Research on Adaptive Decision Strong Tracking Cubature Kalman Filter in Sensor Networks

An adaptive decision strong tracking cubature Kalman filter (ADSTCKF) was proposed to solve the problem of nonlinear and non-Gaussian noise interference in sensor networks, and the traditional cubature Kalman filter was greatly affected by the noise covariance. Firstly, the suboptimal solution is obtained by introducing the suboptimal fading factor according to the strong tracking cubature Kalman filter (STCKF) algorithm. Secondly, the fading factor is taken as the decision condition, and the initial value is adjusted adaptively to reduce the weight of the measured value and the estimation error. Finally, according to the adjusted suboptimal solution, the measurement update is calculated. Simulation results show that ADSTCKF algorithm can be applied to nonlinear and non-Gaussian noise interference conditions. After applying various algorithms to the constant turn rate motion model (CT), ADSTCKF algorithm can significantly optimize the accuracy and stability of state estimation compared with CKF and STCKF algorithm.