On-Line Measurement Of Ceramic Paste Inner Stress Based On The Soft-Measurement Technology

Ceramic is extensively used in modern life, ceramic tiles, as a cera -mic products, is closely interrelated to people's life. But, the ceramic tiles exists the problem of appearing cracks easily plastic extrusion molding production process. This problem seriously influences the density, mechanical strength, surface smoothness and yield of ceramic paste, etc. At the same time, the application and dissemination of ceramic plastic extrusion molding clay materials production technology and production equipment is hindered. The main causes of this problem are due to the uneven distribution of inner stress when the mud in the vacuum pug mill extrusion process. In order to solve this problem, the first step is to measure the distribution of paste inner stress, and then taking compensatory refine to the mud.In this issue, the soft-measurement technique is used in realizing the online estimation of paste inner stress for the problem that paste inner stress can't direct measurement. Based on rheological theory and motion characteristics of mud in vacuum pug mill, the reasons and a variety of factors caused inner stress uneven distribution are analyzed in the extrusion forming process of ceramic tiles. The inwall pressure of vacuum pug mill head is selected as the auxiliary variable for indirect measurement of ceramic inner stress. Because neural network can appro -ximate any nonlinear function and has the ability of self-learning and self-adaptation, neural network is used to build soft-measurement model in this issue. Taking laboratory measurement data for the training sample, a BP (back-propagation) network and RBF (radial basis function) network is separately used for soft-measurement modeling. Simulation training is carried out on Matlab, and simulation effect and generalization ability of the two neural network models are compared. The simulation results show that the established soft-measurement model based on neural network has certain reliability.In order to solve the problem that ceramic products can't predict on line, regard ceramic production process as a grey system, a new diagnosis method of ceramic crack is proposed. For the measurement data of inwall pressure of vacuum pug mill head, takes wavelet packet decomposition, extracts its characteristic vectors and then the possibili -ty of cracking can be predicted timely in the refining process through gray relation analysis. Experimental studies show that the ceramic crack diagnosis method based on wavelet packet and grey system theory has a certain effectiveness, meanwhile, it has the characteristics of don't need a large number samples and simple calculation.In this issue, to achieve above functions, the whole soft-measure -ment system is set up on the LabVIEW platform of NI company. This system realize the real-time data acqui -sition and display of inwall pressure of vacuum pug mill and ceramic paste inner stress. Thereby guide the monitoring process online, reduce waste of resources, and improve the quality of products.