Multi-Array Fusion Localization Based on the Improved PSO Algorithm

In the sound localization system with the Time Difference of Arrival, due to the complex array structure and the numerous sensors, etc., the localization accuracy of single-array targets deteriorates significantly in the presence of environmental noise and a considerable distance. Aiming at this problem, a novel multi-array fusion localization is proposed in this letter. This approach initially estimates the target locations by each single array to obtains the spatial lines separately. Subsequently, the suboptimal target locations is obtained based on the geometric relationship between the spatial lines. Finally, through solving the equation constructed by the shortest distance between suboptimal target locations based on the least square method and obtains the optimal target location is obtained. Furthermore, this paper introduces an improved particle swarm optimization algorithm to solve the data fusion equation. Simulations and experiments using a self built localization platform are conducted to validate the proposed method. The results indicate that the fusion localization method proposed in this letter possesses higher accuracy and stability.