Research On Data Fusion Model And Algorithm Of Grassland Environmental Monitoring Based On WSN

Inner Mongolia Autonomous Region has the largest grassland ecosystem in China and is one of the regional characteristics of Inner Mongolia.However,in recent years,due to excessive grazing,fire,industrial pollution and other factors,desertification has become increasingly serious,so protecting the grassland environment is an urgent issue.This paper uses WSN technology to monitor grassland environment,but in the actual monitoring process,WSN often has a series of problems,such as low accuracy of data collection,short life cycle,unstable transmission,and so on.To solve the above problems,this paper introduces the neural network data fusion algorithm into WSN,designs the non-uniform clustering algorithm to transmit network data,and constructs a WSN-based data fusion model through the above two algorithms.The main work of this paper is as follows.First,in order to ensure the reliability of WSN data transmission in grassland environment and to ensure that wireless sensor network can efficiently send monitoring data to sink nodes,this paper presents a non-uniform clustering routing protocol(IFA-UCR),which improves the Firefly algorithm.The algorithm is divided into two phases: in the nonuniform clustering phase,the fitness function is built to iteratively find the optimal network clustering structure by combining the update principle of fluorescein in the improved firefly algorithm with factors such as cluster head density,cluster head neighbor distance,cluster head location,node and sink node distance to solve the energy heat zone problem in wireless sensor network.In the data forwarding phase,the cost forwarding function is designed by the index of inter-cluster transmission energy consumption and cluster head residual energy.The cluster head with high power is selected as the next hop to construct the optimal data transmission path.At the same time,the secondary cluster head is set to share the data transmission of the primary cluster head node,while the primary cluster head is only responsible for data collection and fusion tasks,so as to improve data transmission stability,balance network energy consumption and extend network life.Secondly,in order to improve the accuracy of WSN data collection in environmental monitoring and reduce redundant data transmission,a large amount of similar data in wireless sensor network needs to be de-redundant.Based on the characteristics of nonuniform clustering routing structure,a data fusion algorithm(UCR-IPSOBP)based on nonuniform clustering and particle swarm neural network is proposed in this paper.The algorithm first improves the particle swarm algorithm with the chemotaxis and migration operators of the bacterial foraging algorithm(BFO),then optimizes the initial value of the BP network with the improved particle swarm algorithm,and finally introduces it into the wireless sensor network.The collected data is fused with the optimized BP network,and data is transmitted with the non-uniform clustering routing algorithm.Finally,the cluster head node sends a small amount of feature data representing the monitoring area to the sink node,which can effectively reduce the redundant data of the network,improve the accuracy of data collection,and extend the network life.Finally,because the neural network and WSN have similar topological structure,the second-order neural network data fusion model can be designed by combining the BP network data fusion technology with the wireless sensor clustering routing structure.Then the second-order neural network data fusion model is applied to WSN,and then the UCRIPSOBP algorithm is used to build the WSN-based grassland environment monitoring data fusion model.Finally,using the historical data of the grassland environmental monitoring site in Saihantara city of Baotou as the experimental data,the grassland environmental monitoring data fusion model based on WSN is analyzed experimentally.The simulation results show that the monitoring data after two-level neural network fusion can improve the reliability of data collection,greatly reduce redundant data transmission,and prolong the network life.