Design And Experiment Of Dedicated Mobile Terminal In Spectral Sensing Internet Of Things

Spectral sensing IoT is an Internet of Things based on spectral sensing nodes,which provides users with the possibility to obtain material spectral information in real time.The spectral sensing node and the environmental information sensing node can acquire the spectral data of the substance and the surrounding environment information data which will be stored in the database.Users can access data through dedicated mobile terminals and use spectral analysis models for data analysis to obtain material’s composition information in real time.This paper proposes a design of the mobile sensing software for the spectrum sensing IoT based on the Android system,which can realize the data interaction and spectrum analysis functions of the user and the spectral sensing IoT.The main research contents include the design and development of mobile terminals,the deployment of spectral analysis experiments and spectral analysis models in mobile terminals,and the spectral sensing IoT joint test.The mobile terminal includes four modules: user information,node information,data information,and component sensing.The ViewPager class is used to construct a dual-list interface to display node information,and the data interaction with the cloud server is performed by using the SOAP protocol.The resulting spectral and environmental data can be displayed in the form of graphic and it can be saved in the phone.Further,on the mobile terminal,normalization of spectral data,first-order differential,second-order differential,peak search and other data pre-processing functions are realized,and the spectral analysis model is deployed in the mobile terminal to realize the component analysis function of the substance.The environmental information data acquired by the mobile terminal includes temperature,humidity,carbon dioxide concentration,and light intensity data,and the latest environmental information data can be displayed in real time,and the environmental data can be searched and plotted in a time period.In order to verify the feasibility of component sensing function integration in mobile terminals,an experiment was conducted to quantitatively and qualitatively analyze the spectral data of tea sugar-inducing detection.163 pieces of summer-autumn sugar tea data collected by Shimadzu IRTracer-100 Fourier spectrometer,of which the modeling set contained 110 pieces and the test set contained 53 pieces.Due to the difference in spectral range and resolution between the spectral sensing node and the Fourier spectrometer,the intersection of the spectral sensing node and the Fourier spectrometer wavelength range was selected as the experimental data on the original data set to verify the feasibility of spectral sensing node for modeling tea sugar-inducing detection.The original data has a wavelength range of 1-2.5μm which has 12446 points in each piece of spectral data while the filter bands is 1-1.7μm,and the volume of each spectral data is 178.Using multivariate linear regression algorithm for quantitative analysis of tea sugar incorporation,the model correlation coefficient was 0.93 in the original data wavelength range,and the model correlation coefficient was 0.82 in the filter bands.In the qualitative analysis experiment,the qualitative analysis results are first obtained based on the comparison between the quantitative analysis results and the threshold values.The prediction accuracy of the original data wavelength range modeling is 92.5%,and the screening band modeling prediction accuracy is 90.6%.Secondly,in the one-dimensional convolutional neural network modeling experiment,the Keras framework is used to construct and train the neural network.The original data is reduced by the equal interval sampling method.The optimal interval is 6 and the model accuracy is 90.6%.The quantitative qualitative analysis model is deployed on the mobile terminal.After testing,the mobile terminal achieves the same result as the computer.It verifies the feasibility of the spectral analysis model to implement the component sensing function in the mobile terminal,which lays a foundation for the development of the subsequent spectral sensing Internet of Things.In order to verify the actual operation effect of the dedicated mobile terminal in the spectrum sensing Internet of Things,a spectral sensing Internet of Things joint test was conducted.The instruments used in the experiment and the experimental procedure were introduced.Finally,the experimental results were presented and compared with the results of the PC-side software which processing the same spectral data.From the experimental results,the mobile terminal successfully requests the spectral data from the cloud server for processing and draws an image to find the peak.During the period,the network request is smooth,the image is drawn accurately,and the spectral analysis is processed without errors.The mobile terminal completes communication with the server,and the data acquisition,drawing,saving,and spectral analysis functions are implemented normally,and the expected effect of the design is realized,and the requirements for realizing the spectrum sensing IOT dedicated mobile terminal at present are verified.