Abnormality diagnosis device and abnormality diagnosis method for passenger conveyor

Abstract

To provide an abnormality diagnostic device and an abnormality diagnostic method of a passenger conveyor capable of properly specifying an abnormal part.SOLUTION: An abnormality diagnosis device for a passenger conveyor includes an operation sound recording part, an abnormality determination part, and an abnormal part determination part. The operation sound recording part records the operation sound of the passenger conveyor device in a storage part of the mobile terminal through a sound collecting part of the mobile terminal or an external sound collecting part. The abnormality determination part compares the recorded operation sound with the reference sound stored in the storage part, determines the presence / absence of abnormality of the passenger conveyor device based on a change amount of the recorded operation sound to the reference sound, and displays a determination result on a display part of the mobile terminal. The abnormal part determination part determines an abnormal part corresponding to the recorded operation sound by comparing the recorded operation sound when the presence of the abnormality is determined with an abnormality model stored in a storage part and learning the abnormal part of the passenger conveyor device corresponding to the operation sound of the passenger conveyor device, and displays a determination result on a display part.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] An embodiment of the present invention relates to an abnormality diagnosis device and an abnormality diagnosis method for a passenger conveyor. [Background technology]

[0002] An escalator, a type of passenger conveyor device, is a device in which steps connected endlessly between lower and upper landings move in a circular motion between each floor. Escalators drive the steps by transmitting the torque of a drive motor to the steps via a chain wound around a drive sprocket on the upper floor and a driven sprocket on the lower floor. Escalators also drive the handrails by transmitting the torque of a drive motor to the handrails via a handrail drive device.

[0003] In escalators with this type of configuration, the drive parts of the steps and handrails are subjected to a load required to drive the steps and handrails. This causes deterioration of the drive parts due to usage and aging, resulting in the generation of abnormal noises. Typical examples of abnormal noises resulting from deterioration of the drive parts include rolling noises caused by deterioration of bearings installed on the rotating shafts of the drive motor, drive sprocket, or driven sprocket, and meshing noises between the chain and sprocket due to stretching or sagging of the chain that transmits the rotational force of the drive motor. Other typical abnormal noises include the rubbing noise of the handrail in the handrail drive mechanism caused by changes in handrail tension or wear, and the rubbing noise around the turn-backs of the handrails at the ends of the handrails on the upper and lower floors.

[0004] These abnormal noises often occur periodically in conjunction with the operation of the escalator, causing discomfort to users. Not only that, but if left unchecked, in the worst case scenario, it could lead to damage to the equipment. Therefore, it is important to detect signs of escalator abnormalities at the abnormal noise stage and take action early. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent No. 6748044 Summary of the Invention [Problem to be solved by the invention]

[0006] In order to improve the efficiency of escalator maintenance and inspection, it is desirable to not only identify whether or not there is an abnormality in the escalator, but also to properly identify the abnormal part and prompt maintenance personnel to take action.

[0007] One method for identifying an abnormal part in an escalator is to identify the abnormal part based on data accumulated through accelerated testing using testing equipment. In accelerated testing, the testing equipment is used to reproduce failure modes for each abnormal part (i.e., for each piece of equipment), and data on abnormal noises generated for each failure mode is accumulated. Then, when the escalator is actually operated, the accumulated abnormal noise data is compared with abnormal noises generated by actual failure modes in the escalator to identify the abnormal part in the escalator. However, because accelerated testing is a reproduction test conducted under assumed conditions, it is not necessarily possible to reproduce the same failure mode that occurs in actual equipment. Furthermore, it is possible that multiple failure modes exist in a single piece of equipment. Therefore, it is difficult to appropriately identify the abnormal part based on accelerated testing.

[0008] Another method is to identify the abnormal part of an escalator based on whether or not there is a change in the operating sounds, including abnormal sounds, when the operating direction or speed of the escalator is changed. This method makes it possible to identify the abnormal part for abnormal events that are known in advance. However, it does not make it possible to identify the abnormal part for unexpected abnormal events.

[0009] The embodiments aim to provide an abnormality diagnosis device and an abnormality diagnosis method for a passenger conveyor that can appropriately identify an abnormal portion. [Means for solving the problem]

[0010] An abnormality diagnosis device for a passenger conveyor according to an embodiment is provided in a mobile terminal and diagnoses an abnormality in a passenger conveyor that circulates a plurality of endlessly connected steps. The abnormality diagnosis device includes an operation sound recording unit, an abnormality determination unit, and an abnormality part determination unit. The operation sound recording unit records operation sounds of the passenger conveyor in a memory unit of the mobile terminal via a sound collector of the mobile terminal or an external sound collector of the mobile terminal. The abnormality determination unit compares the operation sounds recorded by the operation sound recording unit with reference sounds stored in the memory unit, determines whether or not there is an abnormality in the passenger conveyor based on the amount of change in the recorded operation sounds relative to the reference sounds, and displays the determination result on a display unit of the mobile terminal. The abnormality part determination unit compares the recorded operation sounds when the abnormality determination unit determines there is an abnormality with an abnormality model stored in the memory unit that has learned abnormal parts of the passenger conveyor corresponding to the operation sounds of the passenger conveyor, and determines the abnormal part corresponding to the recorded operation sounds, and displays the determination result on the display unit. [Brief explanation of the drawings]

[0011] [Figure 1] FIG. 1 is a side view of a passenger conveyor showing how an abnormality diagnostic device according to this embodiment is used. [Figure 2] FIG. 2 is an enlarged view of part A in FIG. [Figure 3] FIG. 3 is a block diagram showing an abnormality diagnosis device according to this embodiment. [Figure 4] FIG. 4 is a flowchart showing an example of the operation of the abnormality diagnostic device according to this embodiment. [Figure 5] FIG. 5 is a diagram showing an example of a process for determining the amount of change in an example of the operation of the abnormality diagnostic device according to this embodiment. [Figure 6] FIG. 6 is a diagram showing another example of the step of determining the amount of change in the example of operation of the abnormality diagnostic device according to this embodiment. [Figure 7] FIG. 7 is a diagram showing a process of comparing a recorded sound with an abnormality model in an example of operation of the abnormality diagnosis device according to this embodiment. [Figure 8]FIG. 8 is a flowchart showing an example of the operation of the abnormality diagnostic device according to this embodiment, following FIG. [Figure 9] FIG. 9 is a diagram showing a sending process to a data center in an example of the operation of the abnormality diagnostic device according to this embodiment. [Figure 10] FIG. 10 is a flowchart showing an example of the operation of the abnormality diagnostic device according to this embodiment, following FIG. [Figure 11] FIG. 11 is a flowchart showing an example of the operation of the abnormality diagnostic device according to this embodiment, following FIG. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. The present invention is not limited to this embodiment. In addition, in the drawings referred to in this embodiment, identical parts or parts having similar functions are designated by the same or similar reference numerals, and repeated description thereof will be omitted. Below, abnormality diagnosis of an escalator, which is an example of a passenger conveyor system, will be described, but this embodiment can also be applied to passenger conveyor systems other than escalators, such as moving walkways.

[0013] The abnormality diagnostic device according to this embodiment can be used to diagnose abnormalities in the escalator 1 shown in Fig. 1. First, before describing the configuration of the abnormality diagnostic device, the configuration of the escalator 1 will be described.

[0014] As shown in Figures 1 and 2, escalator 1 is a device that moves between floors by circulating steps 7 that are endlessly connected between lower-floor landings 4 and upper-floor landings 5. Passengers can move to their destination floors by stepping on treads 8 of the steps 7. Note that in Figure 1, part of the side cover panel 3 is omitted to illustrate the internal structure of the side cover panel 3. Escalator 1 is equipped with a drive motor 16 on the upper floor side within truss 6. Drive motor 16 transmits its rotation to a motor-side sprocket 17 via a reducer or the like. A drive chain 19 is wound around motor-side sprocket 17. The drive chain 19 is also wound around drive sprocket 14. When motor-side sprocket 17 rotates, the rotation is transmitted to drive sprocket 14 via drive chain 19, causing drive sprocket 14 to rotate. A roller chain 13 is wound around drive sprocket 14. A front wheel roller 9 of the step 7 is attached to roller chain 13. When the drive sprocket 14 rotates, the steps 7 move together with the roller chain 13. A driven sprocket 15 is installed on the lower floor side. The roller chain 13 circulates by folding back between the drive sprocket 14 and the driven sprocket 15. In addition, the rollers of the handrail drive device 18 are attached in a pressed state to the handrail 2. The rotation from the drive motor 16 is transmitted to the handrail drive sprocket 21. The rotation of the handrail drive sprocket 21 is transmitted to the handrail drive device 18 via the handrail drive chain 20. The rotation of the handrail drive sprocket 21 is transmitted to the handrail drive device 18, causing the rollers of the handrail drive device 18 to rotate. The rotation of the rollers of the handrail drive device 18 drives the handrail 2. The drive sprocket 14 and handrail drive sprocket 21 are installed on the same axis. This allows the handrail 2 to move at the same speed as the steps 7.

[0015] The front wheel rollers 9 of the steps 7 are connected to a roller chain 13. The front wheel rollers 9 of the steps 7 are connected to every third roller of the roller chain 13. A front wheel rail 11 is laid on the roller chain 13. The rollers move while rolling on the front wheel rail 11, with the vertical load of the steps 7 being supported by the front wheel rail 11. In addition, a rear wheel rail 12 is laid on the rear wheel rollers 10 of the steps 7. The rollers move while rolling on the rear wheel rail 12, with the vertical load of the steps 7 being supported by the rear wheel rail 12. The front wheel rollers 9 and rear wheel rollers 10 are installed on both the left and right sides of the traveling direction.

[0016] Next, the configuration of the abnormality diagnosis device will be described. As shown in Fig. 1, the abnormality diagnosis device is provided in a mobile terminal 30 that can be carried by a maintenance worker M. The mobile terminal 30 can also be called a portable terminal.

[0017] As shown in FIG. 3 , the mobile terminal 30 includes a microphone 31 constituting a sound collection unit, an input unit 32, a display unit 33, a storage unit 34, and an abnormality diagnosis device 35. The mobile terminal 30 is communicably connected to a data center 40. The abnormality diagnosis device 35 includes an operation sound recording unit 351, an abnormality determination unit 352, an abnormal part determination unit 353, a transmission processing unit 354, a reception processing unit 355, and a threshold update unit 356. The abnormality diagnosis device 35 is configured, for example, by a processor. In this case, the processor functions as the operation sound recording unit 351, the abnormality determination unit 352, the abnormal part determination unit 353, the transmission processing unit 354, the reception processing unit 355, and the threshold update unit 356 by reading and executing a program stored in the storage unit 34. The processor may be, for example, a central processing unit (CPU), a graphics processing unit (GPU), or an application-specific integrated circuit (ASIC). The storage unit 34 is configured by, for example, semiconductor memory elements such as a read-only memory (ROM), a random access memory (RAM), and a flash memory. The abnormality diagnosis device 35 may include hardware other than a processor. Furthermore, at least a part of the abnormality diagnosis device 35 may be configured by software.

[0018] The microphone 31 is a microphone provided in the mobile terminal 30. The microphone 31 may also be an external microphone. Basically, the microphone 31 provided in the mobile terminal 30 can sufficiently collect abnormal sounds, but collecting abnormal sounds in a higher frequency range makes it easier to detect signs of an abnormality. Therefore, an external microphone with higher accuracy than the microphone 31 provided in the mobile terminal 30 may be used as needed.

[0019] The operation sound recording unit 351 records the operation sound of the escalator 1 in the storage unit 34 via the microphone 31. The operation sound recording unit 351 records the operation sound at the landings 4 and 5 (i.e., boarding and alighting boards) on the upper and lower floors of the escalator 1 until the steps 7 make one revolution. Specifically, as shown in FIG. 1 , the operation sound recording unit 351 records the operation sound until the steps 7 make one revolution when the maintenance worker M is standing at the landings 4 and 5 on the upper and lower floors while carrying the mobile terminal 30. The operation sound recording unit 351 also records the operation sound while the maintenance worker M is riding the escalator 1 while carrying the mobile terminal 30 until he arrives at the destination floor.

[0020] The internal equipment of the escalator 1 is housed in a machine room near the feet of the maintenance worker. For this reason, it is preferable that the operation sound recording unit 351 record the operation sound near the feet of the maintenance worker M. Furthermore, from the perspective of appropriately updating the abnormality model described below using the recorded operation sound, and from the perspective of appropriately diagnosing abnormalities in the escalator 1 using the abnormality model, the recording position of the operation sound should be reproducible. For this reason, it is preferable to determine in advance the recording position of the operation sound, for example, at the position of the tip of the handrail 2 in the direction of travel of the handrail 2, or at a position at the height of the tip of the handrail 2 in the height direction.

[0021] Here, near the landing 5 on the upper floor of the escalator 1, devices such as drive parts such as the drive motor 16, handrail drive device 18, handrail 2, and steps 7 are arranged. If there is an abnormality in these devices, abnormal sounds indicating an abnormality will begin to be generated, such as rolling sounds due to deterioration of bearings installed in the drive motor 16 and drive sprocket 14, meshing sounds in the motor-side sprocket 17 and drive sprocket 14 due to stretching or sagging of the drive chain 19, rubbing sounds between the rollers of the handrail drive device 18 and the handrail, rubbing sounds from near the turn-back at the tip of the handrail 2, and rolling sounds due to deterioration of the bearings of the front wheel rollers 9 and rear wheel rollers 10 of the steps 7. Therefore, by recording operating sounds at the landing 5 on the upper floor, the operating sound recording unit 351 can record operating sounds that are useful for determining whether or not there is an abnormality in the equipment on the upper floor.

[0022] Furthermore, at landing 4 on the lower floor of escalator 1, abnormal sounds that indicate a malfunction begin to occur, mainly from rolling sounds caused by deterioration of the bearings of driven sprocket 15 and rolling sounds caused by deterioration of the bearings of front wheel rollers 9 and rear wheel rollers 10 of step 7. Therefore, by recording operating sounds at landing 4 on the lower floor, operation sound recording unit 351 can record operating sounds that are useful for determining whether or not there is a malfunction in the equipment on the lower floor.

[0023] Furthermore, there is a possibility that scraping noises may occur when the step 7 interferes with the side wall. Therefore, the operation sound recording unit 351 can record operation sounds that are useful for determining whether or not there is an abnormality in the equipment between the lower and upper floors by having the maintenance worker M step on the step 7 while carrying the mobile terminal 30 and recording the operation sounds from the time he or she arrives at the upper floor, for example.

[0024] Reference sounds to be compared with the operating sounds are stored in the storage unit 34. The reference sounds include a first reference sound for determining whether or not there is an abnormality due to aging, and a second reference sound for determining whether or not there is a sudden abnormality.

[0025] The first reference sound is, for example, an operation sound recorded by operation sound recording unit 351 when escalator 1 is first installed (i.e., when it is first set up). The first reference sound may be stored in storage unit 34 with a label indicating normality attached (i.e., associated with the label). The label is an example of additional information.

[0026] The second reference sound is, for example, the operation sound previously recorded by the operation sound recording unit 351. In other words, the second reference sound is the most recently recorded sound (i.e., the most recently recorded sound) going back in time from the time when the operation sound to be judged was recorded (i.e., the current recording). The second reference sound may be stored in the memory unit 34 with a label indicating normality attached. The second reference sound may be an operation sound recorded by the operation sound recording unit 351 after an action to resolve an abnormality, such as part replacement or repair, has been taken on the escalator 1.

[0027] The storage unit 34 also stores an abnormality model that learns abnormal parts of the escalator 1 that correspond to the operating sounds of the escalator 1. The abnormality model is a model generated by accumulating data correlating operating sounds, including abnormal sounds, recorded on the escalator 1 with the abnormal parts at the time the operating sounds were recorded, and classifying the accumulated data by machine learning based on the characteristics of the operating sounds. The characteristics of the operating sounds are, for example, the frequency characteristics of the operating sounds. The abnormality model is generated by the data center 40 (i.e., a server in the data center 40). The mobile terminal 30 can download the abnormality model from the data center 40 in accordance with an input operation by the maintenance worker M, for example, before the maintenance worker M goes for maintenance inspection. The abnormality model downloaded from the data center 40 is stored in the storage unit 34. The abnormality model is a model that can be updated according to the progress of accumulation of data correlating recorded operating sounds with abnormal parts and the accompanying progress of machine learning.

[0028] The abnormality determination unit 352 compares the operation sound recorded by the operation sound recording unit 351 with the reference sound stored in the storage unit 34 and determines whether or not there is an abnormality in the passenger conveyor apparatus 1 based on the amount of change in the recorded operation sound relative to the reference sound. Specifically, the abnormality determination unit 352 compares the recorded operation sound with the first reference sound stored in the storage unit 34 and determines whether or not there is an abnormality due to aging based on the amount of change in the recorded operation sound relative to the first reference sound. The abnormality determination unit 352 also compares the recorded operation sound with the second reference sound stored in the storage unit 34 and determines whether or not there is a sudden abnormality based on the amount of change in the recorded operation sound relative to the second reference sound. The abnormality determination unit 352 performs determinations using the first reference sound and the second reference sound in parallel. By using the first reference sound, it is possible to appropriately detect abnormalities in the escalator 1 that gradually progress due to aging during operation. By using the second reference sound, any abnormality that suddenly occurs in escalator 1 can be appropriately detected.

[0029] The abnormality determination unit 352 determines that an abnormality exists when the amount of change in the recorded operation sound relative to the reference sound exceeds a predetermined threshold. Specifically, the abnormality determination unit 352 determines that an abnormality exists in at least one of the following cases: when the amount of change in the recorded operation sound relative to the first reference sound exceeds a predetermined first threshold; and when the amount of change in the recorded operation sound relative to the second reference sound exceeds a predetermined second threshold. The first threshold and the second threshold may be different values. The abnormality determination unit 352 displays the determination result of whether or not an abnormality exists in the passenger conveyor apparatus 1 on the display unit 33. The abnormality determination unit 352 may display the result of the determination using the first reference sound and the result of the determination using the second reference sound on the display unit 33 in a distinguishable manner.

[0030] If it is determined that there is no abnormality, abnormality determination unit 352 assigns a label indicating normality to the recorded operating sound. That is, abnormality determination unit 352 stores the label indicating normality in association with the recorded operating sound in storage unit 34. The operating sound that is determined to be free of abnormality and assigned a label indicating normality when escalator 1 is first installed may be used as the first reference sound in subsequent determinations.

[0031] Even if the abnormality determination unit 352 determines that an abnormality exists, the maintenance worker M may not find any abnormality after investigating the abnormality. In such cases, it is desirable to correct the abnormality determination result after the fact so that the abnormality can be utilized for the next determination. To correct the abnormality determination result after the fact, the abnormality determination unit 352 assigns a label indicating normality to the recorded operation sound in accordance with an input operation to the mobile terminal 30 via the input unit 32, even if an abnormality is determined to exist. That is, the abnormality determination unit 352 stores the label indicating normality in the storage unit 34 in association with the recorded operation sound. The operation sound labeled as normal can be used as a second reference sound for the next determination. The input unit 32 is, for example, an interface that accepts input operations such as a GUI (Graphical User Interface) displayed on the display unit 33 or a mechanical switch. The input unit 32 may include a microphone 31 that accepts voice input.

[0032] The abnormal part determination unit 353 compares the recorded operation sound when the abnormality determination unit 352 determines that there is an abnormality with the abnormality model stored in the memory unit 34, and determines the abnormal part corresponding to the recorded operation sound. The abnormal part determination unit 353 displays the abnormal part determination result on the display unit 33. As described above, the abnormality model is an updatable model. The abnormal part determination unit 353 can increase the number of identifiable abnormal parts as the abnormality model is updated. Specific aspects of updating the abnormality model will be described below.

[0033] Even if the abnormal part determination unit 353 cannot determine an abnormal part because there is no abnormality model similar to the recorded operation sound, the maintenance person M may be able to identify the source of the abnormal sound, i.e., the abnormal part, by conducting an investigation. In this case, the abnormal part identified by the maintenance person M can be used to update the abnormal model. Specifically, when the abnormal part determination unit 353 cannot determine an abnormal part corresponding to the recorded operation sound, the abnormal part determination unit 353 assigns a label indicating the abnormal part identified by the investigation by the maintenance person M to the recorded operation sound in accordance with an input operation on the mobile terminal 30 via the input unit 32. That is, the abnormal part determination unit 353 stores the label indicating the abnormal part in the storage unit 34 in association with the recorded operation sound. As will be described later, the operation sound to which the label indicating the abnormal part has been assigned (i.e., data in which the recorded operation sound is associated with the abnormal part) is transmitted to the data center 40 and used to update the abnormal model. Then, by receiving the updated abnormal model from the data center 40, the updated abnormal model can be used in subsequent abnormal part determinations. This increases the number of identifiable abnormal parts, making it possible to identify abnormal parts even for abnormalities that were not anticipated when the escalator 1 was first installed.

[0034] Furthermore, even when the abnormality determination unit 352 determines that there is no abnormality, there are cases in which the maintenance person M confirms the abnormality and identifies the abnormal part as a result of his / her investigation. In this case, too, the abnormal part identified by the maintenance person M can be used to update the abnormality model. Specifically, even when it is determined that there is no abnormality, the abnormal part determination unit 353 assigns a label indicating the abnormal part identified by the investigation by the maintenance person M to the recorded operating sound in accordance with an input operation on the mobile terminal 30 via the input unit 32. The operating sound to which the label indicating the abnormal part has been assigned is transmitted to the data center 40 and used to update the abnormality model. This makes it possible to further increase the number of identifiable abnormal parts.

[0035] Even if the abnormal part determination unit 353 is able to determine an abnormal part, the maintenance technician M may find, as a result of his / her investigation, that the source of the abnormal sound is different from the abnormal part indicated in the determination result. In this case, too, the abnormal part identified by the maintenance technician M can be used to update the abnormality model. Specifically, even if the abnormal part determination unit 353 is able to determine an abnormal part corresponding to the recorded operation sound, the abnormal part determination unit 353 assigns a label to the recorded operation sound indicating an abnormal part different from the determined abnormal part identified by the maintenance technician M's investigation, in accordance with an input operation on the mobile terminal 30 via the input unit 32. That is, the abnormal part determination unit 353 stores the label indicating the abnormal part different from the determined abnormal part in the storage unit 34 in association with the recorded operation sound. The operation sound to which the label indicating the abnormal part different from the determined abnormal part is assigned is transmitted to the data center 40 and used to update the abnormality model. This improves the accuracy of determining an abnormal part.

[0036] When a label with content different from the determination result by the abnormality determination unit 352 is stored in the storage unit 34, the threshold update unit 356 updates the threshold based on the label with content different from the determination result. For example, when an operation sound determined to have an abnormality by the abnormality determination unit 352 is assigned a label indicating normality and stored in the storage unit 34, the threshold update unit 356 may change (e.g., increase) at least one of the first threshold and the second threshold so that an abnormality determination is less likely to be made from the next time. Furthermore, when an operation sound determined to have no abnormality by the abnormality determination unit 352 is assigned a label indicating an abnormal part and stored in the storage unit 34, the threshold update unit 356 may change (e.g., decrease) at least one of the first threshold and the second threshold so that an abnormality determination is more likely to be made from the next time.

[0037] The transmission processing unit 354 and the reception processing unit 355 are communicably connected to the data center 40 via, for example, a wireless communication module (not shown) provided in the mobile terminal 30. The transmission processing unit 354 transmits to the data center 40 operation sounds labeled with an abnormal part in accordance with an input operation via the input unit 32 described above. The data center 40 learns the abnormal part of the escalator 1 corresponding to the operation sound of the escalator 1, based on the operation sounds labeled with an abnormal part and transmitted from the transmission processing unit 354. That is, the server of the data center 40 updates the abnormality model stored therein based on the operation sounds labeled with an abnormal part. The update of the abnormality model is, for example, at least one of adding, modifying, and deleting the abnormality model. The reception processing unit 355 receives the updated abnormality model from the data center 40 and stores it in the storage unit 34.

[0038] Next, we will explain an example of the operation of the above-mentioned abnormality diagnosis device 35. As shown in Fig. 4, first, the operation sound recording unit 351 records the operation sound until the steps 7 of the escalator 1 make one revolution (step S1).

[0039] After the operation sounds recorded until the steps 7 make one revolution, the abnormality determination unit 352 compares the operation sounds recorded by the operation sound recording unit 351 with the reference sound stored in the storage unit 34 and determines whether or not there is an abnormality in the escalator 1 based on the amount of change in the recorded operation sound relative to the reference sound. Hereinafter, the operation sounds recorded by the operation sound recording unit 351 will also be referred to as "recorded sounds." Specifically, the abnormality determination unit 352 determines whether or not there is an abnormality due to aging by determining whether or not there is an abnormality in the recorded sound relative to the reference sound (step S2). More specifically, the abnormality determination unit 352 determines whether or not there is an abnormality due to aging by determining whether or not there is an abnormality in the recorded sound relative to the first reference sound relative to the first reference sound relative to the first threshold. At this time, the abnormality determination unit 352 further determines whether or not there is an abnormality by determining whether or not there is an abnormality in the recorded sound relative to the second reference sound relative to the second threshold.

[0040] For example, as shown in FIG. 5, the anomaly determination unit 352 may use an autoencoder, which is one type of machine learning, to determine whether the amount of change in the recorded sound relative to a reference sound exceeds a threshold. An autoencoder is an algorithm that compresses input data and then restores it to output data. The anomaly determination unit 352 makes the determination using a trained autoencoder model. The trained autoencoder model is a neural network model of an autoencoder created by inputting a reference sound as input data to the autoencoder and training the autoencoder so that the restored sound as output data is the same as the reference sound. The trained autoencoder model is stored in the storage unit 34. The anomaly determination unit 352 reads the trained autoencoder model stored in the storage unit 34 and uses it for the determination. As shown in FIG. 5(a), when a reference sound (i.e., the first reference sound or the second reference sound) is input to the trained autoencoder model, the restored sound output from the trained autoencoder model has a small error (i.e., a small degree of anomaly) from the reference sound. 5(b), when a recorded sound containing an abnormal sound is input to the trained autoencoder model, the restored sound output from the trained autoencoder model has a large error from the recorded sound. The abnormality determination unit 352 uses the error between the recorded sound and the restored sound as the amount of change in the recorded sound relative to the reference sound, and determines whether the error exceeds a threshold (i.e., the first threshold or the second threshold).

[0041] Furthermore, for example, as shown in FIG. 6 , the abnormality determination unit 352 may determine whether the amount of change in the recorded sound relative to the reference sound exceeds a threshold (i.e., the first threshold or the second threshold) by comparing the difference between the reference sound (i.e., the first reference sound or the second reference sound) and the recorded sound in terms of frequency components. In the example shown in FIG. 6 , the abnormality determination unit 352 determines whether the square root of the sum of the squares of the differences between the reference sound and the recorded sound exceeds a threshold in each frequency band frequency-analyzed using 1 / 3 octave analysis. That is, the abnormality determination unit 352 makes the determination using the square root of the sum of the squares of the differences between the reference sound and the recorded sound as the amount of change in the recorded sound relative to the reference sound. Alternatively, the abnormality determination unit 352 determines whether the maximum value of the difference between the reference sound and the recorded sound in each frequency band exceeds a threshold. That is, the abnormality determination unit 352 makes the determination using the maximum value of the difference between the reference sound and the recorded sound as the amount of change in the recorded sound relative to the reference sound.

[0042] If the amount of change in the recorded sound relative to the reference sound exceeds a preset threshold (step S2: Yes), the abnormality determination unit 352 determines that an abnormality exists (step S3), as shown in Fig. 4. Specifically, the abnormality determination unit 352 determines that an abnormality exists in at least one of the following cases: if the amount of change in the recorded operation sound relative to the first reference sound exceeds a first threshold; or if the amount of change in the recorded operation sound relative to the second reference sound exceeds a second threshold.

[0043] On the other hand, if the amount of change in the recorded sound relative to the reference sound does not exceed the preset threshold (step S2: No), the abnormality determination unit 352 determines that there is no abnormality (step S6). Specifically, the abnormality determination unit 352 determines that there is no abnormality in at least one of the following cases: if the amount of change in the recorded operation sound relative to the first reference sound does not exceed the first threshold, or if the amount of change in the recorded operation sound relative to the second reference sound does not exceed the second threshold.

[0044] After determining that there is an abnormality, the abnormal part determination unit 353 determines whether an abnormality model similar to the recorded sound is stored in the storage unit 34 (step S4). In the example shown in FIG. 7, the abnormal part determination unit 353 compares the recorded sound with abnormality models for each abnormal part (motor bearing, handrail drive device, driven bearing, etc.) stored in the storage unit 34. The abnormality models in the storage unit 34 are models in which the characteristics of operation sounds are classified for each abnormal part according to the results of learning the characteristics of the operation sounds for each abnormal part through machine learning. In the abnormality model, the horizontal axis represents frequency and the vertical axis represents sound pressure level. The abnormality model is, for example, a trained model that has been trained in advance by a server in the data center 40, and then periodically downloaded from the data center 40 via the reception processing unit 355 and stored in the storage unit 34. The abnormal part determination unit 353 determines whether there is an abnormality model similar to the recorded sound, for example, based on whether an abnormality model having characteristics common to the recorded sound is stored in the storage unit 34. In the example shown in FIG. 7, abnormal sound group 2), which is one of the abnormal models, has common characteristics (i.e., frequency distribution) with the recorded sound, so the abnormal part determination unit 353 determines that there is an abnormal model similar to the recorded sound.

[0045] If there is an abnormal model similar to the recorded sound (step S4: Yes), as shown in Figure 4, the abnormal part determination unit 353 displays an inspection instruction and the abnormal part corresponding to the similar abnormal model on the display unit 33 (step S5).

[0046] On the other hand, if there is no abnormal model similar to the recorded sound (step S4: No), the abnormal part determination unit 353 displays an inspection instruction and an instruction to investigate the sound source on the display unit 33 (step S8).

[0047] After determining that there is no abnormality (step S6), the abnormality determination unit 352 assigns a label indicating normality to the recorded sound and stores the recorded sound in the storage unit 34 (step S7). The recorded sound that has been assigned a label indicating normality may be used as a reference sound in the next determination by the abnormality determination unit 352.

[0048] After the inspection instruction and the abnormal part are displayed on the display unit 33 (step S5 in FIG. 4), the maintenance worker M investigates the displayed abnormal part (step S11), as shown in FIG.

[0049] If an abnormality is confirmed as a result of the investigation (step S12: Yes) and the displayed abnormal part is correct (step S13: Yes), the abnormal part determination unit 353 assigns a label indicating the abnormal part to the recorded sound in accordance with an input operation to the mobile terminal 30 via the input unit 32, and stores the label in the storage unit 34 (step S14). Note that the label is set to normal by default, and may be changed to indicate the abnormal part in accordance with an input operation.

[0050] On the other hand, if no abnormality is found as a result of the investigation (step S12: No), the abnormality determination unit 352 assigns a label indicating normality to the recorded sound in accordance with the input operation to the mobile terminal 30 via the input unit 32, and stores the recorded sound in the storage unit 34 (step S17). Note that if the label indicating normality is set as a default, the input operation for the label indicating normality can be simplified.

[0051] If the displayed abnormal part is incorrect (step S13: No), the abnormal part determination unit 353 assigns a label indicating an abnormal part different from the determined abnormal part to the recorded sound in accordance with an input operation to the mobile terminal 30 via the input unit 32, and stores the label in the storage unit 34 (step S18). The abnormal part different from the determined abnormal part is an abnormal part that has become apparent through an investigation by the maintenance worker M.

[0052] After the recorded sound is assigned a label indicating an abnormal part and saved, maintenance person M takes measures such as repairing the abnormal part or replacing parts so that the recorded sound no longer contains abnormal sounds. After taking measures, maintenance person M assigns a label indicating normality to the recorded sound recorded by the operation sound recording unit 351 and saves it in the storage unit 34 (step S15). The recorded sound that has been assigned a label indicating normality may be used as a second reference sound in the next determination by the abnormality determination unit 352.

[0053] After the treatment, the recorded sound is assigned a label indicating normality and stored, and then the transmission processing unit 354 transmits the recorded sound assigned a label indicating an abnormal portion (S14, S18) or the recorded sound assigned a label indicating normality (S15, S17) to the data center 40 (step S16). The transmission to the data center 40 may be performed in accordance with an input operation on the mobile terminal 30 at any timing, such as when the maintenance worker M returns to the office after completing the maintenance inspection.

[0054] As shown in FIG. 9, the data center 40 (i.e., the server) updates the abnormality model stored in the data center 40 (i.e., the database) using the recorded sound with the label attached and transmitted from the mobile terminal 30. For example, the data center 40 adds the recorded sound with the label indicating the abnormal part as a new abnormality model. By updating the abnormality model in the data center 40, the abnormality model downloaded from the data center 40 to the mobile terminal 30 can be updated. This makes it possible to identify an abnormal part that was not anticipated when the escalator 1 was first installed, using the latest abnormality model that reflects the actual abnormal state.

[0055] After the inspection instruction and the instruction to investigate the sound source are displayed on the display unit 33 (step S8 in FIG. 4), the maintenance worker M investigates the source of the abnormal sound (step S21) and identifies the abnormal part (step S22), as shown in FIG. 10.

[0056] After the abnormal portion is identified, the abnormality determination unit 352 assigns a label indicating the abnormal portion to the recorded sound in accordance with an input operation to the mobile terminal 30 via the input unit 32, and stores the recorded sound in the storage unit 34 (step S23).

[0057] After the recorded sound is assigned a label indicating an abnormal part and saved, maintenance person M takes measures such as repairing the abnormal part or replacing parts so that the recorded sound no longer contains abnormal sounds. After taking measures, maintenance person M assigns a label indicating normality to the recorded sound recorded by operation sound recording unit 351 and saves it in storage unit 34 (step S24). The saved recorded sound may be used as a second reference sound in the next determination by abnormality determination unit 352.

[0058] After the recorded sound after treatment is assigned a label indicating normal and saved, the transmission processing unit 354 transmits the operating sound with a label indicating the abnormal area and the recorded sound with a label indicating normal after treatment to the data center 40 (step S25).

[0059] As shown in Fig. 11, even when the abnormality determination unit 352 determines that there is no abnormality (step S6 in Fig. 4), the maintenance person M may determine that there is an abnormality as a result of the maintenance inspection. In this case, the maintenance person M investigates the source of the abnormal noise (step S31) and identifies the abnormal part (step S22). After that, the same processes as steps S23 to S25 in Fig. 10 are performed.

[0060] As described above, in this embodiment, the operation sound recording unit 351 records the operation sound of the escalator 1 in the storage unit 34 of the mobile terminal 30 via the microphone 31 of the mobile terminal 30 or an external microphone. The abnormality determination unit 352 compares the recorded sound with the reference sound stored in the storage unit 34, determines whether or not there is an abnormality in the escalator 1 based on the amount of change in the recorded sound relative to the reference sound, and displays the determination result on the display unit 33. The abnormal part determination unit 353 compares the recorded sound when the abnormality determination unit 352 determines that there is an abnormality with the abnormality model stored in the storage unit 34, which has learned the abnormal parts of the escalator 1 corresponding to the operation sound of the escalator 1, to determine the abnormal part corresponding to the recorded sound, and displays the determination result on the display unit 33. This makes it possible to appropriately identify the abnormal part using the abnormality model that has learned the abnormal part corresponding to the operation sound. If the abnormality model is updated, the number of identifiable abnormal parts can be increased. Furthermore, the identification of the abnormal part can be achieved using a simple and low-cost configuration provided in the mobile terminal 30.

[0061] Furthermore, in this embodiment, when the abnormality determination unit 352 determines that there is no abnormality, it stores a label indicating normality in association with the recorded sound in the storage unit 34. This allows the recorded sound associated with the label indicating normality to be used as a reference sound for abnormality determination, thereby improving the accuracy of abnormality determination.

[0062] Furthermore, in this embodiment, even when it is determined that an abnormality exists, abnormality determination unit 352 stores a label indicating a normal state in memory unit 34 in association with the recorded sound in accordance with an input operation on mobile terminal 30. Furthermore, even when it is determined that no abnormality exists, abnormal portion determination unit 353 stores a label indicating an abnormal portion in memory unit 34 in association with the recorded sound in accordance with an input operation on mobile terminal 30. This allows the recorded sound associated with a label indicating a normal state to be used as a reference sound for abnormality determination, thereby improving the accuracy of abnormality determination. Furthermore, since the recorded sound associated with a label indicating an abnormal portion can be used to update the abnormality model, it becomes possible to determine the abnormal portion even for abnormalities that were not anticipated when escalator 1 was first installed, thereby improving the accuracy of abnormal portion determination.

[0063] Furthermore, in this embodiment, when the abnormal part determination unit 353 is unable to determine an abnormal part corresponding to the recorded sound, a label indicating the abnormal part is stored in the storage unit 34 in association with the recorded sound in accordance with an input operation on the mobile terminal 30. This allows the recorded sound associated with the label indicating the abnormal part to be used to update the abnormal model, thereby improving the accuracy of determining the abnormal part.

[0064] Furthermore, in this embodiment, even if the abnormal part determination unit 353 can determine an abnormal part corresponding to the recorded sound, a label indicating an abnormal part different from the determined abnormal part is stored in the storage unit 34 in association with the recorded sound in accordance with an input operation on the mobile terminal 30. This allows the recorded sound associated with a label indicating an abnormal part different from the determined abnormal part to be used to update the abnormal model, thereby improving the accuracy of determining the abnormal part.

[0065] Furthermore, in this embodiment, transmission processing unit 354 transmits the operation sound, to which a label indicating the abnormal part has been attached, to data center 40. Data center 40 learns the abnormal part of escalator 1 corresponding to the operation sound of escalator 1, based on the operation sound, to which a label indicating the abnormal part has been attached, transmitted by transmission processing unit 354. In this way, by learning the abnormal part on the data center 40 side, it is possible to reduce the load on the mobile terminal 30 side. Furthermore, abnormal part determination unit 353 can identify the abnormal part with high accuracy based on the abnormal part learned by data center 40.

[0066] Furthermore, in this embodiment, data center 40 updates the abnormality model based on the operation sound to which a label indicating the abnormal part has been added. Furthermore, reception processing unit 355 receives the updated abnormality model from data center 40 and stores it in storage unit 34. This allows the abnormality model to be updated based on a combination of recorded sounds accumulated during the operation of escalator 1 and labels indicating the abnormal part, making it possible to identify the abnormal part even when responding to an unexpected abnormality.

[0067] In this embodiment, the abnormality determination unit 352 determines that an abnormality exists when the amount of change in the recorded operation sound relative to the reference sound exceeds a preset threshold. Furthermore, when a label with content different from the determination result by the abnormality determination unit 352 is stored in the storage unit 34, the threshold update unit 356 updates the threshold based on the label with content different from the determination result. This can further improve the accuracy of the abnormality determination.

[0068] In this embodiment, the reference sound includes a first reference sound for determining whether or not there is an abnormality due to aging, and a second reference sound for determining whether or not there is a sudden abnormality. Furthermore, the abnormality determination unit 352 performs determination using the first reference sound and determination using the second reference sound in parallel. This allows two types of abnormality determination to be performed simultaneously, thereby improving determination accuracy and determination efficiency.

[0069] In this embodiment, the first reference sound is the sound recorded when escalator 1 was first installed, and the second reference sound is the sound recorded the previous time. This allows abnormality determination unit 352 to use appropriate operating sounds as reference sounds for determining whether or not there is an abnormality due to aging and for determining whether or not there is a sudden abnormality.

[0070] In this embodiment, the abnormality determination unit 352 determines that an abnormality exists in at least one of the following cases: when the amount of change in the recorded sound relative to the first reference sound exceeds a first threshold value set in advance; and when the amount of change in the recorded sound relative to the second reference sound exceeds a second threshold value set in advance. This allows for appropriate abnormality determination.

[0071] In this embodiment, the operation sound recording unit 351 records the operation sounds of the steps 7 at the boarding and alighting boards on the upper and lower floors of the escalator 1 until they make one rotation, and also records the operation sounds of the maintenance worker carrying the mobile terminal 30 while riding the escalator 1 until he or she arrives at the destination floor. This makes it possible to more appropriately identify the abnormal part.

[0072] Although several embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included within the scope and spirit of the invention, as well as within the scope of the invention described in the claims and their equivalents. [Explanation of symbols]

[0073] 1 escalator, 30 mobile terminal, 31 microphone, 33 display unit, 34 memory unit, 35 abnormality diagnosis device, 351 operation sound recording unit, 352 abnormality determination unit, 353 abnormal part determination unit, 354 transmission processing unit, 355 reception processing unit, 356 threshold update unit, 40 data center

Claims

1. An abnormality diagnosis device provided in a mobile terminal for diagnosing an abnormality in a passenger conveyor device that circulates a plurality of endlessly connected steps, an operation sound recording unit that records operation sounds of the passenger conveyor device in a storage unit of the mobile terminal via a sound collecting unit of the mobile terminal or an external sound collecting unit of the mobile terminal; an abnormality determination unit that compares the operation sound recorded by the operation sound recording unit with a reference sound stored in the storage unit, determines whether or not there is an abnormality in the passenger conveyor device based on an amount of change in the recorded operation sound relative to the reference sound, and displays the determination result on a display unit of the mobile terminal; an abnormality portion determination unit that compares the recorded operation sound when the abnormality determination unit determines that there is an abnormality with an abnormality model that is stored in the storage unit and that has learned abnormal portions of the passenger conveyor corresponding to the operation sound of the passenger conveyor, to determine the abnormal portion corresponding to the recorded operation sound, and displays the determination result on the display unit; A passenger conveyor abnormality diagnosis device comprising:

2. The abnormality diagnosis device according to claim 1 , wherein the abnormality determination unit, when determining that no abnormality exists, causes the storage unit to store additional information indicating normality in association with the recorded operating sound.

3. the abnormality determination unit causes the storage unit to store additional information indicating normality in association with the recorded operation sound in accordance with an input operation on the mobile terminal, even when it is determined that the abnormality exists; 2. The passenger conveyor abnormality diagnosis device according to claim 1, wherein the abnormality part determination unit causes the memory unit to store additional information indicating the abnormal part in association with the recorded operating sound, in accordance with an input operation on the mobile terminal, even if it is determined that there is no abnormality.

4. 2. The abnormality diagnosis device according to claim 1, wherein, when the abnormality portion determination unit is unable to determine the abnormal portion corresponding to the recorded operation sound, the abnormality portion determination unit stores additional information indicating the abnormal portion in the storage unit in association with the recorded operation sound, in accordance with an input operation on the mobile terminal.

5. 2. The abnormality diagnosis device according to claim 1, wherein even if the abnormality portion determination unit is able to determine the abnormal portion corresponding to the recorded operation sound, the abnormality portion determination unit causes the storage unit to store additional information indicating an abnormal portion different from the determined abnormal portion in association with the recorded operation sound, in accordance with an input operation on the mobile terminal.

6. a transmission processing unit that transmits the operation sound associated with the additional information indicating the abnormal portion to a data center; 6. The abnormality diagnosis device according to claim 3, wherein the data center learns the abnormal portion of the passenger conveyor corresponding to the operation sound of the passenger conveyor, based on the operation sound associated with the additional information indicating the abnormal portion transmitted by the transmission processing unit.

7. the data center updates the abnormality model based on the operation sound associated with the additional information indicating the abnormal portion; The abnormality diagnosis device according to claim 6 , further comprising a receiving and processing unit that receives the updated abnormality model from the data center and stores it in the storage unit.

8. the abnormality determination unit determines that the abnormality exists when a change amount of the recorded operation sound relative to the reference sound exceeds a preset threshold value; 4. The abnormality diagnosis device according to claim 3, further comprising a threshold updating unit that, when additional information different from the determination result by the abnormality determination unit is stored in the storage unit, updates the threshold based on the additional information different from the determination result.

9. the reference sound includes a first reference sound for determining whether or not there is an abnormality due to aging deterioration, and a second reference sound for determining whether or not there is a sudden abnormality, The abnormality diagnosis device according to claim 1 , wherein the abnormality determination unit performs a determination using the first reference sound and a determination using the second reference sound in parallel.

10. 10. The abnormality diagnosis device according to claim 9, wherein the first reference sound is an operation sound recorded by the operation sound recording unit when the passenger conveyor device is first installed, and the second reference sound is an operation sound recorded previously by the operation sound recording unit.

11. 11. The abnormality diagnosis device according to claim 10, wherein the abnormality determination unit determines that the abnormality exists in at least one of a case where an amount of change in the recorded operation sound relative to the first reference sound exceeds a predetermined first threshold and a case where an amount of change in the recorded operation sound relative to the second reference sound exceeds a predetermined second threshold.

12. 6. The abnormality diagnosis device according to claim 1, wherein the operation sound recording unit records the operation sounds from the time the steps make one rotation at the boarding and alighting boards on the upper and lower floors of the passenger conveyor device, and also records the operation sounds from the time a maintenance worker, while holding the mobile terminal, rides the passenger conveyor device until he or she arrives at the destination floor.

13. 1. A method for diagnosing an abnormality in a passenger conveyor device that circulates a plurality of endlessly connected steps, using a mobile terminal, comprising: a step of recording the operation sound of the passenger conveyor device in a storage unit of the mobile terminal via a sound collection unit of the mobile terminal or an external sound collection unit of the mobile terminal; a step of comparing the recorded operation sound with a reference sound stored in the storage unit, and determining whether or not there is an abnormality in the passenger conveyor device based on an amount of change in the recorded operation sound relative to the reference sound; a step of displaying a determination result of the presence or absence of an abnormality in the passenger conveyor device on a display unit of the mobile terminal; a step of comparing the recorded operation sound when it is determined that there is an abnormality with an abnormality model stored in the storage unit and which has learned an abnormal part of the passenger conveyor corresponding to the operation sound of the passenger conveyor, thereby determining the abnormal part corresponding to the recorded operation sound; a step of displaying the abnormality region determination result on the display unit; A method for diagnosing abnormalities in a passenger conveyor, comprising:

Images