JP2013200143A - Abnormal sound diagnosis device and abnormal sound diagnosis system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve the problem that variation due to difference in various conditions between determination sound data characteristics and sound data characteristics during diagnosis is erroneously detected as abnormality because differences in variation of time-frequency characteristics of sound of a vehicle as a determination object device among conditions such as temperature, atmosphere, humidity, speed, acceleration, pressure, tension, and load are not considered to determine abnormal sound by comparing a corrected determination value resulting from correcting an abnormal sound determination value of the vehicle by vehicle characteristics such as a size of the vehicle, irregularities of a road surface, an ambient temperature, and an atmosphere, with sound data of the vehicle.SOLUTION: A waveform of sound from a determination object device or an analysis result of the waveform is acquired as sound data by sound data acquisition means, and sensing data of one of temperature, atmosphere, humidity, speed, acceleration, pressure, tension, and load relating to the device is acquired by sensing data acquisition means, and reference sound data is generated on the basis of the sensing data by reference sound data generation means, and the reference sound data and the sound data acquired during diagnosis are compared with each other to detect abnormality of sound of the device by determination means.

Description

  The present invention relates to an abnormal sound diagnosis technique that collects sound generated from a determination target device and determines the possibility of occurrence of abnormal sound in a determination target device during operation by time-frequency analysis of the collected sound.

  Regarding the abnormal sound diagnosis apparatus, an abnormal sound diagnosis apparatus as shown in Patent Document 1 is known. The abnormal sound diagnosis apparatus disclosed in Patent Document 1 stores a determination value for determining an abnormal sound included in a sound generated by a vehicle as a determination target device, and relates to characteristics of the vehicle that is the determination target device. Accepts user settings, corrects stored judgment values based on accepted characteristics, and calculates sound pressure levels in a predetermined frequency band by frequency analysis of sound data obtained from the vehicle that is the judgment target device Then, the abnormal sound is determined by comparing the calculated sound pressure level with the corrected determination value. Here, as the setting by the user, the size of the vehicle that is the determination target device, the road surface and the unevenness, the temperature around the vehicle, the atmospheric pressure around the vehicle, and the like are disclosed. In addition, it is disclosed that the determination value is corrected so that the sound pressure level of the corrected determination value becomes higher as the temperature is lower, by setting the temperature around the vehicle by the user.

JP 2010-096547 A

  However, the conventional abnormal sound diagnosis device is based on various conditions such as temperature, atmospheric pressure, and humidity that are environmental conditions of the determination target device, and speed, acceleration, pressure, tension, and load that are operation conditions of the determination target device. The time frequency characteristics of the sound produced by the sound, especially the difference in changes at a specific time or frequency, is not considered, so the time frequency characteristics of the reference sound data and the time frequency characteristics of the input sound data at the time of diagnosis are There has been a problem that a change in a specific time and frequency due to a difference in conditions is erroneously detected as an abnormality.

An abnormal sound diagnosis apparatus for diagnosing abnormality of sound generated from a determination target device according to the present invention,
Sound data acquisition means for acquiring sound data generated from the determination target device or sound data that is an analysis result of the sound waveform;
Sensing data acquisition means for acquiring any of temperature, pressure, humidity, speed, acceleration, pressure, tension, and load related to the determination target device as sensing data;
Reference sound data generating means for generating reference sound data based on the sensing data;
A determination unit that detects an abnormality of a sound generated from the determination target device by comparing the reference sound data generated by the reference sound data generation unit and the sound data acquired by the sound data acquisition unit;
Is provided.

According to the abnormal sound diagnosis apparatus of the present invention, the environmental conditions of the determination target device such as temperature, atmospheric pressure, and humidity, and the speed, acceleration, and pressure that may change the time-frequency characteristics of the sound emitted by the determination target device. In order to take into account differences in the time-frequency characteristics of sound data, particularly at specific times and frequencies, due to various operating conditions of the target device such as tension and load,
Because the related sound data is generated based on the sensing data of any of the environmental conditions and driving conditions, and the related sound data of the sensing data most similar to the sensing data in the diagnostic mode is used as the reference sound data There is an effect of reducing the possibility of erroneously detecting a change in a specific time or frequency due to a difference in the above conditions as an abnormality.

It is a block block diagram which shows the abnormal sound diagnostic apparatus in Embodiment 1 of this invention. 6 is a block configuration diagram showing an abnormal sound diagnosis apparatus in Embodiment 2. FIG. FIG. 10 is a block configuration diagram showing an abnormal sound diagnosis apparatus in a third embodiment. FIG. 10 is a block configuration diagram showing an abnormal sound diagnosis apparatus in a fourth embodiment. FIG. 10 is a block configuration diagram showing an abnormal sound diagnosis apparatus in a fifth embodiment. 3 is a flowchart of processing from acquisition of a measurement signal to determination of a learning mode or a diagnosis mode in the first embodiment. 5 is a flowchart of processing in a diagnosis mode in the first embodiment. It is a data storage block diagram explaining the structure of a data storage part. 6 is an explanatory diagram of a predetermined partial region of a time frequency distribution in Embodiment 2. FIG. 14 is a flowchart of processing in a monitoring mode in the third embodiment. 3 is a flowchart of processing in a learning mode in the first embodiment.

Embodiment 1 FIG.
The present embodiment is implemented as software on a personal computer (hereinafter referred to as a PC) as a device for diagnosing abnormal sounds generated by a determination target device, and includes a learning mode for capturing a normal waveform and a waveform at the time of diagnosis. And a diagnostic mode for capturing. The measurer installs a sound collector such as a microphone, acoustic sensor, and acceleration sensor on the target device, and connects the sound collector such as a microphone, acoustic sensor, and acceleration sensor to the input terminal of the USB (Universal Serial Bus) interface of the PC. Then, the operation in the learning mode and the diagnosis mode is performed.

The device to be determined is, for example, a device made up of a plurality of operating parts such as an elevator. In the case of an elevator, the sound collector is installed in or out of the elevator car, and the signal from the sound collector is taken into the PC placed in the machine room via the control cable, and the car is reciprocated up and down. By driving, the operation sound of each device of the elevator is diagnosed.
Since the elevator is composed of a plurality of parts, the sound collected by the sound collector is a mixed sound from these parts. Moreover, the frequency characteristics of the sound generated from each component are different. In addition, since the position of the sound collector moves with time, the time frequency component of the time frequency component varies depending on the time.

Furthermore, the intensity of sound generated from each component shows different changes depending on the component due to differences in various conditions such as temperature, atmospheric pressure, humidity, speed, acceleration, pressure, tension, and load. For example, the rail noise generated when the rail and the guide shoe slide slides, and the viscosity of oil applied to the rail decreases as the temperature rises. Therefore, when the temperature rises, the friction decreases and the sound pressure decreases. On the contrary, the sound generated when the rope winds around the sheave does not show much correlation with temperature. In addition, the air-conditioning noise tends to increase in intensity as the fan speed increases with increasing temperature.
In this way, the time frequency distribution of the sound collected by the sound collector varies depending on the specific time and frequency of the time frequency characteristics due to changes in various conditions. Due to differences in conditions, the time-frequency characteristics of the sound collected by the sound collector may change at specific times and frequencies, and the change component due to these differences in conditions may be misjudged as a component due to abnormal sound. Becomes higher.

FIG. 1 is a block configuration diagram showing an abnormal sound diagnosis apparatus according to Embodiment 1 of the present invention.
In FIG. 1, 1 is a sound collector such as a microphone, an acoustic sensor, or a vibration sensor, 2 is a waveform acquisition unit that samples a signal from the sound collector 1 and converts it into a digital signal and outputs waveform data 3, and 4 is a waveform. A time frequency distribution 5 comprising spectral values indicating time and intensity with respect to time and frequency is obtained by performing time-frequency analysis on the waveform data 3 by fast Fourier transform (hereinafter referred to as FFT) while shifting the time window in the time direction by multiplying the data 3 with a time window. This is a time frequency analysis unit that outputs, and the waveform acquisition unit 2 and the time frequency analysis unit 4 form sound data acquisition means.
In the following description of the embodiment, a case will be described in which the time frequency distribution 5 output from the time frequency analysis unit 4 is sound data. The sound data may be waveform data or other feature quantities obtained by analyzing the waveform data.

  Reference numeral 101 denotes a sensor that is provided in the determination target device and detects any one of temperature, atmospheric pressure, humidity, speed, acceleration, pressure, tension, and load related to the determination target device. In this embodiment, the sensor 101 detects the temperature. Reference numeral 102 denotes a sensing data acquisition unit that processes a signal from the sensor 101 to acquire temperature sensing data 103, and 105 associates the time frequency distribution 5 that is sound data in the learning mode with the sensing data 103 in the learning mode. And a related sound data generation unit stored in the data storage unit 104.

  A selection unit 201 compares the sensing data 103 input in the diagnosis mode with the sensing data stored in the data storage unit 104, selects sound data having similar sensing data, and outputs the normal time frequency distribution 6. . This normal time frequency distribution 6 becomes reference sound data at the time of diagnosis, and in this embodiment, the related sound data generation unit 105, the data storage unit 104, and the selection unit 6 constitute a reference sound data generation unit.

  A time frequency distribution 5 output from the time frequency analysis unit 4 at the time of diagnosis is a time frequency distribution 7 at the time of diagnosis, and the normal time frequency distribution 6 is referred to to indicate the degree of possibility of occurrence of abnormal sound from the time frequency distribution. An abnormal time calculation unit 17 that calculates the degree of abnormality and outputs the degree of abnormality 16, and 17 is a determination unit that determines the possibility of occurrence of abnormal sound based on the degree of abnormality 16 and outputs a determination result 18.

The operation of the present embodiment will be described below with reference to the processing flowcharts of FIGS. 6 to 7 and 11 and the data storage configuration diagram of FIG.
In either the learning mode or the diagnostic mode, the waveform acquisition unit 2 acquires a measurement signal output from the sound collector 1, amplifies it, and performs AD conversion, thereby sampling and sampling a 16-bit linear PCM (with a sampling frequency of 48 kHz). Pulse signal modulation) is converted into waveform data 3 of a digital signal (step S1).
The time frequency analysis unit 4 extracts a frame from the waveform data 3 output from the waveform acquisition unit 2 while shifting the time window of 1024 points in the time direction at intervals of 16 ms, and performs frequency calculation on each frame by FFT calculation. Is obtained as time-frequency distribution 5 (step S2). Here, t is an index of time corresponding to the shift interval for shifting the analysis window, and f is an index indicating the frequency of the result of the FFT operation. The time t and the frequency f satisfy the relations 0 ≦ t ≦ T and 0 ≦ f ≦ F, respectively. Here, T is the number of frames in the time direction of the time frequency distribution 5, and F is an index indicating a frequency corresponding to the Nyquist frequency that is ½ of the sampling frequency fs of the waveform data 3 (F = fs / 2).

  The sensing data acquisition unit 102 processes the signal from the sensor 101 and acquires the sensing data 103 (step S3).

  When the sensing data 103 is acquired by the sensing data acquisition unit 102, the abnormal sound diagnosis apparatus determines whether it is in the learning mode or the diagnosis mode (step S4).

  In the learning mode, the related sound data generation unit 105 associates the sensing data 103 and the time frequency distribution 5 and stores them in the data storage unit 104 (step S101). Specifically, as shown in FIG. 8, the data storage unit 104 is configured to store data of the maximum Nmax items, with the index part and the data part as one related item. The number of items before the start of learning is 0, and the number of items increases by 1 each time learning progresses once. In the n-th learning, the sensing data 103 at the time of learning is stored in the n-th index part G (n). In the n-th learning, the time frequency distribution 5 at the time of learning is stored in the nth sound data part X (n). Thus, in each learning, the sensing data 103 and the time frequency distribution 5 are associated with each other and stored in the data storage unit 104.

Next, the operation of the diagnosis process in the diagnosis mode will be described with reference to FIG.
The time frequency distribution 5 is output as a diagnosis time frequency distribution 7 in the diagnosis mode (step S201).

  The selection unit 201 finds sound data having sensing data similar to the sensing data 103 at the time of diagnosis from the data storage unit 104, and outputs the sound data as the normal time frequency distribution 6 (step S202). Specifically, the similarity between the n-th sensing data G (n) stored in the index part of the data storage unit 104 and the sensing data F at the time of diagnosis is obtained, and the sensing data G (n *), The time frequency distribution X (n *) stored in the data storage unit 104 associated with G (n *) is obtained, and the time frequency distribution X (n *) is determined as the normal time frequency distribution 6. Output as.

  The abnormality degree calculation unit 15 receives the time frequency distribution 7 at diagnosis and the time frequency distribution 6 at normal time, calculates the difference between them, and outputs the difference as an abnormality degree 16 (step S203). Specifically, the degree of abnormality 16 is calculated from the respective spectrum values of the normal time frequency distribution x (t, f) and the diagnostic time frequency distribution y (t, f). That is, the degree of abnormality a (R) 16 is calculated as Equation (1) as a mean square value of the difference between the spectral values in a common specific region of each time frequency distribution 5.

Here, in Equation (1), Ψ (x) is a nonlinear mapping function of the variable x, and for example, logarithmic transformation or a generalized Box-Cox transformation (also called generalized logarithmic transformation) or the like can be used.
The region R is a set of a set {t, f} of the time t and the frequency f included in the time frequency region in which the degree of abnormality is calculated. The region R is determined in advance depending on the object to be diagnosed, and is determined based on prior knowledge about the time and frequency at which an abnormal component expected in advance appears for each diagnosis object. The region R may be different depending on the object to be diagnosed.

  The determination unit 17 determines whether or not there is a possibility of abnormal sound by comparing the abnormality degree 16 output from the abnormality degree calculation unit 15 with a predetermined threshold, and outputs the result as a determination result 18. (Step S204). Specifically, the abnormality level a (R) 16 calculated by the equation (1) is compared with a threshold value, and when the abnormality level 16 is equal to or higher than the threshold value, it is determined that there is a possibility that an abnormal sound is generated. “Abnormal” is output as the determination result 18. When the degree of abnormality 16 is less than the threshold, it is determined that there is a low possibility that abnormal sound has occurred, and “normal” is output as the determination result 18.

  As described above, according to the first embodiment, the time frequency distribution at the time of diagnosis is selected and used as the criterion for determination using the time frequency distribution obtained under conditions such as temperature that are most similar to the conditions at the time of diagnosis. Compared with the distribution, it is configured to determine the possibility of abnormal sounds occurring at the time of diagnosis, so even if conditions such as temperature at the time of diagnosis and learning change, it is caused by the change in conditions This has the effect of reducing the possibility of erroneous determination.

Embodiment 2. FIG.
In the present embodiment, a sound change model representing a correspondence relationship between the sensing data 103 and the time frequency distribution 5 that is sound data is obtained from the time frequency distribution, and a determination criterion is obtained from the sensing data and sound change model at the time of diagnosis. To generate sound data.

Only differences from the first embodiment will be described.
FIG. 2 is a configuration diagram of the second embodiment. In the figure, 202 is a learning means that is connected to the data storage unit 104 and outputs a sound change model 203, and 204 is a generation means that generates the normal time frequency distribution 6 from the sensing data 103 and the sound change model 203 at the time of diagnosis. .

  The data storage unit 104 stores n pieces of sensing data G (1: n) and sound data X (1: n) obtained by predetermined n times of learning. In this state, the learning unit 202 learns the sound change model 203. The sound change model 203 obtains the slope and intercept of the least square approximation line representing the relationship between temperature and logarithmic intensity for each predetermined partial region of the sensing data (temperature) and time frequency distribution. As the partial area, for example, as shown in FIG. 9, a rectangular area divided into N in the time direction and M in the frequency direction can be given.

  Assuming that the slope and intercept of the least square approximation line estimated in the partial region R (I, J) are a (I, J) and b (I, J), respectively, the relationship between temperature and intensity is expressed by the equation (2). ).

  The generation unit 204 substitutes the sensing data (temperature) ρ at the time of diagnosis into Expression (2) representing the sound change model, and obtains the intensity f (ρ | I, J) of the region R (I, J). The above processing is repeated / implemented for all the partial areas, the values of the entire time frequency distribution are formed, and output as the normal time frequency distribution 6.

  As described above, according to the second embodiment, a model representing the relationship between sensing data and intensity is estimated for each partial region of the time frequency distribution based on N sensing data and sound data associated with each other. However, since it is configured to generate a time-frequency distribution that is a criterion for determination based on sensing data at the time of diagnosis, even if sound data adjacent to the stored sensing data is not stored, it is appropriate The temporal frequency distribution can be used as a criterion for determination, and there is an effect that false detection due to a difference in measurement conditions is reduced.

Embodiment 3 FIG.
In the present embodiment, in the case of learning driving, when sound data similar to sensing data whose sensing data is already stored in the data storage unit 104 is already stored, the sound data is not added to the data storage unit 104. Thus, the number of data stored in the data storage unit 104 is reduced.

Only differences from the second embodiment will be described.
FIG. 3 is a configuration diagram of the third embodiment.
In the figure, 301 determines whether to add the sensing data 103 at the time of learning and the sound data at the time of learning to the data storage unit 104 with reference to the sensing data already stored in the data storage unit 104, and adds Is a determination adding means for adding sensing data and sound data to the data storage unit 104 in association with each other.

  As shown in the flow of processing in FIG. 10, learning operation is performed in the monitoring mode, and sensing data 103 is acquired (step S301).

Next, the degree of similarity between the acquired sensing data and the sensing data stored in the index part of the data storage unit 104 is obtained, and the sensing data having the highest degree of similarity is obtained. (Step S302).
Finally, the highest degree of similarity of the obtained sensing data is compared with a predetermined value, and if the highest degree of similarity is greater than or equal to a predetermined value, the process ends without doing anything. If the largest similarity is smaller than the predetermined value, the acquired sensing data and sound data are associated with each other and stored in the data storage unit 104.

  As described above, according to the third embodiment, as data stored in the data storage unit 104, sound data related to sensing data similar to the sensing data already stored in the data storage unit 104 is not added. Therefore, the number of data stored in the data storage unit 104 can be reduced.

Embodiment 4 FIG.
In this embodiment, when a plurality of devices are to be determined, in order to reduce the number of sensors installed, a management server is placed and environmental conditions are distributed from the management server to each device.

FIG. 4 is a configuration diagram of the fourth embodiment.
In the figure, 1001-1 to 1001-N are diagnostic units provided for each of the determination target devices 1 to N, and 1-1 to 1-N are installed in the determination target devices 1 to N, respectively. A sound collector 101 connected to ˜1001-N is a sensor installed only in the determination target device 1. A management server 3001 includes a sensing data acquisition unit 102 and a management unit 302. The management unit 302 is connected to each diagnostic unit via a network and manages the operation of each diagnostic unit 1001-1 to 1001-N. In addition, the sensing data acquisition unit 102 of the management server 3001 is connected to the sensor 101.

Here, each of the diagnosis units 1001-1 to 1001-N has the same configuration as that of the abnormal sound diagnosis apparatus as shown in FIG. However, as shown in FIG. 4, the sensor 101 is provided in the determination target device 1, and the sensing data acquisition unit 102 is provided in the management server 3001. The sensor 101 and the sensing data acquisition unit 102 are Shared by diagnostic units 1001-1 to 1001-N. Also, the sensing data 103, which is the output of the sensing data acquisition unit 102 shown in FIG. 4, is the abnormal sound diagnosis shown in FIG. 1 as the diagnostic units 1001-1 to 1001-N via the management unit 302 and the network. It is transmitted as sensing data 103 in the apparatus.
Each diagnosis unit 1001-1 to 1001-N issues a request to acquire the sensing data 103 from the sensor 101 to the management server 3001 in the learning mode and the diagnosis mode. On the other hand, in the management server 3001, based on a request from the diagnostic unit, the sensing data acquisition unit 102 acquires sensing data from the sensor 101, and transmits the sensing data 103 to the requested diagnostic unit. Finally, the diagnosis unit executes learning or diagnosis based on the sensing data 103 acquired from the management server 3001.

  As described above, according to the fourth embodiment, the number of sensors can be reduced because the sensor is installed in a part of the devices to be determined and is shared by the diagnostic units that are a plurality of abnormal sound diagnosis apparatuses. There is an effect.

Embodiment 5 FIG.
In this embodiment, when a plurality of devices of the same model, for example, a plurality of elevators installed in a building are targeted for determination, a management server is installed, and the management server associates sensing data with sound data. It memorizes and learns a sound change model based on the memorized data.

FIG. 5 is a configuration diagram of the fifth embodiment.
The management server 3001 includes at least a management unit 302, a data storage unit 104, a related sound data generation unit 105, and learning means 202. Each of the diagnostic units 1001-1 to 1001-N generates sound data from the signals of the sound collectors 1-1 to 1-N and acquires sensing data from the sensors 101-1 to 101-N. These generated / acquired sensing data and sound data are configured to be transmitted to the management server 3001.

  Here, each diagnosis unit has a configuration of an abnormal sound diagnosis apparatus as shown in FIG. 2, for example, but the data storage unit 104, the related sound data generation unit 105, and the learning means 202 are as shown in FIG. Are provided in the management server 3001. Also, the sound change model 203 output from the learning means 202 shown in FIG. 5 is transmitted as the sound change model 203 in the abnormal sound diagnosis apparatus shown in FIG. 2 as a diagnosis unit via the management unit 302 and the network. Is done.

  In the learning mode, each of the diagnostic units 1001-1 to 1001-N includes sound data generated by the waveform acquisition unit and the time frequency analysis unit from the signals acquired by the sound collectors 1-1 to 1-N, Sensing data acquired from each of the sensors 101-1 to 101-N and acquired by the sensing data acquisition unit is transmitted to the management server 3001. The management server 3001 associates the sound data transmitted from each diagnostic unit and the sensing data with the related sound data generation unit 105 and stores them in the data storage unit 104. When the number of sound data and sensing data from each diagnostic unit stored in the data storage unit 104 reaches a predetermined number, the management server 3001 uses the learning unit 202 to collect all the data of all the sound data from each diagnostic unit. Is used to learn the sound change model 203.

  In addition, each diagnosis unit 1001-1 to 1001-N requests a sound change model from the management server 3001 in the diagnosis mode. The management server 3001 transmits the sound change model based on the request of each diagnostic unit. Each diagnosis unit refers to the sound change model transmitted from the management server 3001, and obtains sound data acquired from each sound collector and generated by the waveform acquisition unit and the time frequency analysis unit, and each sensor in each sensing data acquisition unit. Diagnosis is performed based on the sensing data acquired from.

  As described above, according to the fifth embodiment, it is configured to learn a sound change model using sound data and sensing data associated with each other obtained from a plurality of devices. Since the number of data used for the sound can be increased, it is expected that the estimation accuracy of the sound change model is improved and the false detection is reduced.

  The abnormal sound diagnosis apparatus of the present invention may be used as a detection apparatus that detects an abnormal state in a device whose use conditions are greatly changed, for example, an elevator.

  DESCRIPTION OF SYMBOLS 1; Sound collector, 2; Waveform acquisition part, 3; Waveform data, 4; Time frequency analysis part, 5; Time frequency distribution, 6: Time frequency distribution at normal time, 7: Time frequency distribution at diagnosis, 15; Calculation unit, 16; degree of abnormality, 17; determination unit, 18; determination result, 101; temperature sensor, 102; sensing data acquisition unit, 103; sensing data, 104; data storage unit, 105; 201; Selection means; 202; Learning means; 203; Sound change model; 204; Generation means; 301; Judgment addition means; Management unit 302, 1001-1 to 1001-N;

Claims (7)

In the abnormal sound diagnosis device for diagnosing abnormal sound generated from the device to be determined,
A sound data acquisition means for acquiring a sound waveform generated from the determination target device or an analysis result of the sound waveform as sound data;
Sensing data acquisition means for acquiring, as sensing data, an output from a sensor that detects any of temperature, atmospheric pressure, humidity, speed, acceleration, pressure, tension, and load related to the determination target device;
Reference sound data generating means for generating reference sound data based on the sensing data;
A determination unit for detecting an abnormality of the sound generated from the determination target device by comparing the reference sound data generated by the reference sound data generation unit and the sound data acquired by the sound data acquisition unit;
An abnormal sound diagnosis apparatus comprising: The reference sound data generating means
Data storage means for associating and storing the sensing data and the sound data;
Selecting means for selecting sound data having sensing data similar to the sensing data acquired by the sensing data acquiring means from the data storage means, and outputting the selected sound data as reference sound data;
The abnormal sound diagnosis apparatus according to claim 1, further comprising: The reference sound data generating means
Data storage means for associating and storing the sensing data and the sound data;
Learning means for learning a sound change model representing a mapping relationship between sensing data and sound data from sensing data and sound data stored in association with each other in the data storage means;
Generating means for generating reference sound data based on the mapping relationship represented by the sound change model learned by the learning means by inputting the sensing data;
The abnormal sound diagnosis apparatus according to claim 1, further comprising:   The sensing data acquired by the sensing data acquisition means is compared with the sensing data stored in the data storage means, and the sound associated with the acquired sensing data whose change value is larger than a predetermined value. 4. The abnormal sound diagnosis apparatus according to claim 2, further comprising an adding unit that adds data to the data storage unit. There are a plurality of devices to be determined, and for each determination target device, sound data acquisition means for acquiring the sound waveform generated from the determination target device or the analysis result of the sound waveform as sound data;
Reference sound data generating means for generating reference sound data based on any sensing data of temperature, atmospheric pressure, humidity, speed, acceleration, pressure, tension, and load related to the determination target device acquired from the sensing data acquisition means;
A determination unit that detects an abnormality of a sound generated from the determination target device by comparing the reference sound data generated by the reference sound data generation unit and the sound data acquired by the sound data acquisition unit;
An abnormal sound diagnosis device comprising
Sensing data acquisition means for processing a signal from a sensor provided in any one of the plurality of determination target devices to obtain sensing data of the abnormal sound diagnosis device, and each abnormal sound diagnosis device are collectively managed. An abnormal sound diagnosis system comprising a management server having a management unit, wherein sensing data from the sensing data acquisition means is distributed to each abnormal sound diagnosis device via the management unit. An abnormal sound diagnosis system including an abnormal sound diagnosis device that determines abnormal sound during operation of a determination target device provided in each of a plurality of determination target devices, and a management server that manages each abnormal sound diagnosis device. And
The abnormal sound diagnosis device
Sound data acquisition means for acquiring the sound waveform generated from the determination target device or the analysis result of the sound waveform as sound data;
Sensing data acquisition means that is provided in each determination target device and acquires sensing data from a sensor that detects any of temperature, pressure, humidity, speed, acceleration, pressure, tension, and load related to the determination target device;
Generating means for generating reference sound data serving as a reference for abnormal sound determination in the diagnosis mode;
Based on the sound data acquired from each sound collector and the reference sound data from the generating unit, a diagnosis is performed, and a determination unit that detects an abnormality of the sound generated from the determination target device;
The management server inputs the sensing data from the sensing data acquisition unit and the sound data from the sound data acquisition unit via the management unit, which controls transmission / reception of signals between each abnormal sound diagnosis device and the management server Related sound data generating means for associating sensing data with sound data;
Data storage means for storing associated sensing data and sound data;
Learning means for learning a sound change model representing a mapping relationship between sensing data and sound data from sensing data and sound data stored in association with each other in the data storage means;
Further, the generation unit of the abnormal sound diagnosis apparatus requests the management server to distribute a sound change model of the learning unit to the management mode in a diagnosis mode, and the sensing data from the sensing data acquisition unit and a mapping represented by the sound change model An abnormal sound diagnosis system that generates reference sound data based on a relationship.   The abnormal sound diagnosis apparatus or abnormal sound diagnosis system according to claim 1, wherein the sensing data includes at least a temperature.

Images