Development Of Portable Soil Organic Matter Tester Based On Spectral Characteristics

In modern agricultural production,in order to achieve scientific fertilization,soil nutrient content needs to be accurately detected to achieve balanced soil use.In this paper,based on the principle of soil spectroscopy measurement,the characteristics of soil organic matter spectral detection were summarized and analyzed,and the portable soil organic matter tester was studied.The main contents of this study include:(1)characteristic wavelength selection of soil organic matter tester.In close emperor mountain in Shanxi Province,the table mountain,Zhong Yang,Lou Fan,Ning Wu,Tai Gu collected 50 six in soil samples as the research object,the soil sample was taken to the air drying,sieving,drying and other processing,using Field Spec3 spectrometer for spectrum data collection,after use ASCII Export software to calculating mean and splicing correction data pretreatment,and Export to Excel spreadsheets,finally drawn by Matlab software absorbance spectra and analyzed.By using the two-dimensional spectral correlation technique,the soil spectral signal map was extended to the second dimension to distinguish the small peaks and weak peaks that could not be accurately analyzed in the one-dimensional spectrum.The comparison and analysis of the two-dimensional spectra shows that there is a strong autocorrelation peak near 600 nm,which well contains the spectral information of soil organic matter.600 nm can be used as the characteristic wavelength of the organic matter tester,which can be used to select the light source of the tester.(2)development of soil organic matter tester.A portable soil organic matter tester with low cost and low power consumption based on spectral characteristics was developed.Instruments near the band at 600nm-610 nm high brightness,small volume,long service life of the SMD LED as light source,with high sensitivity,strong anti-jamming capability,In Ga As photoelectric diode as A detector,AT89C51 as the data processor,using optical fiber as signal transduction,through amplifier and A/D converter to complete optical signals sent,acquisition and processing.Instrument after completion of production,in a dark room to repetitive inspection instruments,select three groups of soil sample is measured,measured in each group,10 times test interval to 1 minute,records of test results is performed,the result shows that the first set of test signal maximum value is 1.202,the minimum value of 1.067,with a mean of 1.117,a second set of test signal maximum value is 1.435,the minimum value of 1.334,with a mean of 1.393,the third group test signal maximum value is 1.349,the minimum value of 1.245,with a mean of 1.280,shows that the instrument has good repeatability.After that,10 groups of soil samples were selected for soil organic matter test,the test value and actual value were recorded and the model was established.Instrument reliability test,for the reduction of field soil acquisition conditions,select without drying,the experiments of soil sample observation test results show that the measured value of the instrument and the average error is 17.6%between the real value,analysis of measurement results may be affected by soil moisture factor,so need to seek a method to eliminate the influence of soil moisture detection of organic matter to improve the instruments.(3)research on eliminating the influence of soil moisture on instrument measurement accuracy.The soil samples collected in the above tests were configured with different moisture contents(5%,10%,15%and 17%),and the spectral data were collected and processed.The absorbance spectra were drawn and analyzed,and it was found that there was a strong absorption effect of water in the soil absorbance spectra,which seriously affected the detection of soil organic matter content.In this study,Nonnegative matrix factor(NMF)was used to decompose and reconstruct the soil spectral data,and Partial least-squares(PLS)was used to establish a prediction model to quantitatively predict the soil organic matter,and the predicted value of the organic matter content in the wet soil samples before and after the reconstruction was analyzed.The results showed that the absorbance spectra of the reconstructed wet soil had a higher fitting degree with that of the dry soil samples,and the Predicted correlation coefficient(R)of the reconstructed spectral data for soil organic matter was increased by 0.059,the Predicted standard deviation(SEP)was decreased by 0.154,and the root-mean square error(RMSEP)was decreased by 0.718.The results show that the method can effectively reduce the influence of water on soil organic matter spectral detection,and can be applied to portable soil organic matter tester to improve the reliability of the instrument.However,due to the large amount of data required by this method and the large number of iterations,it cannot be realized on the AT89C51 single chip microcomputer adopted by the instrument.It is considered to use single chip microcomputer to complete data collection and transmission in the subsequent research,and the upper computer develops data processing software to process the data,so as to improve the reliability and applicability of the tester.