A Novel Algorithm for Quantized Particle Filtering With Multiple Degrading Sensors: Degradation Estimation and Target Tracking

Abstract

This paper addresses the particle filtering problem for a class of nonlinear/non-Gaussian systems with quantized measurements and multiple degrading sensors. A degradation variable described by the Wiener process is proposed to describe the phenomenon of sensor degradation that is often encountered in engineering practice. The measurement output of each sensor is quantized by a uniform quantizer before being sent to the remote filter. An augmented system is constructed which aggregates the original system state and the degradation variables. In the presence of the sensor degradation and the quantization errors, a new likelihood function at the remote filter is calculated by resorting to all the transmitted measurements. According to the mathematical characterization of the likelihood function, a novel particle filtering algorithm is developed where the parameters of both the degradation processes and the quantization functions are exploited to obtain the modified importance weights. Finally, the effectiveness of the proposed method is shown via a target tracking example with bearing measurements.

Hybrid YOLOv3 and ReID intelligent identification statistical model for people flow in public places

“…In terms of object tracking, Liu et al proposed an object tracking model for particle images, which adjusted the core network structure of the DeepSORT algorithm and greatly reduced the model parameters. The results showed that the target tracking accuracy on a particle image dataset was 98.6%, which was much higher than the improved DeepSORT algorithm 15 . Jeong et al discussed a real-time computer vision perception method using deep learning to solve the object tracking uncertainty caused by lighting changes and field of view occlusion, and maintained the real-time processing power of the model.…”

Finite-time ℋ∞ filtering for Markov jump systems with uniform quantization

“…which, together with inequalities (38), (39), and (47), results in E ℒ𝒱 ( χ(t), m s (t) = r 1 ) − α𝒱 ( χ(t), m s (t) = r 1 ) − γ 2 ‖ ω(t)‖ 2 + ξ (t) 2 < 0, which means FES (8) has the FTHP index γ = γ e αT f . The specific derivation process is identical to the proof of inequality (36) ensuring inequality (10) in Theorem 1 and we omitted here.…”

“…In this case, the existing literature on filtering under uniform quantization (see, e.g., Refs. [38][39][40]) cannot be applied to find finite-time filters, while the design approaches proposed in Theorems 1 and 2 are applicable. In what follows, we take transition rate matrix as ϒ = −0.5 0.5 0.8 −0.8 .…”

Bayesian Estimation and Reconstruction of Marine Surface Contaminant Dispersion

“…The finite word length of quantised signals naturally leads to some information loss. To handle the quantisation-induced loss, many estimation strategies have been developed, see, for example, Liu et al (2023); Li et al (2021) and the references therein. In environment monitoring studies, the quantisation issue has also garnered some research attention: In the experimental investigation on the effects of polyethylene microplastics on mammals (Sun et al, 2021), the measurement of colon mucin density has been quantised with the Alcian blue periodic acid Schiff.…”