Acoustic diagnosis device and acoustic diagnosis method

Abstract

An acoustic diagnosis device according to the present invention comprises: an operation sound collection unit that collects an operation sound of a facility; a facility start time point acquisition unit that identifies a start time of the facility; an operation section operation sound clipping unit that clips an operation sound for each operation section of the facility from the operation sound collected by the operation sound collection unit on the basis of the start time of the facility identified by the facility start time point acquisition unit and the predetermined operation section of the facility; a sound pressure level acquisition unit that acquires a sound pressure level for each frequency band that is predetermined for each operation section in regard to the operation sound for each operation section of the facility clipped by the operation section operation sound clipping unit; a deviation degree calculation unit that calculates a deviation degree of the sound pressure level for each operation section and each frequency band, acquired by the sound pressure level acquisition unit; and an abnormality detection unit that detects an abnormality of the facility on the basis of the deviation degree calculated by the deviation degree calculation unit.

Description

Acoustic diagnostic device and acoustic diagnostic method

[0001] The present invention relates to an acoustic diagnostic device and an acoustic diagnostic method.

[0002] Patent Document 1 describes a diagnostic device that diagnoses a diagnostic target by comparing an observation waveform observed from the diagnostic target with a reference waveform. Patent Document 2 describes a state estimation device that estimates a change in the state of equipment based on waveforms and frequency information detected by a plurality of sensors provided in the equipment. Patent Document 3 describes a diagnostic device that detects chatter of a rolling roll using the frequency of an acoustic signal detected by an acoustic sensor provided in a rolling mill.

[0003] Japanese Patent Application Laid-Open No. 2008-164490 International Publication No. 2019 / 017345 International Publication No. 2020 / 157818

[0004] In the devices described in Patent Documents 1 to 3, especially for equipment composed of a plurality of parts, there is a possibility that the abnormal part of the equipment cannot be accurately specified. Specifically, when the equipment is composed of a plurality of parts such as bearings, speed reducers, cranks, molds, rolls, and blocks, the timing of the operating sound generation differs depending on the parts. Therefore, if the timing and frequency of the abnormal sound generation are not considered, it is impossible to know from which part the abnormal sound is generated, and there is a possibility that the abnormal part of the equipment cannot be accurately specified.

[0005] Further, there is rolling equipment in a steel process as a group of equipment composed of a plurality of equipment. For example, hot rolling equipment is composed of various equipment such as a device for removing scale before rolling a sheet bar, a crop shear for cutting the crop at the head and tail ends of the sheet bar, a heating device for the sheet bar, a plurality of rolling mills provided thereafter, a large number of steel plate cooling facilities installed downstream of the rolling mill, and a coiler for winding up the rolled steel plate. In such a group of equipment composed of a plurality of equipment, in the devices described in Patent Documents 1 to 3, there is a possibility that it is impossible to accurately specify which equipment has an abnormality, as described above.

[0006] The present invention was made to solve the above problems, and its objective is to provide an acoustic diagnostic device and acoustic diagnostic method that can accurately identify abnormalities in each piece of equipment constituting a group of equipment, as well as abnormalities in the components constituting each piece of equipment.

[0007] The acoustic diagnostic device according to the present invention is an acoustic diagnostic device that identifies abnormal parts of equipment for a plurality of pieces of equipment constituting a group of equipment using acoustic signals during the operation of the equipment and start signals indicating the start of the equipment, and comprises: an operation sound collection unit that collects the operation sounds of the equipment; an equipment start time acquisition unit that acquires the start signals of the equipment and identifies the start time of the equipment based on the acquired start signals; an operation section operation sound extraction unit that extracts the operation sounds of the equipment for each operation section from the operation sounds collected by the operation sound collection unit based on the start time of the equipment identified by the equipment start time acquisition unit and predetermined operation sections of the equipment; a sound pressure level acquisition unit that acquires the sound pressure level for each predetermined frequency band for each operation section of the equipment extracted by the operation section operation sound extraction unit; a deviation degree calculation unit that calculates the deviation degree of the sound pressure levels for each operation section and each frequency band acquired by the sound pressure level acquisition unit; and an abnormality detection unit that detects abnormalities in the equipment based on the deviation degree calculated by the deviation degree calculation unit.

[0008] The deviation calculation unit may calculate the degree of deviation of the sound pressure levels for each operating section and each frequency band acquired by the sound pressure level acquisition unit from the sound pressure levels for each operating section and each frequency band acquired during the normal operation of the equipment.

[0009] It is preferable to include an abnormal part identification unit that identifies the abnormal part of the equipment based on the deviation degree calculated by the deviation degree calculation unit.

[0010] The abnormal part identification unit may identify the abnormal part of the equipment based on a pattern of deviations, which is composed of the deviations in each frequency band determined for each operating section.

[0011] The system includes a database unit that stores data necessary for collecting operating sounds of the equipment and identifying abnormal parts of the equipment. The abnormal part identification unit identifies the abnormal parts of the equipment by referring to a table containing information that shows the relationship between the operating interval and deviation degree patterns stored in the database unit and the abnormal parts of the equipment.

[0012] The table includes information showing the relationship between the operating interval, the deviation pattern, and the operating conditions of the equipment and the abnormal part of the equipment. The abnormal part identification unit may refer to the table and identify the abnormal part of the equipment, taking into account the operating interval, the deviation pattern, and the operating conditions of the equipment.

[0013] The system may include: a normal sound pressure level acquisition unit that acquires the sound pressure level for each predetermined frequency band for each operating section of the operating sound extracted by the operating section sound extraction unit; and a normal distribution determination and recording unit that classifies the sound pressure levels acquired by the normal sound pressure level acquisition unit for each piece of equipment, each operating section, and each frequency band, and records them as the sound pressure levels during normal operation of the equipment.

[0014] The normal distribution determination and recording unit may record a table for each piece of equipment that includes information showing the relationship between the deviation pattern, which is determined by the degree of deviation of the sound pressure level from the normal sound pressure level in each frequency band obtained for each operating interval, and the abnormal part of the equipment.

[0015] If the equipment involves rotational movement, the abnormal part identification unit may identify the abnormal part of the equipment by classifying the sound pressure level by the rotational order obtained by normalizing the frequency by the rotational speed.

[0016] The aforementioned group of equipment is a hot rolling mill, and may include a rolling mill reduction device, rolling rolls, rolling roll motor, rolling work roll shift device, scale removal device, roll coolant device, and strip coolant device.

[0017] The acoustic diagnostic method according to the present invention is an acoustic diagnostic method for identifying abnormal parts of equipment for a plurality of equipment constituting a group of equipment, using acoustic signals during equipment operation and start signals indicating the start of equipment, and includes: an operation sound collection step for collecting the operating sounds of the equipment; an equipment start time acquisition step for acquiring the start signals of the equipment and identifying the start time of the equipment based on the acquired start signals; an operation section operation sound extraction step for extracting the operating sounds of the equipment for each operation section from the operation sounds collected in the operation sound collection step, based on the start time of the equipment identified in the equipment start time acquisition step and predetermined operating sections of the equipment; a sound pressure level acquisition step for acquiring the sound pressure level for each predetermined frequency band for each operating section of the equipment extracted in the operation section operation sound extraction step; a deviation degree calculation step for calculating the deviation degree of the sound pressure level for each operating section and each frequency band acquired in the sound pressure level acquisition step; and an abnormality detection step for detecting an abnormality of the equipment based on the deviation degree calculated in the deviation degree calculation step.

[0018] The deviation calculation step may include a step to identify an abnormal part of the equipment based on the deviation calculated in the deviation calculation step.

[0019] The system may include: a normal sound pressure level acquisition step, in which the sound pressure levels for each predetermined frequency band are acquired for each operating section of the operating sound extracted in the operating section sound extraction step; and a normal distribution determination and recording step, in which the sound pressure levels acquired in the normal sound pressure level acquisition step are classified by each piece of equipment, each operating section, and each frequency band, and recorded as the sound pressure levels when the equipment is operating normally.

[0020] According to the acoustic diagnostic device and acoustic diagnostic method of the present invention, it is possible to accurately identify abnormalities in each piece of equipment constituting a group of equipment, as well as abnormalities in the components constituting each piece of equipment.

[0021] Figure 1 is a block diagram showing the configuration of an acoustic diagnostic device according to one embodiment of the present invention. Figure 2 is a flowchart showing the data recording process flow according to one embodiment of the present invention. Figure 3 is a flowchart showing the abnormality detection and abnormality location identification process flow according to one embodiment of the present invention. Figure 4 is a flowchart showing the time measurement process flow according to one embodiment of the present invention. Figure 5 is a diagram showing an example of a data table. Figure 6 is a diagram illustrating the abnormality location identification process using deviation patterns. Figure 7 is a diagram showing the sound pressure level distribution when the operating conditions of the rolling mill are not limited and when they are limited. Figure 8 is a diagram showing the change in frequency and rotational order with respect to the rotational speed of the rotating body. Figure 9 is a diagram showing the sound pressure levels under normal and abnormal conditions in the embodiment.

[0022] The configuration and operation of an acoustic diagnostic device, which is one embodiment of the present invention, will be described below with reference to the drawings.

[0023] [Configuration] First, with reference to Figure 1, the configuration of an acoustic diagnostic device, which is one embodiment of the present invention, will be described. Figure 1 is a block diagram showing the configuration of an acoustic diagnostic device, which is one embodiment of the present invention.

[0024] As shown in Figure 1, the acoustic diagnostic device 1, which is one embodiment of the present invention, is a device that identifies abnormal parts of each piece of equipment using acoustic signals during operation and sensor signals indicating the start of each piece of equipment that constitute the manufacturing equipment A to be diagnosed, and is composed of an information processing device such as a computer. Each piece of equipment that constitutes the manufacturing equipment A is composed of multiple parts, and examples of such manufacturing equipment A include a collection of equipment in the steelmaking process, such as a hot rolling mill, a cold rolling mill, and a continuous annealing mill. Examples of equipment that constitute the hot rolling mill include the rolling mill's reduction device, rolling rolls, rolling roll motor, rolling work roll shift device, scale removal device, roll coolant device, and strip coolant device.

[0025] The acoustic diagnostic device 1 includes an operating sound collection unit 1a, an equipment startup time acquisition unit 1b, an operating section operating sound extraction unit 1c, a normal sound pressure level acquisition unit 1d, a normal distribution determination and recording unit 1e, a sound pressure level acquisition unit 1f, a deviation degree calculation unit 1g, an abnormality detection unit 1h, and an abnormality location identification unit 1i. Each of these units is realized by a computing processing unit such as a CPU within the information processing device constituting the acoustic diagnostic device 1 executing a computer program. The functions of each of these units will be described later.

[0026] The acoustic diagnostic device 1 includes a database unit 1j. The database unit 1j is composed of a storage device such as an HDD that constitutes the information processing device, and includes an extraction condition database 1j1, a frequency band database 1j2, a normal distribution database 1j3, an operating condition database 1j4, and a deviation pattern database 1j5. Details of each database will be described later. The database unit 1j may be located outside the acoustic diagnostic device 1, and the acoustic diagnostic device 1 may be configured to acquire data from the database unit 1j via a telecommunications line.

[0027] The acoustic diagnostic device 1, having this configuration, detects abnormalities in equipment B, which constitutes manufacturing equipment A, and identifies the abnormal parts by performing the data recording process and abnormality detection / abnormal part identification process described below. The operation of the acoustic diagnostic device 1 when performing the data recording process and abnormality detection / abnormal part identification process will be described below.

[0028] [Data Recording Process] First, the operation of the acoustic diagnostic device 1 when performing data recording processing will be explained with reference to Figure 2. Figure 2 is a flowchart showing the flow of data recording processing, which is one embodiment of the present invention. The flowchart shown in Figure 2 starts when an execution command for data recording processing is input to the acoustic diagnostic device 1, and the data recording processing proceeds to step S1.

[0029] In step S1, the sound collection unit 1a collects acoustic signals of the operating sounds of multiple pieces of equipment B that constitute the manufacturing equipment A at predetermined time intervals, using acoustic sensors C (C1, C2, C3, ..., CN), such as microphones, installed in their vicinity, during a predetermined period of normal operation that includes the k types (k≧1) of operating sections of each piece of equipment that have been classified in advance. With this, the process of step S1 is completed, and the data recording process proceeds to step S2.

[0030] In step S2, the equipment start time acquisition unit 1b identifies the start time of each piece of equipment by using equipment signals or sensor signals that can determine the start time of each piece of equipment as trigger signals to know when each piece of equipment starts. Examples of signals that can determine the start time of each piece of equipment include, for a scale removal device using high-pressure water spray in a hot rolling mill, a spray start signal indicating the start of spray water discharge; for a rolling mill, a metal-in signal indicating the timing when the material to be rolled is loaded into the rolling mill; a work roll shift signal for shifting the work rolls of the rolling mill; a work roll peripheral speed or acceleration signal to know the acceleration timing of the work rolls; a roll coolant signal; a strip coolant signal, etc. The equipment start time acquisition unit 1b identifies the start time of each piece of equipment as the timing when these signals exceed predetermined values. With this, the process in step S2 is completed, and the data recording process proceeds to the process in step S3.

[0031] In step S3, the operating section sound extraction unit 1c extracts k types of acoustic signals for each operating section from the sound pressure signal collected in step S1. Specifically, the extraction condition database 1j1 records the association of each acoustic sensor C with the target equipment, and further records the operating section from which the acoustic signal is extracted for each acoustic sensor C. It also records the correspondence between each piece of equipment and the sensor signal or equipment signal used to determine the start time of each piece of equipment. The operating section sound extraction unit 1c refers to the extraction condition database 1j1 to determine the start time and extraction section for each acoustic signal from the start time of each piece of equipment identified in step S2, and extracts the acoustic signal based on the identified start time and extraction section. The start time of the operating section does not necessarily have to coincide with the start time of each piece of equipment; the operating section may be the time before or after the start time of the equipment. Furthermore, if a fixed time width cannot be determined in advance as the operating interval, the sensor signal for determining the operating interval and its determination criteria are recorded in the extraction condition database 1j1, and the operating interval operating sound extraction unit 1c extracts the acoustic signal by referring to this information. With this, the processing of step S3 is completed, and the data recording process proceeds to the processing of step S4.

[0032] The processes in steps S2 to S4 are executed sequentially each time sensor signals and acoustic signals related to equipment B are acquired, but they may be executed simultaneously for multiple pieces of equipment B if their start-up timings are simultaneous.

[0033] In step S4, the normal sound pressure level acquisition unit 1d first converts the acoustic signals for each operating section extracted in step S3 into sound pressure levels with respect to frequency using a Fourier transform, octave band filter, etc. Next, the normal sound pressure level acquisition unit 1d reads information (minimum and maximum frequencies, etc.) for l types (l≧1) of frequency bands used for anomaly detection, predetermined for each operating section, from the frequency band database 1j2. Then, the normal sound pressure level acquisition unit 1d uses a bandpass filter, etc., to acquire the sound pressure levels for each of the l types of frequency bands (sound pressure levels of normal operating sounds) from the converted sound pressure level data. With this, the process of step S4 is completed, and the data recording process proceeds to step S5.

[0034] In step S5, the normal distribution determination and recording unit 1e classifies the sound pressure levels of the normal operating sounds acquired in step S4 for each operating section and frequency band, and records the sound pressure level distribution of the normal operating sounds and its feature quantities (mean value, standard deviation, etc.) in the normal distribution database 1j3. With this, the process of step S5 is completed, and the series of data recording processes is finished.

[0035] In the data recording process, in addition to recording data such as the sound pressure level of normal operation sounds, a data table summarizing the relationship between deviation patterns from normal operation and abnormal parts, as described later, is pre-stored in the deviation pattern database 1j5. The deviation patterns and their registration details will be explained in the abnormality detection and abnormal part identification process below.

[0036] [Anomaly Detection and Abnormal Location Identification Processing] Next, the operation of the acoustic diagnostic device 1 when executing the anomaly detection and abnormal location identification processing will be explained with reference to Figures 3 to 6. Figure 3 is a flowchart showing the flow of the anomaly detection and abnormal location identification processing, which is one embodiment of the present invention. The flowchart shown in Figure 3 starts when an execution command for the anomaly detection and abnormal location identification processing is input to the acoustic diagnostic device 1, and the anomaly detection and abnormal location identification processing proceeds to step S11.

[0037] In step S11, the sound collection unit 1a collects acoustic signals of the operating sounds of multiple pieces of equipment B that constitute the manufacturing equipment A at predetermined time intervals, using acoustic sensors installed near each piece of equipment, within a predetermined period that includes the k types (k≧1) of operating sections of each piece of equipment that have been classified in advance. With this, the process of step S11 is completed, and the abnormality detection and abnormality location identification process proceeds to step S12.

[0038] In step S12, the equipment start time acquisition unit 1b identifies the start time of each piece of equipment using equipment signals or sensor signals that can determine the start time of each piece of equipment. The flow of this step S12 is the same as the flow of step S2 described above, so a detailed explanation is omitted. With this, the process of step S12 is completed, and the abnormality detection and abnormality location identification process proceeds to step S13.

[0039] In the process of step S13, the operation section operation sound extraction unit 1c extracts acoustic signals for each of the k types of operation sections from the acoustic signals collected in the process of step S11. In a hot rolling facility, a plurality of rolling mills are arranged in tandem, and the material to be rolled is sequentially loaded into each rolling mill and rolled. Hereinafter, referring to FIG. 4, a method for determining the time length, that is, the extraction condition, for extracting an acoustic signal for capturing the biting sound of the material to be rolled in adjacent rolling mills will be described.

[0040] FIG. 4 is a flowchart showing the flow of the time measurement process after metal-in at each rolling mill. In the time measurement process shown in FIG. 4, first, the operation section operation sound extraction unit 1c uses the detection of the metal-in signal at the rolling mill F1 as a trigger detection (step S21: Yes), and measures the time T1 when the metal-in signal is detected at the rolling mill F1 as the start time (step S22). Next, the operation section operation sound extraction unit 1c uses the detection of the metal-in signal at the rolling mill F2 as a trigger detection (step S23: Yes), and measures the time T2 when the metal-in signal is detected at the rolling mill F2 as the start time (step S24). Similarly hereinafter, the operation section operation sound extraction unit 1c measures the time when the metal-in signal is detected as the start time for all the rolling mills. When the time measurement process shown in FIG. 4 is completed, next, the operation section operation sound extraction unit 1c acquires information regarding the operation section (time length) Ts from the extraction condition database 1j1 until the operation sound (for example, reduction operation sound, roll coolant sound, work roll peripheral speed acceleration sound, etc.) is generated and disappears. Then, the operation section operation sound extraction unit 1c extracts the acoustic signal within the range of the operation section Ts from the time T (= T1, T2,...) measured in the time measurement process.

[0041] The start time of the operating interval does not necessarily have to coincide with the start time of each piece of equipment; the operating interval may be the time before and after the start time of the equipment. Also, the trigger signal may be set to capture a change in the state of the equipment (operating interval). For example, in hot rolling, the rolling speed may be accelerated midway through. In that case, the rotational speed of the rolling roll motor, such as the peripheral speed of the work roll, can be monitored, and its acceleration can be used as a trigger signal. By setting a predetermined threshold for the acceleration, the timing of the change in the acceleration state can be determined. In this case, a different operating interval can be set based on the acceleration sound of the rolling roll motor during acceleration and used for abnormality diagnosis. With this, the processing in step S13 is completed, and the abnormality detection and abnormality location identification process proceeds to the processing in step S14.

[0042] In step S14, the sound pressure level acquisition unit 1f acquires the sound pressure level for each of k operating sections and each of l frequency bands from the acoustic signal of the operating sound extracted in step S13. The flow of this step S14 is the same as the flow of step S4 described above, so a detailed explanation is omitted. The sound pressure level acquisition unit 1f may also read information on m types of operating conditions that affect the operation of each piece of equipment from the operating condition database 1j4 and acquire the sound pressure level for each of k operating sections, each of l frequency bands, and each of m types of operating conditions. With this, the processing of step S14 is completed, and the abnormality detection and abnormality location identification process proceeds to step S15.

[0043] In the process of step S15, the deviation degree calculation unit 1g reads out information on the sound pressure level distribution and its characteristic quantities during normal operation from the normal distribution database for each of the k types of operation sections and each of the l types of frequency bands. Next, the deviation degree calculation unit 1g uses the read information to calculate two or more deviation degrees of the sound pressure level of the operation sound obtained in the process of step S13 with respect to the sound pressure level during normal operation for each of the k types of operation sections and each of the l types of frequency bands. The deviation degree can be calculated, for example, by the following formula (1). The calculation formula for the deviation degree is not limited to formula (1), and depending on the characteristics of the detection signal, for example, the denominator may be a power such as the square of the standard deviation, or other adjustment coefficients may be multiplied throughout. Also, the sound pressure level measurement value of the target operation section and the target frequency band may be a statistical value such as the average value of the measurement values within the operation section.

[0044]

[0045] In the process of step S14, when considering m types of operating conditions, the deviation degree calculation unit 1g calculates the deviation degree by referring to a data table as shown in FIG. 5. The data table shown in FIG. 5 shows information on the sound pressure level distribution and its characteristic quantities during normal operation for each of the k types of operation sections, each of the l types of frequency bands, and each of the m types of operating conditions. The data table shown in FIG. 5 corresponds to the contents of the cut-out condition database 1j1, the frequency band database 1j2, the normal distribution database 1j3, and the operating condition database 1j4. Thereby, the process of step S15 is completed, and the abnormality detection / abnormality location identification process proceeds to the process of step S16.

[0046] In step S16, the anomaly detection unit 1h calculates the degree of deviation and determines that if the degree of deviation exceeds a predetermined threshold, it is abnormal (1), and otherwise it is normal (0). The threshold can be set based on, for example, 3σ, which is about three times the standard deviation σ, but it is not limited to this and can be adjusted and determined while looking at the characteristics of the operational data. The anomaly detection unit 1h uses the calculated degree of deviation to determine whether the equipment is normal (0) or abnormal (1) for each degree of deviation, and generates an array pattern vector y' of the determination results (0,1) for each of the l types of frequency bands in each operating section as the deviation pattern y'. With this, the processing of step S16 is completed, and the anomaly location identification process proceeds to step S17.

[0047] In step S17, the abnormal part identification unit 1i identifies the abnormal part of each piece of equipment constituting the manufacturing equipment A based on the deviation pattern y' generated in step S16. Specifically, as shown in Figures 6(a) and (b), the abnormal part identification unit 1i identifies the operating section and deviation pattern y' which are registered in the deviation pattern database 1j5 in advance. ｋ From the data table showing the relationship between the abnormal parts of each piece of equipment constituting manufacturing equipment A, deviation pattern y' is similar to deviation pattern y ｋ The system searches for abnormalities. The data table shown in Figure 6(b) records deviation patterns obtained when equipment malfunctions occur and organizes them in association with the abnormal parts. This data table is created in advance and stored in the deviation pattern database 1j5 by the normal distribution determination and recording unit 1e. In addition, the data table can be configured to identify abnormal parts based on deviation patterns that take into account k types of operating intervals and l types of frequency bands, as well as m types of operating conditions. Therefore, the abnormal part identification unit 1i may also identify abnormal parts of each piece of equipment constituting manufacturing equipment A by taking m types of operating conditions into account.

[0048] The abnormal area identification unit 1i uses the Euclidean distance to identify the deviation pattern y, as shown in the following formula (2). ｋ The similarity of the deviation pattern y' to is calculated, and similar deviation patterns y are found based on the calculated similarity. ｋ It is advisable to explore this. The similarity shown in formula (2) is given by the same result as the deviation pattern y' being the same as the deviation pattern yｋ The closer it is to 1, the more similar it is to the original. The abnormal area identification unit 1i then identifies the discovered deviation pattern y ｋ Information regarding the abnormal part of the equipment corresponding to is read out. For example, in the example shown in Figures 6(a) and (b), the deviation pattern y' and deviation pattern y generated in the process of step S14 are read out. ２ Since they match, the abnormal area identification unit 17 identifies the deviation pattern y ２ The corresponding part B is identified as the abnormal part of the equipment. This completes the process in step S15, and the series of abnormality detection and abnormal part identification processes are finished.

[0049]

[0050] As is clear from the above description, in the acoustic diagnostic device 1, which is one embodiment of the present invention, first, the operating sound collection unit 1a collects the operating sound of equipment B. Next, the equipment start time acquisition unit 1b acquires the start signal of equipment B and identifies the start time of equipment B based on the acquired start signal. Next, the operating section operating sound extraction unit 1c extracts the operating sound for each operating section of equipment B from the operating sound collected by the operating sound collection unit 1a, based on the start time of equipment B identified by the equipment start time acquisition unit 1b and the predetermined operating sections of equipment B. Next, the sound pressure level acquisition unit 1f acquires the sound pressure level for each predetermined frequency band for each operating section of equipment B, for the operating sound extracted by the operating section operating sound extraction unit 1c. Next, the deviation degree calculation unit 1g calculates the deviation degree of the sound pressure level for each operating section and each frequency band acquired by the sound pressure level acquisition unit 1f. Then, the abnormality detection unit 1h detects an abnormality in equipment B based on the deviation degree calculated by the deviation degree calculation unit 1g. This allows for the accurate identification of abnormalities in each piece of equipment that makes up the equipment group, as well as abnormalities in the components that make up each piece of equipment.

[0051] [Modification] In rolling mills, there are operations where the rotational speed of the rolling rolls changes as the rolling process progresses. The rolling rolls are rotated by the rolling motor, but the frequency of the operating noise generated changes depending on the rotational speed. Therefore, if abnormality diagnosis of the rolling mill is performed by classifying conditions by frequency band, it becomes necessary to classify conditions finely according to the rotational speed, which increases the number of deviation patterns and may hinder abnormality diagnosis. Accordingly, when the equipment to be diagnosed is equipment that involves rotational movement, it is better to classify conditions by rotational order (frequency / rotational speed), which is obtained by dividing the frequency by the rotational speed of the rolling motor (e.g., rpm), rather than by frequency. This normalizes the frequency change of the noise generated due to the rotational change, minimizes the number of classifications, and improves the accuracy of abnormality detection. Note that for equipment that involves specific rotation, a classification of rotational order can be used instead of a classification of each of the frequency bands.

[0052] [Example 1] In this example, acoustic sensors were installed around the rolling mill, and the sound of steel plates getting caught in the work rolls was extracted based on the metal-in signal, and a deviation pattern of the normal sound pressure distribution after frequency conversion was obtained. The sound extraction interval was set to t1 seconds before and t2 seconds after the time the metal-in signal was turned on, and sound with a data length of t1 + t2 seconds was acquired. In addition, times t1 and t2 were adjusted so that the jamming sounds of adjacent rolling mills were not mixed. Specifically, when extracting the jamming sound from rolling mill F1, times t1 and t2 were adjusted to a time when the jamming sound from rolling mill F2 was not included. The frequency band of the jamming sound was set to be within the lower and upper limits of the frequency in which the jamming sound exists, centered on the frequency at which the sound pressure when the steel plate gets caught is maximum. Figure 7(a) shows the sound pressure level distribution when there are no limitations on the operating conditions of the rolling mill, and Figure 7(b) shows the sound pressure level distribution when the operating conditions of the rolling mill are limited. As shown in Figures 7(a) and (b), it can be seen that by limiting the operating conditions, the number of coils with the assumed abnormal sound pressure level (83 dB) can be clearly separated. The operating conditions were limited to one of the following categories: steel type p (= 1, 2, 3, ...) category × reduction ratio q (= 1, 2, 3, ...) category. In addition, the change in sound pressure level associated with changes in steel type material and reduction ratio was investigated, and the locations where changes were observed were reflected as categories. However, the operating conditions may be divided into multiple categories. Then, work roll splitting was assumed by dividing into two deviation patterns, and the degree of deviation when the sound pressure level in a certain band becomes 83 dB was calculated. The following formula (3) was used to calculate the degree of deviation. As a result, the degree of deviation when there is no limitation of operating conditions was 1.7, and the degree of deviation when operating conditions are limited was 4.0, confirming that the accuracy of abnormality diagnosis is improved by limiting the operating conditions. Specifically, by setting the threshold for the degree of abnormality to be judged as abnormal to 4, abnormalities can be judged with high accuracy.

[0053]

[0054] [Example 2] In this example, the frequency of the sound pressure level that changes in response to the rotational speed of the rotating body and the changes in the sound pressure level in each frequency band were organized by rotational order. The rotational order was defined as the first rotational component, where one rotation is considered one period, and the nth rotational component is the nth rotational component. Figures 8(a) and 8(b) show the changes in frequency and rotational order with respect to the rotational speed of the rotating body. In the example shown in Figure 8(a), the frequency changes according to the rotational speed, but as shown in Figure 8(b), by converting to the rotational order, the rotational order does not change even if the rotational speed changes. Therefore, by monitoring the sound pressure level by rotational order rather than the sound pressure level with respect to frequency through rotational order conversion, equipment abnormalities can be detected accurately without increasing the number of abnormal patterns. Furthermore, by further classifying according to the rotational speed, even more accurate abnormality diagnosis can be performed.

[0055] [Example 3] In this example, an abnormality was detected in a high-pressure spray water supply system for scale removal of sheet bars, which is installed on the entry side of the finishing rolling mill of a hot rolling mill. In a high-pressure spray water supply system under normal operation, high-pressure spray water is injected in conjunction with the insertion of the sheet bar, and then an acoustic signal with a high sound pressure level is obtained for a predetermined time in a frequency band that depends on the diameter of the spray nozzle. Therefore, in this example, an acoustic signal was acquired using the signal for the start of high-pressure spray water injection in conjunction with the insertion of the sheet bar as a trigger signal, and sound pressure data for a predetermined time corresponding to the operating section for each predetermined piece of equipment was acquired, and the sound pressure level for each frequency band was obtained. As a result, as shown in Figure 9, in the initial normal operation, the sound pressure level around 3 kHz was high, but in the time period of 0.5 seconds or more when a minute hole occurred in the spray water piping, the sound pressure level around 8 kHz was also high. In this case, if we divide the frequency band into five bands as follows: 1.25 Hz or less, over 1.25 Hz and under 2.5 Hz, over 2.5 Hz and under 5 Hz, over 5 Hz and under 10 Hz, and over 10 Hz, then the sound pressure level vector under normal conditions will be

[00100] T (where T represents transposition), and the sound pressure level vector under abnormal conditions will be

[00110] T. Therefore, the deviation pattern will be

[00010] T, and the deviation pattern database 1j5 was searched to show that the abnormality corresponding to this deviation pattern is the occurrence of minute holes in the spray water piping of the scale removal device. The deviation pattern database 1j5 contained data in which the frequency band of the abnormal sound was changed according to the size of the minute holes.

[0056] Although embodiments applying the invention made by the present inventors have been described above, the present invention is not limited by the descriptions and drawings that constitute part of the disclosure of the present invention in this embodiment. That is, all other embodiments, examples, and operational techniques made by those skilled in the art based on this embodiment are included in the scope of the present invention.

[0057] According to the present invention, it is possible to provide an acoustic diagnostic device and an acoustic diagnostic method that can accurately identify abnormalities in each piece of equipment constituting a group of equipment, as well as abnormalities in the components constituting each piece of equipment.

[0058] 1 Acoustic diagnostic device 1a Operating sound collection unit 1b Equipment start time acquisition unit 1c Operating section operating sound extraction unit 1d Normal sound pressure level acquisition unit 1e Normal distribution determination and recording unit 1f Sound pressure level acquisition unit 1g Degree of deviation calculation unit 1h Anomaly detection unit 1i Anomaly location identification unit 1j Database unit 1j1 Extraction condition database 1j2 Frequency band database 1j3 Normal distribution database 1j4 Operating condition database 1j5 Deviance pattern database A Manufacturing equipment B, B1, B2, B3, BN Equipment C, C1, C2, C3, CN Acoustic sensor

Claims

1. An acoustic diagnostic device for identifying abnormal parts of equipment for a group of equipment, using acoustic signals during equipment operation and start signals indicating equipment startup, comprising: an operating sound collection unit for collecting operating sounds of the equipment; an equipment start time acquisition unit for acquiring the start signals of the equipment and identifying the start time of the equipment based on the acquired start signals; an operating section operating sound extraction unit for extracting operating sounds for each operating section of the equipment from the operating sounds collected by the operating sound collection unit, based on the start time of the equipment identified by the equipment start time acquisition unit and predetermined operating sections of the equipment; a sound pressure level acquisition unit for acquiring the sound pressure level for each operating section of the equipment extracted by the operating section operating sound extraction unit, for each predetermined frequency band; a deviation degree calculation unit for calculating the deviation degree of the sound pressure level for each operating section and each frequency band acquired by the sound pressure level acquisition unit; and an abnormality detection unit for detecting abnormalities in the equipment based on the deviation degree calculated by the deviation degree calculation unit.

2. The acoustic diagnostic apparatus according to claim 1, wherein the deviation degree calculation unit calculates the degree of deviation of the sound pressure levels for each operating section and each frequency band acquired by the sound pressure level acquisition unit from the sound pressure levels for each operating section and each frequency band acquired during the normal operation of the equipment.

3. The acoustic diagnostic apparatus according to claim 1 or 2, further comprising an abnormality location identification unit that identifies an abnormal location of the equipment based on the deviation degree calculated by the deviation degree calculation unit.

4. The acoustic diagnostic device according to claim 3, wherein the abnormal location identification unit identifies the abnormal location of the equipment based on a pattern of deviations, the deviations in each frequency band determined for each operating section.

5. The acoustic diagnostic device according to claim 3 or 4, comprising a database unit that stores data necessary for collecting operating sounds of the equipment and identifying abnormal parts of the equipment, wherein the abnormal part identification unit identifies abnormal parts of the equipment by referring to a table containing information showing the relationship between the operating interval and the deviation degree pattern stored in the database unit and the abnormal parts of the equipment.

6. The acoustic diagnostic device according to claim 5, wherein the table includes information showing the relationship between the operating interval, the deviation pattern, and the operating conditions of the equipment and the abnormal part of the equipment, and the abnormal part identification unit identifies the abnormal part of the equipment by referring to the table and considering the operating interval, the deviation pattern, and the operating conditions of the equipment.

7. An acoustic diagnostic device according to any one of claims 1 to 6, comprising: a normal sound pressure level acquisition unit that acquires the sound pressure level for each predetermined frequency band for each operating section of the operating sound extracted by the operating section operating sound extraction unit; and a normal distribution determination and recording unit that classifies the sound pressure levels acquired by the normal sound pressure level acquisition unit for each piece of equipment, each operating section, and each frequency band, and records them as the sound pressure level when the equipment is operating normally.

8. The acoustic diagnostic device according to claim 7, wherein the normal distribution determination and recording unit records a table for each piece of equipment that includes information showing the relationship between a pattern of deviations, which is an element of the degree of deviation of the sound pressure level from the sound pressure level during normal operation in each frequency band determined for each operating interval, and the abnormal part of the equipment.

9. The acoustic diagnostic device according to any one of claims 3 to 6, wherein the abnormal part identification unit identifies an abnormal part of the equipment by classifying the sound pressure level by the rotational order obtained by normalizing the frequency by the rotational speed when the equipment is equipment that involves rotational movement.

10. The acoustic diagnostic device according to any one of claims 1 to 9, wherein the equipment group is a hot rolling equipment, and the equipment includes a rolling mill reduction device, rolling rolls, rolling roll motor, rolling work roll shift device, scale removal device, roll coolant device, and strip coolant device.

11. An acoustic diagnostic method for identifying abnormal parts of equipment for a group of equipment, using acoustic signals during equipment operation and start signals indicating equipment startup, comprising: an operation sound collection step for collecting the operating sounds of the equipment; an equipment start time acquisition step for acquiring the start signals of the equipment and identifying the start time of the equipment based on the acquired start signals; an operation section operation sound extraction step for extracting the operating sounds of the equipment for each operation section from the operation sounds collected in the operation sound collection step, based on the start time of the equipment identified in the equipment start time acquisition step and predetermined operating sections of the equipment; a sound pressure level acquisition step for acquiring the sound pressure level for each predetermined frequency band for each operating section of the equipment extracted in the operation section operation sound extraction step; a deviation degree calculation step for calculating the deviation degree of the sound pressure level for each operating section and each frequency band acquired in the sound pressure level acquisition step; and an abnormality detection step for detecting an abnormality of the equipment based on the deviation degree calculated in the deviation degree calculation step.

12. The acoustic diagnostic method according to claim 11, further comprising a step of identifying an abnormal part of the equipment based on the deviation calculated in the deviation calculation step.

13. The acoustic diagnostic method according to claim 11 or 12, comprising: a normal sound pressure level acquisition step, which acquires the sound pressure level for each predetermined frequency band for each operating sound extracted in the operating sound extraction step for each operating sound section; and a normal distribution determination and recording step, which classifies the sound pressure levels acquired in the normal sound pressure level acquisition step for each piece of equipment, each operating section, and each frequency band, and records them as the sound pressure levels when the equipment is operating normally.

Images