Download PDFOpen PDF in browserData Fusion Algorithms for Wireless Sensor Networks Based on BP Neural Network with Improved Particle Swarm OptimizationEasyChair Preprint 13926 pages•Date: August 12, 2019AbstractData fusion algorithm of wireless sensor network based on Improved Particle Swarm Optimization BP neural network Aiming at the problem of slow convergence speed, sensitive to initial value and easy to fall into local optimal solution of traditional back propagation (BP) neural network in data fusion algorithm of wireless sensor network, this patent proposes WSN data fusion algorithm of improved particle swarm optimization BP neural network. Law. The particle swarm optimization (PSO) algorithm is improved by Beetle Antennae Search algorithm (BAS). The improved PSO algorithm is used to optimize the weights and thresholds of BP neural network. In WSN data fusion, the cluster head node extracts the features of the data through the optimized BP neural network and sends the fused features. Information to the sink node reduces redundant data transmission and prolongs the network life cycle. Keyphrases: BP neural network, Beetle antennae search, clustering routing, data fusion
|