Research And Application Of Load Soft Measurement Of Semi-automatic Grinding Machine

With the rapid development of computer technology,the integration of big data,cloud computing,artificial intelligence and other cutting-edge technologies with traditional process industry has attracted more and more attention.Data mining,information processing and data transmission have become an important direction of research,and relevant data platforms such as Hadoop have also developed rapidly.At the same time,with the rapid development of science and technology,the continuous improvement of human comprehensive quality,the concept of green,energy saving,environmental protection has gradually become popular.In this paper,the semi-autogenous mill which is commonly used in grinding process is taken as the research object,and how to make the production process run continuously and stably in the best condition is discussed,so as to improve the utilization rate of resources and achieve the purpose of energy saving and capacity coordination.The main reason why the semi-autogenous mill cannot achieve economic operation is that its load is difficult to be directly detected in real time.In order to ensure the continuous operation of grinding production and prevent the occurrence of empty grinding and full grinding,the staff can only carry out conservative feeding through experience,which leads to the unbalanced operation efficiency.Therefore,it is of great significance to solve the load monitoring problem of semi-autogenous mill,whether for realizing automatic control or for economy and stable operation.Aiming at the problem of load monitoring of semi-autogenous mill,soft sensor technology is used to study in this paper,which involves the key points of data transmission,data pretreatment,data mining and model building process.The main contents are as follows:(1)The research and development of semi-autogenous mill load at home and abroad are reviewed,and the background and significance of this research are discussed.(2)The semi-autogenous mill process which has good application prospect in mine is briefly summarized,and the process mechanism and working characteristics of the semi-autogenous mill are studied,and the influence of load and detection characteristics are deeply discussed.(3)After analyzing the characteristics of the influence of semi-auto mill load,starting from the soft sensor technology,the elements of soft sensor modeling were elaborated,and the support vector machine(SVM)algorithm and the principle of support vector regression machine were described in detail.The soft sensor model based on SVM was applied to the actual prediction of the semi-auto mill running load.(4)To conduct theoretical analysis and simulation research on the problems encountered in the load detection of SVM soft sensor model,involving the selection of auxiliary variables in the process of soft sensor modeling and data processing.Pearson correlation coefficient method was used to analyze the correlation coefficient between the dominant variable and the auxiliary variable,and the correlation degree between each variable and the dominant variable was determined.For the Phase Space of signal reconstruction related to the load of semi-autogenous mill,the selection dimension of auxiliary variables was preliminarily determined.Principal component analysis(PCA)is used to fuse some auxiliary variables to prevent information redundancy.By combining Pearson correlation coefficient,Phase Space reconstruction and PCA technology,an improved algorithm for selecting auxiliary variables is proposed,which is Pearson Phase Space PCA algorithm(PSPCA)to determine final auxiliary variables.The auxiliary variable data samples were filtered by Wavelet Denoising to remove the noise in the original data,and the data set was dimensionless.Finally,PSPCA algorithm,wavelet denoising algorithm and SVM algorithm were combined to form a composite PSPCAW-SVM soft sensor algorithm,and the PSPCAW-SVM soft sensor model was established to monitor the mill load.(5)Establish a soft measurement model of semi-autogenous mill load based on PSPCAW-SVM,and evaluate the accuracy and stability of the model through simulation prediction results and performance index calculation.Finally,the actual running process and operation mode of the model are introduced to realize energy saving and efficient production.