Abnormality determination apparatus, abnormality determination model generation method, and abnormality determination method

An abnormality determination method of generating a model and determining an abnormality of a facility for performing a predetermined operation, the abnormality determination method comprising:

• a time-series signal clipping step of performing during normal operation of the facility, K times of clipping from two or more time-series signals indicating an operation state of the facility, where K is a positive integer;
• a normal vector registration step of setting types of two or more time-series signals clipped in the time-series signal clipping step as M types, constructing an M-dimensional vector including M types of variables at a same time, and registering the constructed vector as a normal vector, where M is a positive integer;
• an abnormality determination model registration step of setting an abnormality determination flag to a first type in a case where a maximum value of correlation between variables is less than a predetermined value, setting the abnormality determination flag to a second type in a case where the maximum value of correlation between variables is the predetermined value or more, and in a case where the abnormality determination flag is of the second type, performing a principal component analysis on a registered normal vector group to calculate a transform coefficient of a principal component and registering one or more calculated transform coefficients of principal components as an abnormality determination model,
• a time-series signal collection step of collecting a time-series signal indicating an operation state of the facility and a trigger candidate time-series signal that decides a condition for clipping the time-series signal indicating the operation state from a predetermined monitoring target section;
• a trigger condition decision model generation step of specifying beforehand a start time of the monitoring target section at which clipping regarding the time-series signal indicating the operation state of the facility is performed, generating label data having ON for a label of the start time and OFF for other times, and generating, by machine learning, a trigger condition decision model having each value of one or more of trigger candidate time-series signals at each of one or more times as an input and having the label data at each of the one or more times as an output,
• wherein the time-series signal clipping step clips the time-series signal based on the trigger condition decision model,
• a time-series signal clipping out step of clipping two or more time-series signals indicating an operation state of the facility;
• and
• an abnormality determination step of determining the abnormality of the facility;
• wherein the abnormality determination step includes:
• determining whether the abnormality determination flag of the two or more time-series signals clipped in the time-series signal clipping out step is of a first type or a second type,
• calculating, when the abnormality determination flag is of the first type, a distance from a registered normal vector, extracting a predetermined number of the one or more registered normal vectors as neighboring data in ascending order of a plurality of distances that include the distance, calculating a distance between a centroid vector of the neighboring data and the M-dimensional vector as a target of abnormality determination, and performing abnormality determination of the facility based on the distance between the centroid vector of the neighboring data and the M-dimensional vector,
• calculating, when the abnormality determination flag is of the second type, a deviation from a principal component based on a transform coefficient of the principal component, which has been calculated in advance, and performing abnormality determination of the facility based on the deviation, and
• determining necessity of repair of the facility and performing the repair of the facility based on a frequency of determinations at which the facility has been determined to be abnormal in a predetermined period,
• wherein the facility is a facility of at least one of an iron making process, a manufacturing process for petroleum products or chemicals, and a research institution.

An abnormality determination apparatus for determining an abnormality of a facility performing a predetermined operation, the abnormality determination apparatus comprising:

• a time-series signal clipping unit configured to perform clipping of a time-series signal from two or more time-series signals of M types of variables indicating an operation state of the facility, where M is a positive integer;
• and
• an abnormality determination unit configured to, for the time-series signals clipped by the time-series signal clipping unit,
• set an abnormality determination flag as a first type when a maximum value of correlation between the variables is less than a predetermined value,
• set the abnormality determination flag as a second type when the maximum value of correlation between the variables is the predetermined value or more, and
• construct an M-dimensional vector including the M types of the variables at a same time,
• wherein when the abnormality determination flag is of the first type, the abnormality determination unit is configured to
• calculate a distance from a plurality of registered normal vectors clipped as time-series signals indicating the operation state of the facility during normal operation of the facility,
• extract a predetermined number of the normal vectors as neighboring data in ascending order of a plurality of distances which includes the distance,
• calculate a distance between a centroid vector of the neighboring data and the M-dimensional vector as a target of abnormality determination, and
• perform abnormality determination on the facility based on the distance between the centroid vector of the neighboring data and the M-dimensional vector, and
• wherein when the abnormality determination flag is of the second type, the abnormality determination unit is configured to
• calculate a deviation from a principal component based on a transform coefficient of a registered principal component calculated by performing a principal component analysis on a group of the registered normal vectors, and
• perform abnormality determination on the facility based on the deviation, wherein
• the time-series signal clipping unit is configured to clip the time-series signal based on a trigger condition decision model, and
• the trigger condition decision model has been generated by machine learning, for the time-series signal indicating the operation state of the facility, by specifying beforehand a start time of a monitoring target section at which clipping is performed, generating label data having ON for a label of the start time and OFF for other times, having each value of one or more of trigger candidate time-series signals at each of one or more times as an input, and having the label data at each of the one or more times as an output,
• wherein the abnormality determination unit is further configured to determine necessity of repair of the facility and cause the repair the facility based on a frequency of determinations at which the facility has been determined to be abnormal in a predetermined period, and
• wherein the facility is a facility of at least one of an iron making process, a manufacturing process for petroleum products or chemicals, and a research institution.