GPS/INS组合导航系统的研究

讨论了飞机惯性导航系统(INS)与全球卫星导航系统(GPS)的利与弊以及卡尔曼滤波方法在组合定位中的应用情况,进一步提出了基于神经网络数据融合方法的GPS/INS组合导航系统.系统神经网络结构采用单隐层的三层神经网络,输入输出神经元数目是4个,基于256个训练样本由经验公式求得隐层神经元数目为8个,同时还建立了惯导系统的数学模型和数据融合的数学模型.给出了利用MATLAB编制的神经网络训练程序并对这一神经网络进行了训练和仿真.实验表明,组合导航系统经度误差可达9m,纬度误差可达8m,与单独GPS定位和INS定位相比精度得到了提高.

GPS/INS integrated navigation system

Through the discussions on the advantages and disadvantages of inertial navigation system (INS) and global positioning system (GPS) for aircraft, and the application of Kalman filtering method in the integrated positioning system, GPS/INS integrated navigation system, a data fusion method based on neural network is proposed. The neural network framework of the integrated navigation system is a single hidden layer three-layered neural network with 4 input/output neurons, and the number of hidden layer neurons based on 256 training samples and obtained using empiric formula is 8, and mathematical models have been established for inertial navigation system and data fusion. Training and simulation were made with this neural network using the neural network training program compiled with MATLAB, and experimental results indicate that the integrated navigation system has a longitude error of 9 m, and a latitude error of 8 m, and in comparison with separate GPS and INS, the accuracy of integrated GPS/INS has been improved.