Based On GNSS/SINS Radar Seeker Research On Integrated Navigation System

Beidou Satellite Navigation System is a dual-use satellite navigation system independently developed and operated by our country.At present,it has formed a regional networking capability and can provide all-weather for China and more than 30 countries along "Belt and Road" for civilian use.All-terrain free high-precision navigation and positioning services.Inertial Navigation System(INS)plays an irreplaceable role in various navigation and positioning applications by virtue of its concealment reliability and completeness of information.However,not only the single navigation system itself has some shortcomings,even the most widely used GNSS/SINS integrated navigation system also has some problems: Kalman filter needs accurate system model to predict the error.The precise modeling of SINS requires hundreds of dimensions and the sensor error is difficult to model.The navigation error of GNSS / SINS integrated navigation system will accumulate rapidly with time when GNSS is lost.This subject is a problem encountered in the development of the integrated navigation system of a certain type of cruise missile midcourse guidance seeker during my internship,especially when the GNSS signal of the missile cruise section is easily lost.The research focuses on the combination of Beidou satellite navigation system and strapdown inertial navigation system,which is based on window weight update.The main work is as follows:1.The mathematical error model of integrated navigation system is established.The unscented Kalman filter with better nonlinear adaptability is selected as the information fusion algorithm of integrated navigation,and the velocity error and position error are selected as feedback correction variables to design the GNSS/SINS integrated navigation system.The principle and error model of SINS are analyzed,and the positioning and navigation principle of Beidou Satellite Navigation System in China is studied.The state equation and measurement equation of integrated navigation system are derived.2.A neural network aided integrated navigation system based on window weight update is designed.This paper discusses the basic principle of artificial neural network and the current mainstream training process,especially compares the basic principle,training steps,advantages and disadvantages of forward neural network BPNN and RBF neural network.RBF neural network is selected to assist the integrated navigation system when the GNSS signal is out of lock.In this paper,the information processing and weight updating of navigation data in neural network are discussed.The strategy of updating weights based on window can update offline weights of neural networks and update weights of neural networks in fixed time windows.Once the GNSS is lost,the SINS can be aided by the neural network with the latest network weights.3.A simulation experiment was designed to verify the effect of neural network aided GNSS/SINS integrated navigation system.According to the previous chapters,the mathematical model of integrated navigation system is designed and simulated.The simulation results show that the GNSS neural network can suppress the error accumulation of SINS well within 5 minutes of the lost lock of SINS.