Condition detection device and condition detection system

Abstract

To provide a condition detection device and condition detection system, which improve detection accuracy of various conditions including abnormality of diagnosis targets when performing diagnosis on facilities, such as wind power generation equipment, or components thereof such as blades.SOLUTION: A condition detection device 10 is provided, comprising an analysis unit 12 for detecting the condition of a detection target. The analysis unit 12 acquires acoustic information corresponding to sound generated by the detection target, detects the condition of the detection target on the basis of a pattern included in an image showing temporal change in a frequency component of the acoustic information, and outputs a detection result of the condition of the detection target.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present disclosure relates to a condition detection device and a condition detection system. [Background technology]

[0002] Conventionally, when determining abnormalities in wind turbine blades, a method is known in which the presence of Doppler shift components, in which frequency components with high sharpness change over time, is detected in the analysis results of the acoustic information (wind noise) emitted by the blades (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2010-281279 Summary of the Invention [Problem to be solved by the invention]

[0004] When diagnosing equipment such as a wind power generation system or its constituent elements such as blades, it is necessary to detect various conditions such as abnormalities of the equipment to be diagnosed.

[0005] The present disclosure has been made in consideration of the above points, and aims to provide a condition detection device and a condition detection system that can improve the accuracy of detecting various conditions, such as abnormalities, of the equipment being diagnosed when diagnosing equipment such as a wind power generation device or its components such as blades. [Means for solving the problem]

[0006] A condition detection device according to some embodiments includes an analysis unit that detects the condition (such as an abnormality) of a detection object (such as the blades of a wind turbine generator). The analysis unit acquires measurement results of sound generated by the detection object as acoustic information, detects the condition of the detection object based on a pattern contained in an image that represents time-varying frequency components of the acoustic information, and outputs the detection result of the condition of the detection object. By detecting the condition of the detection object as described above, the condition detection device can improve the accuracy of detecting various conditions, such as an abnormality, of the detection object when diagnosing equipment such as a wind turbine generator or its component blades.

[0007] In one embodiment of the condition detection device, the analysis unit may analyze a part of the detection object where the sound represented by the pattern is generated. This makes it easier to inspect or repair the analyzed part. As a result, the safety of the detection object is improved.

[0008] In one embodiment of the condition detection device, the analysis unit may output that the condition of the part where the sound represented by the pattern is generated is abnormal if a numerical value representing the difference between the pattern and a normal pattern is equal to or greater than a difference threshold. This clarifies the criteria for determining the condition of the detection target. As a result, the accuracy of condition detection is improved. Furthermore, this threshold may be set not only to a fixed value, but also to a value that utilizes the operating state or wind conditions. This also improves the accuracy of condition detection.

[0009] In one embodiment of the condition detection device, the analysis unit may detect the condition of the detection object based on a database that associates the patterns with the conditions of the detection object. The analysis unit may generate the normal pattern based on acoustic information when the condition of the detection object is normal and register the normal pattern in the database. This improves the accuracy of pattern classification. As a result, the accuracy of detecting the condition of the detection object is improved.

[0010] A condition detection system according to some embodiments may include the condition detection device and a sound collection device that measures sounds generated by a detection target and outputs the measurement results as acoustic information to the condition detection device. By including the condition detection device, the condition detection system can improve the accuracy of detecting various conditions, such as abnormalities, of a target when diagnosing equipment such as a wind power generation device or its component blades.

[0011] In one embodiment of the condition detection system, the sound collection device may include a plurality of microphones. The sound collection device may generate acoustic information representing sound components arriving at the sound collection device from a predetermined direction based on sounds detected by each of the plurality of microphones. This reduces the influence of sounds generated outside the detection target, such as background sounds. As a result, the accuracy of detecting the condition of the detection target is improved.

[0012] In one embodiment of the condition detection system, the analysis unit of the condition detection device may acquire a plurality of pieces of acoustic information representing components of sound arriving at the sound collection device from a plurality of directions. The analysis unit may detect the condition of the detection target based on the plurality of pieces of acoustic information. By detecting the condition of the detection target based on the plurality of pieces of acoustic information, the influence of sounds generated outside the detection target, such as background sounds, can be reduced. As a result, the accuracy of detecting the condition of the detection target is improved. [Effects of the Invention]

[0013] According to the present disclosure, a condition detection device and a condition detection system are provided that can improve the accuracy of detecting various conditions, such as abnormalities, of the equipment being diagnosed when diagnosing equipment such as a wind power generation equipment or its components such as blades. [Brief explanation of the drawings]

[0014] [Figure 1] FIG. 10 is a diagram showing an example of analysis of acoustic information according to a comparative example. [Figure 2] FIG. 1 is a schematic diagram illustrating an example of the configuration of a state detection system according to an embodiment. [Figure 3]1 is a block diagram illustrating an example of the configuration of a state detection system according to an embodiment. [Figure 4] FIG. 1 is a diagram showing an example of time variation of frequency components of a sound as a grayscale image. [Figure 5] 1 is a diagram showing, as a grayscale image, an example of a change over time in frequency components of acoustic information corresponding to components of a sound arriving from a first direction. [Figure 6] 10 is a diagram showing, as a grayscale image, an example of a change over time in frequency components of acoustic information corresponding to components of sound arriving from a second direction. FIG. [Figure 7] FIG. 10 is a diagram showing an example of a periodic pattern contained in a grayscale image. [Figure 8] FIG. 10 is a diagram showing an example of a granular pattern included in a grayscale image. [Figure 9] FIG. 10 is a diagram illustrating an example of a linear pattern included in a grayscale image. [Figure 10] 10 is a flowchart illustrating an example of a procedure of a state detection method according to an embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0015] (Comparative Example) The system according to the comparative example detects wind noise generated by the rotating blades of a wind turbine using a microphone installed at a fixed position. As shown in the graph in FIG. 1, the sound is expressed as a change in frequency components over time. The horizontal axis of the graph in FIG. 1 represents time, and the vertical axis represents frequency. The frequency components of the sound of the wind turbine blades are expressed as region 90. Furthermore, if the blades have characteristic features such as scratches, the sound generated at those characteristic features is distinguished as a distinct region within region 90.

[0016] Here, the rotating wind turbine blades have a speed relative to a fixed microphone. When the sound source moves along with the blades, the abnormal unevenness of the blades changes the airflow, generating sound that is detected by the microphone along with the sound of the blades. The system according to the comparative example determines a blade abnormality by detecting the presence of a component in the acoustic signal in which a frequency component with high sharpness shifts over time (Doppler shift). The system according to the comparative example also determines a blade abnormality when the shift in the frequency component with high sharpness has a predetermined slope.

[0017] However, the system according to the comparative example can only detect a state in which the sharpness of a frequency component in an acoustic signal appears as a time transition or gradient.

[0018] Therefore, the present disclosure describes a state detection system 1 (see FIGS. 2 and 3) that can detect various states of a detection object even if it cannot detect the time progression or slope of high-sharp frequency components in an acoustic signal based on the sound generated by the detection object, such as the blades 42 of a wind turbine 40 (see FIG. 2).

[0019] (Embodiments of the present disclosure) As shown in FIGS. 2 and 3 , a condition detection system 1 according to one embodiment includes a condition detection device 10, a sound collection device 20, and a display device 30. The condition detection system 1 detects the condition of a detection target based on sound generated by the detection target. In this embodiment, the detection target is assumed to be a blade 42 of a wind turbine 40 shown in FIG. 2. In addition to the blade 42, the wind turbine 40 includes a nacelle 44 and a support 46. In this embodiment, it is assumed that an abnormality has occurred in the blade 42. The abnormality occurring in the blade 42 may include various aspects such as scratches, cracks, breaks, protrusions, or depressions on the surface of the blade 42, or cavities, cracks, or breaks inside the blade 42. The condition detection system 1 may detect an abnormality occurring in the blade 42 using the blade 42 as the detection target. The portion of the blade 42 where the abnormality has occurred is also referred to as an abnormal portion 48.

[0020] (Configuration example of condition detection system 1) An example of the configuration of the condition detection system 1 will be described below.

[0021] <Status detection device 10> The condition detection device 10 detects the condition of the blade 42 based on acoustic information corresponding to the waveform of sound generated by the blade 42, which is the detection target. The detection target is not limited to the blade 42, but may include other parts such as the nacelle 44. The condition detection device 10 may detect the condition of the blade 42 without based on the frequency components of the acoustic information or the time changes in the frequency components of the acoustic information. The condition detection device 10 may acquire acoustic information, the frequency components of the acoustic information, or the time changes in the frequency components of the acoustic information from a sound collection device 20, which will be described later. The condition detection device 10 may acquire the frequency components of the acoustic information or the time changes in the frequency components of the acoustic information by acquiring the acoustic information from the sound collection device 20 and performing frequency analysis of the acoustic information. The condition detection device 10 includes an analysis unit 12, a storage unit 14, and an interface 16.

[0022] The analysis unit 12 controls each component of the state detection device 10. The analysis unit 12 may be configured to include a processor such as a CPU (Central Processing Unit). The analysis unit 12 may realize a predetermined function by causing the processor to execute a predetermined program.

[0023] The storage unit 14 may store various types of information used in the operation of the analysis unit 12, or programs for realizing the functions of the analysis unit 12. The storage unit 14 may function as a work memory for the analysis unit 12. The storage unit 14 may be configured, for example, as a semiconductor memory. The storage unit 14 may be configured to include a volatile memory or a non-volatile memory. The storage unit 14 may be configured as a non-transitory computer-readable storage medium. The storage unit 14 may be included in the analysis unit 12.

[0024] The interface 16 includes a communication device that communicatively connects the state detection device 10 to the sound collection device 20 or the display device 30. The communication device may be configured to be capable of communication based on a mobile communication standard such as 4G (4th Generation), LTE (Long Term Evolution), or 5G (5th Generation). The communication device may be configured to be capable of communication based on a LAN (Local Area Network) communication standard. The communication device may be configured to be capable of wired or wireless communication.

[0025] The interface 16 may be configured to include a display device. The display device may include various displays such as a liquid crystal display. The interface 16 may be configured to include an audio output device such as a speaker. The interface 16 is not limited to these, and may be configured to include various other output devices.

[0026] The interface 16 may be configured to include an input device that accepts input from a user. The input device may include, for example, a keyboard or physical keys, a touch panel or touch sensor, or a pointing device such as a mouse. The input device is not limited to these examples and may include various other devices.

[0027] The state detection device 10 may be configured as a PC (Personal Computer) or as at least one server device. The state detection device 10 may also be realized in a cloud computing system.

[0028] The condition detection device 10 may be configured to allow input of parameters and the like for detecting the condition of a detection object including, for example, the blade 42 and the like.

[0029] <Sound pickup device 20> The sound collection device 20 includes one or more microphones. The sound collection device 20 detects sounds generated from the detection target or its surrounding environment using one or more microphones. In other words, the sound collection device 20 detects sounds generated in the environment in which the detection target is present using one or more microphones. The sound collection device 20 outputs the detection results of the sounds detected by the one or more microphones to the state detection device 10 as acoustic information. The acoustic information may be expressed as a waveform representing the change in sound pressure over time. The acoustic information may also be expressed by the amplitude and phase of the sound pressure.

[0030] When the sound collection device 20 includes multiple microphones, the multiple microphones may be configured as an array microphone arranged in an array. The sound collection device 20 may output measurement results of sound waveforms detected by each microphone constituting the array microphone as acoustic information to the state detection device 10. The state detection device 10 may generate acoustic information corresponding to the waveform of a sound component arriving at the sound collection device 20 from a predetermined direction by analyzing the amplitude and phase of the waveform of the measurement results of each microphone acquired as acoustic information. For example, the state detection device 10 may generate acoustic information corresponding to the waveform of a sound component arriving from a predetermined part of the detection target. The direction from the predetermined part of the detection target toward the sound collection device 20 is also referred to as the first direction. Acoustic information representing the sound component arriving at the sound collection device 20 from the first direction is also referred to as first acoustic information. The state detection device 10 may generate acoustic information corresponding to the waveform of a sound component arriving from a predetermined part other than the detection target. The direction from the predetermined part other than the detection target toward the sound collection device 20 is also referred to as the second direction. The acoustic information representing the components of the sound arriving at the sound collection device 20 from the second direction is also referred to as second acoustic information. The state detection device 10 may generate acoustic information corresponding to the waveforms of the components of the sound arriving from each of the first direction and the second direction. The state detection device 10 may generate a plurality of pieces of acoustic information corresponding to the waveforms of the components of the sound arriving from each of a plurality of directions.

[0031] The sound collection device 20 may be arranged around the support 46 of the wind turbine 40 that is the detection target, as exemplified in Fig. 2. The sound collection device 20 may be arranged near the point where an extension of a line (vertical line) extending from the rotation axis of the blade 42 to the tip of the blade 42 intersects with the ground when the blade 42 has rotated to its lowest position. The sound collection device 20 is not limited to the position exemplified in Fig. 2, and may be arranged in various other positions, such as the nacelle 44. The sound collection device 20 may be arranged near the point where a line extending vertically from the tip of the blade 42 intersects with the ground when the tip of the blade 42 has rotated from the rotation axis to a horizontal position.

[0032] <Display device 30> The display device 30 displays the detection result of the state of the detection target by the state detection device 10. The display device 30 may display various data such as acoustic information, frequency components of the acoustic information, or changes over time in the frequency components of the acoustic information. The display device 30 may include various displays such as a liquid crystal display. The state detection system 1 may also include a speaker that outputs the sound measured by the sound collection device 20 itself, or a sound such as a warning sound generated in accordance with the detection result of the state of the detection target.

[0033] (Example of operation of condition detection system 1) In the condition detection system 1, the condition detection device 10 can detect the condition of the detection target based on acoustic information corresponding to the waveform of a sound generated in the environment in which the detection target exists. An example of the configuration of the condition detection system 1 will be described below.

[0034] <Analysis of acoustic information> The analysis unit 12 of the state detection device 10 may digitize the sound pressure of each frequency of the acoustic information using a method such as FFT (Fast Fourier Transform) or 1 / 1 or 1 / 3 octave analysis. The analysis unit 12 may digitize the amount of change in sound pressure at each frequency. The analysis unit 12 may calculate the frequency at which the sound pressure reaches a maximum value within a specified fixed time width as the peak frequency. By tracking the peak frequency, the analysis unit 12 may calculate the duration of the peak, the amount of movement of the peak frequency, the magnitude of the sound pressure at the peak frequency, and the degree of dispersion of the sound pressure in the frequency band including the peak frequency. The analysis unit 12 may store the analysis results of the acoustic information together with the acoustic information in the storage unit 14.

[0035] The analysis unit 12 may detect the state of the detection target based on changes over time in the frequency components of the acoustic information. The changes over time in the frequency components of the acoustic information may be represented as an image including pixels that represent, in grayscale, the magnitude of a certain frequency component of the acoustic information at a certain time, as illustrated in FIG. 4 . Each column of pixels (pixels aligned vertically) included in the image illustrated in FIG. 4 represents the magnitude of each frequency component of the acoustic information at the same time. Each row of pixels (pixels aligned horizontally) included in the image illustrated in FIG. 4 represents the magnitude of each frequency component of the acoustic information at the same time. In the image illustrated in FIG. 4 , the closer to black (darker) the color of a point representing a certain frequency component at a certain time, the greater the sound pressure of that frequency component at that time. Conversely, the closer to white (lighter) the color of a point representing a certain frequency component at a certain time, the greater the sound pressure of that frequency component at that time. In other words, the image illustrated in FIG. 4 essentially represents, in a three-dimensional graph, the relationship between time, the frequency of the acoustic information, and the magnitude of each frequency component at each time.

[0036] The grayscale image illustrated in Fig. 4 represents the change over time in the frequency components of acoustic information that combines sounds arriving from various directions toward the sound collection device 20 in an environment where the blade 42 is present as a detection target. In other words, the grayscale image illustrated in Fig. 4 represents the change over time in the frequency components of acoustic information detected by the sound collection device 20 without directionality. The change over time in the frequency components of acoustic information is not limited to being represented by a grayscale image, and may be represented by an RGB image or in various other ways.

[0037] As described above, the sound collection device 20 detects sound using a microphone array, thereby generating acoustic information corresponding to the waveform of a sound component arriving from a specific direction. In other words, the sound collection device 20 can detect sound in a directional manner and generate acoustic information.

[0038] The grayscale image illustrated in Fig. 5 represents the change over time in the frequency components of acoustic information corresponding to the waveform of the components of sound arriving at the sound collection device 20 from a first direction. The grayscale image illustrated in Fig. 6 represents the change over time in the frequency components of acoustic information corresponding to the waveform of the components of sound arriving at the sound collection device 20 from a second direction. In the grayscale images illustrated in Figs. 5 and 6, the pixels in each column (pixels lined up in the vertical direction) represent the magnitude of each frequency component of the acoustic information at the same time. The pixels in each row (pixels lined up in the horizontal direction) included in the image illustrated in Fig. 4 represent the magnitude of the same frequency component of the acoustic information at each time.

[0039] <Filtering> The analysis unit 12 may perform filtering on the acoustic information. The analysis unit 12 may perform frequency filtering on high frequencies or low frequencies, or a combination of high and low frequencies. The analysis unit 12 may perform filtering to remove short-term, sudden increases (spikes) in sound pressure. The analysis unit 12 may perform digital filtering using a first-order lag or a moving average, etc. The analysis unit 12 may perform background sound filtering using sound pressure data of recorded or defined background sounds. The analysis unit 12 may perform filtering before analyzing the acoustic information described above.

[0040] <Detecting characteristic patterns> The image representing the time change in the frequency components of the acoustic information includes a characteristic pattern. The analysis unit 12 may recognize acoustic information synchronized with the passage of the blade 42 as the acoustic information of the blade 42. A frequency in the acoustic information of the blade 42 where the sound pressure is significantly higher or lower than that of neighboring pixels or nearby pixels on the time axis (horizontal direction of the image representing the time change in the acoustic information) or frequency axis (vertical direction of the image representing the time change in the acoustic information) is also referred to as a peak frequency. The analysis unit 12 may detect a pattern including the peak frequency as a characteristic pattern. The analysis unit 12 may detect a range including a point that is the peak frequency in the image of the acoustic information and where the sound pressure is equal to or higher than the sound pressure at the point that is the peak frequency multiplied by a predetermined ratio as a characteristic pattern.

[0041] 7, the analysis unit 12 may detect pattern 51, pattern 52, and pattern 53 as patterns corresponding to the wind noise of each of the three blades 42 to be detected. The analysis unit 12 may calculate the rotation speed of the blade 42 based on the time intervals at which patterns 51, 52, and 53 are detected.

[0042] The analysis unit 12 may calculate the frequency range in which each of the patterns 51, 52, and 53 appears. The analysis unit 12 may determine whether the state of each blade 42 has changed based on the change over time in the frequency range in which each pattern appears.

[0043] The analysis unit 12 may detect a characteristic pattern 54 in the pattern 52 as a characteristic pattern that does not exist in the patterns 51 and 53. By detecting the characteristic pattern 54, the analysis unit 12 may detect the possibility that the state of the blade 42 in which the characteristic pattern 54 was detected is abnormal. The analysis unit 12 may analyze which part of the blade 42 may be in an abnormal state based on the amount of change in frequency per unit time of the characteristic pattern 54. By analyzing the part that may be in an abnormal state, inspection or repair, etc., of that part is more likely to be performed. As a result, the safety of the detection target is improved.

[0044] The characteristic sound changes depending on the size of a foreign object or a scratch at the part of the blade 42 that produces the characteristic sound. The analysis unit 12 may detect a change in the condition of the part of the blade 42 that produces the characteristic sound by monitoring a change in the frequency of the characteristic sound, or the magnitude of the sound pressure at the peak frequency, or the spread of a characteristic pattern including the peak frequency. The frequency of the characteristic sound, or the magnitude of the sound pressure at the peak frequency, or the spread of a characteristic pattern including the peak frequency, may change due to a change in the relative speed between the blade 42 and the wind, or the occurrence of turbulence. When the sound collection device 20 detects the same characteristic sound multiple times within one rotation of the blade 42, the analysis unit 12 compares the acoustic information of the characteristic sound detected each time, thereby reducing the influence of changes in wind speed or the occurrence of turbulence from the characteristic sound and improving the accuracy of detecting a change in the condition of the part that produces the characteristic sound.

[0045] The analysis unit 12 may detect pattern 55, which exists in a low frequency range, as another pattern. When the analysis unit 12 detects pattern 55, it may determine that it has detected the sound of a substation facility present in the vicinity of the detection target, or the sound of waves on the coast or at sea, etc. The analysis unit 12 may detect pattern 56, which exists in a high frequency range, as another pattern. When the analysis unit 12 detects pattern 56, it may determine that it has detected the sound of wind blowing in the vicinity of the detection target, or the noise of an aircraft, etc.

[0046] The analysis unit 12 may detect a granular feature pattern 57 as exemplified in Fig. 8. When the analysis unit 12 detects the granular feature pattern 57 in the pattern corresponding to the blade 42, the analysis unit 12 may detect that fine vibrations (so-called chatter) are occurring in the blade 42. The analysis unit 12 may detect a linear feature pattern 58 as exemplified in Fig. 9. When the analysis unit 12 detects the linear feature pattern 58 in the pattern corresponding to the blade 42, the analysis unit 12 may detect that unevenness is occurring on the surface of the blade 42.

[0047] The analysis unit 12 may detect the state of the blade 42 based on factors other than the shape, such as granular or linear. For example, the analysis unit 12 may detect the state of the blade 42 based on the frequency band of the acoustic information and the duration for which a characteristic pattern appears. The analysis unit 12 may calculate the frequency band of the acoustic information and the duration for which a characteristic pattern appears as the area of ​​the characteristic pattern in an image of the acoustic information, and detect the state of the blade 42 based on the calculated area of ​​the characteristic pattern. The analysis unit 12 may detect the state of the blade 42 based on the magnitude of the frequency component included in the characteristic pattern (the magnitude of the sound pressure, or the color density of the pixel when the magnitude of the component is expressed in grayscale).

[0048] The analysis unit 12 may recognize acoustic information that is not synchronized with the passage of the blade 42 as a sound occurring in the surrounding environment other than the blade 42, or may recognize it as the sound of a bird or flying object colliding with at least a part of the blade 42 as the detection target.

[0049] <Classification of characteristic patterns> Patterns included in an image expressing acoustic information are classified into categories representing the state of the detection target and categories unrelated to the state of the detection target. The categories representing the state of the detection target are further classified into categories representing a normal state of the detection target and categories representing an abnormal state of the detection target. The categories representing an abnormal state of the detection target are further classified into categories representing abnormal aspects occurring in the detection target. The categories unrelated to the state of the detection target are further classified into categories representing natural sounds such as wind or waves in the environment in which the detection target exists, and categories representing artificial sounds occurring around the detection target. The categories representing artificial sounds may be further classified into categories representing various artificial sounds such as the sound of a car driving, the sound of a ship sailing, or the sound of an aircraft flying. The categories are not limited to these examples.

[0050] The analysis unit 12 may set categories in advance and classify the detected pattern into the set category. When the category into which the pattern is classified represents the state of the detection object, the analysis unit 12 may detect the state represented by the category as the state of the detection object when the characteristic pattern is detected. For example, when the analysis unit 12 classifies the characteristic pattern into a category representing a hole present on the surface of the blade 42, the analysis unit 12 may detect the presence of a hole on the surface of the blade 42 as the state of the detection object. For example, when the analysis unit 12 classifies the characteristic pattern into a category unrelated to the state of the detection object or a category representing that the state of the detection object is normal, the analysis unit 12 may detect the state of the detection object as being normal.

[0051] The analysis unit 12 may register data associating characteristic pattern categories with the state of the detection target in a database in advance. The analysis unit 12 may classify the characteristic patterns into the categories registered in the database. By classifying the characteristic patterns using the database, the classification accuracy of the characteristic patterns is improved. The analysis unit 12 may detect the state associated with the classified category as the state of the detection target. By detecting the state of the detection target after the classification accuracy of the characteristic patterns has been improved, the detection accuracy of the state of the detection target is improved. The database may be stored in the memory unit 14 or in an external storage device connected to the state detection device 10.

[0052] The database may be created for each individual blade 42 of the wind turbine 40. Depending on factors such as the repair history of each individual blade 42 or errors in parts during manufacturing or errors during assembly, the pattern in which each blade is abnormal may differ for each blade 42. The analysis unit 12 may register information on repair marks for each blade 42 as a normal pattern in the database by measuring the sound immediately after repair for each blade 42 using the sound collection device 20 to obtain a waveform, or by extracting a characteristic pattern from an image of the acoustic information by the analysis unit 12.

[0053] The database may be generated in common for multiple blades 42. When a common database is generated for multiple blades 42, unless the individual differences between the blades 42 described above are taken into account in the common database, the analysis unit 12 may determine that the state of the blade 42 is abnormal even when a characteristic pattern included in an image representing the acoustic information of each blade 42 corresponds to the normal state of the blade 42. Conversely, the analysis unit 12 may determine that the state of the blade 42 is normal even when a characteristic pattern corresponds to the abnormal state of the blade 42. In other words, the individual differences between the blades 42 may reduce the rate at which states corresponding to characteristic patterns can be correctly recognized (recognition rate). To maintain the recognition rate, when the wind turbine 40 starts operating for the first time or when operation of a blade 42 is resumed after repair, the analysis unit 12 may group blades 42 whose identical or similar patterns correspond to normal or abnormal states into one group, and generate a common database using a common characteristic pattern for the blades 42 included in that group as a standard pattern.

[0054] The analysis unit 12 may generate a trained model for classifying characteristic patterns by performing machine learning using training data that indicates which category the characteristic patterns should be classified into. The analysis unit 12 may acquire the trained model for classifying characteristic patterns from an external device. The analysis unit 12 may classify the characteristic patterns using the trained model.

[0055] The normal state of the blade 42 as the detection target includes various states. For example, the normal state may include a state when the blade 42 is repaired or replaced. The analysis unit 12 may acquire acoustic information obtained by measuring sounds generated from the blade 42 when the blade 42 as the detection target is in a normal state. The analysis unit 12 may register various patterns included in the acoustic information when the blade 42 is in a normal state in the database in association with a normal state or an abnormal state. A pattern included in the acoustic information when the blade 42 as the detection target is in a normal state is also referred to as a normal pattern. The analysis unit 12 may register the normal pattern in the database in association with the normal state of the blade 42. The analysis unit 12 may update the database by adding a new normal pattern while leaving the normal patterns already registered in the database as they are. The analysis unit 12 may update the database by replacing the normal patterns already registered in the database with the new normal patterns. Even if the condition of the blade 42 has changed, by registering a pattern contained in the acoustic information when the blade 42 is known to be in a normal condition as a new normal pattern in the database, a pattern corresponding to a sound generated due to, for example, repair marks will not be mistakenly detected as an abnormality, thereby improving the accuracy of detecting the condition of the object to be detected.

[0056] The analysis unit 12 may extract a pattern that has a difference from a normal pattern from among the patterns included in the acoustic information of the blade 42. When the analysis unit 12 is able to extract a pattern that has a difference from the normal pattern, it may detect that the state of the blade 42 is not normal. The state of the blade 42 being not normal may include, for example, scratches on the surface of the blade 42.

[0057] To determine whether a pattern extracted from the acoustic information of the blade 42 corresponds to a normal pattern, the analysis unit 12 may calculate a numerical value representing the difference between the extracted pattern and the normal pattern. If the numerical value representing the difference is less than a difference threshold, the analysis unit 12 may detect that the condition of the area where the sound represented by the characteristic pattern is generated is normal. If the numerical value representing the difference is equal to or greater than the difference threshold, the analysis unit 12 may detect that the condition of the area where the sound represented by the characteristic pattern is generated is abnormal. This clarifies the criteria for determining the condition of the detection target. As a result, the accuracy of condition detection is improved.

[0058] The analysis unit 12 may determine whether the state of the blade 42 is normal without calculating the difference between a pattern included in the acoustic information of the blade 42 and a normal pattern. For example, the analysis unit 12 may register, in the database as an abnormal pattern, a pattern included in the acoustic information when the blade 42 as the detection target is not in a normal state, or a pattern included in the acoustic information when the blade 42 is in an abnormal state. When the pattern included in the acoustic information of the blade 42 matches the abnormal pattern, the analysis unit 12 may determine that the state of the blade 42 is not normal or that the state of the blade 42 is abnormal.

[0059] Artificial sounds are generated for various reasons. The analysis unit 12 may register characteristic patterns, such as operational sounds that are known to occur under normal circumstances, in a database as normal patterns. When the analysis unit 12 detects a characteristic pattern that does not correspond to a normal pattern, it may detect, for example, lightning in the vicinity of the detection target. When the analysis unit 12 detects a characteristic pattern that does not correspond to a normal pattern, it may determine that a bird, flying object, or the like has collided with the blade 42. When the analysis unit 12 detects a characteristic pattern that does not correspond to a normal pattern, it may determine that an abnormal event, such as the intrusion of a suspicious person, has occurred in the vicinity of the wind turbine 40 having the blade 42 as the detection target.

[0060] The trained model may be generated so that characteristic patterns corresponding to various states when the blade 42 to be detected is in a normal state are classified into categories representing normal states.

[0061] The analysis unit 12 may detect the state of the detection target based on first acoustic information representing sound components arriving at the sound collection device 20 from a first direction and second acoustic information representing sound components arriving at the sound collection device 20 from a second direction. For example, the analysis unit 12 may extract noise components from the second acoustic information and detect the state of the detection target based on information obtained by removing the extracted noise components from the first acoustic information. Detecting the state of the detection target based on acoustic information from which the noise components have been removed improves the accuracy of detecting the state of the detection target. The analysis unit 12 may detect a characteristic pattern based only on the first acoustic information representing sound components arriving from the detection target. In this way, the influence of sounds generated outside the detection target, such as background sounds, can be reduced. As a result, the accuracy of detecting the state of the detection target is improved.

[0062] The analysis unit 12 may use a band-pass filter to remove or attenuate acoustic information in a predetermined frequency band from the acoustic information. For example, the analysis unit 12 may use a band-pass filter to remove or attenuate acoustic information in a frequency band other than the frequency of the sound of the blade 42. By doing so, the accuracy of detecting the state of the blade 42 based on the acoustic information of the blade 42 is improved.

[0063] The analysis unit 12 may perform filtering using a reference on data of time changes in frequency components of the acoustic information. Specifically, the analysis unit 12 may calculate a difference spectrum by subtracting a frequency spectrum from a frequency spectrum at a certain time by a predetermined time in the past. In the difference spectrum, the frequency spectrum of the background sound is removed or reduced. As a result, the accuracy of detecting sudden sounds contained in the background sound is improved.

[0064] <Tracking> The analysis unit 12 may monitor the peak frequency that changes over time and interpret it as a series of events. From the series of events, the analysis unit 12 may calculate a sound pressure threshold to detect a characteristic pattern including the peak frequency, the peak frequency at the start or end of the series of events, the duration of the series of events, or the extent of the characteristic pattern. The extent of the spread refers to the width of frequencies that are equal to or greater than the sound pressure threshold.

[0065] <Status notification> The analysis unit 12 may output the state represented by the classification of the characteristic patterns to the display device 30 as the detection result of the state detection device 10. The display device 30 may notify an entity that manages the wind turbine 40 that has the blades 42 as the detection target by displaying the detection result of the state detection device 10. The analysis unit 12 may output the detection result of the state detection device 10 to various other devices, not limited to the display device 30, such as a speaker or a light. The speaker may notify an entity that manages the wind turbine 40 that has the blades 42 of the detection target by outputting audio. The light may notify an entity that manages the wind turbine 40 that has the blades 42 of the detection target of the state of the blades 42 by emitting light or changing its light emission state.

[0066] <Example of procedure for detecting the status> The analysis unit 12 of the condition detection device 10 may execute a condition detection method including the steps of the flowchart illustrated in Fig. 10. The condition detection method may be realized as a condition detection program executed by a processor constituting the analysis unit 12. The condition detection program may be stored in a non-transitory computer-readable medium.

[0067] The analysis unit 12 acquires acoustic information of the blade 42 as a detection target (step S1). The analysis unit 12 analyzes the frequency components of the acoustic information or an image representing changes in the frequency components over time (step S2). The analysis unit 12 determines whether a pattern has been detected from the image of the acoustic information (step S3). If the analysis unit 12 does not detect a pattern from the image of the acoustic information (step S3: NO), the analysis unit 12 ends the execution of the procedure of the flowchart in FIG. 10.

[0068] If the analysis unit 12 detects a pattern from the image of the acoustic information (step S3: YES), it determines (step S4) whether the detected pattern corresponds to a pattern corresponding to the sound of the blade 42. In other words, the analysis unit 12 determines whether the detected pattern is classified into a category representing the sound generated by the blade 42.

[0069] If the detected pattern corresponds to a pattern corresponding to the sound of the blade 42 (step S4: YES), the analysis unit 12 determines whether the detected pattern matches a normal pattern (step S5). In other words, the analysis unit 12 determines whether the pattern is classified into a category indicating that the state of the blade 42 as the detection target is normal. If the detected pattern matches a normal pattern (step S5: YES), the analysis unit 12 determines that the state of the blade 42 as the detection target is not abnormal and ends the execution of the procedure in the flowchart of FIG. 10. If the detected pattern does not match the normal pattern (step S5: NO), the analysis unit 12 determines that the state of the blade 42 as the detection target is abnormal and notifies the display device 30 that the state of the blade 42 as the detection target is abnormal (step S6). After executing the procedure in step S6, the analysis unit 12 ends the execution of the procedure in the flowchart of FIG. 10.

[0070] If the detected pattern does not correspond to a pattern corresponding to the sound of the blade 42 (step S4: NO), the analysis unit 12 classifies the detected pattern into an environmental sound category and notifies the type of sound associated with the classified category via the display device 30 (step S7). After executing the procedure of step S7, the analysis unit 12 ends execution of the procedure of the flowchart in FIG.

[0071] <Summary> As described above, in the condition detection system 1 according to this embodiment, the condition detection device 10 detects characteristic patterns from an image representing a time change in the frequency components of acoustic information generated based on a sound generated by a detection target. The condition detection device 10 classifies the detected characteristic patterns into categories representing various conditions, and detects the conditions represented by the categories. In this way, various conditions are detected while being distinguished from one another. As a result, the accuracy of detecting various conditions is improved.

[0072] The condition detection system 1 allows the condition of the detection target to be ascertained even when an entity (e.g., a worker) managing the wind turbine 40 equipped with the blade 42 that is the detection target is not present on-site. As a result, the frequency and cost of workers visiting the site of the detection target are reduced. In addition, continuous monitoring is realized. Continuous monitoring improves the safety of the detection target.

[0073] The condition detection system 1 clarifies the criteria for determining the condition. Furthermore, the analysis results of the acoustic information are clarified as numerical values. As a result, the safety of the detection target is improved.

[0074] The condition detection system 1 can store acoustic information. A worker or the like can check the stored acoustic information together with the detection result of the condition of the detection target by the condition detection system 1, thereby improving the skill or knowledge of the worker or the like.

[0075] (Other embodiments) Other embodiments are described below.

[0076] <Detection of conditions based on instantaneous frequency components> As described above, the analysis unit 12 may detect the state of the detection target based on an image representing a change over time in the frequency components of the acoustic information. The analysis unit 12 may also detect the state of the detection target based on a frequency spectrum representing the instantaneous frequency components of the acoustic information at a certain time. For example, the analysis unit 12 may determine that the state of the detection target is in a specific state when the sound pressure of a specific frequency component in the frequency spectrum exceeds a threshold.

[0077] <Other examples of detection targets> The detection target of the condition detection system 1 is not limited to the blades 42 of the wind turbine 40 described above, and may detect the status of various other devices or facilities. The condition detection system 1 may detect, for example, the status of a device or facility that generates sound as the detection target. The condition detection system 1 may detect, for example, the status of a fan or blower that generates sound in a periodic pattern as the detection target. The condition detection system 1 may detect, for example, the status of a compressor that generates sound in a discrete or sudden pattern as the detection target.

[0078] Although the embodiments of the present disclosure have been described based on the drawings and examples, it should be noted that those skilled in the art can make various modifications or alterations based on the present disclosure. Therefore, it should be noted that these modifications or alterations are included in the scope of the present disclosure. For example, the functions included in each component can be rearranged so as not to cause logical inconsistencies, and multiple components can be combined or divided into one. [Explanation of symbols]

[0079] 1. Condition detection system 10 Anomaly detection device (12: analysis unit, 14: memory unit, 16: interface) 20 Sound collection device 30 Display device 40 Detection target (42: Blade, 44: Nacelle, 46: Strut, 48: Abnormal part) Patterns 51-53, 55, 56 54, 57, 58 Feature patterns

Claims

1. an analysis unit for detecting the state of the detection target; The analysis unit a sound collection device positioned away from the detection target is used to measure the sound generated by the detection target, and the result is acquired as acoustic information; detecting a state of the detection object based on a pattern included in an image representing a time change in frequency components of the acoustic information, and analyzing a part of the detection object from which a sound represented by the pattern is generated; outputting a detection result of the state of the detection object and an analysis result of the part of the detection object where the sound represented by the pattern is generated; Condition detection device.

2. A condition detection device as described in Claim 1, wherein the detection object is a wind turbine blade.

3. The condition detection device according to claim 2 , wherein the analysis unit outputs an indication that the condition of the part where the sound represented by the pattern originates is abnormal when a numerical value representing a difference between the pattern and a normal pattern is equal to or greater than a difference threshold value.

4. The analysis unit detecting a state of the detection object based on a database in which the pattern is associated with the state of the detection object; 4. The state detection device according to claim 3, wherein the normal pattern is generated based on acoustic information when the state of the detection target is normal and registered in the database.

5. A condition detection system comprising: the condition detection device according to claim 1; and a sound collection device that measures a sound generated by a detection target and outputs the measurement result as acoustic information to the condition detection device.

6. The sound collection device is Equipped with multiple microphones, The state detection system according to claim 5 , further comprising: generating acoustic information representing components of sounds arriving at said sound collection device from a predetermined direction based on sounds detected by each of said plurality of microphones.

7. The state detection system according to claim 6, wherein the analysis unit of the state detection device acquires a plurality of pieces of acoustic information representing components of sounds arriving at the sound collection device from each of a plurality of directions, and detects the state of the detection target based on the plurality of pieces of acoustic information.

Images