Research On Multi-sensor Data Fusion Based On Neural Network

In recent years,with the rapid development of science and technology,new and higher requirements have been put forward for data processing technology.In the multi-sensor system,the data in the multi-sensor system can be divided into uncertainty,imperfection and correlation due to the influence of the test accuracy of the sensor,the cost of data acquisition,the multiple links of the system and the external environment.The traditional data processing methods can not meet the requirements of the engineering application on the speed and precision of data processing.Therefore,this paper conducts research on multi-sensor data fusion based on neural network.The specific research contents are as follows:(1)In this paper,the model based on RS-BPNN is used to study the problem of BP neural network processing multi-sensor data.Firstly,the attribute reduction algorithm of rough set(RS)is used to mine the data in the system,which reduces the data dimension of the system without affecting the fusion effect.Then the BP neural network is used to fuse the minimum reduction after attribute reduction.Finally,the rapidity and accuracy of the model are verified by a case study of water quality grade evaluation.(2)In this paper,PSO-RBFNN model is used to study the problem of multi-sensor data fusion with RBF neural network.Particle swarm optimization(PSO)algorithm is used to find the optimal parameters of RBF neural network,and the updating speed and accuracy of network parameters are improved through parameter optimization,and the RBF neural network model based on PSO algorithm is established.Finally,the validity of this method is verified in the case of air quality index.(3)The neural network is used for multi-sensor data fusion,and then the neural network fusion model is used for sensor fault diagnosis.Based on the nonlinear fitting ability of neural network and the data correlation between sensors in the multi-sensor system,the accurate predictive value is used to diagnose the fault of the sensor and the fault location,and the fault sensor data is repaired.Finally,the effectiveness of the proposed method is verified by an actual engine test.