An Evolutionary Enhanced Particle Filter-Based Fusion Localization Scheme for Fast Tracking of Smartphone Users in Tall Complex Buildings for Hazardous Situations

Harun Jamil,Yang Jian

DOI: https://doi.org/10.1109/jsen.2024.3352599

IF: 4.3

2024-03-02

IEEE Sensors Journal

Abstract:Precise indoor tracking of smartphone users in intricate multistory structures is an instructive challenge, particularly in dangerous scenarios. To precisely identify users, we present the enhanced particle filter-weighted differential evolution (EPF-WDE) scheme in this study. This novel fusion-based approach uses smartphone sensors such as accelerometers, gyroscopes, magnetometers, barometers, and Bluetooth low-energy (BLE) beacons with their coordinates information. The proposed approach manages nonlinear and multidimensional systems with a focus on convergence. We incorporate the weighted differential evolution (WDE) method, which is well known for its global optimization capabilities, into the resampling stage of the EPF to improve convergence by providing multiple key smartphone-based sensor information as input. Our findings demonstrate notable gains in accuracy and convergence. Scenario 1 uses the evolutionary particle filter with four inputs, obtaining an astounding root mean square error (RMSE) of 0.099 compared to Scenario 2 and the basic particle filter (PF) with an RMSE of 0.14 and 0.29, respectively. These results establish our method as a superior alternative for practical applications, especially in emergency scenarios and tall, complex buildings where accurate user localization is critical.

engineering, electrical & electronic,instruments & instrumentation,physics, applied