Research On The Date Fusion Algorithm Of Wireless Sensor Network For Train Cabin Environment Monitoring

With the rapid development of China's railway industry and the continuous improvement of people's living standards,the comfort of trains has been gradually received great attention.The environmental quality of train carriages directly affects the comfort experience of passengers.Therefore,how to accurately and continuously monitor the environment of train carriages has become one of the hot topics of current research.Nowadays,wireless sensor network technology can achieve comprehensive and accurate continuous monitoring of the cabin environment,but it has the disadvantages of limited data storage and transmission in applications,so it becomes more and more meaningful to process the monitored environmental data.Based on the classic matching and tracking algorithms,this thesis proposes an improved matching and tracking algorithm and applies it to the reconstruction of train cabin environment monitoring data.The main research contents of the paper are as follows:(1)The research and simulation on the sparsity variable step-size adaptive compressive sampling matching pursuit algorithm based on the compressed sensing.Firstly,six classic matching pursuit algorithms in compressed sensing are compared and studied.Secondly,based on the compressed sensing theory and the sampling adaptive matching pursuit algorithm,the sparsity variable step-size adaptive compressive sampling matching pursuit algorithm is proposed.Finally,the reconstruction performance of the improved CSVssAMP algorithm is evaluated by one-dimensional Gaussian random sparse signal reconstruction experiments.The experiment is simulated from three aspects of different sparsity,observation value and step length.The simulation results show that the improved CSVssAMP algorithm has the best reconstruction performance under the same conditions.(2)Reconstruction of cabin environmental monitoring data based on CSVssAMP algorithm.In order to solve the problem of limited storage and transmission of cabin environmental monitoring data,the sparse representation method of K-SVD over-complete dictionary is used to establish the data processing system model.The reconstruction simulation experiment of a mechanism simulating real train environment attribute data is carried out by using SAMP algorithm and the improved CSVssAMP algorithm.The simulation results show that the compressed sensing matching tracking algorithm is effective and feasible in the reconstruction of train carriage environmental monitoring data.In addition,under the same simulation conditions,the reconstruction error of the improved CSVssAMP algorithm is less than the SAMP algorithm,that is,the improved CSVssAMP algorithm has better reconstruction performance.(3)Environmental quality evaluation of train carriage based on the fuzzy comprehensive evaluation method.In this thesis,50 groups of original data and CSVssAMP algorithm reconstruction data are randomly selected to form 100 sets of evaluation factors,which are composed of temperature,relative humidity,carbon dioxide,air velocity,and total root mean squared acceleration of vibration.According to a series of standards such as UIC553 of International Railway Federation,evaluation levels are established,and the fuzzy relation matrix and weight set of influence factors are determined.Through the composite operation of the fuzzy matrix,the environmental quality evaluation results of 50 groups of carriages are obtained,and the evaluation results of the original data in each group of experiments and the reconstruction data of CSVssAMP algorithm are compared and analyzed The results show that the improved CSVssAMP algorithm can not only reduce the amount of transmitted data,but also change the results of environmental quality assessment of train carriage.