Study On The Key Technology Of Fast Startup Of GPS Receiver Based On Satellite Orbit Prediction

The Global Positioning System(GPS)is currently one of the key methods for obtaining accurate location and time information.The GPS receiver is an important part of the GPS system,and the time to first fix(TTFF)is one of the important indexes to measure the performance of the GPS receiver.Currently,under the warm/cold start mode,the satellite position information required for GPS receiver positioning needs to be obtained by receiving satellite broadcast ephemeris,which leads to the TTFF of GPS receiver may be as long as several minutes or even more than ten minutes in complex environments.Among all existing methods to reduce TTFF,the satellite orbit prediction,which can greatly accelerate the start-up speed of the receiver,and has many advantages such as no network connection,no increase in the cost of the receiver,and simple and convenient implementation has become the preferred method to reduce TTFF of the receiver without a network connection.However,the current studies on satellite orbit prediction methods applicable to the receiver still remains shortcomings including imperfect orbital dynamics smoothing strategy,inaccurate orbital dynamics model,inefficient orbital integration,limitations in dynamics orbit prediction methods,and lack of experimental verification in the impact analysis of orbit prediction results on the start-up performance of the receiver.In response to the above-mentioned shortcomings,the research on efficient and high-precision satellite orbit prediction methods applicable for receivers was conducted,and the relevant research results were used for transformation of GPS receiver and the startup performance of the receiver after transformation was analyzed,which has important practical significance for providing better GPS user services.The research contents and innovations of this article are as follows:1.Based on the processing methods of the initial state of the orbit,the solar radiation pressure model parameters,and the earth orientation parameters in the orbit dynamics smoothing,an optimized smoothing strategy for orbital dynamics was proposed to meet the needs of high-precision satellite orbit prediction at the GPS receiver.The simulation results show that when the fitting arc length is long enough(more than 18 hours),the optimized smoothing strategy for orbital dynamics proposed in this paper is adopted to predict the satellite orbit of 1 day,3 days,6 days,and the maximum prediction errors are under 5 m,13 m,and 26 m respectively.The accuracy is approximately 3 to 4 times as high as that of the predicted orbit of U-blox 8 receiver Assist Now Autonomous(the satellite orbit prediction function built into the U-blox receiver)provided by U-blox's.2.A high-performance dynamics orbit prediction method that can be used for long-term(14 days)satellite orbit prediction at the receiver was proposed.By studying the influence of small perturbation factors on the accuracy of orbit prediction,a high-precision dynamics model was constructed to improve the accuracy of orbit prediction.On this basis,a numerical integration method,the "symplectic perturbation method",was introduced to solve the satellite orbit dynamics equation,and thus improving the efficiency of orbit prediction.The simulation results show that,compared with the traditional orbit prediction method(using the basic dynamics model and Adams-Cowell integration method),the high performance dynamics orbit prediction method proposed in this paper improves the accuracy and efficiency of orbit prediction.3.Two orbit prediction methods,namely,orbit prediction method based on Artificial Neural Network(ANN)(used for 1 to 7 days prediction)and orbit prediction method based on Convolutional Neural Networks(CNN)(used for 7 to 14 days prediction),were proposed,which improve the accuracy of satellite orbit prediction within 14 days.These methods use the neural network model to model and compensate the error of dynamics orbit prediction,and greatly improves the accuracy of satellite orbit prediction.The simulation results show that compared with the dynamics orbit prediction method,the orbit prediction method based on ANN and the orbit prediction method based on CNN have significantly improved the orbit prediction accuracy.4.According to the actual experimental scenes,the test,analysis,and comparison of the effects of various satellite orbit prediction methods on the receiver's startup performance when applied to the receiver were completed for the first time.Aiming at the problem of lack of experimental verification when analyzing the influence of satellite orbit prediction methods on receiver startup performance in existing research,this paper designed the system architecture of the receiver with satellite orbit prediction function,built actual experimental scenarios,and completed the receiver startup performance test.The experimental results show that the satellite orbit prediction method significantly reduced the TTFF of the receiver in cold start mode.In the dynamics orbit prediction methods,the high-precision dynamic model constructed in this paper improves the positioning accuracy of the receiver.Compared with the dynamics orbit prediction method,the satellite orbit prediction method based on ANN and the satellite orbit prediction method based on CNN proposed in this paper further improve the positioning accuracy of the receiver.