An adaptive particle filter based on PSO and fuzzy inference system for nonlinear state systems

Abstract

To cite this article: Ramazan Havangi (2018) An adaptive particle filter based on PSO and fuzzy inference system for nonlinear state systems, Automatika, 59:1, 94-103, DOI: 10.1080/00051144.2018 ABSTRACTParticle filters have been widely used in nonlinear/non-Gaussian Bayesian state estimation problems. However, the particle filter (PF) is inconsistent over time. The inconsistency of PF mainly results from the particle depletion in resampling step and an incorrect priori knowledge of process and measurement noise. To cope with this problem and enhance the accuracy and consistency of the state estimation, an adaptive particle filter(APF) is proposed in this paper. In APF, an adaptive fuzzy square-root unscented Kalman filter (AFSRUKF) is used to generate the proposal distribution. This causes that beside the merit of reducing the computational cost, APF has some other advantages such as increasing consistency that leads to more numerical stability and better performance. Moreover,APF can work in unknown statistical noise behaviour and is more robust. This is why the fuzzy inference system (FIS) supervises the performance of square-root unscented particle filter (SRUPF) using tuning statistical noises. In APF, to increase the diversity of particles, the resampling process is done based on the particle swarm optimization (PSO). With this resampling strategy, the small-weight particles are modified to the large-weight ones without duplication and elimination of particles. The effectiveness of APF is demonstrated by using two experiment examples through Monte Carlo simulations. The simulation results demonstrate the effectiveness of the proposed method. ARTICLE HISTORY

Numerical simulation of groundwater in an unconfined aquifer with a novel hybrid model (case study: Birjand Aquifer, Iran)

“…where N, x (i) t and w (i) t are the number, locations and weights of particles. The PF approximates the posterior function of p(x t jy 1:t ) as follows (Manoli et al 2015;Havangi 2018):…”

“…On the other hand, according to the Bayes law, the posterior PDF can be written as follows (Havangi 2018): So, the particle weight is calculated as the following recursive relation (Arulampalam et al 2002):…”

“…This method is known as the post probability of significant sequential important sampling (SIS) (Arulampalam et al 2002;Havangi 2018). In order to avoid the corrosion phenomenon, with resampling, the SIS algorithm is introduced (Ruknudeen & Asokan 2017).…”

“…Based on the references presented in this field, the best value for this parameter is N=2 (Manoli et al 2015). The following equation shows the condition that resampling technique must be carried out (Havangi 2018):…”

Prediction‐discrepancy based on innovative particle filter for estimating UAV true position in the presence of the GPS spoofing attacks

“…In this case, the accuracy of positioning solution is approximately equal to the accuracy of the Loran‐C system that is about 100 m or more [14]. Another solution consists of data fusion algorithms such as unscented Kalman filter (UKF), unscented particle filter (UPF), particle swarm optimisation (PSO), generic algorithm (GA) and any other heuristic and non‐heuristic algorithms [15–20]. Usually, these algorithms are used in the state estimation process in the INS/GPS integrated systems.…”