Research on Automatic Titration of Rare Earth Acidity Based on Convolutional Neural Network Algorithm

Abstract:Nowadays, with the development of rare earth industry, the market demand is getting bigger and bigger, and the scope of application is getting more and more wide. It is necessary to establish a simple and suitable method for the determination of all kinds of rare earth acidity. In view of the current problems of low efficiency, low accuracy and large difference of titration endpoints, it is difficult to meet the needs of real-time online detection. This paper presents an online analyzer of rare earth acidity based on convolutional neural network algorithm. Convolutional neural network algorithm is recorded through the HD industrial camera samples in the process of the solution color change, the solution of real-time image feature extraction and learning, so as to effectively and accurately realize the solution color automation of chemical reaction, with stepper motor and injection pump components to realize automatic titration process. Image recognition is essentially a feature extraction of image information, and the convolutional neural network algorithm has the advantages that the traditional recognition methods do not have, such as self-training, faster recognition speed, and fewer required features. This instrument combines automatic titration with convolutional neural network to realize the integration of automatic sampling and pretreatment of titration process, titration process and end point judgment, and the instrument can carry out the titration test of five samples at the same time, which greatly improves the titration efficiency and accuracy.