Research On Cluster Routing And Data Fusion Algorithm Optimization In Wireless Sensor Network

As a momentous application of the perception layer of the Cyber-physical system,wireless sensor network is one of the technologies that have a huge impact on human life in the 21 st century and is widely used.The primary condition for the survival of wireless sensor networks is energy.When some nodes fail or the energy is exhausted and the information cannot be collected in time,it will lead to the blind area in the monitoring area that cannot be perceived,affecting the integrity and accuracy of the information in the whole monitoring area.Therefore,under the premise of finite energy of sensor nodes,for the sake of decreasing node power consumption,improving energy utilization efficiency,and extend the life span of the network,this thesis carried out optimization research on clustering routing and data fusion algorithm,shortened communication distance and balanced network load through the network topology of clustering structure.In addition,data fusion can effectively reduce redundant data in the network,extract data features,reduce network data transmission volume,and further reduce network energy expenditure.The main contents are as follows:1.Aiming at problem that the quality of the cluster heads selected by the traditional and various improved fuzzy logic algorithms are not high and do not balance the cluster heads and the node load in the cluster effectively,this thesis proposes a clustering routing algorithm for wireless sensor networks according to fuzzy logic optimized by whale Optimization algorithm.Firstly,a clustering threshold is designed to select candidate cluster heads according to the possessed energy and distance,so that the energy and position of the selected candidate cluster heads are more reasonable.Secondly,when using fuzzy logic for cluster head selection,the fuzzy rules set based on empirical knowledge in the Mamdani model cannot more reasonably construct the model between input and export variables,and the number of fuzzy rule combinations is too large,and the existing intelligent algorithm optimization of fuzzy logic rules lack the analysis of influencing factors of clustering on the macro level of sensors.The whale algorithm was introduced to optimize the fuzzy rules of the fuzzy logic Mamdani inference model,and three independent input variables of the fuzzy inference model,node energy,node neighbor number,node distance to the base station,were designed for fuzzy logic input,so that the quality of the final cluster head was higher.At the same time,the competition radius and optimized node clustering mechanism are used to balance the energy consumption of the cluster head and improve the energy utilization efficiency of the nodes around the base station.2.In order to effectively reduce the data exchange between sensor nodes and further enhance the energy efficiency of the network,and to solve the problem of BP neural network initial value sensitivity in WSNs data fusion algorithm based on BP neural network,this thesis proposes a data fusion for wireless sensor networks based on BP neural network optimized by improved sparrow search algorithm.Firstly,reverse learning,golden sine algorithm and Cauchy mutation operator are introduced to improve the basic sparrow search algorithm,which improves the search accuracy,global optimization ability and stability.Then,the improved sparrow search algorithm is used to ameliorate the parameters of the BP neural network,which surmounts the defects of the traditional BP neural network,such as poor convergence,large oscillation of training results and easy to fall into the local optimal solution due to the randomness of the initial weights selection.It improves the generalization ability,convergence ability and solving accuracy of the BP neural network.Finally,the optimized BP neural network is applied to the data fusion of the wireless sensor network.The original data collected by the sensor node is processed by the BP neural network optimized by the ameliorated sparrow search algorithm for data fusion processing,and the processing results are sent to the base station.