Adaptive state updating particle filter tracking algorithm based on cubic spline interpolation

In the particle filter tracking process, the prior information largely determines the accuracy of the filtering. We often need to obtain the prior information of the target at the next moment through point trace prediction, but the actual motion state of the target is usually unknown, so how to find a state model that matches the actual moving target becomes a thorny problem. Therefore, this paper proposes an adaptive state estimation method, which interpolates the target motion state adaptively through cubic spline, and the state equation changes adaptively with the change of target motion state. Through simulation test, the algorithm ensures the prediction accuracy of particles, and solves the problems of reduced filter tracking accuracy and tracking failure caused by target motion state transition, variable acceleration or irregular curve in the process of particle filter tracking.