Research On Process Monitoring Method Based On Improved KPCA

As the scale of modern industrial processes is expanding and the complexity of processes is increasing, how to ensure the safety of process operation and improve product quality are two issues need to be solved in industrial production enterprises. Process monitoring technology is an effective way to solve these two issues. Since the complexity and fluctuation of industrial processes, accurate process models are difficult to build and apply. Therefore, application of traditional process monitoring methods based on qualitative or quantitative models subjects to certain limitation.Because of the development of intelligent instrumentations and computer technology in industrial process applications. A large number of high-dimensional and strongly correlated process data is collected and stored. It is difficult to remove redundancy and interference to extract useful information. As an efficient way to deal with the correlation between multivariate, multivariate statistical process monitoring technology is suffered from sustained concerned and developed. In this paper, the following work is done focus on the process change caused by aging equipment, process drift and sensor measurement errors in nonlinear industrial process:(1) Focus on the issues of process data gradually increasing and process change, an adaptive kernel component analysis method is proposed in this paper. This method combines two existing methods, that is moving window kernel component analysis method and exponential weighted kernel component analysis method. When a new sample is available, moving window kernel component analysis method is firstly used. This method judges whether the new sample satisfies the conditions of model updating. If it satisfies the conditions, then the new sample will be accepted to update the process monitoring model, and exponential weighted kernel principal component analysis method is used in model updating. On the contrary, the process monitoring model will not update until the next normal sample is available. Simulation study using the proposed method is carried on fused magnesia furnace process monitoring. Simulation study results demonstrate the feasibility of adaptive kernel principal component analysis method.(2) In practical industrial process, outliers are contained in the collected data, while traditional kernel principal component analysis method is based on the assumption that no outliers in the sample data. Outliers still exist even after mapping them into feature space. Even if the sample data contains only a small amount of outliers, great negative affect will be applied on process model. Therefore, an advanced kernel principal component analysis method is proposed in this paper, which defines a loss function in feature space in the sense of minimum reconstruction error. Then iteration with penalty will be carried to obtain the principal components, which can eliminate the adverse effects of outliers. Whenever a new sample is available, reconstruct it with the previous transfer matrix and calculate the reconstruct error. If the new sample is an outlier, then update model with the reconstructed sample, otherwise update model with the original sample. Simulation results show that the advanced KPCA method can reduce the impact of outliers, and improve the accuracy of the process monitoring model as well.