JP2024025095A - Trend calculation device, trend calculation system, and trend calculation method - Google Patents

Abstract

An object of the present invention is to provide a trend calculation device and the like that can obtain information for carrying out inspections at appropriate timing. [Solution] A trend calculation device includes an acquisition unit (server communication unit) that acquires the measurement result of measuring the current value of the supplied power to a drive source driven by the supplied power, and a past first Based on the measurement results obtained during the time point and the second time point after the first time point, input based on the current value shown in each obtained measurement result and when the drive source is driving normally. Calculate a trend indicating the change tendency between the first time point and the second time point when the degree of change corresponding to the magnitude of the difference from the standard based on the electric current value measured in advance increases or decreases over time. and a calculation unit (one function of the server processing unit) that outputs the output. [Selection diagram] Figure 7

Description

The present invention relates to a trend calculation device, a trend calculation system, and a trend calculation method.

In recent years, as robot control technology has improved, technology has been developed that uses machine learning to allow robots to automatically control movement in the same way as when operated by a human. In such devices such as robots, it is necessary to determine whether there is an abnormality (trouble) in the device and to perform inspections or the like according to the results. For example, a diagnostic system is known that can improve the accuracy of determining the degree of abnormality in equipment even when there is no period during which the rotational speed of a motor, etc. that serves as the equipment's drive device (drive source) is maintained constant. (See Patent Document 1).

International Publication No. 2019/244599

By the way, when inspecting each drive source of a device after a problem occurs and restoring it to a state where the problem has been resolved, the problem is that the device must be stopped, and the service itself that uses the device will also be stopped. There is. On the other hand, in order to inspect equipment sufficiently before a problem occurs, frequent inspections that almost certainly do not occur are necessary and complicated.

Therefore, the present disclosure provides a trend calculation device and the like that can obtain information for inspecting a drive source used for operating a device at an appropriate timing.

A trend calculation device according to one aspect of the present invention includes an acquisition unit that acquires a measurement result of measuring a current value of the supplied power to a drive source driven by the supplied power, and a first time point in the past. and based on the measurement results obtained during a second time point after the first time point, an input based on a current value indicated in each of the obtained measurement results, and whether the drive source is normally driven. A change trend between the first time point and the second time point when the degree of change corresponding to the magnitude of the difference from the reference based on the current value of the electric power measured in advance when the electric power is in use increases or decreases over time. and a calculation unit that calculates and outputs the trend shown.

Further, a trend calculation system according to one aspect of the present invention includes the trend calculation device described above and a current sensor used for measuring a current value that the acquisition unit acquires as the measurement result, and the current sensor Measures a current value of power supplied from outside the power line connecting the drive source and a power supply source that supplies power to the drive source via the power line.

Further, in the trend calculation method according to one aspect of the present invention, a measurement result obtained by measuring a current value of the power supplied to a drive source driven by the supplied power is obtained, and Based on the measurement results obtained during the second time point after the first time point, the current value indicated in each of the obtained measurement results and the current value when the drive source is normally driven. A trend is calculated that indicates a tendency of change between the first time point and the second time point when the degree of change corresponding to the magnitude of the difference between the electric power and the current value increases or decreases over time.

Note that these comprehensive or specific aspects may be realized by a system, a device, an integrated circuit, a computer program, or a computer-readable recording medium such as a CD-ROM, and the system, device, integrated circuit, computer program and a recording medium may be used in any combination.

According to the trend calculation device and the like of the present disclosure, it is possible to obtain information for inspecting the drive source at an appropriate timing.

FIG. 1 is a diagram schematically showing an overview of a trend calculation system according to the present embodiment. FIG. 2 is a block diagram schematically showing the functional configuration of the trend calculation system according to this embodiment. FIG. 3 is a diagram for explaining measurement characteristics of the current sensor according to this embodiment. FIG. 4 is a sequence diagram for explaining an example of the operation of the trend calculation system according to this embodiment. FIG. 5 is a sequence diagram for explaining an example of the operation of the trend calculation system according to the present embodiment. FIG. 6 is a diagram for explaining trends calculated in the trend calculation system according to the present embodiment. FIG. 7 is a diagram for explaining notification information output in the trend calculation system according to the present embodiment.

(Summary of disclosure)
A trend calculation device according to a first aspect of the present disclosure includes an acquisition unit that acquires a measurement result of measuring a current value of power supplied to a drive source driven by the supplied power, and a first time point in the past. Based on the measurement results obtained during the second time point after the first time point, input based on the current value shown in each obtained measurement result, and input in advance when the drive source is driving normally. Calculates and outputs a trend indicating the change tendency between the first and second time points when the degree of change corresponding to the magnitude of the difference from the reference based on the current value of the measured power increases or decreases over time. and a calculation unit.

Such a trend calculation device calculates the difference from the first point in time based on the degree of change corresponding to the difference between the input based on the current value shown in the measurement result and the reference based on the current value during normal driving. A trend indicating a tendency of increase/decrease in the degree of change between the second time points can be calculated and output. The magnitude of the difference between the input based on the current value shown in the measurement results and the reference based on the current value during normal driving indicates how far the current value is from the normal state. Since it includes information on how abnormal the state is, it can be used to estimate the occurrence of an abnormality in the drive source. Here, the degree of change at one point includes cases where an abnormality actually occurs and cases where the abnormality appears normal due to external factors or the like. On the other hand, by calculating a trend that shows the change tendency between the first and second time points, the trend can be used to detect the occurrence of abnormalities in the drive source from the tendency of increase/decrease in the degree of change at multiple points in time. It becomes possible to estimate. That is, according to the above aspect, it is possible to obtain a trend that can be used to inspect the drive source at an appropriate timing.

Further, for example, in the trend calculation device according to the second aspect, the calculation unit includes a learned trend calculation model that outputs a trend in response to input of measurement results obtained between the first time point and the second time point. , the trend calculation device described in the first aspect may be used.

According to this, the trend can be calculated as an output according to the input of the measurement result using the learned trend calculation model.

Further, for example, in the trend calculation device according to the third aspect, the calculation unit has a learned change calculation model that outputs the change degree according to the input of the acquired measurement results, and Described in the first aspect, in which the trend is calculated based on an approximation formula in the time domain that approximates each degree of change output when each measurement result obtained between time points is input into a degree of change calculation model. It may be a trend calculation device.

According to this, the degree of change can be calculated as an output according to the input of the measurement result using the learned degree of change calculation model. Then, a trend can be calculated based on an approximation formula that approximates the calculated degree of change in the time domain.

Further, for example, in the trend calculation device according to the fourth aspect, the acquisition unit acquires the current value measured by the current sensor as the measurement result, and the current sensor has a higher sensitivity as the frequency is higher in the frequency region of the current value. The trend calculation device according to any one of the first to third aspects may have measurement characteristics.

According to this, when an abnormality occurs in the drive source, it is possible to measure with high sensitivity the harmonic frequency region where the difference from the reference based on the current value during normal driving tends to be more noticeable. Therefore, the occurrence of an abnormality in the drive source can be clearly estimated.

Further, for example, in the trend calculation device according to the fifth aspect, when the calculated trend indicates an upward trend in which the change tendency of the degree of change between the first time point and the second time point becomes larger than a predetermined threshold value, the drive The trend calculation device according to any one of the first to fourth aspects may further include a diagnosis unit that diagnoses that the operation of the source is abnormal and outputs a diagnosis result.

According to this, it is possible to diagnose the drive source based on the trend and output the diagnosis result.

Further, for example, in the trend calculation device according to the sixth aspect, the diagnosis result includes the calculated trend when an upward trend is indicated and a message indicating that the operation of the drive source to be measured is abnormal. The trend calculation device according to the fifth aspect may include at least one of the above.

According to this, at least one of the calculated trend when an upward trend is indicated and a message indicating that the operation of the drive source to be measured is abnormal can be output as the diagnosis result.

Further, a trend calculation system according to a seventh aspect of the present disclosure includes the trend calculation device according to any one of the first to sixth aspects, and is used for measuring a current value that the acquisition unit acquires as a measurement result. A current sensor is provided, and the current sensor measures a current value of power supplied from outside the power line connecting the drive source and a power supply source that supplies power to the drive source via the power line.

According to this, a trend calculation system including a trend calculation device and a current sensor can be realized. The current sensor in the trend calculation system is configured to measure the current value of the power supplied from the outside of the power line connecting the drive source and the power supply source that supplies power to the drive source via the power line. The trend calculation system can be installed after the drive source, the power supply source, and the power line are installed without removing the power line connecting the drive source and the power supply source, that is, as a retrofit.

Further, the trend calculation method according to the eighth aspect of the present disclosure acquires the measurement result of measuring the current value of the supplied power to a drive source driven by the supplied power, and Based on the measurement results obtained during the second time point after the first time point, the current value shown in each obtained measurement result and the current value of the electric power when the drive source is normally driven. A trend indicating a change tendency between the first time point and the second time point is calculated when the degree of change corresponding to the magnitude of the difference increases or decreases over time.

According to this, the same effects as the trend calculation device described above can be achieved.

Note that these comprehensive or specific aspects may be realized by a system, a device, an integrated circuit, a computer program, or a computer-readable recording medium such as a CD-ROM. Alternatively, it may be realized using any combination of recording media.

Hereinafter, embodiments will be specifically described with reference to the drawings.

Note that the embodiments described below are all inclusive or specific examples. The numerical values, shapes, materials, components, arrangement positions and connection forms of the components, steps, order of steps, etc. shown in the following embodiments are merely examples, and do not limit the present invention. Further, among the constituent elements in the following embodiments, constituent elements that are not described in the independent claims indicating the most significant concept will be described as arbitrary constituent elements.

(Embodiment)
First, the configuration of the trend calculation system in this embodiment will be explained.

FIG. 1 is a diagram schematically showing an overview of a trend calculation system according to the present embodiment. FIG. 2 is a block diagram schematically showing the functional configuration of the trend calculation system according to this embodiment.

In the present embodiment, the trend calculation system measures the current value of the power supplied to drive the drive source 99 before an abnormality occurs in the drive of the drive source 99, and detects a sign that an abnormality will occur. , is configured so that it can be understood by calculating trends. In other words, it is possible to know the signs of an abnormality based on the trend before it occurs, so the timing of inspection can be determined. Therefore, the drive source 99 can be stopped and inspected in a planned manner, or a substitute drive source can be used in place of the drive source 99 while the drive source 99 is inspected. A trend is a reference based on a pre-measured current value (normal value) of the supplied power when the drive source 99 is operating normally, and an input based on a current value measured during trend calculation. This is an index showing the tendency of change over time in the degree of change corresponding to the magnitude of the difference. Note that a standard based on a normal value may be the normal value itself, a differential value obtained by differentiating the normal value current with respect to time, or a value obtained by further performing another calculation process on the above normal value or differential value. Often, for example, it is the power of each frequency in the frequency domain of normal values. Furthermore, the input based on the current value may be the current value itself, a differential value obtained by differentiating the current with respect to time, or a numerical value obtained by further performing another calculation process on the above-mentioned current value or differential value. For example, , is the power of each frequency in the frequency domain of the current value.

For the reference based on the normal value and the input based on the current value, numerical values obtained by processing measured values in another system may be used. An autoencoder may be used as this other system. The autoencoder is a dimension reduction algorithm using a neural network. To explain an example using an autoencoder, a trained model is constructed by performing unsupervised learning using multiple normal values as input and output. If a normal value is input to this trained model, the output value will be the same as the input value, but if a value that is not normal is input, the output value will be a value different from the input value.

The standard based on the normal value may be the difference between the input value and the output value when the normal value is input to the model of the above-described other system (that is, in the learning phase). In this case, even if the normal value is input to the model of another system, the difference between the input value and the output value will be 0. As the input based on the current value, the difference between the input value and the output value when the measured current value is input to the model of the above-described other system (that is, in the inference phase) may be used. In this case, when a numerical value that is not a normal value is input to a model of another system, the difference between the input value and the output value becomes a predetermined numerical value that is not zero. Then, as the difference between the reference based on the normal value and the input based on the current value, the difference between 0 and a predetermined numerical value (that is, the predetermined numerical value itself) is calculated as the degree of change. Note that another system may use, for example, Mahalanobis distance, or may use other unsupervised learning.

The degree of change becomes larger as the input based on the current value deviates from the standard based on the normal value. When the trend shows an upward trend in which the degree of change increases over time, it can be estimated that there is a sign that trouble such as a failure will occur in the drive source 99 that is the measurement target. By the way, the degree of change itself also contains information such as the size of the difference between the input based on the current value and the standard based on the normal value, so the degree of change can be used to directly estimate signs of trouble. It is also possible to do so. However, if the degree of change is estimated only at one point in time in the time domain, a "pseudo" sign will be estimated even when there is no actual sign of trouble due to noise or the like. Such false signs cause unnecessary inspections, which are not only complicated but also cause an increase in inspection costs. Therefore, in this embodiment, by calculating a trend using the degree of change at multiple points in time in the time domain, estimation of such pseudo signs can be suppressed, and complexity and cost can be reduced. becomes possible.

In this embodiment, drive source 99 includes a motor 99a and a rotation conversion mechanism 99b connected to motor 99a. If a sign of trouble occurs in the motor 99a, physical drive resistance will occur, and an abnormality will occur in the current value of the supplied power. The rotation conversion mechanism 99b is a mechanism that converts the direction of the drive shaft of the rotation drive by the motor 99a, the position of the drive shaft, the amount of drive, the drive form, etc. For example, in the figure, a belt is shown that changes the position and drive amount of the drive shaft by a belt drive. The rotation conversion mechanism 99b may be any other structure that converts the rotational drive of the motor 99a, such as a ball screw, gear box, or rack.

If any sign of trouble occurs in the drive (drive conversion) of the rotation conversion mechanism 99b, this will become a physical drive resistance of the motor 99a, and an abnormality will occur in the current value of the supplied power. Therefore, by measuring the current value of the power supplied to the motor 99a, it is possible to estimate the occurrence of a sign of trouble in the entire drive source 99 by calculating a trend.

The current value of the power supplied to the drive source 99 is measured by the current sensor 98. The current sensor 98 measures the current value of the power supplied from the outside of the power line 99d that connects the drive source 99 and the power supply source 99c that supplies power to the drive source 99 via the power line 99d. Specifically, the current sensor 98 has a magnetic core that goes around the outside of the insulating coating of the power line 99d in a plane of intersection that intersects the power line 99d. Then, the magnetic flux generated by the current flowing inside the magnetic core is measured as a voltage value by a coil element wound around the magnetic core. The electric current value of electric power is measured according to such a principle.

By the way, if the current sensor 98 is configured to be installed outside the power line 99d as described above, the current sensor 98 can be installed without disconnecting any of the connections between the power supply source 99c, the drive source 99, and the power line 99d. can be installed. In other words, since the current sensor 98 can be retrofitted to existing equipment, it is possible to further reduce the barrier to introducing the trend calculation system. The measurement results measured by the current sensor 98 are transmitted to the edge device 200. The edge device 200 includes a master device 200a and a slave device 200b, and one or more current sensors 98 are connected to the slave device 200b.

The slave device 200b aggregates measurement results from several current sensors 98 (for example, multiple current sensors 98 for measuring the drive sources 99 of each of the multiple joints of one robot arm) and outputs the results to the parent device 200a. send to. The master device 200a transmits the measurement results of the current values thus collected to the server device 100. The drive source 99, etc., the current sensor 98, and the edge device 200 are arranged in the building where the drive source 99 is installed. On the other hand, the server device 100 is realized, for example, as a cloud server built on the network 150 (see FIG. 2).

Note that since the functions of the server device 100 are optimized for each measurement result obtained from the drive source 99, they may be provided individually for each drive source 99, each edge device 200, or the like. This has the advantage of, for example, making it easier to visualize the usage rate of processing resources for each user of the trend calculation system, and making it easier to calculate system usage fees in accordance with the usage rate. Another benefit is that it makes it easier to restrict access to data among users of the trend calculation system, which is advantageous from a security perspective. Further, the server device 100 may be placed as an edge server in a building in which the drive source 99 is installed. In this case, the server device 100 may be integrated with some other components such as the base device 200a.

As shown in FIG. 2, the trend calculation system includes a server device 100, an edge device 200, a current sensor 98, and a terminal device 300. Note that as long as the trend calculation system includes at least some functions of the server device 100 and the current sensor 98, the trend calculation system does not need to include other configurations.

The server device 100 includes a server control section 110, a server processing section 120, a server storage section 130, and a server communication section 140.

The server control unit 110 is a functional unit for controlling various functions of the server device 100. The server control unit 110 has a processor and a memory, and controls various functions of the server device 100 by executing a predetermined program.

The server processing unit 120 functions as a trend calculation unit that calculates a trend from measurement results, and a diagnostic unit that diagnoses whether or not the operation of the drive source 99 is abnormal based on the calculated trend and outputs the diagnosis result. and features. The server processing unit 120 is realized by executing programs corresponding to each function using a processor and memory.

The server storage unit 130 is a storage module that stores acquired measurement results, programs related to each function of the server device 100, and the like. The server storage unit 130 is realized by a non-temporary computer-readable recording medium or the like.

The server communication unit 140 is a communication module that communicates with the edge device 200 and the terminal device 300 via the network 150. The server communication unit 140 receives measurement results from the edge device 200 and outputs them to the server storage unit 130 and the server processing unit 120. The server communication unit 140 is an example of an acquisition unit.

The edge device 200 includes a control section 210, a communication section 220, a storage section 230, and a measurement section 240.

The control unit 210 is a functional unit for controlling various functions of the edge device 200. The control unit 210 has a processor and a memory, and controls various functions of the edge device 200 by executing a predetermined program.

The communication unit 220 is a communication module that communicates with the server device 100 via the network 150. The communication unit 220 transmits the measurement results obtained from the current sensor 98 to the server device 100.

The storage unit 230 is a storage module that stores measurement results obtained from the current sensor 98, programs related to each function of the server device 100, and the like. The storage unit 230 temporarily stores the measurement results obtained from the current sensor 98, and stores a certain amount of measurement results in order to transmit the measurement results to the server device 100 every time a certain amount of measurement results are accumulated. Output as a combined data set. The storage unit 230 is realized by a non-temporary computer-readable recording medium or the like.

The basic configurations of the current sensor 98 and the drive source 99 are the same as described above, so the explanation here will be omitted.

By the way, the current sensor 98 measures the current value, and the current value indicated by the measurement result is used to estimate a sign of trouble in the drive source 99. Therefore, from the viewpoint of estimating signs of trouble, it is preferable to use an appropriate current sensor 98. For example, in this embodiment, a current sensor 98, also called a harmonic sensor, is used. The difference between a harmonic sensor and a general current sensor will be explained with reference to FIG. 3. FIG. 3 is a diagram for explaining measurement characteristics of the current sensor according to this embodiment. In Figure 3, the upper row shows the spectrum obtained by performing Fourier transform on the current value measured by the harmonic sensor, as shown as the power for each frequency component, and the lower row shows the spectrum obtained by performing a Fourier transform on the current value measured by the harmonic sensor. This shows the spectrum obtained when Fourier transform was performed on the measurement results of the current values measured in . Note that FIG. 3 shows measurement results for the same measurement target.

As the harmonic sensor here, a current sensor having a measurement characteristic in which the higher the frequency is, the higher the sensitivity is used in the frequency domain of the current value. More specifically, a current sensor whose sensitivity increases in proportion to frequency is used as a harmonic sensor. Therefore, the spectral power in the high frequency region (on the right side of the paper) is measured to be higher than that of a general current sensor. Normally, it is known that when there is a problem with the drive source 99, a change in current value appears in a relatively high-frequency region, and by using a harmonic sensor with the measurement characteristics described above, it is possible to prevent signs of trouble from occurring. It becomes possible to improve detection accuracy. In addition, in the above example, a current sensor whose sensitivity increases in proportion to frequency is used as a harmonic sensor, but as a harmonic sensor, by directly measuring the value obtained by differentiating the current with respect to time, Current sensors with similar effects may also be used. Note that it is not essential to use such a harmonic sensor; for example, the same function can be achieved even with the measurement characteristics of a general current sensor shown in the lower row, although the detection accuracy is low.

Returning to the explanation of FIG. 2, the terminal device 300 includes a screen generation section 310, a terminal display section 320, a terminal control section 330, and a terminal communication section 340.

The screen generation unit 310 is a processing unit that generates an image signal to be displayed on the terminal display unit 320 from image information. Screen generation section 310 is realized by a processor and memory. In particular, the processor of the screen generation unit 310 uses a processor specialized in simple calculations such as image processing, such as a so-called graphics processing unit.

The terminal display section 320 is a display module having a display element such as a liquid crystal, an organic EL, or a micro LED. An image is displayed on the screen (display surface) of the terminal display section 320 according to the image signal generated by the screen generation section 310.

The terminal control unit 330 is a functional unit for controlling various functions of the terminal device 300. The terminal control unit 330 has a processor and a memory, and controls various functions of the terminal device 300 by executing a predetermined program.

Terminal communication unit 340 is a communication module that communicates with server device 100 via network 150. The terminal communication unit 340 outputs information received from the server device 100 to the screen generation unit 310 in order to display it on the terminal display unit 320, and inputs input from the user via an input unit (not shown) to the server device. It has functions such as transmitting to 100.

Next, the operation of the trend calculation system will be explained with reference to FIGS. 4 to 7. 4 and 5 are diagrams for explaining trends calculated in the trend calculation system according to the present embodiment. In particular, FIG. 4 shows the operation in the phase of learning the trend calculation model included in the trend calculation system, and FIG. 5 shows the operation in the phase of inference using the trained trend calculation model.

When a trend calculation system is introduced, first, as shown in FIG. 4, an operation for learning a trend calculation model is performed. First, the current sensor 98 measures the current value of the power supplied to the drive source 99 (S101). The measurement results are sequentially transmitted to the edge device 200 and stored in the storage unit 230 (S102). After the current sensor 98 measures (S101) and accumulates in the storage unit 230 (S102) repeatedly, when the amount of measurement results (number of data pieces) required for learning the trend calculation model is accumulated, these The measurement results are transmitted to the server device 100 (S103). The server communication unit 140 receives (obtains) the measurement results (S104). The server storage unit 130 stores the received measurement results (S105), and when the required amount (number of data sets) of measurement results is accumulated, outputs the data set of these measurement results to the server processing unit 120. .

The server processing unit 120 trains the trend calculation model using the output measurement result data set (S106). Since the trend calculation model is configured to calculate trends from measurement results, correct data of trends corresponding to each data set, etc. are used for learning. Note that the configuration of the trend calculation model is not limited to one that calculates trends from measurement results, but may also be one that calculates the degree of change from measurement results, or simply calculates standard data of normal values. But that's fine.

When the trend calculation model is configured to calculate the degree of change from measurement results (that is, the degree of change calculation model), the operation of calculating the trend from a plurality of degrees of change is performed using a predetermined algorithm. In addition, if the trend calculation model simply calculates standard data of normal values, there are operations to calculate the degree of change from the generated standard data, and operations to calculate the trend from multiple degrees of change. are each performed using a predetermined algorithm.

Subsequently, the server processing unit 120 evaluates the trend calculation model corresponding to the learning results using the accumulated measurement results (S107). The learning results are stored in the server storage unit 130 in a state where they are linked to the evaluation results (S108). In order to determine a trend calculation model corresponding to the learning results used for inference, the server processing unit 120 determines whether the evaluation results linked to the learning results are good (S109). If the evaluation result is not good (No in S109), the process returns to step S101 and the same process is repeated starting from measuring the current value.

If the evaluation result is good (Yes in S109), the process proceeds to step S110, and the server processing unit 120 sets the trend calculation model corresponding to the learning result for trend calculation by inference.

Next, as shown in FIG. 5, an inference operation using the trend calculation model is performed. First, the current sensor 98 measures the current value of the power supplied to the drive source 99 (S201). The measurement results are sequentially transmitted to the edge device 200 and stored in the storage unit 230 (S202). After the current sensor 98 measures (S201) and accumulates in the storage unit 230 (S202) repeatedly, and the measurement results are accumulated in the amount (number of data) required for trend calculation, these measurement results are It is transmitted to the server device 100 (S203). The server communication unit 140 receives (obtains) the measurement result (S204).

The server storage unit 130 stores the received measurement results (S205), and when a necessary amount (data number) of measurement results is accumulated, it outputs a data set of these measurement results to the server processing unit 120. The required number of data here corresponds to one data set during learning of the trend calculation model, but may be a different number. However, from the perspective of evaluating the long-term tendency of changes in current values by trend calculation, between two different points in time (i.e., a first point in the past and a second point in time after the first point in time) It is necessary to accumulate measurement results that complement the above. The period between the first time point and the second time point varies depending on how long it takes to transition from a stable period (a period of stable operation) to a failure period (a period in which the possibility of failure due to aging increases). For example, periods of days, weeks, or months may apply. Further, trend calculation may be performed for all periods in the past. In this case, the first point in time is the first point in time when the inference phase of the trend calculation system is started, and the second point in time is the point in time when the most recent measurement result is obtained.

After accumulating measurement results within a predetermined period between a first time point in the past and a second time point after the first time point, the server processing section 120 uses the function as a calculation section to collect the measurement results. is input into the trend calculation model to output (calculate) the trend (S206). At this time, first, it is determined whether or not the trend calculation system is set to perform determinations based on trends in subsequent steps S208 and subsequent steps (S207). For example, setting information in which the user sets whether or not to make a judgment based on trends is stored in the server storage unit 130 or the like, and the server processing unit 120 is set to read out the setting information and make judgments based on trends. Determine whether or not. This determination of whether or not to make a determination based on the trend is not essential, and if the trend determination system is set to always make a determination based on the trend, step S207 may be omitted.

If it is set in the setting information that determination based on trends is not performed (No in S207), the process does not proceed to determination based on trends. In that case, for example, the process may be terminated as shown in the figure, or the process may be terminated using information of the nature of the difference between the input based on the current value and the standard based on the normal value, which is included in the degree of change itself. Processing (not shown) for directly estimating signs of trouble may be started. In the setting information, if it is set to perform determination based on the trend (Yes in S207), the server processing unit 120 uses the function of the diagnosis unit to determine whether the output trend is determined between the first time point and the second time point. It is determined whether or not the change tendency of the degree of change shows an upward trend that becomes greater than a predetermined threshold value (S208). According to this determination result, the server processing unit 120 diagnoses whether or not the operation of the drive source 99 is abnormal, and outputs (notifies) the diagnosis result.

Here, FIG. 6 is a diagram for explaining trends calculated in the trend calculation system according to the present embodiment. FIG. 6 shows a graph of the degree of change versus time (indicated by white circles) (if the degree of change is calculated). The upper part of FIG. 6 shows the notification timing when a threshold value is simply determined for the degree of change, and the lower part of FIG. 6 shows the notification timing when determining whether an upward trend is indicated. There is. As shown in the upper part of Figure 6, when a threshold value is simply determined for the degree of change, notifications are issued at three points when the degree of change on the left side of the graph rises rapidly, but these are false notifications due to noise. Therefore, this is an inappropriate notification. On the other hand, as shown in the lower part of Figure 6, when determining whether or not an upward trend is indicated, the trend does not indicate an upward trend in the period within the rectangle on the left that includes the three points at which the rate of change rises rapidly. Therefore, false alarms like the one described above do not occur. When it is determined whether or not an upward trend is indicated, it can be seen that the trend indicates an upward trend after entering the period within the rectangle on the right side. In this manner, in this embodiment, it is possible to calculate trends as information for performing inspections at appropriate timing.

Returning to FIG. 5, if the server processing unit 120 determines that there is no upward trend (no upward trend is detected) (No in S208), the process returns to step S201 and starts measurement for the next period. On the other hand, if the server processing unit 120 determines that an upward trend is indicated (an upward trend is detected) (Yes in S208), the process proceeds to step S208, and the server processing unit 120 generates notification information. Note that the threshold value for determining whether the trend is an upward trend may be set empirically or experimentally.

The generated broadcast information is transmitted to the terminal device 300 by the server communication unit 140 (S2010). Upon receiving the broadcast information (S211), the terminal communication unit 340 outputs it to the screen generation unit 310. The screen generation unit 310 generates image information indicating the content of the notification information, and causes the terminal display unit 320 to display the image information (S212).

FIG. 7 is a diagram for explaining notification information output in the trend calculation system according to the present embodiment. As shown in FIG. 7, on the terminal device 300, an image showing the content of the broadcast information is displayed. Specifically, the displayed image shows the calculated trend (broken line curve and solid line tangent to the curve) when an upward trend is shown, the degree of change used to calculate the trend, etc. Information indicating the period during which plots and trend calculations were performed (rectangular frame), information identifying the drive source 99 to be measured (pre-registered name, etc.), and information indicating that the operation of the drive source 99 is abnormal. Messages such as the following are displayed.

As described above, from the current value of the power supplied to the drive source 99, a trend indicating the tendency of change over time in the degree of change is calculated, and depending on whether or not the trend indicates an upward trend, the current value of the drive source 99 is determined. This makes it possible to appropriately estimate signs of trouble.

Although the trend calculation device and the like according to one or more aspects have been described above based on the embodiment, the present disclosure is not limited to this embodiment. As long as it does not depart from the spirit of the present disclosure, various modifications that can be thought of by those skilled in the art to the above embodiments, and forms constructed by combining components of different embodiments are also within the scope of one or more embodiments. may be included within.

For example, in the above embodiments, each component is configured with dedicated hardware, but it may be realized by executing a software program suitable for each component. Each component may be realized by a program execution unit such as a CPU or a processor reading and executing a software program recorded on a recording medium such as a hard disk or a semiconductor memory. Here, the software that implements the information device of the above embodiment is the following program.

That is, this program is a program that causes a computer to execute a trend calculation method.

The present disclosure can be used when determining the inspection timing of the drive source.

98 Current sensor 99 Drive source 99a Motor 99b Rotation conversion mechanism 99c Power supply source 99d Power line 100 Server device 110 Server control section 120 Server processing section 130 Server storage section 140 Server communication section 150 Network 200 Edge device 200a Master device 200b Child device 210 Control Section 220 Communication section 230 Storage section 240 Measurement section 300 Terminal device 310 Screen generation section 320 Terminal display section 330 Terminal control section 340 Terminal communication section

Claims (8)

an acquisition unit that acquires a measurement result of measuring a current value of the power supplied to a drive source driven by the supplied power;
Based on the measurement results acquired between a first time point in the past and a second time point after the first time point, an input based on a current value indicated in each of the acquired measurement results and the drive source are input. from the first point in time to the second point in time when the degree of change corresponding to the magnitude of the difference from the reference based on the current value of the electric power measured in advance during normal driving increases or decreases over time. A calculation unit that calculates and outputs a trend indicating a trend of change over time. The trend calculation device according to claim 1, wherein the calculation unit includes a trained trend calculation model that outputs the trend in response to input of the measurement results obtained between the first time point and the second time point. . The calculation unit is
It has a trained degree of change calculation model that outputs the degree of change in accordance with the input of the acquired measurement results,
Based on an approximation formula that approximates, in the time domain, each of the degrees of change output when each of the measurement results obtained between the first time point and the second time point is input into the degree of change calculation model. , the trend calculation device according to claim 1, wherein the trend is calculated. The acquisition unit acquires a current value measured by a current sensor as the measurement result,
The trend calculation device according to claim 1, wherein the current sensor has a measurement characteristic in which the higher the frequency, the higher the sensitivity in the frequency domain of the current value. If the calculated trend indicates an upward trend in which a tendency of change in the degree of change between the first time point and the second time point is larger than a predetermined threshold value, the operation of the drive source is diagnosed as abnormal. The trend calculation device according to any one of claims 1 to 4, further comprising a diagnosis section that outputs a diagnosis result. The diagnosis result is
the calculated trend when the upward trend is shown;
The trend calculation device according to claim 5, further comprising at least one of the following: a message indicating that the operation of the drive source to be measured is abnormal. The trend calculation device according to claim 1;
a current sensor used to measure the current value that the acquisition unit acquires as the measurement result,
The current sensor measures a current value of power supplied from outside the power line connecting the drive source and a power supply source that supplies power to the drive source via the power line. Obtaining a measurement result of measuring the current value of the power supplied to a drive source driven by the supplied power,
Based on the measurement results obtained between a first time point in the past and a second time point after the first time point, the current value indicated in each of the obtained measurement results and the drive source are normally driven. calculate a trend indicating a change tendency between the first time point and the second time point when the degree of change corresponding to the magnitude of the difference between the electric power and the current value when the electric power is running increases or decreases over time. Trend calculation method.

Images