Unscented Kalman filtering for target tracking systems with packet dropout compensation

An adaptive nonlinear filter with missing measurements compensation for manoeuvring target tracking

“…There are two main descriptions for the sensor missing measurements problem, that only measurement noise v k+1 is received [29] and that nothing (zero) is received [30]. The latter scheme is chosen to describe the missing measurements as follows: [21]…”

“…In the latter case, the one-step prediction value can be used as the compensated measurement [16,17,21] We define ẑk+1|k as the measurement one-step prediction. The model of the compensator is as follows:…”

“…Step 3: Model-conditioned filtering We consider the stochastic hybrid system model described by Equations ( 5) and ( 6), and the fact that the UKF has high estimation accuracy for nonlinear systems by approximating the mean and variance of the state using the unscented sampling method [21].…”

“…In addition, an optimal UKF algorithm with packet dropout compensation is proposed in ref. [21] by the unscented sampling method and one-step prediction compensation framework. It should be noted that the investigation of the aforementioned optimal filters for solving the missing measurements problem has been limited in the field of target tracking.…”

A survey on distributed filtering, estimation and fusion for nonlinear systems with communication constraints: new advances and prospects

“…Compared with the EKF, the UKF has many advantages and has been widely used in the engineering field. For example, the UKF algorithm has the same accuracy as second-order Gaussian filter and does not need to calculate the Jacobian matrix (Li & Xia, 2015;K. Ma et al, 2019;Zheng et al, 2019).…”