diagnostic device

Abstract

Diagnostic device (100, 100A), comprising: a sensor (110) that detects diagnostic target information generated by a diagnostic target device (1); a threshold setting unit (127) that sets a threshold for the diagnostic target information; and a diagnostic unit (129) that diagnoses the diagnostic target device (1) based on the diagnostic target information detected by the sensor (110) and the threshold value, wherein the threshold setting unit (127) sets the threshold based on the diagnostic target information in a predetermined period before a diagnostic time, and the diagnostic unit (129) performs a diagnosis based on a current time threshold set at the time of diagnosis and at least one past threshold set in the past.

Description

Technical field

[0001] The present invention relates to a diagnostic device. State of the art

[0002] In the prior art, a system was proposed that diagnoses an abnormality by detecting diagnostic target information (e.g., vibration) generated by a device, apparatus, or the like (see, for example, PTL 1).

[0003] In such a diagnostic system in the prior art, a device, instrument or the like is a diagnostic target, diagnostic target information is continuously detected by a sensor, the moving mean of the detected data is calculated, the standard deviation is obtained from the value, and a threshold for determining an abnormality is calculated from the values ​​of the moving mean and the standard deviation and continuously recorded.

[0004] Furthermore, an abnormality is diagnosed by comparing the currently detected data with the threshold calculated at a time close to the past operating conditions.

[0005] Furthermore, PTL 2 discloses a diagnostic device for a heating device, comprising an index calculation section that calculates an index obtained from a transition state of the temperature control, and a threshold determination section that determines an equilibrium threshold based on the index. Additionally, PTL 3 discloses a device for monitoring and diagnosing a rotating machine, comprising a vibration detector, an operating parameter detector, and a threshold setting device that sets a preliminary threshold based on stored detection values. PTL 4 discloses a method for diagnosing a sensor that provides a first signal and a second signal, taking into account a past history of the first signal and / or a past history of the second signal if the second signal has a time delay relative to the first signal. List of citations from patent literature [PTL 1] Japanese Unexamined Patent Publication No. 2017-181500 [PTL 2] Japanese Unexamined Patent Publication No. 2019-40439 [PTL 3] Japanese Unexamined Patent Publication No. H07-286892 [PTL 4] German unexamined patent application No. 101 45 485 A1 Summary of the invention: Technical problem

[0006] In the state of the art described above, there may be a case where the trend of the diagnostic target information, such as vibration, fluctuates even under the same operating conditions, and it is difficult to set a suitable threshold in such a case.

[0007] One object of the present invention is to provide a diagnostic device that enables a more suitable diagnosis. Solution to the problem

[0008] A diagnostic device according to the present invention comprises: a sensor that detects diagnostic target information generated by a diagnostic targeting device; a threshold setting unit that sets a threshold for the diagnostic target information; and a diagnostic unit that diagnoses the diagnostic target device based on the diagnostic target information detected by the sensor and the threshold value, wherein the threshold setting unit sets the threshold based on the diagnostic target information in a predetermined period before a diagnostic time, and The diagnostic unit performs a diagnosis based on a current-time threshold set at the time of diagnosis and at least one past threshold set in the past. Advantageous effects of the invention

[0009] According to the present invention, it is possible to provide a diagnostic device that is capable of suitably diagnosing a target device. Brief description of the drawings Fig. Figure 1 is a cross-sectional view showing a reduction gear, which is a positional example of a diagnostic target device. Fig. Figure 2 is a functional block diagram showing a diagnostic device according to one embodiment. Fig. Figure 3 is a flowchart showing a diagnostic process performed by a computer of the diagnostic device. Fig. Figure 4 is a graph showing a diagram depicting a transition from numerical values ​​of maximum value feature sets and a diagram depicting a transition from numerical values ​​of thresholds for the current time, with the diagrams superimposed. Fig. Figure 5 is a graph showing the diagram showing the transition of numerical values ​​from maximum value feature sets and a diagram showing a transition of numerical values ​​from second past thresholds, with the diagrams superimposed. Fig. Figure 6 is a block diagram showing another example of the diagnostic device according to the embodiment. Description of embodiments

[0010] Embodiments of the present invention are described in detail below with reference to the drawings.

[0011] This embodiment illustrates a diagnostic device that uses a reduction gear as a diagnostic target device and vibration as diagnostic target information.

[0012] Fig. Figure 1 is an axial sectional view of an eccentric oscillating reduction gear 1, which is a diagnostic targeting device. As shown in this figure, the reduction gear 1 includes a drive shaft 12 with three eccentric bodies 11, three external gears 13, an internal gear 14 engaging with the three external gears 13, a flange body 16 with an internal pin 15, a housing 17 holding the internal gear 14, and bearings 18 to 21. In the following description, a “radial direction” is the radial direction of the circumference centered on a centerline O of the drive shaft 12.

[0013] Since this reduction gear 1 has the same configuration as the power transmission device disclosed in Japanese unexamined patent publication No. 2006-263878, a detailed description of each configuration is omitted.

[0014] Fig. Figure 2 is a functional block diagram showing a diagnostic device 100 according to one embodiment.

[0015] The diagnostic device 100 includes an acceleration sensor 110 as a sensor provided in the reduction gear 1, a computer 120 which receives an output of vibration as diagnostic target information detected by the acceleration sensor 110 and performs the diagnostic processing of the reduction gear 1, a display unit 140 which displays and outputs information, and an input device 150, such as a pointing device or a keyboard, which can input data from outside into the computer 120.

[0016] The acceleration sensor 110 is located on the outer circumference of the housing 17 of the reduction gear 1, outside the internal gear 14 in the radial direction. The arrangement of the sensor is not limited to the above and can be changed as required.

[0017] The accelerometer 110 detects the vibration of the reduction gear 1 and outputs vibration waveform data to the computer 120 in response to the detection. For example, the accelerometer 110 can be a contact-type accelerometer that uses a piezoelectric element.

[0018] The computer 120 contains a central processing unit (CPU), a storage device that stores control data and programs, a random access memory (RAM), an interface that receives the vibration waveform data from the accelerometer 110, and the like, and is also connected to the display unit 140 and the input device 150.

[0019] The computer 120 diagnoses an abnormality of the reduction gear 1 by using the vibration waveform data of the reduction gear 1 according to a preset program.

[0020] As in Fig. As shown in Figure 2, the computer 120 contains a maximum value feature set calculation unit 121 as a feature set calculation unit, a feature set storage unit 122, a calculation unit 123 for moving mean, a standard deviation calculation unit 124, a threshold value calculation unit 125, a threshold value storage unit 126, a threshold value setting unit 127, a threshold value determination unit 128 and a diagnostic unit 129.

[0021] These are functional configurations implemented by the CPU of computer 120, which is running a program.

[0022] The following describes the processing activities performed by the various functional configurations of computer 120, with reference to the flowchart of Fig. 3 described. The specific numerical values ​​and quantities shown below are examples and can be changed.

[0023] First, the accelerometer 110 periodically detects vibration and generates vibration waveform data while the reduction gear 1 is operating. For example, a change in vibration is detected every 20 minutes for 10 seconds, and vibration waveform data is generated. Vibration detection can be performed continuously instead of periodically, and the continuous vibration waveform data can be separated by a predetermined time unit.

[0024] When the vibration waveform data is input from the accelerometer 110, the maximum value feature set calculation unit 121 calculates the current maximum value feature set from the vibration waveform (step S1).

[0025] When, as described above, the vibration waveform data indicating the change in vibration from 10 seconds ago to the present are acquired by the accelerometer 110, the maximum value feature set calculation unit 121 divides the vibration waveform into several sections, for example for 10 seconds, or extracts several sections that meet predetermined conditions, such as the magnitude of the rate of change, obtains the maximum value of vibration for each section as the feature set, and averages the feature sets, thereby calculating the maximum value feature set at the current time.

[0026] The maximum value feature set calculation unit 121 can pre-calculate the maximum value feature set after performing envelope processing on the vibration waveform data input by the accelerometer 110.

[0027] The feature set storage unit 122 performs a process every 20 minutes of storing the value of the maximum value feature set calculated by the maximum value feature set calculation unit 121 in the storage device of the computer 120.

[0028] The moving average calculation unit 123 calculates the mean (moving average) of several maximum value feature sets from 6 hours ago to the present (a predetermined period before the time of diagnosis), which are stored in the storage device and are recorded every 20 minutes by the maximum value feature set calculation unit 121 (step S3).

[0029] This means that the moving average calculation unit 123 calculates the moving average every 20 minutes according to the detection cycle of the accelerometer 110.

[0030] Next, the standard deviation calculation unit 124 calculates a standard deviation using the multiple maximum value feature sets from 6 hours ago to the present, which are stored in the storage device and are captured every 20 minutes by the maximum value feature set calculation unit 121, and using the mean calculated by the moving average calculation unit 123 (step S5).

[0031] That is, the standard deviation calculation unit 124 calculates the standard deviation every 20 minutes according to the detection cycle of the accelerometer 110.

[0032] Next, the threshold calculation unit 125 calculates a threshold value (denoted as T) using the mean (denoted as Xave) calculated by the moving average calculation unit 123 and the standard deviation (denoted as σ) from the standard deviation calculation unit 124 (step S7). This threshold value is used for comparison with the maximum value feature set to diagnose the abnormality of the reduction gear 1.

[0033] This means that the threshold calculation unit 125 calculates the threshold every 20 minutes according to the detection cycle of the accelerometer 110.

[0034] For example, the threshold calculation unit 125 calculates the threshold T via T = Xave + 3σ. The calculation formula is an example, and the threshold T can be calculated by changing the coefficient of σ or by another formula with Xave and σ as parameters.

[0035] The threshold storage unit 126 performs a process of storing the threshold calculated by the threshold calculation unit 125 in the storage device of the computer 120 every 20 minutes.

[0036] The threshold setting unit 127 performs a process of reading a current-time threshold, which is the last threshold calculated by the threshold calculation unit 125 at the current time, a first past threshold calculated 12 hours ago (past threshold), and a second past threshold calculated 24 hours ago (past threshold) from the computer's storage device 120 (step S9).

[0037] The threshold determination unit 128 compares the maximum value feature quantity at the current time, calculated by the maximum value feature quantity calculation unit 121, with the threshold for the current time, the first past threshold and the second past threshold, set by the threshold setting unit 127, and determines whether each of the thresholds is exceeded or not.

[0038] During the determination, the threshold determination unit 128 performs a process of resetting a counter variable to 0 in advance to count how many of the above three thresholds exceed the maximum value feature set at the current time (step S11).

[0039] Furthermore, the threshold determination unit 128 individually compares the maximum value feature set at the current time with the threshold for the current time, the first past threshold and the second past threshold.

[0040] This means that it is determined whether the current maximum value feature set is greater than the threshold for the current time (step S13). If the threshold for the current time is not exceeded, the process continues unchanged to step S25. If the threshold for the current time is exceeded, one is added to the counter variable (step S15), and then the process continues to step S25.

[0041] Similarly, it is determined whether the maximum value feature set is currently larger than the first past threshold or not (step S17). If the first past threshold is not exceeded, the process continues unchanged to step S25. If the first past threshold is exceeded, one is added to the counter variable (step S19), and then the process continues to step S25.

[0042] Similarly, it is determined whether the maximum value feature set is greater than the second past threshold at the current time (step S21). If the second past threshold is not exceeded, the process continues unchanged to step S25. If the second past threshold is exceeded, one is added to the counter variable (step S23), and then the process continues to step S25.

[0043] The determination of the size between the maximum value feature set at the current time and the threshold for the current time at step S13, the determination of the size between the maximum value feature set at the current time and the first past threshold at step S17, and the determination of the size between the maximum value feature set at the current time and the second past threshold at step S21 do not have to be carried out in parallel, but can be carried out sequentially.

[0044] The diagnostic unit 129 diagnoses, based on the determination results of steps S13 to S23 by the threshold determination unit 128, whether an abnormality has occurred in the reduction gear 1 or not (step S25).

[0045] In particular, diagnostic unit 129 determines whether an abnormality is present or not, depending on whether the aforementioned count variable is 50% or more of the maximum total count value of all determination devices. The determination devices consist of three devices, including one at step S13, where the maximum feature set at the current time is compared with the threshold value for the current time; one at step S17, where the maximum feature set at the current time is compared with the first past threshold value; and one at step S21, where the maximum feature set at the current time is compared with the second past threshold value. The maximum total count value is 3.Therefore, in a case where the count variable is greater than 1.5, the occurrence of an abnormality is diagnosed.

[0046] While the case is illustrated where the occurrence of an abnormality of reduction gear 1 is diagnosed in a case where the ratio of abnormalities in the determination results of the threshold for the current time and the first and second past thresholds is 50% or more, the ratio is an example and may be increased or decreased taking into account the number of determination devices.

[0047] Alternatively, the diagnostic unit can determine 129 weighting coefficients in advance for some or all of the counts of the current time threshold determination device, the first past threshold determination device, and the second past threshold determination device, multiply the weighting coefficients, and then diagnose an abnormality from the ratio to the maximum total value.

[0048] Preferably, the weighting coefficients are pre-stored in the storage device. In addition, the weighting coefficients can optionally be set via the input device 150.

[0049] In a case where the count variable is 1 or less and it is diagnosed that there is no abnormality, the diagnostic unit 129 controls the display unit 140 to indicate a normal state of the reduction gear 1 (step S27) and terminates a series of processes.

[0050] In a case where the counting variable is 2 or more and an abnormality is diagnosed, the display unit 140 is controlled to indicate the occurrence of an abnormality in the reduction gear 1 (step S29), and a series of processes is terminated.

[0051] The diagnostic unit 129 controls the display unit 140 at the time of each diagnosis to display a diagram B, showing the transition of numerical values ​​from maximum value feature sets captured every 20 minutes, and a diagram Tn, showing the transition of numerical values ​​from threshold values ​​captured every 20 minutes for the current time, so that they are superimposed on a graph (see Fig. 4).

[0052] Similarly, the diagnostic unit 129 controls the display unit 140 to display diagram B, which shows the transition of numerical values ​​from maximum value feature sets captured every 20 minutes, and a diagram showing the transition of numerical values ​​from first past thresholds captured every 20 minutes, so that they are superimposed on a graph (not shown).

[0053] Furthermore, the diagnostic unit 129 controls the display unit 140 to display diagram B, which shows the transition of numerical values ​​from maximum value feature sets captured every 20 minutes, and a diagram Ta, which shows the transition of numerical values ​​from second past thresholds captured every 20 minutes, in such a way that they are superimposed on a graph (see Fig. 5).

[0054] The diagnostic unit 129 then performs control to display on the three graphs above, shown on the display unit 140, the time at which the maximum value feature set exceeds each threshold and the time at which an abnormality is diagnosed.

[0055] Fig. Figure 5 shows an example of displaying the time (comparison result) at which the maximum value feature set exceeds the threshold, as line L2, and Fig. 4 and Fig. Figure 5 shows an example of displaying the time at which an abnormality is diagnosed as line L1.

[0056] The time at which an abnormality is diagnosed coincides with the time at which the maximum value feature set exceeds the threshold, but the line L1, which indicates the diagnosis of an abnormality, is preferentially displayed.

[0057] As in Fig. As shown in Figure 4, in a case where line L1 is selected by clicking or the like, various types of information regarding the diagnosis can be displayed in a pop-up display or the like, or outside the frame of the graph. Fig. The information displayed includes the detection date and time, the characteristic quantity value, the past threshold calculated 12 hours ago, the number of detections, and similar data. However, the information is not limited to this, and any related information can be displayed.

[0058] The number of detections shown here is a numerical value in a case where the number of times an abnormality is diagnosed within a predetermined period is counted and recorded. As described above, the diagnostic unit 129 can record the number of times an abnormality is diagnosed.

[0059] Displaying the different types of information regarding the above diagnosis is not limited to line L1 and can be done in the same way for line L2.

[0060] Furthermore, as in Fig. Figure 5 shows that the graph Tn, which shows the transition of the numerical values ​​of thresholds for the current time, and the graph Ta, which shows the transition of the numerical values ​​of second past thresholds, are displayed in such a way that they are superimposed on one screen (the same applies to the graph showing the transition of the numerical values ​​of first past thresholds). [Technical effects of embodiment of the invention]

[0061] As described above, in the diagnostic device 100, the threshold setting unit 127 sets the threshold based on the oscillation (maximum value feature set) in the predetermined period before the diagnosis time, and the diagnostic unit 129 performs diagnosis based on the current time threshold set at the diagnosis time and one or more past thresholds (the first past threshold and the second past threshold) that were set in the past.

[0062] Since diagnosis is performed using the threshold value for the current time, even in cases where there is a change in the oscillation trend, which is the diagnostic target information, the change is reflected in the threshold value for the current time. Accordingly, it is possible to set a suitable threshold and perform diagnosis.

[0063] In a case where diagnosis is performed solely using the threshold value for the current time, there is a concern that a threshold value influenced by an external factor at the time of diagnosis could be set, and there could be cases where an abnormality is determined by both an external and an internal factor. However, since the diagnostic unit 129 of the diagnostic device 100 also performs determinations based on the past threshold value, the influence of an external factor can be reduced, and the occurrence of an abnormality that is an internal factor of the device can be detected through this determination.

[0064] Similarly, in the detection of vibration, even in a case where fluctuations occur that cause noise, there is no impact on past threshold values ​​at different times. Therefore, the influence of noise on the diagnosis is reduced, and it is possible to perform a more appropriate diagnosis.

[0065] As a result, the diagnostic device 100 can more appropriately diagnose an abnormality of the device while responding to the fluctuation of the tendency of vibration, which is the diagnostic target information.

[0066] Furthermore, diagnostic unit 129 diagnoses that an abnormality is present in a case where the ratio of abnormalities in the diagnostic result of the threshold for the current time and the first and second past thresholds is a predetermined value (50% in the above example) or more.

[0067] By adjusting the ratio and determining an abnormality, as described above, even in a case where no abnormality is determined based on some of the threshold values, an abnormality can be determined based on the other threshold values. Therefore, it is possible to make a more appropriate diagnosis without omissions.

[0068] In a case where the diagnostic unit 129 is configured to weight all or some of the threshold for the current time and the first and second past thresholds, and then performs a diagnosis of an abnormality from the ratio of the abnormalities, it is possible to perform a more appropriate diagnosis according to the tendency of the diagnostic target.

[0069] Since the diagnostic device 100 has the display unit 140, which displays the transition of the maximum value characteristic set of vibration in a predetermined period, the threshold set in the predetermined period for the current time and each of the first and second past thresholds, it is also possible to visually recognize the correlation between the maximum value characteristic set of vibration, which changes over time, and each of the thresholds, and it is not only possible to diagnose an abnormality, but also to easily identify the operating state of the reduction gear 1.

[0070] Since the point in time at which an abnormality is diagnosed is marked by line L1 at the transition of the maximum value feature set of the oscillation within a predetermined period, as displayed on the display unit 140, not only the occurrence of the current abnormality but also the progression of past abnormalities can be visually verified. Therefore, changes in the maximum value feature set of the oscillation before and after the occurrence of an abnormality, as well as changes in each of the threshold values, can be simultaneously monitored. This is thus also useful for analysis regarding the diagnosis of an abnormality.

[0071] Since the display unit 140 shows the comparison result for each of the threshold values ​​for the current time and the first and second past threshold values ​​used for diagnosis with the line L2, it is also possible to visually and quickly recognize the magnitude of each of the threshold values ​​when the maximum value feature set of vibration exceeds each of the threshold values.

[0072] Furthermore, by displaying the information regarding the diagnosis at the times of lines L1 and L2 on the display unit 140, it is possible to visually and quickly recognize more detailed information at a specific time. [Further]

[0073] While the embodiment of the present invention has been described above, the present invention is not limited to this embodiment. That is to say, the details shown in the embodiment can be modified appropriately without departing from the concept of the invention.

[0074] For example, in this embodiment, the diagnostic target device is a reduction gear, but this is not limited to it. For example, this embodiment can be applied to any machine, instrument, and device that requires diagnostics, such as a geared motor, an injection molding machine, and a cryogenic cooler.

[0075] Furthermore, the diagnostic target information is not limited to vibration and can be any information that can be detected by various sensors and used for diagnosis, such as current and temperature.

[0076] Furthermore, the content of the diagnosis is not limited to the presence or absence of an abnormality and can include any events that need to be predicted in advance, such as predicting life expectancy.

[0077] There may be cases where the trend of the diagnostic target information fluctuates due to the influence of the surrounding environment. For example, in a case where the temperature or similar changes within a fixed time window, and thus the trend of the diagnostic target information fluctuates within that fixed time window, the diagnostic device 100 may be equipped with a timer or similar device to change the cycle in which the sensor performs sampling within that fixed time window (for example, to shorten the period to increase the detection frequency).

[0078] Furthermore, the threshold values ​​from 12 hours ago and 24 hours ago are used as the past threshold values, but these times are examples and can be modified as needed according to the characteristics of the vibration target information. Additionally, the number of past threshold values ​​used for diagnostics can optionally be increased or decreased.

[0079] Furthermore, the embodiment illustrates a case in which the maximum value feature set calculation unit 121, as the feature set calculation unit of the computer 120, calculates the maximum value feature set by dividing the vibration waveform data input by the acceleration sensor 110 into several sections and by averaging the maximum values ​​of vibration in any section as the feature set, but the feature set calculation unit is not limited to this.

[0080] The feature set calculation unit can obtain the feature set from the processing result obtained by performing predetermined numerical processing on the detected vibration waveform data, and further preferably performs conversion into a numerical value that has a correlation with the original vibration waveform data, or into another numerical value that retains the tendency of a change in numerical value.

[0081] As an example, if the feature set calculation unit acquires the vibration waveform data from the accelerometer 110, the feature set calculation unit can divide the vibration waveform for 10 seconds into several sections or extract several sections that meet predetermined conditions, such as the magnitude of the rate of change, obtain the minimum or mean value of vibration for each section as the feature set, and further average the feature sets, thereby calculating the feature set at the current time.

[0082] Furthermore, the feature set calculation unit can individually calculate several types of feature sets via several different operations. For example, the minimum, mean, and maximum values ​​of vibration are obtained for each section, and the values ​​are averaged for each section. The feature set at the current time is then calculated based on the minimum value, the mean value, and the maximum value.

[0083] Then, the threshold calculation unit 125 calculates the threshold for each of the multiple sets of characteristics, and the threshold determination unit 128 compares the multiple sets of characteristics with the threshold for the current time and one or more past thresholds, obtaining the count of the determination result. For example, in a case where three types of sets of characteristics are obtained and compared with the threshold for the current time, the first past threshold, and the second past threshold, the maximum total count is 3 × 3 = 9.

[0084] Therefore, the diagnostic unit 129 can diagnose the occurrence of an abnormality in the reduction gear 1 in a case where the ratio of abnormalities in the determination result to the maximum total value 9 is a predetermined ratio (for example, 50%) or more.

[0085] Furthermore, in the embodiment, the configuration which includes the accelerometer 110, the computer 120, the display unit 140 and the input device 150 is illustrated as the diagnostic device 100, but the configuration can be changed as long as the configuration has the same function as the diagnostic device.

[0086] For example, as in Fig. Figure 6 shows a diagnostic device configured by several sensors (illustrated by four accelerometers 110) that are individually provided at several diagnostic target devices, or by several sensors that are provided at several locations of one or more diagnostic target devices, several diagnostic units 120A that are individually provided at the several sensors, and a host computer 130A that is connected to all the diagnostic units 120A and has a display unit and an input device.

[0087] In the case of a diagnostic device 100A, in a configuration where the diagnostic unit 120A and the host computer 130A contain a CPU, RAM, auxiliary storage device, and the like, various functions of the maximum value feature set calculation unit 121, the feature set storage unit 122, the moving average calculation unit 123, the standard deviation calculation unit 124, the threshold calculation unit 125, the threshold storage unit 126, the threshold setting unit 127, the threshold determination unit 128, and the diagnostic unit 129 can be shared and executed by the diagnostic units 120A and the host computer 130A.

[0088] Alternatively, each of the functional units 121 to 129 can be executed by each of the diagnostic units 120A, and the host computer 130A can be configured to perform a process of collecting and displaying the diagnostic results and transmitting various settings to each of the diagnostic units 120A.

[0089] Furthermore, the configuration is not limited to the above, and each of the functional units 121 to 129 can be shared by multiple processing devices. Industrial applicability

[0090] The present invention is industrially applicable to a diagnostic device. Reference symbol list 1 Reduction gear (diagnostic target device) 100,100A diagnostic device 110 Accelerometer (Sensor) 120 computers 120A diagnostic unit 121 Maximum value feature set calculation unit 122 Feature set storage unit 123 Unit of calculation for moving average 124 Standard deviation calculation unit 125 Threshold calculation unit 126 Threshold storage unit 127 Threshold setting unit 128 Threshold determination unit 129 Diagnostic Unit 130A Host Computer 140 display unit 150 Input device

Claims

[1] Diagnostic device (100, 100A), comprising: a sensor (110) that detects diagnostic target information generated by a diagnostic target device (1); a threshold setting unit (127) that sets a threshold for the diagnostic target information; and a diagnostic unit (129) that diagnoses the diagnostic target device (1) based on the diagnostic target information detected by the sensor (110) and the threshold value, wherein the threshold setting unit (127) sets the threshold based on the diagnostic target information in a predetermined period before a diagnostic time, and the diagnostic unit (129) performs a diagnosis based on a current time threshold set at the time of diagnosis and at least one past threshold set in the past. [2] Diagnostic device (100, 100A) according to claim 1, wherein the diagnostic target information includes a set of features obtained from a processing result obtained by performing a predetermined numerical processing on a detected value of the sensor (110). [3] Diagnostic device (100, 100A) according to claim 1 or 2, wherein the diagnostic unit (129) diagnoses that an abnormality is present in a case where a ratio of abnormalities in a diagnostic result of the threshold for the current time and the past threshold is a predetermined value or more. [4] Diagnostic device (100, 100A) according to claim 1, wherein the diagnostic target information contains several types of feature sets obtained from processing results obtained by performing several types of predetermined numerical processing on a detected value of the sensor (110), and The diagnostic unit (129) individually receives a diagnostic result from each of the thresholds for the current time and the past thresholds for the multiple types of feature sets and diagnoses that an abnormality is present in a case where a ratio of abnormalities in the diagnostic results is a predetermined value or more. [5] Diagnostic device (100, 100A) according to claim 3 or 4, wherein the diagnostic unit (129) weights some or all of the diagnostic results from the threshold for the current time and the past threshold, respectively, and diagnoses that an abnormality is present in a case where the ratio of abnormalities is a predetermined value or more. [6] Diagnostic device (100, 100A) according to any one of claims 1 to 5, further comprising: a display unit (140) that indicates a transition of the diagnostic target information in a predetermined period and, respectively, the threshold for the current time and the past threshold set in the predetermined period. [7] Diagnostic device (100, 100A) according to claim 6, wherein, during the transition of the diagnostic target information in the predetermined period displayed on the display unit (140), a time point is marked at which it is diagnosed that an abnormality is present. [8] Diagnostic device (100, 100A) according to any one of claims 1 to 7, further comprising: a display unit (140) that shows a comparison result for each of the threshold values ​​for the current time and the past threshold values, which are used for diagnosis. [9] Diagnostic device (100, 100A) according to one of claims 6 to 8, the display unit (140) displays information regarding diagnosis at a specific time.

Images