Study On Key Technologys Of High Precision And Real-time Simulation Of Satellite Navigation System

In the initial stage of building satellite navigation system, a corresponding simulation system is strongly demanded to supply simulate satellite navigation message. The simulate satellite navigation message could be used to satisfy cooperating test requirements of the ground operation control system and test the function and capability of receivers. This simulation systerm is different from general satellite navigation simulation systems. It has three characteristics: (1) High precision demand. The simulate data could be used for orbit determination and ephemeris forecast, which has high precision demand for simulate data. (2) Strong real-time demand. The simulation system needs to commute with hardware under strict controlling of external clock, which has strong real-time demand. (3) Capability of data forecasting. When carrying out together with real satellite in orbit, the simulation systerm is demanded to have the capability of data forecasting in order to ensure the precision of simulate data, which should keep coherence with real data. Much work has been done of satellite navigation simulation system study, but most of them did not discuss the high precision demand and strong real-time demand synchronously. Based on analysis of these simulation system characteristics, this thesis mainly discusses three key points of high precision and real-time satellite navigation simulation system, which includes regional ionospheric delay modeling, ionospheric short-term forecasting and regional tropospheric delay modeling. The main content of this thesis is as follows:1. Firstly, this thesis has discussed the algorithm of establishing the regional ionospheric VTEC model. Using real observation from GPS stations over China, three commonly used VTEC models (Polynomial Model, Generalized Trigonometric Series Function Model and Spherical Harmonic Function Model) are established. The influence of model exponents and basis functions on model accuracy is compared. Then a new method is presented to establish VTEC Polynomial Model for its weak space forecasting capability. Besides real observation data, grid data calculated by Global Ionosphere Model has been used as a background in this method. Availability of the method is proved by mathematic experimentation. Results show that model can improve the space forecasting capability of VTEC Polynomial Model. Lastly, this thesis has proposed a new model for ionospheric VTEC based on generalized regression neural network. Feasibility and availability of the network model are validated based on the comparing analysis between the network model and the Polynomial Model. Results show that the forecasting capability of network model is better than Polynomial Model in the blank data region.2. Secondly, this thesis has try to study regional ionospheric short-term forecasting over China based on time series theory. Using Polynomial Model as the original model to calculate ionospheric VTEC, characters of regional model coefficient are comprehensively detailed by comparing with global model coefficient. Coefficients of Polynomial Model at different times form the time series, which are predicted by time series theory so as that the prediction of VTEC is implemented. Experimental results show that, forecasting model has large errors on the edge of original model effective region. With expanding effective region of original model, forecasting model can reach a good degree of accuracy and keep well long-term stabilization over China.3. Thirdly, regional tropospheric delay modeling has been studied in this thesis. Using real meteorological data over China, effects of different meteorological parameters on calculating tropospheric zenith delay are discussed in different meteorological condition. And some conclusions are reached from experiment results. Zenith delay is the increasing function of meteorological parameters. The sensitivity of different meteorological parameters is different from each other in the same meteorological condition. And the variation of sensitivity is different neither when meteorological condition changes: the sensitivity variation of temperature is more severe than that of relative humidity, and pressure's keeps a constant. Based on these analyses, grid model of meteorological parameter is proposed by inverse distance squared method using real weather data over China. The absolute mean bias of temperature, pressure and relative humidity are 0.96 mbar, 1.60 C and 7.72%. Tropospheric zenith delay error caused by meteorological parameter's inaccuracy could be reduced to a little magnitude using these grid data.4. Finally, real-time simulation algorithm of high precision ionospheric delay model and tropospheric delay model has been studied. For meeting the strongly real-time demand of satellite navigation simulation system, a distributed structure of simulation system is proposed. The structure is divided into negative real-time part, week real-time part and strong real-time part. The calculate flow of regional ionospheric delay model, regional ionospheric delay short-term forecasting model and regional tropospheric delay model are divided into model pretreatment stage, data prepared stage and real-time model calculation satge. Then based on the distributed structure, the three stages are calculated respectively and real-time calculation of satellite navigation simulation system has been realized. Experimental results show that single step calculation time of the system is less than twenty milliseconds, which has validated the validity of the distributed structure.