Research On Performance Evaluation Method Of Inertial Navigation System Based On PSO-SVM

The platform inertial navigation system has the important position in the navigation field, because of its own autonomous navigation information and so on, widely used in our tactical missile field. However, for the inertial platform, the problem of the stability of the inertial navigation platform and the short calibration period are the common problems. For inertial navigation platform, error parameter of the performance is good or bad for the platform inertial navigation system has a decisive influence, when the platform inertial navigation system of the calibration data with time shift, platform performance changes, and even affect the parameters of platform inertial navigation system level, as in known platform used to guide system error source parameters of the case, evaluation of platform inertial navigation system has a is a problem to be solved in the engineering field. Therefore, the establishment of the platform inertial navigation error parameters of the platform inertial navigation system evaluation method is a certain practical significance.In this paper, according to the navigation principle, the error propagation equation of inertial navigation system is studied, and the error model of inertial navigation device is studied. In order to improve the accuracy, to extend the device properties of source parameters, in addition to the common scale factor and zero bias device error source parameters and compensation about the acceleration a, quadratic error term. The error source parameters can focus on the error characteristics of the device and improve the accuracy of the modeling. Afterwards, the propagation equation of the inertial navigation error is established. From this, the model of the error source factor and the final navigation error can be established..Secondly, on the error characteristics of inertial navigation platform analysis, design a practical trajectory, produced a large number of different device parameters error of inertial navigation platform flight data, from a data mining perspective, of the device parameters are established for the inertial navigation platform influence model, and based on this data as indicators of the platform, the establishment of the inertial platform classification.Finally, based on the inertial platform navigation by real-time calculation of the navigation system error because of the existence of the time-consuming. Therefore, the based on support vector machine of the inertial platform performance evaluation method. For support vector machine, the core problem is how to find its optimization parameters. In general, the method of grid partition is time-consuming, and the efficiency is low. In order to solve this problem, the particle swarm optimization algorithm is used to support vector machine optimization, and its optimization efficiency and precision are greatly improved. Through numerical experiments, the accuracy and error of inertial navigation platform were predicted. The results show that the error of the inertial navigation platform based on PSO-SVM is highly accurate and practical.