Font Size:     

In industry, collection of data and its storage is necessary for the analysis and monitoring of the process. Computational and graphical techniques are required for the analysis of this data. The useful information buried in this data is required to exploit and integrate this data with the help of simulators. Problems are involved in the integration of the data including missing and incorrect values, high volume of data, noise and uncertainty, high dimensionality, dynamics and time delays for the data containing the time dimension. Lack of supporting data mining tools is the major reason that most of the data remain unexploited. Visualization is one of this tool in which multi dimensional geometry is used to visualize many variables simultaneously and to extract useful information from the process. Large volume and high dimensional challenging issues can be resolved by this method. This method provides graphical displays and animation on which investigators based their observations on all the human senses; hear, vision and even touch is able to get insight into the complex process.Parallel coordinate plot is an addition to multivariable multidimensional visualization for exploration and interpretation of data set. Parallel coordinate plot is a two dimensional plot in which multidimensional data is able to visualize. Series of unbroken line segments representing each to a different variable pass in between the parallel axis. This representation is used for the detection of clustering of objects, correlation between the variables and their effect on each other when these are classified to dependent and independent variables.For the purpose of safety, product quality and cost minimization process monitoring, early fault detection and diagnosis is an important issue in the chemical industries. Statistical process control charts are the part of total quality management. Quality of the product can be improved by the display of the data measured on the process over time and thus detection of process upset, trends and shift is possible.Shewhart chart, cumulative sum(CUSUM) and exponentially weighted moving average (EWMA) are univariatethe statistical process control graphs and used for controlling one dimensional process. First one is oldest and commonly used for the detection of sudden medium to large mean shifts. Second one detects the fault faster than Shewart provided that the change is relatively small. Third one is designed to detect deviation in the process mean.Collinear multivariate data is generated by the processes .Univariate process control charts are misleading whereas multivariable control charts show process variables simultaneously and extraction of information from their correlation structure is possible. These charts help to handle the process over time and identify to detect the regions and points where the process is operating normal and the points where it started to go out of control. Projection methods such as principal component analysis (PCA) and partial least squares (PLS) are using by many researchers now a days as base in the multi dimensional visualization.