×

Statistical processing methods used in abnormal situation detection

  • US 7,752,012 B2
  • Filed: 04/04/2006
  • Issued: 07/06/2010
  • Est. Priority Date: 04/04/2005
  • Status: Active Grant
First Claim
Patent Images

1. A method of detecting an abnormal situation associated with a process plant, comprising:

  • receiving measured data pertaining to a process parameter sensed by at least one sensor device associated with the process plant;

    collecting a first set of data points for the process parameter from the measured data;

    using a processor for determining a frequency component of the process parameter based on the collected first set of data points for the process parameter;

    using a processor for determining a dominant system time constant from the frequency component and setting a block length for calculating a statistical measure from the measured data based on the dominant system time constant;

    using a processor determining one or more statistical measures associated with the process parameter using the measured data received during a block of time corresponding to the block length; and

    using a processor for detecting an abnormal situation in the process plant based on the one or more statistical measures associated with the process parameter.

View all claims
  • 1 Assignment
Timeline View
Assignment View
    ×
    ×