Modified Smooth Variable Structure Filter for Radar Target Tracking

The Smooth Variable Structure Filter (SVSF) is an effective method for radar target tracking and capable of providing robust target state estimation under model uncertainty. The traditional SVSF utilizing sign function or saturation function as the switching function in the correction gain suffers from the cut-off switching operations and the imprecise correction gain when the uncertainty rises unexpectedly. Therefore, the undesired chattering phenomenon occurs and the posteriori estimation error of the target velocity and acceleration greatly increases. In this paper, a modified form of SVSF using the hyperbolic tangent function as the "soft" switching principle is proposed to suppress the chattering phenomenon. The proposed method, referred to as the Tanh-SVSF, improves the tracking accuracy without increasing any computational burden. Simulation results demonstrate the advantage of the proposed Tanh-SVSF over traditional methods in terms of the accuracy of target motion state estimation.