Elevator diagnostic apparatus and method for diagnosing elevator

Abstract

To improve diagnostic accuracy of an elevator.SOLUTION: A diagnostic apparatus creates a reference space for each elevator, compares multidimensional feature quantity data generated from sound-vibration waveform data acquired for the elevator with a reference space prepared for the elevator among reference spaces prepared for respective elevators, and performs diagnosis regarding an abnormality of the elevator on the basis of the result of the comparison.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates generally to elevator diagnostics. [Background technology]

[0002] An example of an elevator is an elevator. A known diagnostic device for elevators is disclosed in Patent Document 1. The diagnostic device disclosed in Patent Document 1 prepares a reference space for the sound and vibration waveforms of an elevator that is not abnormal for each elevator specification, and diagnoses whether or not there is an abnormality in the elevator based on the Mahalanobis distance between the analysis results of the sound and vibration waveforms detected in the elevator car of the elevator to be diagnosed and the reference space corresponding to the elevator specification. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2013-113775 Summary of the Invention [Problem to be solved by the invention]

[0004] The waveforms of sound / vibration (sound and / or vibration) from elevators and escalators (including horizontal escalators known as walking walkways) depend on the installation accuracy and / or installation environment of the elevator. Therefore, even if multiple elevators with different installation accuracy or environments have the same specifications, the distribution of data related to the elevator sound / vibration waveforms does not necessarily fall within the same reference space. For example, for elevators with the same specifications, the feature distributions of sound / vibration from elevator doors installed in one location generally overlap for each floor, while the feature distributions of sound / vibration from elevator doors installed in another location differ from floor to floor and do not overlap. Thus, even if the specifications are identical, the distributions do not fall within the same reference space, which can result in reduced diagnostic accuracy. [Means for solving the problem]

[0005] The diagnostic device creates a reference space for each elevator. When diagnosing an elevator, the diagnostic device compares multidimensional feature data generated from sound / vibration waveform data acquired for that elevator with a reference space prepared for that elevator from among the reference spaces prepared for each elevator, and diagnoses any abnormalities in that elevator based on the results of the comparison. [Effects of the Invention]

[0006] According to the present invention, the accuracy of diagnosis of an elevator is improved. [Brief explanation of the drawings]

[0007] [Figure 1] 1 shows the overall configuration of a system according to an embodiment. [Figure 2] The configuration of an elevator control device is shown. [Figure 3] 1 shows the configuration of a relay device. [Figure 4] 1 shows the configuration of a diagnostic device. [Figure 5] An example of the reference space and deviation for each elevator is shown below. [Figure 6] 10 shows a flow of an example of processing performed by an elevator control device. [Figure 7] 10 shows an example flow of processing performed by a diagnostic device. [Figure 8] 10 shows an overview of an example of a process for determining a door abnormality. [Figure 9] An overview of the processing for each window is shown below. [Figure 10] An example of a window is shown. [Figure 11] 1 shows an example of a reference space for each window. [Figure 12] 10 shows an example of a maintenance screen. [Figure 13] 10 shows an example of a maintenance screen. DETAILED DESCRIPTION OF THE INVENTION

[0008] In the following description, an "interface apparatus" may refer to one or more interface devices. The one or more interface devices may be at least one of the following: One or more I / O (Input / Output) interface devices. The I / O (Input / Output) interface devices are interface devices for at least one of the I / O device and a remote display computer. The I / O interface device for the display computer may be a communications interface device. The at least one I / O device may be a user interface device, for example, either an input device such as a keyboard and a pointing device, or an output device such as a display device. One or more communication interface devices. The one or more communication interface devices may be one or more homogeneous communication interface devices (e.g., one or more NICs (Network Interface Cards)) or two or more heterogeneous communication interface devices (e.g., an NIC and an HBA (Host Bus Adapter)).

[0009] In the following description, "memory" refers to one or more memory devices, typically a primary storage device. At least one of the memory devices may be a volatile memory device or a non-volatile memory device.

[0010] In the following description, a "persistent storage device" may refer to one or more persistent storage devices, which are an example of one or more storage devices. A persistent storage device may typically be a non-volatile storage device (e.g., an auxiliary storage device), and specifically may be, for example, a hard disk drive (HDD), a solid state drive (SSD), a non-volatile memory express (NVME) drive, or a storage class memory (SCM).

[0011] In the following description, the term "storage device" may refer to at least one of memory and persistent storage device.

[0012] Furthermore, in the following description, a "processor" may refer to one or more processor devices. The at least one processor device may typically be a microprocessor device such as a CPU (Central Processing Unit), but may also be another type of processor device such as a GPU (Graphics Processing Unit). The at least one processor device may be a single-core or multi-core. The at least one processor device may also be a processor core. The at least one processor device may also be a processor device in a broader sense, such as a circuit that is a collection of gate arrays written in a hardware description language that performs some or all of the processing (for example, an FPGA (Field-Programmable Gate Array), a CPLD (Complex Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit)).

[0013] In the following description, data (information) that produces an output in response to an input may be described using expressions such as "xxx table." However, this data (information) may be data of any structure, or may be a learning model such as a neural network, genetic algorithm, or random forest that produces an output in response to an input. Therefore, an "xxx table" may be referred to as "xxx data." In the following description, one table may be divided into two or more tables, or all or part of two or more tables may be one table.

[0014] Hereinafter, embodiments will be described with reference to the drawings. In the present invention, the "elevator" to be diagnosed may be either an elevator or an escalator, and the escalator may be an inclined escalator or a horizontal escalator known as a walking walkway. In the following embodiments, the elevator is an elevator.

[0015] FIG. 1 shows the overall configuration of a system according to an embodiment.

[0016] An elevator control device 101, a relay device 102, a diagnostic device 103, a manager terminal 121, a maintenance worker terminal 122, and a user terminal 123 communicate with each other via a communication network 110. The communication network 110 may include at least one of a dedicated communication network, the Internet, a WAN (Wide Area Network), and a LAN.

[0017] The elevator control device 101 controls multiple elevators 150. In this embodiment, multiple elevators 150 exist in one building, but the multiple elevators 150 may exist in multiple locations, such as multiple buildings, and there may be only one elevator 150 in one location. Also, the elevator control device 101 may be common to multiple locations, but in this embodiment, it is assumed that one elevator control device 101 exists in one building.

[0018] The elevator 150 is equipped with multiple (or one) sensors 151 for detecting sound / vibration. The sensors 151 include a microphone 151A, an acceleration sensor 151B, etc. Waveform data of the sound / vibration detected by each elevator 150 is transmitted to the relay device 102 via (or without) the elevator control device 101 and over the communication network 110. For each elevator 150, the sound / vibration waveform data may include sound / vibration waveform data for each sensor 151.

[0019] The relay device 102 receives and stores the sound / vibration waveform data of each elevator 150. The sound / vibration waveform data received by the relay device 102 may be data processed by the elevator control device 101, or may be raw data that has not been processed by the elevator control device 101. Alternatively, the relay device 102 may be omitted, and the diagnostic device 103 may receive and store the sound / vibration waveform data of each elevator 150 (in other words, the diagnostic device 103 may have the function of the relay device 102).

[0020] The relay device 102 accumulates sound / vibration waveform data for each elevator 150. For each elevator 150, the sound / vibration waveform data includes sound / vibration waveform data at the time of creating the reference space (in other words, sound / vibration waveform data when the elevator 150 is normal) and sound / vibration waveform data at the time of diagnosis.

[0021] The diagnostic device 103 acquires sound / vibration waveform data for the elevator 150 to be diagnosed from the relay device 102, generates multidimensional feature data from the waveform data, and diagnoses the condition of the elevator 150 based on the results of comparing a reference space with the multidimensional feature data. For each of the multiple elevators 150, when creating a reference space, the diagnostic device 103 generates multidimensional feature data from sound / vibration waveform data acquired for the elevator 150 when the elevator is normal, and creates a reference space of multidimensional features for the elevator 150 from the multidimensional feature data. For each of the multiple elevators 150, when diagnosing the elevators 150, the diagnostic device 103 compares the multidimensional feature data generated from the sound / vibration waveform data acquired for the elevator 150 with a reference space prepared for the elevator 150 out of the reference spaces prepared for each elevator 150, and diagnoses an abnormality in the elevator 150 based on the results of the comparison. In this way, for each elevator 150, the reference space that is compared with the multidimensional feature data of that elevator 150 at the time of diagnosis is the reference space of that elevator 150 itself. For this reason, the reference space created for each elevator 150 reflects the installation accuracy and installation environment of that elevator 150, so the reference space is appropriate, and therefore the diagnostic accuracy of the elevator is improved.

[0022] The diagnostic device 103 can communicate with a manager terminal 121 , a maintenance worker terminal 122 , and a user terminal 123 via a communication network 110 .

[0023] The administrator terminal 121 is an information processing terminal of the administrator. In this specification, the term "information processing terminal" refers to a stationary or mobile computer (e.g., a personal computer or a smartphone). The "administrator" may be, for example, a data scientist, who may use the administrator terminal 121 to assign a label indicating whether the result of the diagnosis by the diagnosis device 103 is correct or to instruct the diagnosis device 103 to update the reference space.

[0024] The maintenance personnel terminal 122 is an information processing terminal for the maintenance personnel. The maintenance personnel inputs the record of the maintenance work into the diagnostic device 103 through the maintenance personnel terminal 122.

[0025] The user terminal 123 is an information processing terminal for inputting feedback from users of the elevator 150. The users input feedback about the elevator 150 (for example, their impressions about the ride comfort of the car) to the diagnostic device 103 through the user terminal 123. The feedback does not necessarily have to be input by the users themselves; a maintenance worker may input the feedback from the maintenance worker terminal 122 based on complaints and comments received from users.

[0026] FIG. 2 shows the configuration of the elevator control device 101.

[0027] The elevator control device 101 includes an interface device 201, a memory device 202, and a processor 203 connected to the devices 201 and 202.

[0028] Control of the elevator 150 and detection of sound / vibration of the elevator 150 are performed through the interface device 201. In addition, communication with the relay device 102 and / or the diagnosis device 103 is performed through the interface device 201.

[0029] Data and programs are stored in the storage device 202. The data includes, for example, control data for each elevator 150. The control data for each elevator 150 is data that defines the operation of the elevator 150, and may include, for example, data indicating the floor at which the car starts traveling (e.g., the landing floor) and the floor at which the car ends traveling (e.g., the destination floor), the timing of when the car starts and ends traveling, the timing of when the car doors open and close, the traveling speed of the car, the timing of when the car fan starts and ends driving, and the like.

[0030] The processor 203 executes a program stored in the storage device 202 to realize functions such as an operation control unit 212. The operation control unit 212 controls each elevator 150 based on control data 211 for each elevator 150. The operation control unit 212 selectively performs, for example, diagnostic operation and normal operation for each elevator 150. Normal operation is operation for transporting passengers from a landing floor to a destination floor. Diagnostic operation is operation for diagnosing the elevator 150, and is performed, for example, late at night when there are no passengers.

[0031] FIG. 3 shows the configuration of the relay device 102.

[0032] The relay device 102 may be a physical computer system (e.g., one or more physical computers) or a logical computer system (e.g., a virtual server or a cloud system based on a physical computer system). The relay device 102 may be a server. The relay device 102 has a data lake 301 for each elevator 150. The data lake 301 may be an example of a storage area. The data lake 301 contains sound / vibration waveform data of the corresponding elevator 150.

[0033] FIG. 4 shows the configuration of the diagnostic device 103.

[0034] In this embodiment, the diagnostic device 103 is a physical computer system, but may also be a logical computer system (for example, a virtual server or a cloud system based on a physical computer system). The diagnostic device 103 may be a server. The diagnostic device 103 has an interface device 401, a storage device 402, and a processor 403 connected to the devices 401 and 402.

[0035] Communication with the relay device 102 and / or the elevator control device 101 is performed through the interface device 201. Communication with the manager terminal 121, the maintenance staff terminal 122, and the user terminal 123 is also performed through the interface device 201.

[0036] The storage device 402 stores data and programs. The data includes, for example, a diagnosis database 410 and a correct label database 420. The diagnosis database 410 includes an individual measurement table 411, an individual feature table 412, an individual reference space table 413, an individual deviation table 414, and an individual judgment table 415. The correct label database 420 includes a work record table 421, a feedback record table 422, and an operation status record table 423. The individual measurement table 411 includes data representing the time series of the measurement diagnosis results (whether the measurement was successful or failed) for each elevator 150. The individual feature table 412 includes data representing the time series of the multidimensional feature extracted for each elevator 150. The individual reference space table 413 includes data representing the reference space created for each elevator 150. The individual deviation table 414 includes data representing a time series of deviations (deviations between the multidimensional feature values ​​acquired during diagnosis and the reference space) for each elevator 150. The individual determination table 415 includes data representing the history of diagnosis results for each elevator 150 (for example, whether the elevator 150 is normal or abnormal, or the degree of abnormality of the elevator 150). The work record table 421 includes data representing work records by maintenance personnel for each elevator 150. The feedback record table 422 includes data representing feedback from users. The operation status record table 423 includes data representing a time series of operation statuses for each elevator 150 (operation statuses remotely monitored for each elevator 150).

[0037] The processor 403 executes the programs stored in the storage device 402 to implement functions such as a measurement diagnosis unit 431, a feature extraction unit 432, an individual creation unit 433, an individual evaluation unit 434, a deviation calculation unit 435, an abnormality diagnosis unit 437, and a communication unit 438. The measurement diagnosis unit 431 performs measurement diagnosis for each elevator 150, including determining whether sound / vibration measurement was successful, and records data representing the results of the measurement diagnosis in an individual measurement table 411. The feature extraction unit 432 extracts multidimensional features from the sound / vibration waveform data for each elevator 150, and records data representing the extracted multidimensional features in an individual feature table 412. The individual creation unit 433 identifies multidimensional features for each elevator 150 from the individual feature table 412, creates (updates) a reference space based on the identified multidimensional features, and records data representing the created reference space in an individual reference space table 413. The individual evaluation unit 434 identifies a reference space for each elevator 150 from the individual reference space table 413 and evaluates the identified reference space (the individual creation unit 433 updates the reference space of the elevator 150 according to the evaluation result). The deviation calculation unit 435 identifies a reference space for each elevator 150 from the individual reference space table 413, calculates the deviation between the multidimensional feature at the time of diagnosis and the identified reference space, and records data representing the calculated deviation in the individual deviation table 414. The abnormality diagnosis unit 437 identifies a deviation trend (time series) for each elevator 150 from the individual deviation table 414, performs an abnormality diagnosis, which is a diagnosis of an abnormality in the elevator 150, based on the deviation trend, and records data representing the result of the abnormality diagnosis in the individual determination table 415. The communication unit 438 communicates with the relay device 102 and / or the elevator control device 101, and also communicates with the manager terminal 121, the maintenance staff terminal 122, and the user terminal 123. For example, the communication unit 438 transmits the results of an abnormality diagnosis of the elevator 150 to the maintenance staff terminal 122 (and / or the manager terminal 121). Also, for example, the communication unit 438 remotely monitors the operation status of each elevator 150, and records data representing the operation status as a monitoring result in the operation status record table 423.

[0038] FIG. 5 shows an example of the reference space and the deviation degree for each elevator 150.

[0039] For each of the plurality of elevators 150 (for example, elevators #1 to #4), there is a reference space 501 and a deviation transition 502. Elements related to the reference space 501 include a model type, a deviation threshold, and a feature set.

[0040] The model type may be, for example, a normal distribution model, a Gaussian Mixture Model (GMM), or a machine learning model (e.g., a neural network). Specifically, for example, the model of the reference space 501 of elevators #1 to #3 is a normal distribution model, and the model of the reference space 501 of elevator #4 is a GMM. Note that, depending on the type of model, a determination algorithm for the degree of discrepancy between the reference space and the multidimensional feature amount at the time of diagnosis may differ. For example, when the model of the reference space 501 is a normal distribution model, Mahalanobis discrimination using the Mahalanobis distance between the reference space and the multidimensional feature amount may be adopted as the determination algorithm. Furthermore, when the model of the reference space 501 is a machine learning model, a machine learning method (Variational Auto Encoder: VAE) may be adopted as the determination algorithm.

[0041] A feature set is m-dimensional feature (m is an integer of 2 or more) that constitutes a multidimensional feature, in other words, multiple types of feature (multiple feature items).

[0042] In this way, the model type, deviation threshold, and feature set according to the installation accuracy and installation environment of the elevator 150 are associated with the reference space, so it is expected that diagnostic accuracy using the reference space will be maintained for each elevator 150. According to the example shown in Fig. 5, the number of dimensions of the features defining the reference space 501 is two for all of elevators #1 to #4, and feature 1 and feature 2 are used. Depending on the elevator 150, two or more other types of feature may be used instead of or in addition to feature 1 and / or feature 2.

[0043] In this embodiment, for each of the multiple elevators 150, if the deviation exceeds the deviation threshold value only once, the elevator 150 is not diagnosed as having an abnormality, but rather, a diagnosis of an abnormality for the elevator 150 is made based on the deviation transition. In the illustrated example, of elevators #1 to #4, it is determined that elevators #1 and #4 have an abnormality.

[0044] An example of the processing performed in this embodiment will be described below.

[0045] FIG. 6 shows an example of the flow of processing performed by the elevator control device 101.

[0046] The operation control unit 212 determines whether or not to perform diagnostic operation (S601). For example, if the current time falls within the time period during which diagnostic operation is performed (e.g., late at night), or if a maintenance worker or the like has instructed to perform diagnostic operation, the determination result of S601 may be true.

[0047] If the determination result of S601 is false (S601: NO), the operation control unit 212 performs normal operation. When normal operation is performed, users can use the elevator 150 to move between floors.

[0048] If the determination result of S601 is true (S601: YES), the operation control unit 212 performs a diagnostic operation (S603). The sound / vibration waveform data obtained in the diagnostic operation is used to create a reference space or diagnose an abnormality.

[0049] In the diagnostic operation, the operation control unit 212 refers to the control data 211 for each elevator 150 (S611), extracts a section (S612), and transmits sound / vibration waveform data for the section to the relay device 102 (S613). The data transmitted in S613 is linked to the ID of the elevator 150 and section information (information indicating the section). The relay device 102 identifies the data lake 301 corresponding to the ID linked to the received sound / vibration waveform data, and stores the sound / vibration waveform data in the identified data lake 301.

[0050] Here, a "section" refers to the period from the start to the end of control. Specifically, for example, to diagnose abnormalities related to sound / vibration while the door is opening / closing or while the car is traveling, it is necessary to identify the section, such as the point in time at which the door opened (or the car started traveling) and the point in time at which the door closed (or the car stopped). The control data 211 specifies the timing of the start and end of the door opening / closing operation and the timing of the start and end of the car traveling, so that extraction of the section can be realized in the edge device, i.e., the elevator control device 101.

[0051] Fig. 7 shows an example of the flow of processing performed by the diagnostic device 103. The processing shown in Fig. 7 is performed for each elevator 150. Below, one elevator 150 will be taken as an example.

[0052] The measurement diagnosis unit 431 performs measurement diagnosis, which includes determining whether the measurement of the sound / vibration of the elevator 150 is successful or unsuccessful (S701). Specifically, for example, the measurement diagnosis unit 431 compares the acquired sound / vibration-related values ​​of the elevator 150 in a non-operating state during a time period starting from the time of creation of a reference space or diagnosis of the elevator 150, with a failure threshold, which is a threshold value for the sound / vibration of the elevator 150 in a non-operating state, and determines whether the sound / vibration measurement of the elevator 150 is successful or unsuccessful based on the result of the comparison. The "time period" referred to here may be a certain period of time before and / or after the time of creation of the reference space or diagnosis (for example, 24 hours before and after the time of creation of the reference space).

[0053] The sound / vibration waveform data of the elevator 150 depends on the installation accuracy or installation environment of the elevator 150. The sound / vibration waveform data (sound / vibration waveform data of a normal elevator 150) acquired to create a reference space for the elevator 150 is preferably cleansed data that does not include waveform data of environmental sounds or environmental vibrations. If the sound / vibration waveform data acquired to create a reference space includes waveform data of environmental sounds or environmental vibrations, the distribution will be inaccurate and an appropriate reference space cannot be created, which may result in erroneous determinations in abnormality diagnosis or a deterioration in the accuracy of the abnormality diagnosis.

[0054] Therefore, the measurement diagnosis unit 431 evaluates the sound / vibration waveform data of the elevator 150 during a time period starting from the time of creation of the reference space or the time of diagnosis. For example, when the elevator 150 is not in operation, ideally, neither sound nor vibration is generated, and a failure threshold representing such a state is prepared for the sound / vibration of the elevator 150. A failure threshold is prepared for each elevator 150 for sound (e.g., noise level [dB]) and vibration (e.g., amplitude [g]), and the failure threshold for the elevator 150 may be recorded, for example, in the individual measurement table 411. The measurement diagnosis unit 431 compares the sound / vibration values ​​acquired for the elevator 150 in an inoperable state during a time period starting from the time of creation of the reference space or the time of diagnosis with the failure threshold for sound / vibration of the elevator 150 in an inoperable state, and determines whether the sound / vibration measurement for the elevator 150 was successful or failed based on the result of the comparison. The "time period" referred to here may be a certain period of time before and / or after the creation of the reference space or the diagnosis (for example, 24 hours before and after the creation of the reference space).

[0055] The "sound / vibration acquired from the elevator 150 in a non-operating state" may be, for example, the sound / vibration entering from the hall (building) when the doors are fully open (stationary state), or the sound / vibration in a stationary state before the car starts moving. Ideally, no sound / vibration occurs when the mechanical components of the elevator 150 are not operating. However, if sound / vibration is observed from the elevator 150 in a non-operating state (stationary state), this can be considered to indicate a state in which environmental sounds (e.g., car noise, announcements, etc.) or environmental vibrations (e.g., vibrations from construction work, vibrations from trains, etc.) from around the elevator 150 are loud. In this embodiment, waveform data of such sound / vibration acquired in a non-operating state that is greater than a failure threshold is not used in creating the reference space and diagnosing an abnormality. This is expected to improve the accuracy of the reference space and the accuracy of abnormality diagnosis.

[0056] In the measurement diagnosis, for example, the measurement may be determined to have failed if the detected sound volume (e.g., noise level [dB]) is greater than a sound failure threshold. Also, in the measurement diagnosis, for example, the measurement may be determined to have failed if the peak-to-peak or standard deviation of the detected vibration is greater than a failure threshold.

[0057] If the measurement is unsuccessful (S701: NO), the measurement diagnosis unit 431 refers to the individual measurement table 411 and determines whether the elevator 150 has received a failure judgment result N times in a row (N is an integer equal to or greater than 2) (S731). If the judgment result in S731 is true (S731: YES), the measurement diagnosis unit 431 updates the failure threshold for the elevator 150 in question to reduce the likelihood of obtaining a failure judgment result (S732). N consecutive measurement failures indicate that there is constant loud environmental noise or vibration. Therefore, if a result is obtained that suggests that the elevator 150 in question is in such an environment (i.e., N consecutive measurement failures), the measurement diagnosis unit 431 updates the failure threshold for the elevator 150 in question (specifically, updates the failure threshold to a smaller value to reduce the likelihood of obtaining a failure judgment result). This allows for measurement diagnosis results that are appropriate for the installation environment of the elevator 150.

[0058] The above measurement diagnosis may be performed for a time period based on one or both of the time of creating the reference space and the time of diagnosis. Furthermore, when the elevator 150 is in a non-driving state (stationary state), the elevator control device 101 may transmit sound / vibration waveform data of the non-driving state elevator 150, which is linked to the ID of the elevator 150 and data (label) indicating that the elevator 150 is in a non-driving state, to the relay device 102. This allows the measurement diagnosis unit 431 to acquire the acquired sound / vibration waveform data of the non-driving state elevator 150 from the relay device 102.

[0059] The feature extraction unit 432 acquires sound / vibration waveform data of the elevator 150 in operation from the relay device 102, extracts multidimensional features from the sound / vibration waveform data, and records the multidimensional feature data in the individual feature table 412 (S702).

[0060] If the current time is the time of creating a reference space, the individual-specific creation unit 433 acquires the multidimensional feature data of the elevator 150 from the individual-specific feature table 412, creates a reference space based on the multidimensional feature data, and records the data of the reference space in the individual-specific reference space table 413 (S711). Here, "the time of creating a reference space" may mean a time when sufficient data (for example, a predetermined number of multidimensional features or more) for creating a reference space is not recorded in the individual-specific feature table 412, or in other words, a time when a reference space has not yet been created to the extent that the accuracy of abnormality diagnosis can be expected to be equal to or greater than a certain value.

[0061] The individual evaluation unit 434 evaluates the reference space created in S711 and determines whether the evaluation result satisfies the criteria (S712). Specifically, for example, the individual evaluation unit 434 evaluates the reference space corresponding to the elevator 150 in question based on the distribution of multidimensional features acquired for creating the reference space corresponding to the elevator 150 in question. "Evaluating the reference space" may mean determining whether the model type, deviation threshold, and feature set associated with the current reference space are appropriate based on the distribution of multidimensional features acquired for creating the reference space.

[0062] For example, assume that Mahalanobis discriminant analysis using a reference space is performed for abnormality diagnosis. Mahalanobis discriminant analysis is based on the assumptions that each feature is normally distributed and that the correlation between features is low. In this embodiment, a reference space is automatically created for each elevator 150. The created reference space is based on assumptions regarding the model type, deviation threshold, and feature set associated with the reference space. If the assumptions regarding the reference space are not met, the accuracy of the reference space may be reduced, resulting in a decrease in the accuracy of abnormality diagnosis. Therefore, to determine whether the assumptions regarding the reference space are satisfied, for example, the individual evaluation unit 434 determines whether the multidimensional feature acquired for creating the reference space for the elevator 150 fits within the reference space according to the current assumptions in evaluating the reference space. Furthermore, for example, if there are outliers in the feature, the individual evaluation unit 434 removes the outliers. Furthermore, for example, after creating the reference space for the elevator 150, the individual evaluation unit 434 may evaluate the distribution of the reference space or the correlation between feature values. For example, if the evaluation is an evaluation of a normal distribution, the individual evaluation unit 434 may use a normality test such as a QQ plot or Shapiro-Wilk. Also, if the evaluation is an evaluation of the correlation between feature amounts, the individual evaluation unit 434 may determine whether the feature amounts are excessively correlated (for example, by determining multicollinearity or a threshold value for the correlation coefficient).

[0063] If the evaluation results indicate that the evaluation criteria are not met (the assumptions of the reference space are not met), it is not desirable to continue creating the reference space according to the current assumptions, and the result of S712 is false. On the other hand, if the evaluation results indicate that the evaluation criteria are met (the assumptions of the reference space are met), it is not a problem to continue creating the reference space according to the current assumptions, and the result of S712 is true.

[0064] If the determination result in S712 is false (S712: NO), or if the result of the evaluation performed after the creation of the reference space for the elevator 150 does not satisfy the criteria, the individual-specific creation unit 433 updates the reference space for the elevator 150 (S721). For example, the individual-specific reference space table 413 includes data representing the model type, deviation threshold, and feature set as data representing elements defining the reference space for each elevator 150. The individual-specific creation unit 433 changes at least one of the model type, deviation threshold, and feature set associated with the reference space. As a result, the individual-specific reference space table 413 is updated. For example, the model type may be changed from a first type model (e.g., a normal distribution model) to a second type model (e.g., a GMM), the deviation threshold may be changed to a higher or lower value, or a certain type of feature may be removed from the feature set, replaced with a different type of feature, or added. This is expected to avoid a decrease in the accuracy of the reference space and, thereby, to avoid a decrease in the accuracy of abnormality diagnosis. Furthermore, even if a change occurs that causes the expected feature distribution to be lost due to a change in the installation accuracy or installation environment of the elevator 150, such a change is detected and the reference space is updated appropriately. Therefore, the accuracy of the reference space can be maintained, and it is expected that the accuracy of abnormality diagnosis will be maintained. Note that "after creation" here may be after the creation of the reference space during a period before the reference space creation time is completed, or may be after the reference space creation time is completed (for example, after a sufficient number of multidimensional features have been obtained to create the reference space) (i.e., it may mean "after completion" of the reference space).

[0065] If the determination result of S712 is true (S712: YES), the individual creation unit 433 determines whether or not there exists a reference space of another elevator 150 whose creation has progressed more than the reference space of the elevator 150, based on a distribution that is the same as or similar to the previous distribution (distribution of multidimensional features) of the elevator 150 (S713). If the determination result of S713 is true (S713: YES), the individual creation unit 433 uses the reference space of the other elevator 150 as the reference space of the elevator 150 (S714).

[0066] Specifically, for example, the individual-specific creation unit 433 identifies the previous distribution (distribution of multidimensional features) of the elevator 150 from the individual-specific feature table 412. The individual-specific creation unit 433 searches the individual-specific feature table 412 for a distribution that is composed of more multidimensional features than the multidimensional features that make up the identified distribution and that is identical to or similar to the identified distribution. A reference space created based on a distribution composed of more multidimensional features than the multidimensional features that make up the identified distribution is an example of a reference space that has been created more advancedly than a reference space based on the previous distribution of the elevator 150. Furthermore, a "distribution similar to" the identified distribution may mean, for example, that the feature sets between the distributions are completely or partially the same and the distributions of the same type of feature are similar, or that the distributions are similar for different feature values ​​but highly correlated feature values. When an identical or similar distribution with a larger number of multidimensional features is found, the individual creation unit 433 may use the reference space of another elevator 150 based on the identical or similar distribution as the reference space of the elevator 150. Specifically, for example, the reference space may be used in any of the following ways. Linking the elevator 150 to the reference space of another elevator 150, and making the reference space a common reference space between the other elevator 150 and the elevator 150. - Modify the reference space of another elevator 150 based on the specifications or other correlations between the elevator 150 and the elevator in question, and use the modified reference space as the reference space of the elevator 150 in question.

[0067] Appropriately using the reference space of another elevator 150 in this manner contributes to speeding up the creation of the reference space. That is, the sound / vibration waveform data used to create the reference space is data collected over a long period of time, for example, data acquired only during diagnostic operations performed within a limited time frame. For this reason, it takes a long time to complete the reference space. However, by appropriately using the reference space of another elevator 150 as in S713 and S714, the time required to complete the reference space can be shortened, and the accuracy of the completed reference space is expected to be within an acceptable range. It is not necessary to use the reference space or data acquired by another elevator 150; only the reference space or data acquired by the elevator 150 may be used. When a sufficient amount of data can be acquired using only the elevator 150, or in situations where highly accurate determination is required, it is preferable to use data from only the elevator 150.

[0068] If the determination result of S703 is false (S703: NO), that is, if the current time is the time of diagnosis, the deviation calculation unit 435 identifies the reference space of the elevator 150 from the individual reference space table 413 (S704). The deviation calculation unit 435 calculates the deviation between the multidimensional feature (generated multidimensional feature data) extracted from the sound / vibration waveform data acquired at the time of diagnosis and the identified reference space, and records the calculated deviation (e.g., Mahalanobis distance) in the individual deviation table 414. The abnormality diagnosis unit 437 identifies the deviation transition (temporal transition of the deviation) of the elevator 150 from the individual deviation table 414, and performs an abnormality diagnosis of the elevator 150 based on the identified deviation transition (S705). In this embodiment, "abnormality diagnosis" refers to a determination of whether or not there is an abnormality, but may instead or in addition include a determination of the degree of abnormality. The method (e.g., algorithm) for calculating the deviation in S704 and the method (e.g., algorithm) for diagnosing the abnormality in S705 may depend on the model type and feature set of the identified reference space. Furthermore, the "degree of abnormality" may be a statistical value (e.g., average value or maximum value) of multiple deviations that constitute the deviation transition.

[0069] If it is determined that an abnormality has occurred (S705: YES), the abnormality diagnosis unit 437 performs abnormality processing, for example, issues a report to the maintenance personnel terminal 122 via the communication unit 438 (S707).

[0070] The elevator 150 may experience a temporary increase in noise / vibration due to a sudden malfunction. For example, if a foreign object (e.g., a pebble) gets stuck in the sill of the car door, the object may collide with the car door when the car opens or closes, generating abnormal noise / vibration. However, in such a case, the foreign object is broken into pieces, resulting in few cases of repeated abnormal noise / vibration. Therefore, if an abnormality is determined and an alarm is issued when a multidimensional feature value exceeds a deviation threshold value only once, even if a maintenance worker travels to the site where the elevator 150 is installed to check, the abnormality may not be reproduced, resulting in a waste of time for the maintenance worker. In this embodiment, abnormality diagnosis is performed based on deviation transitions. For example, as illustrated in FIG. 5, an abnormality may be determined when a predetermined percentage or more of the deviations constituting the deviation transition exceed the deviation threshold. In this manner, an abnormality diagnosis based on deviation transitions is expected to result in improved accuracy of abnormality diagnosis.

[0071] As described above, a reference space is created for each elevator 150, and an abnormality diagnosis is performed. In creating the reference space, for example, 14-dimensional features, such as loudness (1-dimensional) and MFCC (13-dimensional), may be acquired as sound-related features, and / or, for example, peak-to-peak (6-dimensional) features may be acquired as vibration-related features. Examples of sound / vibration-related features to be acquired are not limited to these examples, and other features may also be acquired. Furthermore, the number of data items required to create the reference space (the number of components of the reference space) may be any number between 100 and 1000, for example.

[0072] The reference space creation and / or abnormality diagnosis may be performed for each elevator 150 in finer units.

[0073] An example of a subdivided unit for each elevator 150 may be each inspection item, such as each elevator part, such as a car door, a fan installed in the car, or a door for each floor. Another example of a subdivided unit may be each elevator running position (e.g., every half floor). That is, for each of the multiple elevators 150, the individual creation unit 433 may generate multidimensional feature data for each elevator part or position from the sound / vibration waveform data acquired for that elevator part or position, and create a reference space for that elevator part or position from the multidimensional feature data. The abnormality diagnosis unit 437 may diagnose an abnormality for each of the multiple elevators 150 based on the results of a comparison (e.g., a deviation transition) between the multidimensional feature data generated from the sound / vibration waveform data acquired for that elevator part or position and the reference space corresponding to that elevator part or position during diagnosis of that elevator 150. This makes it possible to identify which part or position of each elevator 150 has an abnormality.

[0074] For example, door abnormality diagnosis can be implemented as follows. That is, the individual creation unit 433 may generate multidimensional feature data from the acquired sound / vibration waveform data (and acceleration data) regarding the door opening and closing at each floor, and create a reference space for the door opening and closing at each floor from the multidimensional feature data. For example, as shown in FIG. 8 , in abnormality diagnosis during diagnosis of the elevator 150, the abnormality diagnosis unit 437 may determine the presence or absence of a door abnormality at each floor based on the result of comparing the multidimensional feature data generated from the sound / vibration waveform data acquired regarding the door opening and closing at each floor with the reference space corresponding to the door opening and closing at each floor. If there is an abnormality in the car door, abnormal sound / vibration will occur regardless of the floor at which the car door is opened or closed. Therefore, the abnormality diagnosis unit 437 performs an AND combination on the determination results for all floors. If the combination result is True (if door abnormalities are detected at all floors), the abnormality diagnosis unit 437 determines that there is an abnormality in the car door. On the other hand, if the combination result is False, the abnormality diagnosing unit 437 determines that there is an abnormality in the landing doors of some of the floors where the door abnormality was detected (if there is no door abnormality at any floor, it is determined that there is no door abnormality). This makes it possible to distinguish whether there is an abnormality in the car door or the landing door, improving maintenance efficiency. Note that, in diagnosing an abnormality in the door opening and closing at each floor, a window, which will be described with reference to FIG. 9, may be used. Specifically, for example, the time of door opening and closing may be divided into a plurality of windows, and for each window, the multidimensional feature value may be compared with the reference space.

[0075] An example of a subdivided unit for each elevator 150 may be each of two or more windows, which are two or more time intervals during the time the elevator 150 is in operation. The (x-1)th window and the xth window (x is an integer greater than or equal to 1 and less than the number of windows) may partially overlap. It is desirable that the (x-1)th window and the xth window are not separated from each other. For example, as shown in FIG. 9 , for each of multiple elevators 150, during diagnosis of the elevator 150, the abnormality diagnosis unit 437 may diagnose an abnormality in the operation of the elevator 150 for each of two or more windows during the time the elevator 150 is in operation based on the results of a comparison (e.g., deviation transition) between multidimensional feature data generated from the sound / vibration waveform data of the window and a reference space corresponding to the elevator 150. The abnormality diagnosis unit 437 may OR the abnormality diagnosis results of two or more windows (e.g., abnormal = “1”, normal = “0”) to determine the presence or absence of an abnormality for each time interval, and may identify the abnormal part for each time interval based on the relationship with the parts that were operating during the time interval. The abnormality diagnosis unit 437 may send data indicating the identified abnormal part to the maintenance staff terminal 122 as an alert. This makes it possible to identify which part of each elevator 150 is abnormal. For example, by dividing the waveform data of the sound / vibration while the elevator is traveling from the first floor to the top floor into multiple windows and determining the time period in which the abnormality occurs, it is possible to easily identify which floor the abnormal part is located on. Alternatively, by dividing the waveform data of the sound / vibration when the door is opening and closing into multiple windows and determining the time period in which the abnormality occurs, it is possible to easily identify whether the abnormal part is at the beginning of opening or in the middle of opening or closing, thereby assisting the maintenance staff in identifying the cause of the abnormal part. Such window-by-window determination can be applied, for example, to the diagnosis of ride comfort.

[0076] As illustrated in FIG. 9 , a reference space may also be created for each window (or for each running position, as described above) for each elevator 150. That is, for each of the multiple elevators 150, the individual creation unit 433 may generate multidimensional feature data from the sound / vibration waveform data at each of two or more windows (or for each of multiple running positions of the car) during the operation of the elevator 150, and create a reference space corresponding to the window (or the running position) from the multidimensional feature data. For abnormality diagnosis, the abnormality diagnosis unit 437 may compare the multidimensional feature data generated from the sound / vibration waveform data at the window (or the running position) with the reference space corresponding to the window (or the running position) among the reference spaces corresponding to the elevator 150. This is expected to improve the accuracy of abnormality diagnosis.

[0077] An example of a subdivision unit for each elevator 150 may be a combination of two or more of an elevator part, a running position, and a window. For example, for car parts such as the car door and a fan installed on the car, two or more windows are prepared that divide the time period from the start of the car's run to the end of the car's run by floor, and the unit may be a combination of a car part and a window. This makes it possible to identify information such as when an abnormality occurred, at which running position it occurred, and / or at which running position it occurred. Providing the identified information to maintenance personnel is expected to improve maintenance efficiency. Note that the above-mentioned "window" may be, for example, a division identified from the control data 211.

[0078] If a reference space is prepared for each subdivided unit (for example, each running position), many reference spaces are required according to the number of elevators and the subdivided units, and therefore a lot of data is required to create the reference spaces, which results in a large management burden, consumes a lot of memory capacity, and takes a long time to acquire the data.

[0079] Therefore, an appropriate narrowing of the window is adopted. Specifically, an appropriate running segment is adopted as the window. As shown in FIG. 10, the appropriate running segments are three running segments of the car: "acceleration" (Acc.), "steady" (Steady), and "deceleration" (Dec.). As shown in FIG. 10, two more running segments, "start of running" (Start) and "stop of running" (Stop), may be adopted as running segments. The above three running segments (or five running segments including the above two running segments) may be prepared for each of the car's "upward" (UP) and "downward" (DOWN), as shown in FIG. 10. In this case, ten reference spaces are prepared, as shown in FIG. 11. By preparing a reference space for each appropriately narrowed-down running segment in this way, it is possible to avoid acquiring a huge amount of data for creating the reference spaces.

[0080] Note that, for each of two or more windows, the abnormality diagnosis unit 437 may determine that an abnormality exists when the degree of deviation between the multidimensional feature amount data generated from the sound / vibration waveform data in that window and the reference space corresponding to that window exceeds a threshold value N times consecutively (N is an integer equal to or greater than 2). For example, in each window, the degree of deviation is acquired every second, and if a degree of deviation exceeding the threshold value is acquired N times consecutively in one window, it may be determined that an abnormality exists for that window.

[0081] The results of the abnormality diagnosis are recorded in the individual determination table 415, and the communication unit 438 refers to the individual determination table 415 and transmits visualized information of the abnormality diagnosis results to the maintenance staff terminal 122. As a result, the abnormality diagnosis results are displayed on the maintenance staff terminal 122, and as a result, the maintenance staff can be informed of the abnormality diagnosis results.

[0082] When abnormality diagnosis is performed in subdivided units for each elevator 150, the abnormality diagnosis results are also recorded in the individual judgment table 415 in subdivided units, and therefore the abnormality diagnosis results can be visualized in subdivided units.

[0083] 12, a UI (User Interface) 900 according to the visualization information transmitted to the maintenance staff terminal 122 includes a UI 901 for accepting the designation of an elevator part, a UI 902 for accepting the ID of the elevator 150, a UI 903 for indicating each floor, and a UI 904 for indicating a time series of the abnormality level (an example of an abnormality diagnosis result) of the designated part (the elevator part designated in UI 901) for each floor. In other words, the UI 904 displays the history of the relationship between the abnormality diagnosis result of the elevator 150 corresponding to the ID designated in UI 902 and the elevator part or window.

[0084] The floor where an abnormality is found is highlighted in UI 903. The "floor where an abnormality is found" may be a floor where an abnormality diagnosis result indicating an abnormality is found for a specified part, or a floor where the average abnormality level is equal to or greater than a predetermined value during the period to be displayed in UI 904 (for example, from June 1 to October 31).

[0085] In UI 904, the period to be displayed may be divided into one-month increments, or may be divided into more detailed or coarser increments than one-month increments. Furthermore, in UI 904, the degree of emphasis of the display may differ depending on the degree of abnormality. For example, the higher the degree of abnormality, the darker the color, and the lower the degree of abnormality, the lighter the color.

[0086] As shown in the example of FIG. 13, when the designated part changes in the UI 901, the display of the UI 903 may change, and the display content of the UI 904 may change based on the individual determination table 415.

[0087] 12 and 13, the communication unit 438 transmits visualized information representing the history of the relationship between the abnormality diagnosis result of the elevator 150 and the elevator part or window to the maintenance staff terminal 122. This allows the maintenance staff to grasp the abnormality of the elevator 150 in a subdivided unit such as the elevator part or window, and by understanding the history as well, it is possible to determine whether the abnormality is a sudden failure, a failure that has occurred over time, or a temporary malfunction that will not recur.

[0088] The communication unit 438 may receive, from the maintenance staff terminal 122, the diagnosis results of the elevator 150 by the maintenance staff for each elevator 150, and record the received diagnosis results in the work record table 421. The individual creation unit 433 may update the reference space corresponding to the elevator 150 based on the discrepancy between the abnormality diagnosis result of the elevator 150 by the maintenance staff (e.g., a result indicating whether each inspection item is normal or abnormal) and the abnormality diagnosis result of the elevator 150 by the abnormality diagnosis unit 437. That is, the abnormality diagnosis result of the elevator 150 by the maintenance staff may be set as the correct abnormality diagnosis result, and if the discrepancy between this correct abnormality diagnosis result and the abnormality diagnosis result by the abnormality diagnosis unit 437 is greater than or equal to a threshold, the individual creation unit 433 may update the reference space so as to reduce the discrepancy (so that the abnormality diagnosis result by the abnormality diagnosis unit 437 approaches the correct result). This can improve the accuracy of the reference space. The reference space may be updated for each elevator 150, or may be updated for each unit smaller than the elevator 150 (for example, for each elevator section or window).

[0089] Furthermore, the communication unit 438 may receive feedback on the elevator 150 from a user (or a manager) via the user terminal 123 (or the manager terminal 121). The feedback may be an evaluation of the ride comfort or an evaluation of elevator parts. The individual creation unit 433 may update the reference space corresponding to the elevator 150 based on the discrepancy between the feedback on the elevator 150 from the user or manager and the abnormality diagnosis result of the elevator 150 by the abnormality diagnosis unit 437. That is, the feedback (typically, an evaluation) on the elevator 150 from the user (or manager) may be taken as the correct abnormality diagnosis result, and if the discrepancy between this correct abnormality diagnosis result and the abnormality diagnosis result by the abnormality diagnosis unit 437 is greater than or equal to a threshold, the individual creation unit 433 may update the reference space so as to reduce the discrepancy (so that the abnormality diagnosis result by the abnormality diagnosis unit 437 approaches the correct result). This improves the accuracy of the reference space. The reference space may also be updated in units of elevators 150, or in units smaller than the elevators 150 (for example, for each elevator section or window).

[0090] Although one embodiment has been described above, this is merely an example for explaining the present invention, and the scope of the present invention is not limited to this embodiment. The present invention can be implemented in various other forms. [Explanation of symbols]

[0091] 103...Diagnostic equipment

Claims

1. A diagnostic device that acquires waveform data of sound and / or vibration related to an elevator to be diagnosed, generates multidimensional feature data from the waveform data, and diagnoses the condition of the elevator based on a result of comparing a reference space with the multidimensional feature data, an individual creation unit that generates multidimensional feature data for each of a plurality of elevators from sound / vibration waveform data acquired for the elevators in a normal state, and creates a reference space for multidimensional feature data for the elevators from the multidimensional feature data; an abnormality diagnosis unit that, for each of the plurality of elevators, compares multidimensional feature data generated from sound / vibration waveform data acquired for that elevator with a reference space prepared for that elevator from among reference spaces prepared for each elevator, and performs an abnormality diagnosis that is a diagnosis of an abnormality in that elevator based on the result of the comparison; A diagnostic device comprising:

2. a measurement and diagnosis unit that compares, for each of the plurality of elevators, a sound / vibration-related value acquired for the elevator in a non-operating state during a time period starting from the time of creating a reference space or diagnosing the elevator, with a failure threshold that is a threshold for sound / vibration for the elevator in a non-operating state, and determines whether the sound / vibration measurement for the elevator is successful or unsuccessful based on the result of the comparison; The diagnostic device according to claim 1, further comprising:

3. For each of the plurality of elevators, when a determination result of failure is obtained N times in succession (N is an integer equal to or greater than 2), the measurement and diagnosis unit updates the failure threshold value to reduce the possibility of obtaining a determination result of failure.

3. The diagnostic device according to claim 2.

4. for each of the plurality of elevators, elements defining a reference space corresponding to the elevator include a model type, a deviation threshold, and a feature set that define the reference space; the deviation threshold is a threshold of the deviation between the multidimensional feature amount and the reference space, The feature set is an m-dimensional feature (m is an integer of 2 or more) constituting a multidimensional feature. The diagnostic device according to claim 1 .

5. the abnormality diagnosis unit, for each of the plurality of elevators, performs an abnormality diagnosis for the elevator based on a temporal transition of a degree of deviation between a reference space corresponding to the elevator and multidimensional feature data generated from sound / vibration waveform data acquired for the elevator, during diagnosis of the elevator. The diagnostic device according to claim 1 .

6. for each of the plurality of elevators, the individual creation unit generates multidimensional feature data for one or more elevator parts or positions from the sound / vibration waveform data acquired for the elevator part or the position, and creates a reference space for the elevator part or the position from the multidimensional feature data; the abnormality diagnosis unit, for each of the plurality of elevators, performs an abnormality diagnosis of the elevator part based on a result of a comparison between multidimensional feature data generated from sound / vibration waveform data acquired for the elevator part or position of the elevator and a reference space corresponding to the elevator part or position during diagnosis of the elevator. The diagnostic device according to claim 1 .

7. the plurality of elevators are a plurality of elevators, for each of the plurality of elevators, the individual creation unit generates multidimensional feature data from the acquired sound / vibration waveform data regarding door opening and closing for each floor, and creates a reference space for door opening and closing for each floor from the multidimensional feature data; The abnormality diagnosis unit, for each of the plurality of elevators, in the abnormality diagnosis at the time of diagnosing the elevator, For each floor, the presence or absence of a door abnormality is determined based on the results of comparing multidimensional feature data generated from sound / vibration waveform data acquired regarding the door opening and closing on that floor with a reference space corresponding to the door opening and closing on that floor; If there is a door abnormality on at least one floor, If an abnormality is detected on all floors, it is determined that there is an abnormality in the car door, If an abnormality is detected on a part of floors, it is determined that there is an abnormality in the landing door on the part of floors. The diagnostic device according to claim 1 .

8. the abnormality diagnosis unit, for each of the plurality of elevators, performs an abnormality diagnosis of operation in each of two or more windows, which are two or more time intervals during operation of the elevator, based on a result of comparing multidimensional feature data generated from sound / vibration waveform data of the window with a reference space corresponding to the elevator, during diagnosis of the elevator; The diagnostic device according to claim 1 .

9. the abnormality diagnosis unit determines that an abnormality exists when, for each of the two or more windows, the degree of deviation between the multidimensional feature amount data generated from the sound / vibration waveform data in that window and the reference space corresponding to that window exceeds a threshold value N times consecutively (N is an integer equal to or greater than 2); The diagnostic device according to claim 8 .

10. the plurality of elevators are a plurality of elevators, For each of the plurality of elevators, the two or more windows are three windows: acceleration, constant speed, and deceleration windows for the elevator car; The diagnostic device according to claim 8 .

11. For each of the plurality of elevators, the two or more windows include, in addition to the three windows, two windows for starting and stopping the elevator car's movement. The diagnostic device according to claim 10.

12. For each of the plurality of elevators, the three windows are for each of the up and down car of the elevator; The diagnostic device according to claim 10.

13. for each of the plurality of elevators, when there is a reference space of another elevator that has been created more advancedly than the reference space of the elevator based on a distribution that is identical or similar to the distribution of the multidimensional feature values ​​of the elevator so far, the individual creation unit sets the reference space of the elevator to the reference space that is identical or similar to the reference space of the other elevator. The diagnostic device according to claim 1 .

14. A communication unit that communicates with an information processing terminal Equipped with the information processing terminal is an information processing terminal of a maintenance worker, the communication unit transmits visualized information representing an abnormality diagnosis result of the elevator to the information processing terminal of the maintenance worker. The diagnostic device according to claim 1 .

15. A communication unit that communicates with an information processing terminal Equipped with the information processing terminal is an information processing terminal of a maintenance worker, the communication unit transmits visualization information representing a history of a relationship between an elevator abnormality diagnosis result and an elevator part, a position, or a window to the information processing terminal of the maintenance worker.

10. The diagnostic device according to claim 6, wherein the diagnostic device is a diagnostic device for detecting a change in a state of a blood vessel.

16. A communication unit that communicates with an information processing terminal Equipped with the information processing terminal is an information processing terminal of a maintenance worker, the communication unit receives an elevator abnormality diagnosis result by the maintenance worker from the information processing terminal, the individual creation unit updates the reference space corresponding to the elevator based on a discrepancy between the result of the abnormality diagnosis of the elevator by the maintenance person and the result of the abnormality diagnosis of the elevator by the abnormality diagnosis unit. The diagnostic device according to claim 1 .

17. A communication unit that communicates with an information processing terminal Equipped with the information processing terminal is an information processing terminal of a user or an administrator, the communication unit receives elevator feedback from the user or the manager from the information processing terminal; the individual creation unit updates the reference space corresponding to the elevator based on a discrepancy between the elevator feedback from the user or the manager and the abnormality diagnosis result of the elevator by the abnormality diagnosis unit. The diagnostic device according to claim 1 .

18. an individual evaluation unit that, after creating a reference space for each of the plurality of elevators, evaluates the distribution of the reference space or the correlation between feature quantities; Equipped with for each of the plurality of elevators, the individual creation unit changes at least one of a model type, a deviation threshold, and a feature set associated with the reference space corresponding to the elevator, based on an evaluation result of the reference space corresponding to the elevator.

5. The diagnostic device according to claim 4.

19. A diagnostic method for diagnosing a state of an elevator to be diagnosed, comprising: acquiring waveform data of sound and / or vibration related to the elevator to be diagnosed; generating multidimensional feature data from the waveform data; and diagnosing a state of the elevator based on a result of comparing a reference space with the multidimensional feature data, For each of a plurality of elevators, multidimensional feature data is generated from the sound / vibration waveform data acquired for the elevator in a normal state, and a reference space for the multidimensional feature of the elevator is created from the multidimensional feature data; For each of the plurality of elevators, when diagnosing the elevator, the multidimensional feature data generated from the sound / vibration waveform data acquired for the elevator is compared with a reference space prepared for the elevator from among reference spaces prepared for each elevator, and an abnormality diagnosis is performed, which is a diagnosis of an abnormality in the elevator, based on the result of the comparison. A diagnostic method characterized by:

Images