WO2021229815A1 - Information processing device, evaluation method, and evaluation program - Google Patents

Abstract

An information processing device of one embodiment of the present invention includes: an acquisition unit that acquires a plurality of change scores by finding, in each of a plurality of window periods included in a learning period, a change score having as a component the degree of change, for each of plurality of sensors that measure the state of an object to be evaluated, of a sparse structure model created from time-series measurement data obtained from the sensors that has been measured in a window period after a reference window period, with respect to a sparse structure model created from the time series measurement data measured in the reference window period; a specification unit that clusters a plurality of change scores and specifies the position of the cluster; and an evaluation unit that evaluates deterioration of the object to be evaluated on the basis of movement of the position of the cluster generated by adding to the cluster at least one change score found in at least one window period included in a period to be evaluated.

Description

Information processing equipment, evaluation method, and evaluation program

The present invention relates to an information processing device, an evaluation method, and an evaluation program.

For example, technologies for detecting abnormalities and deterioration of industrial equipment, industrial plants, information processing systems, and other equipment and facilities have been developed. For example, it is conceivable to install a sensor in a device or equipment, detect an abnormality in the device or equipment, or estimate the degree of deterioration based on the time-series data obtained from the sensor.

Regarding this, for example, techniques for monitoring and diagnosing abnormalities and deterioration of equipment and facilities are known (for example, Patent Documents 1 to 3).

Japanese Unexamined Patent Publication No. 2016-95751 Japanese Unexamined Patent Publication No. 2004-86897 Japanese Unexamined Patent Publication No. 2016-105678

However, in addition to equipment abnormalities and deterioration, the measurement data measured by the sensor includes fluctuations based on events that are not closely correlated with equipment abnormalities and deterioration, such as fluctuations in measured values due to seasonal changes. There is. Since there can be innumerable such factors, it is difficult to identify all of them and exclude them from the measurement data. Therefore, it is desired to provide a technique capable of evaluating deterioration of equipment and facilities from measurement data that can fluctuate under the influence of various factors.

In one aspect, the present invention aims to evaluate deterioration of equipment, equipment, etc. from measurement data that can fluctuate under the influence of various factors.

The information processing apparatus according to one aspect of the present invention is a sparse structure model generated from the data measured during the reference window period among the time-series measurement data obtained from a plurality of sensors that measure the state of the evaluation object. For each of the multiple window periods included in the learning period, the change score, which is based on the degree of change for each sensor of the cluster structure model generated from the data measured in the window period after the reference window period, was obtained. An acquisition unit that acquires a plurality of change scores, a specific unit that identifies the position of a cluster by clustering a plurality of change scores, and at least one change score obtained in at least one window period included in the evaluation target period. It includes an evaluation unit that evaluates the deterioration of the evaluation target based on the movement of the cluster position caused by adding to the cluster.

It is possible to evaluate the deterioration of equipment and facilities from measurement data that can fluctuate under the influence of various factors.

It is a figure which illustrates the deterioration of a plurality of types of a boiler tube. It is a figure which illustrates the block structure of the information processing apparatus which concerns on embodiment. It is a figure which illustrates the flow of the evaluation process of deterioration which concerns on embodiment. It is a figure which illustrates the operation flow of the process which generates the sparse structure model of the window period which is the reference which concerns on embodiment. It is a figure which illustrates the operation flow of the deterioration evaluation process which concerns on embodiment. It is a figure which illustrates the flow of the evaluation process of deterioration which concerns on embodiment. It is a figure which illustrates the movement of a plurality of clusters. It is a figure which illustrates the determination information which concerns on embodiment. It is a figure which illustrates the hardware configuration of the computer for realizing the information processing apparatus which concerns on embodiment.

Hereinafter, some embodiments of the present invention will be described in detail with reference to the drawings. The corresponding elements in a plurality of drawings are designated by the same reference numerals.

As described above, for example, it is conceivable to install a sensor that measures the state of equipment or equipment and estimate the state of deterioration of equipment or equipment based on the time-series measurement data obtained from the sensor. In the following, an object for evaluating deterioration of equipment or equipment may be referred to as an evaluation object.

As an example, suppose you want to manage the deterioration and remaining life of the boiler tube group of a thermal power plant. In this case, for example, a temperature sensor such as an optical fiber temperature sensor and a sensor such as a vibration sensor are installed in an arrangement capable of measuring the temperature and vibration of the boiler tube group. Then, it is conceivable to evaluate the deterioration of the boiler tube group of the thermal power plant from the change of the measurement data such as the temperature and vibration of each part of the boiler tube group measured by the sensor, and estimate the remaining life. Then, for example, if deterioration can be evaluated and the remaining life can be estimated, the execution timing of maintenance work such as inspection, replacement, and repair can be optimized. In addition, for example, it is possible to avoid a situation such as an unplanned operation stop due to a leak in the boiler tube.

However, changes in the measurement data of the sensor may include information caused by factors other than deterioration of the evaluation target. As an example, it is assumed that a sensor acquires time-series measurement data for one year. In this case, the measured data may fluctuate due to, for example, differences between day and night, seasons and weather differences. Further, when the evaluation object includes a plurality of operation modes such as high-speed operation, low-speed operation, and stop, the measurement data of the sensor may be changed by changing the operation modes. Furthermore, the measurement data of the sensor may fluctuate due to changes in the state of peripheral equipment such as the operation and stop of the air conditioner arranged around the evaluation target.

It is desirable to eliminate fluctuations in measurement data based on factors that have a weak correlation with the deterioration of the evaluation target when evaluating the deterioration of the evaluation target. However, it is difficult to identify all of these factors, and it is practically difficult to exclude fluctuations caused by factors that are weakly correlated with the deterioration of the evaluation target from the measurement data. Even if it can be eliminated, it takes a lot of time and effort to remove the fluctuation according to the factor having a weak correlation with the deterioration of the evaluation target from the measurement data.

Therefore, it is desired to provide a technology that can evaluate the deterioration of the evaluation target from the measurement data of the sensor that can fluctuate under the influence of various factors.

Furthermore, for example, there may be multiple types of deterioration of the evaluation target. FIG. 1 is a diagram illustrating a plurality of types of deterioration of a boiler tube. In FIG. 1, the factors of deterioration, the classification of deterioration, and the contents of deterioration are associated with each other. Deterioration factors of the boiler tube include creep, fatigue, corrosion / wear, and the like. In addition, the classification of deterioration according to the way of occurrence of each deterioration factor is shown. For example, creep is a long-term creep in which low-speed deformation occurs under long-term stress action at high temperature, and a short-time in which rapid deformation occurs when overheated due to an obstacle in the flow of steam due to scale. It can be classified as creep. In this way, the deterioration that occurs in the boiler tube is wide-ranging. The appearance of deterioration in the measurement data of the sensor may differ depending on the type of deterioration.

Therefore, it is also desired to provide a technique capable of evaluating deterioration according to a plurality of types of deterioration of the evaluation target from the measurement data. Hereinafter, the evaluation of deterioration of the evaluation object according to the embodiment will be described.

FIG. 2 is a diagram illustrating a block configuration of the information processing apparatus 200 according to the embodiment. The information processing device 200 may be, for example, a computer such as a server computer, a personal computer (PC), a mobile PC, or a tablet terminal. The information processing device 200 includes, for example, a control unit 201, a storage unit 202, and a communication unit 203. The control unit 201 includes, for example, an acquisition unit 211, a specific unit 212, an evaluation unit 213, and the like, and may also include other functional units. The storage unit 202 of the information processing apparatus 200 stores information such as determination information 800, which will be described later, for example. The communication unit 203 may communicate with another device such as a sensor according to the instruction of the control unit 201, for example. The details of each of these units and the details of the information stored in the storage unit 202 will be described later.

FIG. 3 is a diagram illustrating the flow of deterioration evaluation processing according to the embodiment. FIG. 3A shows measurement data acquired from a plurality of sensors that measure the state of the evaluation object. The control unit 201 of the information processing apparatus 200 may, for example, directly collect measurement data from a plurality of sensors, or may acquire measurement data measured by a plurality of sensors from another communication device.

Subsequently, the control unit 201 may execute preprocessing on the measurement data, for example. The preprocessing involves, for example, the removal of unwanted data that can be noise. For example, data in a state different from the normal stable operating state, such as measurement data when the evaluation target is being cleaned, inspected, or just started and the operation is not stable yet. If the measurement data of the included period is brought into the evaluation of deterioration, it may become noise. Therefore, the control unit 201 may execute a preprocessing for removing the measurement data during the non-normal operation in advance. In another embodiment, the control unit 201 may arrange the window period described below so as not to include the measurement data during the abnormal operation.

Subsequently, the control unit 201 sets a plurality of window periods for the measurement data. The window period may be, for example, a period of a predetermined length, and in one embodiment, the control unit 201 may process the measurement data in units of the window period.

In FIG. 3A, n window periods of window period 1 to window period n are set in time series. Here, the first window period 1 in the time series may be used as a reference window period. The reference window period may be set to, for example, a period in which the evaluation target has little deterioration and the evaluation target is operating normally. In one example, the reference window period may be set to the early days when the evaluation object has just been installed or replaced and put into operation. Further, the other window period of the plurality of window periods may be set after the reference window period, for example.

Then, the control unit 201 acquires the measurement data 1 of the reference window period 1 and the measurement data of a plurality of other window periods. FIG. 3B illustrates a case where the measurement data 2 of the window period 2 is acquired.

Subsequently, the control unit 201 models the relationship between the sensors in the window period as a sparse structure model for each of the acquired measurement data of the plurality of window periods. The sparse structure model may be, for example, a model having a structure obtained by extracting the essential connection between variables from multivariate data. The sparse structure model may be, for example, a structure that is not a "dense structure" in which all variables are connected by some kind of correlation, and is obtained as a result of omitting weak correlations and leaving only strong correlations. It may be a "sparse structure". The sparse structure model has a structure in which, for example, in a matrix-represented correlation (for example, an accuracy matrix for a variance-covariance matrix of multivariate data), most of the components are sparse matrices (sparse matrix). It can be represented by a model. For example, a sparse structure model that models the relationship between sensors based on the measurement data in the above-mentioned window period is generated using algorithms such as a Gaussian graphical model (GGM) and a graphical lasso. be able to. FIG. 3 (c) illustrates a sparse structure model in a graph structure. The sparse structure model may be represented by other expressions such as an Ising model and a Potts model.

Then, the control unit 201 changes the change score of the collapse of the relationship generated between the sparse structure model obtained from the measurement data 1 of the window period 1 as a reference and the sparse structure model obtained from the measurement data of another window period. Detect as. For example, the control unit 201 may calculate a change score including the degree of change for each sensor from the sparse structure model of the window period 1 and the window period 2 as a reference (FIG. 3 (c)). Since the change score can be obtained from the reference window period 1 and the sparse structure model of another window period, if the number of window periods is n, n-1 change scores can be obtained. .. Further, the change score can be obtained by calculating the difference between distributions using, for example, KL divergence (Kullback-Leibler divergence), JS divergence (Jensen-Shannon divergence), and the like.

Subsequently, the control unit 201 clusters the obtained plurality of change scores. For example, when the change score is obtained by KL divergence, the change score is represented by a vector whose component is the degree of change for each sensor of a plurality of sensors. Therefore, the change score can be plotted as points on a multidimensional space centered on each of the components of the change score. In FIG. 3D, an example in which the points of the change score are plotted in two dimensions is shown, but in reality, the points of the change score may be plotted in a multidimensional space having as many dimensions as the number of sensors. Then, the control unit 201 may cluster the points corresponding to the plurality of change scores plotted on the multidimensional space. For example, the control unit 201 may perform learning using a mixed Gaussian model and perform clustering of change scores. Alternatively, the control unit 201 may execute clustering using other algorithms such as DBSCAN (Density-based spatial clustering of applications with noise) and K-means (k-means). FIG. 3 (d) shows two clusters obtained as a result of clustering.

Subsequently, the control unit 201 further provides a plurality of window periods in the measurement data after the plurality of window periods used for clustering, and adds points of change scores corresponding to each window period to the cluster. The cluster in which the points of the change score to be added are clustered may be, for example, the cluster closest to the points of the change score to be added. Alternatively, in another embodiment, the cluster in which the points of the change score to be added are clustered may be determined by another method such as performing clustering again. Further, when adding a point to the cluster, the control unit 201 responds to the addition of the point to the cluster, for example, by removing the point having the oldest window period among the points of the change score belonging to the cluster. Old points belonging to may be removed.

Then, when a new change score point is added to the cluster in this way, the position of the cluster moves as shown in FIG. 3 (e). In the embodiment, the control unit 201 evaluates the deterioration of the evaluation target based on the movement of the position of the cluster. For example, the control unit 201 may evaluate the deterioration of the evaluation target based on the movement distance, the movement speed, the movement acceleration, and the like at the position of the cluster. The moving distance of the cluster may be, for example, the distance between the cluster specified based on the measurement data at the initial stage of operation and the cluster after the movement specified based on the subsequent measurement data. In one example, the distance between clusters can be expressed as the distance between the centers of gravity of the clusters. Further, as shown in FIG. 3 (e), when movement occurs in a plurality of clusters, the control unit 201 controls the average of the values such as the movement distance, the movement speed, and the movement acceleration in each of the plurality of clusters. Deterioration of the evaluation target may be evaluated based on statistical values such as variance.

As described above, the measurement data measured by the sensor may include fluctuations caused by factors other than deterioration of the evaluation target. However, the fluctuation of the measured value due to factors other than deterioration includes many factors showing reversible changes that can return to the original state. For example, fluctuations in sensor measurements due to day-night differences, weather differences, and whether peripheral devices are operating are reversible changes that can return to their original values in the long run. Is. Therefore, it is presumed that the fluctuation of the measured value due to the factors showing these reversible changes can contribute to the cluster division, but it is a component that does not change the position of the cluster so much in the long term.

On the other hand, the deterioration that occurs in the evaluation target is often an irreversible phenomenon, and it is unlikely that the fluctuation of the measured value of the sensor that once occurred due to the deterioration will return to the original. Therefore, as described above, by clustering the change scores and following the movement of the cluster position over a long period of time, such as months, years, and decades, the evaluation object is based on the movement of the cluster. It is possible to evaluate the deterioration of. Then, when the deterioration of the evaluation target is evaluated based on the movement of the cluster in this way, the deterioration of the evaluation target can be evaluated by suppressing the influence of the fluctuation of the measurement data caused by factors other than the deterioration.

Hereinafter, the deterioration evaluation process according to the embodiment will be described in more detail.

_{In the following, it is assumed that the data matrix X j} including the measurement data in a certain window period j of a plurality of sensors for measuring the state of the evaluation object is represented by the equation 1.
X _j = {x ₁ , x ₂ , ..., x _k , ..., x _N } ... Equation 1

Here, the window period may be a period of a predetermined length, and x ₁ , x ₂ , ..., X _k , ..., X _N (here, k = 1 to N) is, for example, the window period. It is a vector whose elements are the measured values of a plurality of sensors at each time point. For example, x _k is a vector whose element is the measured value of each sensor of a plurality of sensors at the time point k, and can be expressed by the following equation.
x _k = (x _k1 , x _k2 , ..., x _ki , ..., x _kM ) ^T ... Equation 2

Here, M is the total number of sensors. Further, i represents i = 1 to M and represents a number for identifying the sensor. Therefore, x _k1 , x _k2 , ..., x _ki , ..., X _kM are measured values of each sensor i of the plurality of sensors at the time point k within the window period.

FIG. 4 is a diagram illustrating an operation flow of a process for generating a sparse structure model for a window period as a reference according to an embodiment. The control unit 201 may start the process of FIG. 4 when, for example, an execution instruction of an operation flow for generating a sparse structure model of the window period as the reference of FIG. 4 is input.

In S401, the control unit 201 acquires a _{data matrix X ref} including the measured values measured during the reference window period from the measurement data measured by the plurality of sensors that measure the information of the evaluation target. In the following, the reference window period may be referred to as, for example, the reference window period. The reference window period may be set to, for example, a period in which the evaluation target has little deterioration and the evaluation target is operating normally. For example, the control unit 201 may read the _{data matrix X ref} from the measurement data stored in the storage unit 202.

In S402, the control unit 201 generates a sparse structure model from the _{data matrix X ref} including the measurement data of the sensor in the reference window period, and this operation flow ends. For example, the control unit 201 may generate a variance-covariance matrix based on the measurement data of each sensor of the plurality of sensors in the reference window period, and sparse the generated variance-covariance matrix to generate a sparse structure model. In one example, the sparse structure model can be derived by GGM. In the following, an example will be described in which the control unit 201 derives GGM: p (x | X _{ref ) as a sparse structure model from the data matrix X ref.} However, the embodiment is not limited to this, and the sparse structure model may be generated by other methods.

In the above, the sparse structure model (for example, GGM: p (x | X _ref )) obtained from the measurement data of the reference window period is used as a reference model in the state where the evaluation object is not deteriorated, and the change score is used. To calculate.

The change score is, for example, how much the measurement data of the plurality of sensors measured in the window period after the reference window period changes with respect to the measurement data of the plurality of sensors measured in the above-mentioned reference window period. It may be a score that evaluates. The change score can be obtained by calculating the difference between the distributions between the sparse structure model of the reference window period and the sparse structure model of the window period to be evaluated. Differences between distributions can be calculated using, for example, KL divergence, JS divergence, and the like. Hereinafter, an embodiment will be described by taking as an example a case where a change score is obtained using KL divergence.

FIG. 5 is a diagram illustrating an operation flow of deterioration evaluation processing according to an embodiment. For example, the control unit 201 may start the operation flow of FIG. 5 when an execution instruction of the deterioration evaluation process is input.

_{In S501, the control unit 201 acquires the data matrix X j} (j = 1 ... L) of a plurality of window periods in the learning period for performing clustering. For example, j may be a number for identifying the window period, and L may be the number of window periods used for learning. For example, the control unit 201 may read the _{data matrix Xj} from the measurement data stored in advance in the storage unit 202.

In S502, the control unit 201 generates a sparse structure model for each window period j. For example, the control unit 201 may derive GGM: p (x | X _{j ) from the data matrix X j.}

The control unit 201 in S503, in each of the window period j (j = 1 ... L) , calculates the variation degree _{a i} for each sensor i of the plurality of sensors. The control unit 201 uses, for example, the GGM: p (x | X _ref ) of the reference window period and the GGM: p (x | X) of the window period j to be evaluated by using the KL divergence equation shown in the following equation 3. may calculate the change of a _i for each sensor i with the _j). Then, the control unit 201, with respect to the window period j, change scores _{A j = {a 1, a} 2, ..., a i, ..., a M} to the change of _{a i} for each sensor i component get.

The control unit 201 in S504, the clustering resulting change scores _{A j.} For example, the control unit 201, the respective components of the change score A _j on multidimensional space having axes, the point corresponding to change scores A _j for each window period j may be plotted. FIG. 6 is a diagram illustrating the flow of deterioration evaluation processing according to the embodiment, and as shown in FIG. 6A, the control unit 201 has a change score A _{j obtained for each of the window periods j.} May be plotted on a multidimensional space. In FIG. 6, but the point in the plane of two components as the axis of the change score is shown an example which is plotted, actually on multidimensional space whose axes represent the respective components of the change score A _j is The points corresponding to the change score _Aj may be plotted in.

Then, the control unit 201 may cluster the plotted points of the plurality of change scores. In one example, the control unit 201 may perform learning using a mixed Gaussian model and perform clustering on points with a plurality of change scores.

Note that the control unit 201 may determine the number of clusters used for clustering in the mixed Gaussian model using indicators such as Akaike's Information Criterion (AIC) and Bayesian Information Criterion (BIC). Further, the embodiment is not limited to clustering by a mixed Gaussian model, and the control unit 201 may execute clustering by another algorithm such as DBSCAN or K-means clustering.

FIG. 6 (b) illustrates clustering of change scores, and change scores are classified into two clusters.

In S505, the control unit 201 identifies the position of each cluster from the clustering model. For example, the control unit 201, as the position of the cluster, the representative point mu _r representing each cluster _(r = 1 ... s, s is the number of clusters) may calculate the position of the. For example, the center of gravity of the cluster can be used as a representative point representing the cluster. In FIG. 6B, the center of gravity is shown as a representative point of the cluster. The representative point representing the cluster may be a point representing a point in the cluster other than the center of gravity.

By the above processing up to S505, the control unit 201 can specify, for example, the initial position of the cluster. Then, in the subsequent processing after S506, the control unit 201 is, for example, an evaluation target from the measurement data of the sensor during the period of the target for evaluating the deterioration state of the evaluation target (hereinafter, may be referred to as the evaluation target period). Evaluate the deterioration of things.

In S506, the control unit 201 determines whether or not there is measurement data for the evaluation target period. The evaluation target period may be, for example, a period for evaluating the deterioration of the evaluation target, and may be a period including measurement data measured after the learning period for performing clustering. If there is no measurement data for the evaluation target period (S506 is NO), the flow repeats the process of S506. On the other hand, if there is measurement data for the evaluation target period (YES in S506), the flow proceeds to S507.

Subsequently, in S507, the control unit 201 acquires the _{data matrix Xo from the measurement data of the evaluation target period.} o may be o = 1 ... Q, and may be a number that identifies the window period included in the measurement data of the evaluation target period. Further, Q may be the number of window periods included in the measurement data of the evaluation target period. For example, the control unit 201 may read the _{data matrix Xo} from the measurement data stored in advance in the storage unit 202.

The control unit 201 in S508 generates a sparse structure model from the data matrix _{X o} for each window period o. For example, the control unit 201 may derive GGM: p (x | X _{o ) from the data matrix X o.}

In S509, for example, the control unit 201 calculates the degree of change for each sensor i of the plurality of sensors in each window period o of the evaluation target period. The degree of change can be calculated, for example, by using X _{o instead of X j in} the KL divergence formula shown in the above formula 3, and the control unit 201 can thereby calculate the change score A _{o for each window period o.} To get.

In S510, the control unit 201 _{adds the obtained change score Ao} to the cluster. In one example, the control unit 201 may _{cluster the change score Ao} to the cluster closest to the point _{where the change score Ao is plotted on the multidimensional space.} Alternatively, in another embodiment, the control unit 201, again by adding the change score A _o, etc. perform the clustering by the Gaussian mixture model, change scores A _o may determine the cluster to be clustered. Then, for example, _{by adding the change score Ao} to the cluster, the cluster can move. The FIG. 6 (c), the example of adding the change score A _o the cluster are the indicated, the addition of the change scores A _o to the cluster, the cluster is moving.

In S511, the control unit 201 reidentifies the position of the cluster. For example, the control unit 201, as the position of the cluster, the representative point mu _t representing each cluster _(t = 1 ... s, s is the number of clusters) may identify the. As the representative point, the same points as those used in S505, such as the center of gravity of the cluster, can be used. _{Since the cluster includes the point of the change score Ao} obtained from the measurement data of the evaluation target period, the position of the representative point of the cluster may change as compared with S505.

In S512, the control unit 201 evaluates the deterioration of the evaluation target object based on the movement of the position of the cluster, outputs the evaluation result, and the main operation flow ends. The evaluation result may be, for example, the degree of deterioration indicating the degree of deterioration of the evaluation object. For example, in S512, the control unit 201 evaluates a value representing the movement of the cluster from the position of the cluster in the learning period specified in S505 to the position of the cluster including the point of the change score in the evaluation target period specified in S511. It may be output as the degree of deterioration of the object. The values representing the movement may be, for example, the movement distance, the movement speed, the movement acceleration, and the like of the cluster. For example, FIG. 6D shows an example in which the distance between the centers of gravity of two clusters is used as the degree of deterioration, and the user can use the magnitude of the distance between the centers of gravity output as the degree of deterioration to evaluate the object. The state of deterioration can be estimated.

Although FIG. 6D shows an example in which a value representing the movement of one cluster is used as the degree of deterioration, it is possible that a plurality of clusters obtained as a result of clustering move. In this case, the value representing the movement of clusters may be represented by, for example, statistical values such as the average value, the median value, the mode value, and the distribution of the values representing the movement of each cluster of a plurality of clusters. FIG. 7 is a diagram illustrating the movement of a plurality of clusters. For example, as shown in FIG. 7A, it is assumed that there are three clusters A, B, and C, and each cluster is moving. In this case, the control unit 201 may output a statistical value (for example, an average value) of a value representing the movement acquired in each of the three clusters as the degree of deterioration.

Further, in another embodiment, the control unit 201 does not have to use the cluster in which the value representing the movement of the cluster is a predetermined value or less and the movement is small in the evaluation of deterioration. For example, in FIG. 7B, the movement of cluster B is very small. Thus, it is presumed that the clusters with small movement do not contain much component of change due to deterioration of the evaluation object. Therefore, for example, the control unit 201 may acquire the degree of deterioration from the value representing the movement of the remaining clusters, except for the clusters whose movement value is equal to or less than a predetermined value. For example, in the example of FIG. 7B, the control unit 201 has a statistical value (for example, an average value) of values representing the movement of cluster A and cluster C, except for cluster B whose value representing movement is a predetermined value or less. May be output as the degree of deterioration.

Further, as another example of the evaluation result of the deterioration of the evaluation object output by the control unit 201 in S512, the control unit 201 determines and outputs the necessity of maintenance for the evaluation object based on the movement of the cluster. You may. For example, the control unit 201 may refer to the determination information 800 and determine whether or not maintenance is required for the evaluation target.

FIG. 8 is a diagram illustrating the determination information 800 according to the embodiment. In the determination information 800, for example, information for determining the necessity of maintenance work for the evaluation target may be registered. For example, a threshold value is registered in the determination information 800 of FIG. 8A. For the threshold value, for example, a value indicating the movement of the cluster from the position of the cluster in a state where the deterioration of the evaluation target is small to the position of the cluster in a state where it is desirable to perform maintenance work on the evaluation target is examined in advance. , Is registered. As a result, for example, the control unit 201 may output information prompting maintenance work for the evaluation target when the value representing the movement of the cluster specified in S512 is larger than the threshold value. The output of the information prompting the maintenance work may be, for example, displaying information such as a message prompting the maintenance work of the evaluation target object to be displayed on a display device such as a display connected to the information processing device 200. Further, in another example, the output of the information prompting the maintenance work may be to notify the administrator of the evaluation target object of the information prompting the maintenance work by e-mail or the like.

Further, in the determination information 800, for example, information for determining the necessity of maintenance work may be registered for each type of deterioration of the evaluation target object. For example, in the determination information 800 of FIG. 8B, information on the movement direction is registered in addition to the threshold value of the movement distance of the cluster in association with the type of deterioration of the evaluation target object. Further, in the determination information 800 of FIG. 8C, the threshold value of the movement distance of the cluster is associated with the type of deterioration of the evaluation target, and the information for designating the cluster used for specifying the movement distance is registered in association with each other. ing. In the determination information 800 of FIG. 8D, the threshold value of the movement distance of the cluster, the range of the movement speed, and the range of the acceleration of the movement are registered in association with the type of deterioration of the evaluation object.

As described with reference to FIG. 1, there may be various types of deterioration of the evaluation target, and the deterioration process may also differ depending on the type of deterioration. And the tendency of cluster movement may differ depending on the type of deterioration. Also in this case, another information that characterizes the movement of the cluster is used, for example, the movement direction of the cluster shown in FIG. 8 (b) and the information for specifying the cluster used for specifying the movement distance shown in FIG. 8 (c). Therefore, it is possible to specify the type of deterioration. Similarly, for example, as shown in FIG. 8D, by using a plurality of information such as a threshold value of the movement distance of the cluster, a range of movement speed, and a range of acceleration of movement as values representing the movement of the cluster, It is possible to specify the type of deterioration. Then, in the process of S512, the control unit 201 may refer to the determination information 800, determine the necessity of maintenance for the evaluation target according to the type of deterioration, and output the determination result. For example, if the movement of the cluster satisfies the condition of deterioration of long-term creep registered in the determination information 800, the control unit 201 may output information prompting maintenance of deterioration of long-term creep. On the other hand, if the movement of the cluster satisfies the condition of corrosion / wear deterioration registered in the determination information 800, the control unit 201 may output information prompting maintenance of the corrosion / wear deterioration.

As described above, according to the operation flow of FIG. 5, the control unit 201 can evaluate the deterioration of the evaluation target object based on the movement of the clusters obtained by clustering the change scores.

For example, when maintenance work is performed regularly to maintain the state of the evaluation object, it may happen that the deterioration progresses quickly, so the maintenance work execution interval is longer than the average life of the evaluation object. Can also be set shorter. However, according to the embodiment, since the deterioration state of the evaluation target can be estimated with high accuracy based on the movement of the cluster, it is possible to lengthen the execution interval of the maintenance work. As a result, the cost of maintenance work and the work load can be reduced.

Further, even if the deterioration of the evaluation target happens to progress quickly, according to the embodiment, the deterioration of the evaluation target can be detected with high accuracy based on the movement of the cluster, so that the user can evaluate the evaluation target. Maintenance work can be performed before an object breaks. Therefore, it is possible to avoid an unplanned stoppage of the evaluation target. In the above, the deterioration of the boiler tube group is described as an example of the evaluation target, but the embodiment is not limited to this, and the embodiment is also applied to the deterioration of other parts, devices, and equipment. be able to.

Although the embodiments have been illustrated above, the embodiments are not limited to this. For example, the above-mentioned operation flow is an example, and the embodiment is not limited thereto. When possible, the operation flow may be executed by changing the order of processing, may include additional processing, or may omit some processing.

For example, FIG. 5 shows an example in which the processes from S501 to S505 and the processes from S506 to S512 are continuously executed, but the embodiment is not limited to this. In another embodiment, the processes from S501 to S505 and the processes from S506 to S512 may be executed as different operation flows. As an example, the processes from S501 to S505 need only be executed once when the operation of the evaluation target object at the initial stage of operation is stable. Further, for example, in the processing from S506 to S512, when the measurement data for a predetermined period such as once a month or once every six months is accumulated, the measurement data for the predetermined period is executed as the data for the evaluation target period each time. You can do it.

Further, in the above-described embodiment, for example, an example in which the moving distance of the cluster is specified by using the distance between the centers of gravity of the cluster is described, but the embodiment is not limited to this, and other values are used for the cluster. It can also be used as the moving distance of. For example, in another embodiment, the cluster moves the shortest or longest distance between a plurality of change score points in the cluster before the move and a plurality of change score points in the cluster after the move. It may be used as a distance. Alternatively, in another embodiment, the average of the distances at all combinations of the change score points included in the two clusters before and after the movement may be used as the movement distance of the cluster.

Further, in the above-described embodiment, the addition of the points of the change score corresponding to the window period of the evaluation target period to the cluster may be executed in the order according to the time series of the measurement of the measurement data in one example. By adding the points of the change score to the cluster in the order according to the time series, it is possible to capture the time-series change of the movement of the position of the cluster. However, the embodiments are not limited to this. For example, in another embodiment, the control unit 201 may add the points of the change score corresponding to the window period included in the predetermined period to the cluster in chronological order, collectively for each predetermined period. In this case, the points of the change score corresponding to the window period within the predetermined period may be added in any order.

In the above-described embodiment, in the process of S503, the control unit 201 operates as, for example, the acquisition unit 211. Further, in the process of S505, the control unit 201 operates as, for example, the specific unit 212. In the process of S512, the control unit 201 operates as, for example, the evaluation unit 213.

FIG. 9 is a diagram illustrating a hardware configuration of a computer 900 for realizing the information processing apparatus 200 according to the embodiment. The hardware configuration for realizing the information processing device 200 of FIG. 9 includes, for example, a processor 901, a memory 902, a storage device 903, a reading device 904, a communication interface 906, and an input / output interface 907. The processor 901, the memory 902, the storage device 903, the reading device 904, the communication interface 906, and the input / output interface 907 are connected to each other via, for example, the bus 908.

The processor 901 may be, for example, a single processor, a multiprocessor, or a multicore. The processor 901 provides a part or all of the functions of the control unit 201 described above by executing, for example, a program describing the procedure of the operation flow described above using the memory 902. For example, the processor 901 of the information processing device 200 operates as the acquisition unit 211, the specific unit 212, and the evaluation unit 213 by reading and executing the program stored in the storage device 903.

The memory 902 is, for example, a semiconductor memory, and may include a RAM area and a ROM area. The storage device 903 is, for example, a semiconductor memory such as a hard disk or a flash memory, or an external storage device. RAM is an abbreviation for Random Access Memory. ROM is an abbreviation for Read Only Memory.

The reading device 904 accesses the removable storage medium 905 according to the instructions of the processor 901. The removable storage medium 905 is realized by, for example, a semiconductor device, a medium in which information is input / output by magnetic action, a medium in which information is input / output by optical action, and the like. The semiconductor device is, for example, a USB (Universal Serial Bus) memory. Further, the medium to which information is input / output by magnetic action is, for example, a magnetic disk. The medium to which information is input / output by optical action is, for example, a CD-ROM, a DVD, a Blu-ray Disc, or the like (Blu-ray is a registered trademark). CD is an abbreviation for Compact Disc. DVD is an abbreviation for Digital Versatile Disk.

The storage unit 202 may include, for example, a memory 902, a storage device 903, and a removable storage medium 905. For example, the storage device 903 of the information processing device 200 stores, for example, the determination information 800.

The communication interface 906 communicates with another device according to the instruction of the processor 901. In one example, the communication interface 906 may communicate with the sensor 950, which measures the state of the evaluation object, by wire or wireless communication, and collect measurement data from the sensor 950. Further, in another embodiment, the communication interface 906 may communicate with another device for storing the measurement data measured by the sensor 950 by wire or wireless communication to acquire the measurement data. The communication interface 906 is an example of the above-mentioned communication unit 203.

The input / output interface 907 is, for example, an interface between an input device and an output device. The input device is, for example, a device such as a keyboard, a mouse, or a touch panel that receives an instruction from a user. The output device is, for example, a display device such as a display and an audio device such as a speaker.

Each program according to the embodiment is provided to the information processing apparatus 200 in the following form, for example.
(1) It is pre-installed in the storage device 903.
(2) Provided by a removable storage medium 905.
(3) It is provided from a server such as a program server.

The hardware configuration of the computer 900 for realizing the information processing apparatus 200 described with reference to FIG. 9 is an example, and the embodiment is not limited to this. For example, a part of the above configuration may be deleted, or a new configuration may be added. Further, in another embodiment, for example, some or all the functions of the above-mentioned control unit 201 may be implemented as hardware by FPGA, SoC, ASIC, PLD, or the like. FPGA is an abbreviation for Field Programmable Gate Array. SoC is an abbreviation for System-on-a-chip. ASIC is an abbreviation for Application Specific Integrated Circuit. PLD is an abbreviation for Programmable Logic Device.

In the above, some embodiments will be described. However, the embodiments are not limited to the above embodiments and should be understood to include various variants and alternatives of the above embodiments. For example, it will be understood that various embodiments can be embodied by transforming the components within a range that does not deviate from the purpose and scope. It will also be appreciated that various embodiments can be implemented by appropriately combining the plurality of components disclosed in the above-described embodiments. Further, various embodiments may be implemented by removing some components from all the components shown in the embodiments or by adding some components to the components shown in the embodiments. Those skilled in the art will understand.

200 Information processing device 201 Control unit 202 Storage unit 203 Communication unit 211 Acquisition unit 212 Specific unit 213 Evaluation unit 800 Judgment information 900 Computer 901 Processor 902 Memory 903 Storage device 904 Reading device 905 Detachable storage medium 906 Communication interface 907 Input / output interface 908 Bus 950 sensor

Claims (7)

Of the time-series measurement data obtained from multiple sensors that measure the state of the evaluation object, the window period after the reference window period for the sparse structure model generated from the data measured during the reference window period. The acquisition unit that obtains a change score based on the degree of change for each sensor of the sparse structure model generated from the data measured in each of the multiple window periods included in the learning period and acquires a plurality of change scores.
A specific part that clusters the plurality of change scores to identify the position of the cluster, and
An evaluation unit that evaluates deterioration of the evaluation target object based on the movement of the position of the cluster caused by adding at least one change score obtained in at least one window period included in the evaluation target period to the cluster. When,
Information processing equipment, including. The first aspect of the present invention, wherein the evaluation unit evaluates the deterioration of the evaluation object based on at least one of the movement distance, the movement speed, and the movement acceleration of the position of the cluster caused by adding to the cluster. Information processing equipment. The specific part identifies the positions of a plurality of clusters obtained by the clustering, and identifies the positions of the plurality of clusters.
The evaluation unit evaluates the deterioration of the evaluation target based on the movement of the position of at least one of the plurality of clusters.
The information processing apparatus according to claim 1 or 2. The information processing apparatus according to any one of claims 1 to 3, wherein the evaluation unit outputs information prompting maintenance work for the evaluation object when the movement of the position of the cluster satisfies a predetermined condition. .. The evaluation unit
When the movement of the position of the cluster satisfies the first condition, the information prompting the maintenance work of the deterioration of the first type for the evaluation object is output.
When the movement of the position of the cluster satisfies the second condition, the information prompting the maintenance work of the deterioration of the second type for the evaluation object is output.
The information processing apparatus according to any one of claims 1 to 3. Of the time-series measurement data obtained from multiple sensors that measure the state of the evaluation object, the window period after the reference window period for the sparse structure model generated from the data measured during the reference window period. The change score, which is based on the degree of change for each sensor of the sparse structure model generated from the data measured in, is obtained for each of the multiple window periods included in the learning period, and multiple change scores are obtained.
Clustering the multiple change scores to locate the cluster,
Deterioration of the evaluation target is evaluated based on the movement of the position of the cluster caused by adding at least one change score obtained in at least one window period included in the evaluation target period to the cluster.
An evaluation method performed by an information processing device, including the above. Of the time-series measurement data obtained from multiple sensors that measure the state of the evaluation object, the window period after the reference window period for the sparse structure model generated from the data measured during the reference window period. The change score, which is based on the degree of change for each sensor of the sparse structure model generated from the data measured in, is obtained for each of the multiple window periods included in the learning period, and multiple change scores are obtained.
Clustering the multiple change scores to locate the cluster,
Deterioration of the evaluation target is evaluated based on the movement of the position of the cluster caused by adding at least one change score obtained in at least one window period included in the evaluation target period to the cluster.
An evaluation program that causes an information processing device to execute processing.

Images