Research And Development Of PH Detection System Based On Spectral Analysis

The detection of pH value is one of the most important links in many industries and fields,such as environmental protection,drug preparation,food processing and dyeing process control.The traditional glass pH sensor has some shortcomings such as poor anti-interference ability,short life and being easy to be poisoned.Aiming at the problem of continuous detection of pH in industrial automation field,this thesis presented a pH detection system based on spectral analysis.The system has the characteristics of stable signal,strong anti-interference ability,long life and no pollution.It realized the continuous measurement of solution pH in laboratory,and provided a new idea for the products development.The main contents of this thesis are as follows:1.According to the characteristics of the optical pH sensor,the overall scheme of the pH detection system based on spectral analysis was designed.In view of the characteristics of sensitive membranes embedded with various indicators(Congo Red,Bromophenol Blue,Cresol Red and Chlorophenol Red),a spectral signal acquisition system of the pH detection system was designed based on the principle of indicator discoloration and the transmission principle of pH detection,and the spectral signal acquisition of the detection light was realized.A new type of sampling cell for micro pH solution spectrum signal was proposed,and the structure design of the sampling cell was completed.Experiments show that compared with the traditional cuvette sampling cell,the proposed micro-structure sampling cell has small size,less sample consumption,high sampling efficiency and strong data stability.2.A spectrum signal processing system of pH detection system based on ARM architecture was designed.The functions of the signal processing system include: driving of spectrometer,data storage,feature extraction of spectral data,prediction and display of pH value,human-computer interaction and system setting,etc.The hardware circuit design of the spectrum signal processing system was completed,including power module,spectrometer interface circuit,processor peripheral circuit and display circuit.The development and debugging of the spectrum processing software was also finished.The functions of processing the absorption spectrum signal,predicting and displaying the solution pH value were realized.3.The prediction algorithm of the absorbance-pH model was studied.Based on the basic principles of Support Vector Machine Regression(SVR)and Random Forest(SVR),an absorbance-pH model based on RF-SVR was established,and the parameters of SG convolution algorithm and RF-SVR were optimized.Finally,the function of predicting the solution pH value based on the absorbance spectrum data was realized.The prediction results show that the maximum absolute error and minimum absolute error of the model based on RF-SVR were 0.196 and 0.007 respectively.4.The experimental analysis of the system was completed.A series of pH buffers with a pH range of 3.50-9.00 and an interval of 0.5 were prepared.The spectral data of the detected light passing through the buffers was collected at room temperature.The performance of the proposed micro-structure sampling cell and the absorbency-pH model based on RF-SVR were analyzed by comparative experiments.The experimental results show that the maximum absolute error of the model based on RF-SVR was reduced by 0.078 compared with that of the model using traditional cuvette cell.Compared with the absorbance-pH model based on SVR,the maximum absolute error of the model predicted value based on RF-SVR was reduced by 0.075.