Research On INS/GPS Integrated Navigation Fusion Method For Unmanned Ships

With the deepening of economic globalization and the strengthening of strategic policies such as maritime power,marine activities are increasing day by day,and unmanned ships with intelligence and low risk have become important research hotspots.The INS/GPS integrated navigation technology,as a key technology for unmanned ship navigation,can combine the advantages of inertial navigation system(INS)autonomy and stability with the global positioning system(GPS)to provide real-time navigation information for unmanned ships.Due to the accumulation of INS errors over time and the presence of noise interference,Moreover,the poor adaptability of GPS in high dynamic environments,as well as issues such as external environmental influences or deceptive signal interference,result in unstable GPS signals.The main innovative work carried out in this article is as follows:(1)Aiming at the problem of noise interference in inertial measurement unit(IMU),a noise reduction method based on Ensemble Empirical Mode Decomposition(EEMD)is proposed.First,the inertial measurement unit signal is preprocessed,and then empirical mode decomposition is realized.The correlation coefficient is used to judge the correlation degree between the mode eigenfunction of EEMD decomposition and the original signal,so that the useful signal is selected for reconstruction,and the noise reduced signal is obtained,which provides accurate information for subsequent INS/GPS data fusion and model training.(2)The SSA-ELM model training is completed by inputting the denoised specific force and angular velocity information,as well as the position and velocity information difference between GPS and INS.In response to the issue of GPS signals being easily interrupted by external interference,an INS/GPS integrated navigation intelligent structure based on Sparrow Search Algorithm(SSA)improved Extreme Learning Machine(ELM)is proposed.The speed and position difference predicted by the SSA-ELM model,as well as the position and speed of the INS,are used to provide accurate correction values using Extended Kalman Filter(EKF),To some extent,it suppresses the accumulation of INS errors over time after GPS interruption,thereby improving the INS/GPS integrated navigation performance of unmanned ships.(3)Through the design of high-speed steering navigation phase,low-speed steering navigation phase,high-speed straight navigation phase and low-speed straight navigation phase,the prediction performance of multi-layer feedforward neural network(BP),ELM neural network and SSA-ELM model in the case of GPS interruption is compared and analyzed.It is concluded that the proposed SSA-ELM model can effectively improve the predictive performance of integrated navigation in the event of GPS interruption caused by external interference.The combined navigation fusion method of EEMD and SSA-ELM proposed in this paper plays an important supporting role in ensuring the safe navigation of unmanned ships at sea,in response to the problems of noise interference in IMUs and the susceptibility of GPS to interruption caused by external signal interference.