Diversity-preserving quantum-enhanced particle filter for abrupt-motion tracking

Abrupt-motion tracking is challenging due to the target's unpredictable action. Although particle filter is suitable for target tracking of nonlinear non-Gaussian systems, it suffers from the problems of particle impoverishment and sample-size dependency. Inspired by quantum mechanics that one quantum bit could represent a superposition of two states, this paper proposes a diversity-preserving quantum-enhanced particle filter (DQPF). Firstly, we quantized the motion modes of particles into a superposition of several modes of motion, resulting in a quantum particle set that retains diversity. Aiming for the abruption of target motion, we propagate the quantum particles during the prediction stage. The quantum particles will already be in these possible positions even if abruption occurs, which addresses the abrupt-motion issue and reduces the tracking delay. Benefitting from quantum mechanics, the proposed particle filter has better precision and stability with fewer particles than the general particle filter. Compared to state-of-the-art, numerical experimental results demonstrate that the proposed DQPF has higher accuracy and stability under the same conditions, displaying superior performance to traditional modified particle filter methods.