Correlation analysis

Silent Controls correlation analysis improves the understanding of the causal connections in the process. That is useful in many ways, for instance to

  • detect sensor errors
  • improve the control
  • validate process models
  • obtain early warnings from well-known situations
  • identify variables when creating models
  • improve for the operator through multivariate observation systems

Regardless of the cause for the alarm problem, Silent Controls alarm- and correlation analysis helps you to identify the problem and to find effective solutions. Our solutions are often small simple adjustments in the existing system, for example changing the values for the dead band or inserting simple filters.
      Before you rush away to buy expensive add-on systems or create advanced models our much simpler solution should be considered!


The correlation analysis is carried out in four steps:

Data logging. (Normally about 2 weeks of data.)
Correlation analysis with Silent Control software. Characteristics from the signals are extracted and compared.
The analysis results are discussed with the operators and other personal with process knowledge. This takes less then one day to carry out.
The changes are applied into the existing system. No investments in new or add-on systems are necessary. The changes can be carried out during a planned stoppage.