adapted algorithm is applied to steadily control the ATP system on the basis of identification of ATP system construcion model. The model parameters of the controling target are identified online by forgetting factor recursive least squares. The prediction control is robust with rolling optimization and feedback compensation. Simulation experiment results verify that the control algorithm is steady and robust.Finally, an experimant platform for ATP system is built indoor to carry out two-dementional mobile tracking experiments and the coarse tracking precision is 198.5μrad.2.3km and 16km coarse tracking experiments are conducted outdoor and the precisions are 177.3μrad and 186.2μrad respectively when simulated mobile platform rotate at the speed of 0.5°/s. Besides.16km fine tracking together with 16km coarse tracking experiments are conducted and the tracking presision is 24.9μrad after compension. Helical scan acquisition experiments with GPS telephoto lenses are conducted. The results verify that acquisition mean time is less by identifying uncertain areas with GPS. compared with experiments when no uncertain areas are identified.