WO2021049498A1

WO2021049498A1 - Diagnostic system

Info

Publication number: WO2021049498A1
Application number: PCT/JP2020/033999
Authority: WO
Inventors: 豊明渡; 市川　智子
Original assignee: 住友重機械工業株式会社
Priority date: 2019-09-13
Filing date: 2020-09-08
Publication date: 2021-03-18
Also published as: JP7357494B2; JP2021043124A; CN114375389A; DE112020004325T5

Abstract

Provided is a diagnostic system that enables diagnosis with influences between a plurality of components considered. A diagnostic system (1) comprises: a plurality of sensors (11) provided in one-to-one correspondence with a plurality of components (111) of a device (100) being diagnosed for each outputting state detection information of a corresponding one of the components (111); an anomaly determination unit (21) for determining, on the basis of the status detection information of each component (111), whether or not the component (111) has an anomaly; and an anomaly prediction unit (22) for predicting which component (111) is likely to have an anomaly next upon the anomaly determination unit (21) determining that one of the components (111) has an anomaly.

Description

Diagnostic system

The present invention relates to a diagnostic system.

Conventionally, a diagnostic system for detecting a device failure has been proposed (see, for example, Patent Document 1). The diagnostic system of Patent Document 1 includes a plurality of gear motors and the like included in the device as diagnostic target elements, and includes a plurality of sensors and a plurality of processing units corresponding to the plurality of diagnostic target elements. Then, the sensor is attached to each diagnostic target element, and the processing unit repeatedly executes the diagnostic processing of each diagnostic target element based on the output of the sensor.

Japanese Unexamined Patent Publication No. 2018-173333

In a device that has multiple components and operates in cooperation with multiple components, the state change of one component may affect the other components.

An object of the present invention is to provide a diagnostic system capable of diagnosing in consideration of the influence between a plurality of components.

One diagnostic system according to the present invention is
Multiple sensors that are provided corresponding to multiple components of the device to be diagnosed and output the status detection information of each corresponding component, and
An abnormality determination unit that determines whether or not there is an abnormality in the component based on the state detection information of each component, and an abnormality determination unit.
When one component is determined to have an abnormality by the abnormality determination unit, an abnormality prediction unit that predicts the component in which the next abnormality will occur, and an abnormality prediction unit.
To be equipped.

Another diagnostic system according to the present invention is
Multiple sensors that are provided corresponding to multiple components of the device to be diagnosed and output the status detection information of each corresponding component, and
An impact calculation unit that calculates the impact between one designated component specified among the plurality of components and each of the other components.
To be equipped.

According to the present invention, it is possible to provide a diagnostic system capable of diagnosing in consideration of the influence between a plurality of components.

It is a block diagram which shows the diagnostic system and the device to be diagnosed which concerns on embodiment. It is a flowchart of the diagnosis process executed by the abnormality determination unit and the abnormality prediction unit. It is a figure explaining the calculation process of the risk degree of a certain component by an abnormality prediction part. It is a figure explaining the risk degree of each component calculated by an abnormality prediction part. Next, it is an image diagram which shows the output example of the component which is predicted that an abnormality occurs. It is a flowchart of the setting input processing executed by the input processing unit. It is an image figure which shows the setting input screen of a calculation expression. It is an image figure which shows the setting input screen of a coefficient matrix. It is a flowchart of the influence degree output process executed by the influence degree calculation part. It is a figure explaining the calculation process of the influence degree by the influence degree calculation part. It is an image diagram which shows an example of the output screen of the degree of influence.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing a diagnostic system and a device to be diagnosed according to the embodiment.

The diagnostic system 1 of the present embodiment is attached to the diagnostic target device 100 having a plurality of components 111, detects an abnormality of each component 111, notifies the component 111 having a high risk of failure (risk degree), and notifies the component 111. It is a system that calculates and displays the degree of influence among a plurality of components 111. In the present embodiment, the diagnosis target device 100 is a belt conveyor having a plurality of gear motors. The diagnosis target device 100 is not limited to this, and may be any device as long as it is a device including a plurality of components 111 that operate mechanically or electrically in cooperation with each other. Hereinafter, a plurality of gear motors for driving the belt will be described as a plurality of components 111.

The diagnostic system 1 includes a plurality of sensors 11 provided corresponding to the plurality of components 111, a computer 20 that receives outputs (state detection information) of the plurality of sensors 11 and performs diagnostic processing of the diagnostic target device 100, and information. It is provided with a display device 30 for outputting data, and an input device 40 such as a pointing device and a keyboard capable of inputting data to the computer 20 from outside the system.

The sensor 11 attached to each component 111 may be one type of sensor such as a vibration sensor, or may include a plurality of types of sensors such as a temperature sensor, a motor current sensor, and an iron powder concentration sensor in lubricating oil. May be good. Further, sensors of the same type may be provided at a plurality of locations of one component. The type of sensor is not limited to the above specific example.

The computer 20 has an abnormality determination unit 21 that determines the occurrence of an abnormality in any of the components 111 based on the outputs of the plurality of sensors 11, and a risk level that an abnormality will occur next when one component 111 is determined to be abnormal. The anomaly prediction unit 22 that predicts the high component 111, and the influence calculation unit 23 that calculates the degree of influence between any one component 111 and each of the other components 111 based on the request from the operator and outputs it from the display device 30. It also includes an input processing unit 24 that can change the settings related to the calculation of the degree of risk or the degree of influence, a storage device 27 that stores control data, and an interface 28 that captures signals from a plurality of sensors 11. The control data stored in the storage device 27 includes the relationship degree data table T1 used by the abnormality prediction unit 22 and the influence degree calculation unit 23, the state degree calculation control data DB1, the coefficient matrix M3, and the reason database DB2. included. The computer 20 includes a CPU (Central Processing Unit), a storage device 27 for storing control data and programs, and a RAM (Random Access Memory), and includes the above-mentioned abnormality determination unit 21, abnormality prediction unit 22, and impact calculation unit 23. The input processing unit 24 is a software function module realized by the CPU executing a program.

<Diagnostic processing>
Subsequently, the detailed operation of each part will be described with reference to the flowchart. FIG. 2 is a flowchart of diagnostic processing executed by the abnormality determination unit and the abnormality prediction unit.

While the diagnosis target device 100 is in operation, the abnormality determination unit 21 takes in the outputs of the plurality of sensors 11 (S1) and performs a determination process to determine whether or not an abnormality has occurred in each component 111 (step S2). Then, if there is no component 111 determined to be abnormal, the abnormality determination unit 21 repeats the processes of steps S1 and S2.

The determination of the abnormality in step S2 is a determination of a failure in which the component 111 becomes inoperable, a determination of an abnormality close to the failure, a determination of an abnormality notifying the maintenance required, or a determination of some abnormality not reaching the maintenance required level. May be good. That is, the abnormality determination unit 21 may determine whether or not it meets the predetermined abnormality conditions, and the abnormality conditions are not particularly limited.

In step S2, the abnormality determination unit 21 compares, for example, the output value of each sensor 11 with a preset threshold value indicating an abnormality, and when the output value of a certain sensor 11 exceeds the threshold value, the sensor The component 111 to which the 11 is attached is determined to be abnormal. Alternatively, the abnormality determination unit 21 compares the trend of the output value of the sensor 11 with the pattern indicating the abnormality of the trend, and if both match, determines that the component 111 to which the sensor 11 is attached is abnormal. You may. Alternatively, when a plurality of sensors 11 having different types or installation positions are attached to each component 111, the abnormality determination unit 21 may combine the outputs of the plurality of sensors 11, their trends, or both. It may be calculated by a predetermined determination formula or subjected to a predetermined determination process, and the occurrence of an abnormality may be determined based on these results.

When the abnormality determination unit 21 determines that any one of the components 111 is abnormal, the abnormality determination unit 21 notifies the abnormality (step S3). For example, the abnormality determination unit 21 displays and outputs the occurrence of the abnormality and the identification information of the component in which the abnormality has occurred from the display device 30 as notification of the abnormality.

When the abnormality determination unit 21 determines the abnormality and notifies the abnormality, the abnormality prediction unit 22 then causes one component 111 in which the abnormality has occurred (hereinafter referred to as “abnormal component 111”) and the other components. Considering the degree of relationship with the component 111, the risk degree of each of the other components 111 in which the abnormality occurs next is calculated (step S4). Next, the calculation process of the degree of risk will be described in detail.

<Danger calculation processing>
FIG. 3 is a diagram for explaining the calculation process of the risk level of a certain component by the abnormality prediction unit. Hereinafter, in order to identify each of the plurality of components 111, the plurality of components 111 are also represented as component A to component E. FIG. 3 shows an example of calculating the risk of an abnormality occurring in the component B next when an abnormality occurs in the component A.

<Relationship>
The abnormality prediction unit 22 uses the relationship degree data table T1 (FIG. 1) in which each parameter value of the relationship degree between the plurality of components 111 is registered in order to calculate the risk degree in step S4.

In the relationship degree data table T1, a value indicating the degree of relationship between each pair of components 111 by dividing them into a plurality of items is registered. The relationship degree list L1 of FIG. 3 is a list obtained by extracting the value of the relationship degree between the component A and the component B from the relationship degree data table T1. The items of the relationship degree data table include a distance coefficient, a correlation coefficient, a transmission rate, and the like, as shown in the relationship degree list L1 of FIG. In the relationship data table T1, all combinations of components A to E ({A, B}, {A, C}, {A, D}, {A, E}, {B, C}, {B, For D}, {B, E}, {C, D}, {C, E}, {D, E}), the values of the same items as the relationship list L1 in FIG. 3 are registered. Each value of the relationship degree data table T1 is obtained by investigating the diagnosis target device 100, and is registered in advance. The relationship degree list L1 corresponds to an example of the relationship degree information according to the present invention.

The distance coefficient is, for example, a normalized value representing a spatial distance. Not limited to this, the relationship data table T1 contains various types of distances such as a distance along a power transmission path, a distance along a belt, a distance between electric wires connecting each other, and a value related to an insulation creepage distance between control boards. The item of coefficient may be included. Further, as the distance coefficient, a value related to the distance in a certain metric space constructed by an arbitrary theory may be adopted. For example, when the causal relationships of various physical phenomena of all components 111 are expressed in a network, the minimum number of nodes in the network of causal relationships connecting a predetermined physical phenomenon of one component to a predetermined physical phenomenon of another component 111, etc. May be one distance coefficient.

The correlation coefficient is a coefficient indicating a cross-correlation between a predetermined physical quantity of one component 111 (for example, the magnitude of vibration) and a predetermined physical quantity of the other component 111 (for example, the magnitude of vibration). The relationship data table T1 may also include cross-correlation items for various physical quantities such as motor rotation speed, total number of motor rotations, and temperature. The cross-correlation may be a cross-correlation between different physical quantities, such as the temperature of one component 111 and the magnitude of vibration of the other component.

The transmission rate is, for example, the torque transmission rate. Not limited to this, the relationship data table T1 also includes various transmission rate items such as energy transmission rate, power supply power transmission rate, or signal transmission rate when a signal is serially transmitted. May be.

<Condition>
In order to calculate the degree of risk in step S4, the abnormality prediction unit 22 further calculates the current state degree list L2 (FIG. 3) of the component B to be calculated. The state degree list L2 has a plurality of items, and the value of each item for the component B is obtained from the sensor output of one or a plurality of sensors 11 attached to the component B. Since the abnormality prediction unit 22 obtains each state degree, one or a plurality of statistical processing programs and calculation formula data are stored in the state degree calculation control data DB1. The state degree list L2 corresponds to an example of the state degree information according to the present invention.

The items in the state degree list L2 include the degree of abnormality, the degree of load, the degree of failure based on expert knowledge, and the like.

The degree of abnormality is the degree to which the output of the predetermined sensor 11 approaches the value indicating the abnormality. The state degree list L2 may include a plurality of items of abnormality degrees corresponding to a plurality of sensors 11 having different types or installation positions. Further, the degree of abnormality may include the degree of abnormality over time obtained by statistically processing the output values of the sensor 11 from the past to the present. The degree of anomaly obtained by statistical processing corresponds to an example of the degree of statistical state according to the present invention. The degree of abnormality may be expressed as a percentage (percentage) of how close the parameter for determining whether or not the abnormality is to the threshold value indicating the abnormality, or may be expressed by a stepwise display such as high, medium, or low. May be good.

The load degree is a degree indicating the mechanical or electrical load on the component B at the present time calculated from the output of the sensor 11. The state degree list L2 may include a plurality of load degree items for different parts and different objects in one component B.

The degree of failure based on expert knowledge indicates the degree of failure (deterioration degree) that is close to the degree of failure formulated by theory or know-how, and one or more sensors are used in the calculation formula stored in the state degree calculation control data DB1. It is calculated by applying (substituting) the output value of 11. Even if the items of the state degree list L2 include various types of state degree items that can be calculated by substituting the output value of the sensor 11 into the calculation formula with the state quantity formulated by theory or know-how. Good. The degree of failure based on expert knowledge or the degree of state obtained by an arithmetic expression corresponds to an example of the degree of logical state according to the present invention.

<Relation coefficient>
To calculate the risk of step S4, the anomaly prediction unit 22 further uses the coefficient matrix M3. The coefficient matrix M3 contains (i × j) relational coefficients α _fg . Here, f = 1 to i, g = 1 to j, and i are the number of items in the state degree list L2, and j is the number of items in the relationship degree list L1. The relation coefficient α _fg is a value that relatively represents how much the combination of the f-th item of the state degree list L2 and the g-th item of the relation degree list L1 contributes to the failure risk of the component 111. Set. For example, the combination of the kinetic energy transfer coefficient (for example, the fourth item in the relationship list) and the vibration anomaly (the fifth item in the state list) is strongly involved in the risk of failure. The relational coefficient between these items becomes a large value (for example, α ₅₄ = 0.7), and the transfer coefficient of kinetic energy (the fourth item in the relational degree list) and the abnormal degree of the noise amount of the electric signal (for example, in the state degree list) Since the combination with the sixth item) has a weak involvement in the risk of failure, the relational coefficient between these items becomes a small value (α ₆₄ = 0.1).

_{Each relation coefficient α fg} of the coefficient matrix M3 is a value determined by the contents of the item of the relation degree list L1 and the item of the state degree list L2, and depends on the component 111 to be calculated or the component 111 determined to be abnormal. It is a value that does not.

<Danger matrix>
In order to calculate the degree of risk in step S4, the abnormality prediction unit 22 first calculates the degree of abnormality of the component A determined to be abnormal by multiplying the above-mentioned relationship degree list L1, state degree list L2, and coefficient matrix M3. Do. In the calculation, the degree of anomaly of component A is multiplied by each component of the list, the relationship list L1 and the state degree list L2 are tensor producted, and the product of the tensor product result and the coefficient matrix M3 is the corresponding components. Is the multiplication of. By such a calculation, a risk matrix M10 having the number of components (the number of items in the relationship list L1) × (the number of items in the state list L2) is obtained. The value of each item in the relationship list L1 is normalized, the value of each item in the state list L2 is normalized, and the relation coefficients α _fg are set to relative values. Therefore, the value of each component of the risk matrix M10 is a value at which the magnitude of the risk can be compared with each other. That is, the value of each component is an index showing how much the item of the relationship degree list L1 and the item of the state degree list L2 corresponding to each component contribute to the risk of causing the next abnormality.

When the risk matrix M10 is calculated in step S4, the abnormality prediction unit 22 further scalarizes the risk matrix M10, and sets this value as the risk that an abnormality will occur in the component B next. As the scalarization method, for example, the total sum of all the components of the risk matrix M10, the total sum of all the components, and their standardization can be adopted.

In step S4, the abnormality prediction unit 22 similarly calculates the risk matrix M10 and the scalarized risk for each of the other components C, D, and E. FIG. 4 is a diagram for explaining the degree of risk of each component calculated by the abnormality prediction unit. As shown in FIG. 4, the risk level calculated for each component B to E is a value calculated in consideration of the relationship between the component A determined to be abnormal and the other components B to E. , Indicates the possibility that the next abnormality will occur due to the abnormality of the component A. Since the values of the relationship degree list L1 and the state degree list L2 used in the calculation of the risk degree matrix M10 are different for each component to be calculated, the risk degree matrix M10 and the risk degree values are different for each component B to E.

<Reason for risk>
After calculating the risk matrix M10 and the risk of each component B to E, the abnormality prediction unit 22 extracts the reason for the risk from the reason database DB2 (step S5). In the reason database DB2, the notation contents of the reasons corresponding to the positions (number of rows and columns) of each component of the risk matrix M10 are registered. That is, the value of each component of the risk matrix M10 contributes to how much a certain item in the relationship list L1 and a certain item in the state list L2 corresponding to each component contribute to the risk of causing an abnormality next. It is an index showing whether you are doing it. Therefore, when any of the components shows a high value, it means that the synergistic action between the item of the degree of relation and the item of the degree of state corresponding to this component increases the degree of risk. Therefore, the content showing this synergistic action is adopted as the reason corresponding to this component.

In step S5, the abnormality prediction unit 22 searches the risk matrix M10 of each component B to E for a component value exceeding the threshold value, and extracts from the reason database DB 2 the reason corresponding to the component value exceeding the threshold value. To do. If the values of all the components of the risk matrix M10 for a component 111 are less than the threshold, no reason is extracted for that component 111.

<Output of risk ranking table>
FIG. 5 is an image diagram showing an output example of a component that is predicted to cause an abnormality next. Next, the abnormality prediction unit 22 creates the data of the risk ranking table H1 in which the components B to E, the risk calculated in step S4, and the reason extracted in step S5 are associated with each other. Output from the display device 30 (step S6). The risk ranking table H1 may be sorted in order of risk. The risk ranking table H1 does not include information on all components, but may include information on only high-risk components. When outputting the display of the risk ranking table H1, the notification display J1 of the abnormal component in step S3 may also be performed. The notification display J1 may include a display item of the degree of abnormality of the abnormal component. By displaying the degree of abnormality, the severity of the abnormality can be indicated to the user. The risk ranking table H1 of FIG. 5 shows that the abnormality prediction unit 22 predicted the component C because it has the highest risk as the component that causes the next abnormality.

By finding the component C having the highest risk from the risk ranking table H1, the operator can recognize that the component C is the component predicted to have the next abnormality following the abnormality of the component A. .. Furthermore, the operator can recognize from the notation of the reason what has a great influence and increases the risk of occurrence of an abnormality. From these support information, the operator can plan an efficient maintenance schedule or an efficient parts replacement schedule for the device 100 to be diagnosed.

<Setting input processing>
FIG. 6 is a flowchart of the setting input process executed by the input processing unit.

In the diagnostic system 1 of the present embodiment, the operator can change the above-mentioned settings related to the calculation of the degree of risk by the function of the input processing unit 24 (FIG. 1). The objects whose settings can be changed are the calculation formula of the "fault degree by expert knowledge" in the state degree list L2 and each relation coefficient α _fg of the coefficient matrix M3.

FIG. 7 is an image diagram showing a setting input screen of a calculation formula.

When the operator selects execution of the setting input processing from the system menu, the input processing unit 24 starts the setting input processing. In the setting input processing, first, the input processing unit 24 performs a selection input processing for causing the operator to select whether to change the setting of the arithmetic expression or the setting of the coefficient matrix M3 (step S11). As a result, when the setting change of the calculation formula is selected, the input processing unit 24 outputs the setting input screen G1 of FIG. 7 to the display device 30 (step S12). The setting input screen G1 includes an explanatory notation N1 for variables that can be used in the calculation formula, an input box N2 for inputting the calculation formula, and a command button N3 for selecting input completion or input cancellation. Here, the operator inputs a newly applied arithmetic expression in the input box N2 of the setting input screen G1 and selects and operates the command button N3 for completing the input.

Then, the input processing unit 24 determines the selection operation (step S13), and if the input is completed, the input content of the input box N2 is used as a calculation formula of the failure degree based on expert knowledge in the state degree calculation control data DB1. The data of the calculation formula of is rewritten (step S14). Then, the setting input process is completed.

In the above setting input process, an example of changing the setting of the calculation formula for the failure degree based on the expert knowledge among the data included in the state degree calculation control data DB1 is shown. However, when the state degree calculation control data DB 1 includes an arithmetic expression for calculating the state degree of another item, the setting of this arithmetic expression may be changed by the same processing.

FIG. 8 is an image diagram showing a setting input screen of the coefficient matrix.

When the operator selects to change the setting of the coefficient matrix M3 in the selection input process of step S1, the input processing unit 24 first outputs the setting input screen G2 of FIG. 8 to the display device 30 (step S15). The setting input screen G2 includes a table N11 having the number of rows and columns corresponding to the coefficient matrix M3, and a command button N12 capable of selecting input completion or input cancellation. Furthermore, within each cell of the table N11, it includes a display N14 of an item name of the item name and status of the degree of relationship corresponding to the relationship coefficient alpha _fg, and input box N15 relationship coefficient alpha _fg is. _{Here, the operator inputs the updated value of each correlation coefficient α fg} into each input box N15 in the table N11, and selects the command button N12 for completing the input.

Then, the input processing unit 24 determines the selection operation (step S16), and if the input is completed, _{rewrites each relation coefficient α fg} of the coefficient matrix M3 of the storage device 27 with the input contents of the input box N15 (step S17). ). The value of the relationship coefficient alpha _fg is a value that determines the risk equation, setting change of the relationship between the coefficient alpha _fg corresponds to a configuration change of the risk equation. Then, the setting input process is completed.

According to the setting input process of the calculation formula shown in FIG. 7, when an improvement point is found in the calculation formula showing the degree of failure based on expert knowledge or the calculation formula showing the degree of state of some kind, the calculation formula is corrected. Therefore, it becomes possible to express a more accurate degree of failure or state.

According to the setting input process of the coefficient matrix M3 shown in FIG. 8, how much the synergistic action between each item of the relationship degree list L1 and each item of the state degree list L2 is involved in the risk of causing the component 111 to be abnormal. However, when there is new knowledge, or when it is desired to adjust the _{relation coefficient α fg} _{from the actual calculation result, each relation coefficient α fg} of the coefficient matrix M3 can be appropriately modified. Thereby, the accuracy of the risk matrix M10 calculated using the coefficient matrix M3 can be further improved.

<Calculation and output of influence between components>
FIG. 9 is a flowchart of the influence degree output process executed by the influence degree calculation unit 23. FIG. 10 is a diagram illustrating a calculation process of the degree of influence by the degree of influence calculation unit. FIG. 11 is an image diagram showing an example of an influence degree output screen. In the diagnostic system 1 of the present embodiment, a certain component 111 and each of the other components 111 are combined with each other at a stage where no abnormality has occurred in any of the plurality of components 111 by the function of the influence calculation unit 23 (FIG. 1). The degree of influence between them can be calculated and displayed. In the following, a case where component A is designated as an influence source and the degree of influence from component A on each of the other components B to E is calculated and output will be described. The designated component is also referred to as the "designated component".

When the operator selects execution of the influence degree output process from the system menu, the influence degree calculation unit 23 starts the influence degree output process of FIG. Then, the influence degree calculation unit 23 first outputs the screen G3 of the influence degree output process of FIG. 11 from the display device 30 (step S21). The screen G3 of FIG. 11 shows the display output after the calculation of the degree of influence, and at the stage of step S21, the input box N21 and the result display frame N23 are blanked.

The influence degree output processing screen G3 includes an input box N21 for selecting a designated component, a command button N22 for indicating completion or cancellation of designation, and a result display frame N23 for outputting a calculation result. On this screen, the operator first specifies the influencing component 111 using the input box N21. Then, the command button N22 for which the designation is completed is selected and operated. Then, the influence degree calculation unit 23 determines the selection operation (step S22), and if the designation is completed, calculates the influence degree from the designated component A to each of the other components B to E (step S23). Further, when the command button N22 for completing the designation is selected, the abnormality prediction unit 22 or the influence degree calculation unit 23 calculates the abnormality degree of the designated component A, and determines the abnormality degree of the designated component A as the abnormality of the result display frame N23. It may be output to the degree column N23a. The degree of abnormality in the degree of abnormality column N23a may be displayed as a percentage or in stages.

Next, the calculation method of the degree of influence on one component B will be described. The calculation of the degree of impact is similar to the calculation of the degree of risk described above. The influence degree calculation unit 23 multiplies the relationship degree list L1 of FIG. 10, the state degree list L2, and the coefficient matrix M3 to obtain the influence degree matrix M20. The relationship degree list L1 is a list obtained by extracting a portion showing the relationship between the designated component A and the component B to be calculated from the relationship degree data table T1, and is the same as the relationship degree list L1 in FIG. The state degree list L2 is a list showing the state degree of the component B to be calculated, and is the same as the state degree list L2 of FIG. The coefficient matrix M3 is the same as the coefficient matrix M3 of FIG. The multiplication method is also the same, and the influence degree matrix M20 is the same as the value of the risk degree matrix M10 in FIG. 3 except that the value of each component does not include the product of the anomaly degree of the component A.

The influence degree calculation unit 23 further scalarizes the influence degree matrix M20 of FIG. 11 in the same manner as in the case of FIG. 3, and obtains this value as the influence degree from the designated component A to the component B. In step S23, the influence degree calculation unit 23 calculates the influence degree for the other components C to E in the same manner. As a result of the calculation, each influence degree from the designated component A to each of the other components B to E is obtained.

When the degree of influence on the other components B to E is calculated, the degree of influence calculation unit 23 outputs the image Z24 showing the degree of influence from the display frame N23 (step S24). In the image Z24, for example, the symbols P1 to P5 indicating a plurality of components A to E, the correspondence line L indicating a combination of two of the plurality of components A to E, and the correspondence relationship of the combination for which the degree of influence is calculated are shown. The influence degree display V attached to the line L is included. In the example of FIG. 11, since the influence degree from the designated component A to the other components B to E is calculated, the influence degree display V is output along with the correspondence line L between them.

The operator can recognize how much the state change of the designated component A imposes a burden on each of the other components B to E from the calculation result of the degree of influence. Then, the operator can plan an efficient maintenance schedule or an efficient parts replacement schedule of the diagnosis target device 100 by using these as support information.

As described above, according to the diagnostic system 1 of the present embodiment, the abnormality prediction unit 22 or the impact calculation unit 23 considers the relationship between the plurality of components 111 of the device 100 to be diagnosed, and the risk or impact of failure. To calculate. Therefore, based on these calculation results, the diagnosis target device 100 can be diagnosed in consideration of the influence between the plurality of components.

The embodiment of the present invention has been described above. However, the present invention is not limited to the above embodiment. For example, the item of the relationship list L1 used for calculating the risk of the next abnormality is not limited to the example of the embodiment as long as it is information indicating the relationship between the components, and includes various items. It may be. The items in the state degree list L2 are not limited to the examples of the embodiment as long as they indicate the state of the component 111 obtained from the output of the sensor 11 attached to each component 111, and various items are included. You may. In addition, the details shown in the embodiment, such as the calculation formula of the degree of risk and the calculation formula of the degree of influence, can be appropriately changed without departing from the spirit of the invention. Further, in the above embodiment, the belt conveyor and the gear motor have been described as an example of the device to be diagnosed and the plurality of components, but the device to be diagnosed includes various devices such as an injection molding machine, a machine tool, and an industrial robot. May be applied, and as a component, various elements such as various mechanical parts such as a chain sprocket and various electric parts such as a control board may be applied.

It can be used for a diagnostic system that can diagnose the diagnostic target in consideration of the influence between multiple components included in the diagnostic target.

1 Diagnostic system 11 Sensor 20 Computer 21 Abnormality judgment unit 22 Abnormality prediction unit 23 Impact degree calculation unit 24 Input processing unit 27 Storage device T1 Relationship degree data table DB1 Status degree calculation control data M3 Coefficient matrix DB2 Reason database L1 Relationship degree list L2 Status list M10 Risk matrix M20 Impact matrix J1 Abnormal component notification display H1 Risk ranking table G1, G2 Setting input screen G3 Impact output processing screen N21 Input box for designated component Z24 Image 30 display showing impact Device 40 Input device 100 Diagnostic target device 111 component

Claims

Multiple sensors that are provided corresponding to multiple components of the device to be diagnosed and output the status detection information of each corresponding component, and
An abnormality determination unit that determines whether or not there is an abnormality in the component based on the state detection information of each component, and an abnormality determination unit.
When one component is determined to have an abnormality by the abnormality determination unit, an abnormality prediction unit that predicts the component in which the next abnormality will occur, and an abnormality prediction unit.
Diagnostic system with.
The abnormality prediction unit includes relationship degree information indicating the relationship of each of the other components with respect to the abnormality component determined by the abnormality determination unit among the plurality of components, and the state detection information of each of the other components. Based on the state information of the component obtained based on, and predict the component that will cause the next abnormality,
The diagnostic system according to claim 1.
The relationship degree information includes information on the distance between the anomalous component and each of the other components, information on the mutual correlation between the first physical quantity of the anomalous component and the second physical quantity of each of the other components, and the above. Information on the transmission rate of the third physical quantity between the anomalous component and each of the other components, or a plurality of of these information is included.
The diagnostic system according to claim 2.
The state degree information is indicated by a statistical state degree obtained by statistically processing the state detection information, a logical state degree obtained by applying the state detection information to a pre-stored arithmetic expression, and the state detection information. The load degree indicating the current burden, or multiple pieces of these information are included.
The diagnostic system according to claim 2 or 3.
The abnormality prediction unit applies the state degree information and the relationship degree information to a pre-stored calculation formula to calculate the risk of the next abnormality.
An input processing unit capable of changing the calculation formula from outside the system is provided.
The diagnostic system according to any one of claims 2 to 4.
The abnormality prediction unit outputs the prediction result and the reason of the component in which the abnormality occurs next.
The diagnostic system according to any one of claims 1 to 5.
Multiple sensors that are provided corresponding to multiple components of the device to be diagnosed and output the status detection information of each corresponding component, and
An impact calculation unit that calculates the impact between one designated component specified among the plurality of components and each of the other components.
Diagnostic system with.
The influence degree calculation unit is based on the relationship degree information indicating the relationship of each of the other components with respect to the designated component, and the state degree information of the component obtained based on the state detection information of each of the other components. To calculate the degree of influence,
The diagnostic system according to claim 7.