Agrcultural Information Monitoring System Design And Data Fusion Research Based On Wireless Sensor Networks

With the rapid development of sensor technology, integrated circuit and its design, signal detection and processing, and computer networks technology, Wireless Sensor Networks (WSN) and multi-sensor data fusion have been new discipline directions and current hot research spots, and widely applied in the military fields and civilian areas. The main research content is as follow:According to the need analysis of monitoring agricultural information, the model of agricultural information monitoring system based on WSN is proposed, and the main structure is set up.This system can achieve a lot of functions such as acquisition, gathering, storing, analysis and statistics.According to the functional requirementsof nodes, we designed sensor and sink, based on MSP430F149 and LPC2214. The driver of wireless transceiver module is be programmed. And then we proposed a routing protocol based on minimum hop while data packet format in communication was worked out as well. After that we connected wireless network to Ethernet by transplanting the ZLG/IP into sink. It is convenient to carry on centralized management via remote monitoring system.In this dissertation,the Principle and value of data aggregation are studied,and the existent typical approaches of data aggregation are compared. The innovation are showed as followsIn the foundation of introducing the data fusion theory and its application, a multi-sensor circuit fault diagnosis approach is proposed based on Bayes decision theory. The ability to identify failure modes is effectively enhanced, so that the accuracy of fault determination can be improved.Sensor optimal management is an important link affecting multi-sensor goal assignment. Combining with dynamical weight partical swarm optimical (DWPSO) algorithm which has global optimization characteristic, the simulated data demonstrate the feasibility and the validity of new algorithm.When the scale is big, the algorithm shows quicker convergence rate and higher precision.A data fusion algorithm (simpled NNDF) will be proposed based on the clustering routing protocol. The NNDF can be described the following steps: Firstly, the clustering routing algorithm is used to divide the network into several clusters; sencondly, every cluster is act as a neural network model which is have three perceptions; then, the information of cluster head and cluster members is submitted to the processing function of neurons. Finally, all the fusioned results will be collected to the sinks.Using multi-sensor data fusion algorithm based on QDPSO-BP network. Higher accuracy would be gotten by using it in the simulation.It is a potential data fusion method for multi-sensors.