Improved Mahalanobis Distance Based JITL-LSTM Soft Sensor for Multiphase Batch Processes

Intra-Batch Evolution Based Process Monitoring for Multiphase Batch Processes

Batch-wise variations, called intra-batch evolution here, widely exist in batch processes. In this paper, intra-batch evolution is tracked and monitored for multiphase batch processes. First, a batch cycle is divided into multiple phases. Within each phase, sliding windows are constructed for analysis of intra-batch relative variations, based on which different process modes are separated in or...

An LMI Approach to Iterative Learning Control Based on JITL for Batch Processes

Improved Batch Pan Monitoring and Control – a soft sensor approach

Conductivity-based vacuum pan controllers dominate the Australian sugar industry. A new scheme, based on Brix control of the massecuite and mother molasses is proposed. Multi-loop control of the process is demonstrated, whereby the massecuite Brix is controlled by manipulating the liquor/A molasses feed rate to the pan and the mother molasses Brix is controlled by manipulating the steam flow ra...

Normalized Clustering Algorithm Based on Mahalanobis Distance

FCM (fuzzy c-means algorithm) based on Euclidean distance function converges to a local minimum of the objective function, which can only be used to detect spherical structural clusters. The added fuzzy covariance matrices in their distance measure were not directly derived from the objective function. In this paper, an improved Normalized Clustering Algorithm Based on Mahalanobis distance by t...

Anisotropic Superpixel Generation Based on Mahalanobis Distance

Superpixels have been widely used as a preprocessing step in various computer vision tasks. Spatial compactness and color homogeneity are the two key factors determining the quality of the superpixel representation. In this paper, these two objectives are considered separately and anisotropic superpixels are generated to better adapt to local image content. We develop a unimodular Gaussian gene...