Performance Assessment And Improvement Of Process Quality Control

It is very important to improve product quality and competition for an enterprise applying advanced quality technology and quality control methods, which appeal researcher and practicer to design and produce low-cost, short-period, high quality & reliability product. These methods make enterprises win superiority competition. The tendency of current quality engineering is to reduce, restrain and control the variation in the process.Variation is the fundamental cause of poor quality. Reducing and controlling variation has become a core research domain in modem quality engineering science. In this paper, we systematically study some theories, methods and implementation techniques of reducing and controlling variation of a product with different quality characteristics in its forming process by using positive analysis and simulation, which are based on variation theory and process conception.At first, some shortcomings of traditional control charts are pointed out in this paper. For a process that quality characteristic is univariable normal distribution, average product length is proposed as a tool to assess monitoring efficiency of control charts. Sensitivity analysis of each parameter is made by this way. And then, we can get a suitable selection of each parameter to build an optimal design model of traditional control charts.Exponentially weighted moving average chart is often used to detecting small shifts or drifts in the mean of a process. We also use average product length as a tool of monitoring efficiency assessment to build optimal design model of EWMA charts. The optimal design method is an improver of general EWMA chart. The chart made by this method is more sensitive to small shifts in process.In process control, traditional methods are based on the tacit assumption that process output quality characteristics are normal distribution. But in practice, there are numbers of non-normal processes. On the basis of analysis quality control methods used in non-normal process, we propose weight variation method as a tool to split distribution to construct asymmetry control limits. Then average product length is used as a monitoring efficiency assessment tool to build non-normal optimal design models of Shewhart control chart and EWMA chart.Multivariate quality control is an important aspect of product quality assurance. Based on analysis of several general multivariate quality control charts, this paper propose a kind of multivariate control chart under different multiple dots alarm rules. This chart can detect quickly small changes in the mean vector of a multivariate process. At the same time, a diagnostic method is proposed by using dummy variable regression technique. This method can solve effectively the question that multivariate control chart cannot identify which characteristic or group of characteristics are out of control.At last, for a process that quality characteristic is auto-correlated, the residuals are used to construct EWMA chart to monitor little shifts of process mean and variance. Comparing with other methods, we can illustration that this kind EWMA residuals chart has better efficiency for auto-correlated processes. Then, a process quality control method of integrating statistical process control and automatic process control in a reasonably way is discussed in this paper.The improving methods of process quality control proposed in this paper adapt respectively to multiple conditions, such as normal distribution, skewed distribution, small shift, multivariate, and auto-correlated. It is important for organizations to improve monitoring efficiency, enhance product quality, minimize quality loss and strengthen market competitiveness.