Statistical signatures used with multivariate statistical analysis for fault detection and isolation and abnormal condition prevention in a process
First Claim
1. A multivariate statistical analysis system comprising:
- a processor;
a memory coupled to the processor; and
a first set of machine-readable instructions for execution by the processor, stored in the memory, and comprising a first analysis tool that, when executed;
receives a first set of statistical measures indicative of the operation of a process in a process plant, each statistical measure comprising the result of a statistical calculation performed on two or more measured values of a particular parameter of the process;
performs a multivariate statistical analysis on the first set of statistical measures; and
outputs a representation of the operation of the process based on the first set of statistical measures.
1 Assignment
0 Petitions
Accused Products
Abstract
A system and method for monitoring a process in a process plant and detecting an abnormal condition includes collecting data representative of the operation of the process, performing a multivariate statistical analysis to represent the operation of the process in a known state based on a set of collected reference data, where the reference data includes a statistical measure of the operation of the process in the known state. The system and method may further include representing the operation of the process in an unknown state based on a set of monitored data, where the monitored data includes a statistical measure of the operation of the process in an unknown state, and using the output of the multivariate statistical analysis as an input, and comparing the process in the unknown state to the multivariate statistical representation of the operation of the process in the known state to determine the operational state of the process.
68 Citations
49 Claims
-
1. A multivariate statistical analysis system comprising:
-
a processor; a memory coupled to the processor; and a first set of machine-readable instructions for execution by the processor, stored in the memory, and comprising a first analysis tool that, when executed; receives a first set of statistical measures indicative of the operation of a process in a process plant, each statistical measure comprising the result of a statistical calculation performed on two or more measured values of a particular parameter of the process; performs a multivariate statistical analysis on the first set of statistical measures; and outputs a representation of the operation of the process based on the first set of statistical measures. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16)
-
-
17. A method for monitoring a process in a process plant, the method comprising:
-
collecting data representative of an operation of the process; calculating a statistical measure on the collected data representative of the operation of the process in a known state, wherein the statistical measure is the result of a statistical calculation on two or more measured values of a particular parameter of the process; and performing a multivariate statistical analysis to represent the operation of the process in a known state based on a set of collected reference data comprising the statistical measure of the operation of the process in the known state. - View Dependent Claims (18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34)
-
-
35. A system for monitoring a process in a process plant, the system comprising:
-
a first analysis tool for; receiving a first set of statistical measures, each of the first set of statistical measures being the result of a statistical calculation on two or more measured values of a particular parameter of the process, each of the first set of the statistical measures indicative of the operation of the process while the process is in one or more known conditions; and performing a multivariate statistical analysis on the first set of statistical measures to generate a multivariate statistical representation of the operation of the process in the one or more known conditions; and a second analysis tool for; receiving a second set of statistical measures, each of the second set of statistical measures being the result of a statistical calculation on two or more measured values of the particular parameter of the process, each of the second set of statistical measures indicative of the operation of the process while the process is in an unknown condition; receiving the multivariate statistical representation of the operation of the process; projecting the received second set of statistical measures onto the multivariate statistical representation of the operation of the process to represent the operation of the process in the unknown condition; and determining the unknown condition based on the observation of the second set of statistical measures projected onto the multivariate statistical representation as compared with the observation of the first set of statistical measures projected onto the multivariate statistical representation. - View Dependent Claims (36, 37, 38, 43, 44, 45, 46, 47, 48, 49)
-
-
39. A system for monitoring a process in a process plant, the system comprising:
-
a first analysis tool for; receiving a first set of statistical measures, each of the first set of statistical measures being the result of a statistical calculation on two or more measured values of a particular parameter of the process, each of the first set of the statistical measures indicative of the operation of the process while the process is in one or more known conditions; and performing a multivariate statistical analysis on the first set of statistical measures to generate a multivariate statistical representation of the operation of the process in the one or more known conditions; and a second analysis tool for; receiving a second set of statistical measures, each of the second set of statistical measures being the result of a statistical calculation on two or more measured values of the particular parameter of the process, each of the second set of statistical measures indicative of the operation of the process while the process is in an unknown condition; receiving the multivariate statistical representation of the operation of the process; projecting the received second set of statistical measures onto the multivariate statistical representation of the operation of the process to represent the operation of the process in the unknown condition; and determining the unknown condition based on the observation of the second set of statistical measures projected onto the multivariate statistical representation as compared with the observation of the first set of statistical measures projected onto the multivariate statistical representation; wherein the first analysis tool is adapted to assign an observation of the first set of statistical measures projected onto the multivariate statistical representation to an abnormal condition according to a score discriminant; and further wherein the score discriminant comprises a maximum score discriminant calculated as; x=vector of original process variable measurements gi=likelihood that x belongs to abnormal condition class i x i=a mean vector of all observations belonging to class iP =a principal component analysis loading matrix Si=a covariance matrix of all observations belonging to class i pi=a priori probability of an observation belonging to class i. - View Dependent Claims (40)
-
-
41. A system for monitoring a process in a process plant, the system comprising:
-
a first analysis tool for; receiving a first set of statistical measures, each of the first set of statistical measures being the result of a statistical calculation on two or more measured values of a particular parameter of the process, each of the first set of the statistical measures indicative of the operation of the process while the process is in one or more known conditions; and performing a multivariate statistical analysis on the first set of statistical measures to generate a multivariate statistical representation of the operation of the process in the one or more known conditions; and a second analysis tool for; receiving a second set of statistical measures, each of the second set of statistical measures being the result of a statistical calculation on two or more measured values of the particular parameter of the process, each of the second set of statistical measures indicative of the operation of the process while the process is in an unknown condition; receiving the multivariate statistical representation of the operation of the process; projecting the received second set of statistical measures onto the multivariate statistical representation of the operation of the process to represent the operation of the process in the unknown condition; and determining the unknown condition based on the observation of the second set of statistical measures projected onto the multivariate statistical representation as compared with the observation of the first set of statistical measures projected onto the multivariate statistical representation; wherein the first analysis tool is adapted to assign an observation of the first set of statistical measures projected onto the multivariate statistical representation to an abnormal condition according to a score discriminant; and further wherein the score discriminant is calculated as;
gi(x)=(x−
x i)TP(PTSiP)−
1PT(x−
x i)+ln[det(PTSiP)]wherein; x=vector of original process variable measurements likelihood that x belongs to abnormal condition class i gi=a mean vector of all observations belonging to class i x i=a mean vector of all observations belonging to class iP=a principal component analysis loading matrix Si=a covariance matrix of all observations belonging to class i. - View Dependent Claims (42)
-
Specification