Adaptive modeling of changed states in predictive condition monitoring
AssigneeSmartSignal Corporation
InventorsStephan W. Wegerich, David R. Bell, Xiao Xu
Life of this patent
Estimating the term…
Abstract
An improved empirical model-based surveillance or control system for monitoring or controlling a process or machine provides adaptation of the empirical model in response to new operational states that are deemed normal or non-exceptional for the process or machine. An adaptation decision module differentiates process or sensor upset requiring alerts from new operational states not yet modeled. A retraining module updates the empirical model to incorporate the new states, and a pruning technique optionally maintains the empirical model by removing older states in favor of the added new states recognized by the model.
Claims37
A method for adapting an empirical model used in monitoring a system, comprising the steps of:
- receiving actual values of monitored parameters characterizing the system;
- generating from the empirical model estimates of the parameter values in response to receiving actual values;
- calculating a global similarity score for a comparison of a set of estimated parameter values and a related set of actual parameter values;
- and
- adapting the empirical model to account for at least some received actual values based on the global similarity score.
A method for adapting a reference data set of an empirical model used in empirically modeling a system, comprising the steps of:
- receiving actual values of monitored parameters characterizing the system, including at least one control parameter;
- determining that the actual value for the at least one control parameter lies outside a predetermined range;
- adapting the empirical model in response to said determining step, to account for the received actual values.
An apparatus for monitoring an operation of a system characterized by operational parameters, comprising:
- an empirical model for generating estimates of parameter values in response to receiving actual parameter values characterizing the operation of the system;
- a global similarity engine for generating a global similarity score for a comparison of a set of estimated parameter values and a related set of actual parameter values from said system;
- and
- an adaptation module for adapting the empirical model to account for at least some received actual parameter values based on the global similarity score.
A computer program product, said computer program product comprising a computer usable medium having computer readable program code thereon for causing a computer processor to execute a series of instructions, for monitoring an operation of a system characterized by operational parameters, comprising:
- computer readable program code means for receiving actual parameter values characterizing operation of the system, including at least one control parameter;
- computer readable program code means for determining that an actual value for the at least one control parameter lies outside a predetermined range;
- computer readable program code means for storing data representing an empirical model of the system in a data storage means;
- and
- computer readable program code means for adapting the stored empirical model data in response to said determining step, to account for the received actual values.
An apparatus for monitoring a condition of a system selected from the set of a machine and a process, using monitored sensor information from said system, comprising:
- a memory means for storing an empirical model of sensor information characterizing said system;
- a computer processing unit disposed to receive actual values of the monitored sensor information characterizing said system, including at least one control parameter, and generate estimates of at least some of the monitored sensor information for evaluating the condition of said system;
- and
- an adaptation program module executable in said computer processing unit for adapting the empirical model to account for at least some received actual values upon determining that the received actual value for the at least one control parameter lies outside a predetermined range.
Dependencies are inferred from the claim text.