State monitoring device, nacelle monitoring device, monitoring system, and monitoring method

Abstract

In the present invention, a state monitoring device (200) acquires, from a first vibration sensor (111), a second vibration sensor (112), and a third vibration sensor (113), a vibration value for a first blade (101), a vibration value for a second blade (102), and a vibration value for a third blade (103), respectively. When the absolute values for the vibration value for the first blade (101), the vibration value for the second blade (102), and the vibration value for the third blade (103) all exceed a first threshold value, the state monitoring device (200) detects a lightning strike on a wind power generation device.

Description

Condition monitoring device, nacelle monitoring device, monitoring system, and monitoring method

[0001] The present disclosure relates to a condition monitoring device, a nacelle monitoring device, a monitoring system, and a monitoring method.

[0002] For example, Japanese Patent Application Laid-Open No. 2005-62080 (Patent Document 1) discloses a wind turbine generator. This wind turbine generator has multiple blades, which are located at high positions. Therefore, lightning may strike multiple blades. When lightning strikes, the blades of the wind turbine generator may be damaged. Therefore, the wind turbine generator has a lightning detection device for detecting lightning strikes. The lightning detection device has a Rogowski coil and a detection unit that detects the lightning current flowing through the Rogowski coil when lightning strikes.

[0003] Japanese Patent Application Laid-Open No. 2005-62080

[0004] However, the invention described in the above-mentioned Patent Document 1 requires a detection device (lightning detection device) for detecting abnormalities such as lightning strikes, which can cause a problem of increased manufacturing costs for the wind turbine generator.

[0005] The present disclosure has been made to solve the above-mentioned problems, and its object is to detect an abnormality in a wind turbine generator while suppressing manufacturing costs.

[0006] A condition monitoring device disclosed herein monitors the condition of a wind turbine generator having multiple blades. The condition monitoring device includes an interface and a control device. The interface acquires vibration values ​​of the multiple blades from multiple vibration sensors, each of which detects vibration values ​​of the multiple blades. The control device detects a lightning strike to the wind turbine generator when all absolute values ​​of the vibration values ​​of the multiple blades exceed a first threshold.

[0007] A condition monitoring method disclosed herein is a method for monitoring the condition of a wind turbine generator having a plurality of blades. The condition monitoring method includes acquiring vibration values ​​of the plurality of blades. The condition monitoring method also includes detecting that the wind turbine generator has been struck by lightning when absolute values ​​of the vibration values ​​of the plurality of blades exceed a first threshold.

[0008] According to the present disclosure, it is possible to detect an abnormality in a wind turbine generator while suppressing manufacturing costs.

[0009] 6 is a diagram illustrating a condition monitoring system and the like of the present disclosure. FIG. 7 is a diagram illustrating the internal structure of a hub and a nacelle. FIG. 8 is a diagram illustrating an example of the hardware configuration of a condition monitoring device and a nacelle monitoring device. FIG. 9 is a diagram illustrating the waveform of time-series data of vibration values. FIG. 10 is a diagram illustrating the relationship between the rotational position of a first blade and a first vibration value. FIG. 11 is a diagram illustrating the waveform of vibration values ​​when lightning strikes a wind turbine generator. FIG. 12 is an enlarged view of a portion of the waveform in FIG. 6 where the vibration value is 0. FIG. 13 is a diagram illustrating an example of the waveform of vibration values. FIG. 14 is a diagram illustrating vibration data for one rotation of three blades. FIG. 15 is a diagram illustrating a specific example of a correlation coefficient. FIG. 16 is a functional block diagram of a condition monitoring device. FIG. 17 is a flowchart illustrating the main processing flow of a condition monitoring device and the like.

[0010] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. In the embodiments described below, when numbers, quantities, etc. are mentioned, the scope of the present disclosure is not necessarily limited to those numbers, quantities, etc., unless otherwise specified. The same reference numerals are used for the same or equivalent parts, and redundant descriptions may not be repeated. It is intended from the beginning that the configurations in the embodiments may be used in appropriate combinations.

[0011] [Configuration of Wind Power Generation System] Fig. 1 is a diagram showing an example of the configuration of a wind power generation device 100 and a state monitoring system 500 according to the present disclosure. In Fig. 1, a front view of the wind power generation device 100 is shown.

[0012] The wind turbine generator 100 has M blades (M is an integer equal to or greater than 2). The M blades correspond to the "plurality of blades" of the present disclosure. In this embodiment, M=3. That is, the wind turbine generator 100 has three blades: a first blade 101, a second blade 102, and a third blade 103. Note that, as a modification, M may have another value, such as M=2 or M=4.

[0013] The wind turbine generator 100 further includes a hub 104, a nacelle 105, a tower 180, a first down conductor 161, a second down conductor 162, and a third down conductor 163. A first blade 101, a second blade 102, and a third blade 103 are joined to the hub 104. The nacelle 105 is installed behind the hub 104. The tower 180 holds the first blade 101, the second blade 102, the third blade 103, the hub 104, and the nacelle 105.

[0014] The condition monitoring system 500 includes a condition monitoring device 200 and M (in this embodiment, M=3) vibration sensors. In the example of Fig. 1 , the condition monitoring device 200 is installed inside the hub 104. The condition monitoring device 200 is a device that monitors the condition of the wind turbine generator 100, and more typically, a device that monitors the condition of a plurality of blades.

[0015] 1 discloses three vibration sensors: a first vibration sensor 111, a second vibration sensor 112, and a third vibration sensor 113. The first vibration sensor 111, the second vibration sensor 112, and the third vibration sensor 113 are installed on the first blade 101, the second blade 102, and the third blade 103, respectively. The first vibration sensor 111, the second vibration sensor 112, and the third vibration sensor 113 detect the vibration value of the first blade 101, the vibration value of the second blade 102, and the vibration value of the third blade 103, respectively. In this embodiment, the first vibration sensor 111, the second vibration sensor 112, and the third vibration sensor 113 are acceleration sensors. That is, the first vibration sensor 111, the second vibration sensor 112, and the third vibration sensor 113 respectively detect the acceleration of the first blade 101, the acceleration of the second blade 102, and the acceleration of the third blade 103 as vibration values. Note that, as a modified example, the vibration sensors may be other sensors, such as a displacement sensor or a velocity sensor.

[0016] The first down conductor 161, the second down conductor 162, and the third down conductor 163 are respectively installed inside the first blade 101, the second blade 102, and the third blade 103. The first down conductor 161, the second down conductor 162, and the third down conductor 163 are also collectively referred to simply as "down conductors." When lightning strikes the wind turbine generator 100, the down conductors safely discharge the lightning current resulting from the lightning strike to the ground. In other words, when lightning strikes the wind turbine generator 100, the lightning current flows through the down conductors.

[0017] FIG. 2 is a diagram showing the internal structure of the hub 104 and the nacelle 105. As described above, the condition monitoring device 200 is installed inside the hub 104. In the example of FIG. 2, a vibration sensor is installed inside each blade. In the example of FIG. 2, a first vibration sensor 111 is installed inside the first blade 101, and a third vibration sensor 113 is installed inside the third blade 103. Although not specifically shown, a second vibration sensor 112 is installed inside the second blade 102. The vibration values ​​(accelerations) detected by the first vibration sensor 111, the second vibration sensor 112, and the third vibration sensor 113 are transmitted to the condition monitoring device 200 as time-series data.

[0018] 2 also illustrates a portion of the first down conductor 161 inside the first blade 101. Since the vibration sensor and the down conductor are installed inside the blade, the wiring 141 from the vibration sensor (first vibration sensor 111) and the first down conductor 161 are arranged close to each other (see point a in FIG. 2). In this way, in each blade, the wiring 141 from the vibration sensor and the down conductor are arranged close to each other.

[0019] The condition monitoring device 200 executes a pre-lightning strike anomaly detection process, a lightning strike detection process, and a blade anomaly determination process based on the vibration values ​​detected by the first vibration sensor 111, the second vibration sensor 112, and the third vibration sensor 113. The pre-lightning strike anomaly detection process is a process for detecting anomalies in the first blade 101, the second blade 102, and the third blade 103. In the pre-lightning strike anomaly detection process, for example, if a vibration value falls within an abnormal range, the condition monitoring device 200 determines that the blade on which the vibration sensor that output the vibration value is installed is abnormal. Details of the lightning strike detection process and the blade anomaly determination process will be described later. In this embodiment, the first vibration sensor 111, the second vibration sensor 112, and the third vibration sensor 113 are used in the pre-lightning strike anomaly detection process, the lightning strike detection process, and the blade anomaly determination process.

[0020] The main shaft 121 holds the hub 104 and the nacelle 105 so that the hub 104 is rotatably held in the nacelle 105. Furthermore, a main bearing 122, a gearbox 123, a generator 124, and a nacelle monitoring device 300 are installed inside the nacelle 105.

[0021] The main bearing 122 rotatably holds the main shaft 121. The gearbox 123 amplifies the rotation of the hub 104 to a rotational speed required for generating electricity. The generator 124 converts the rotational force of the hub 104 (the rotational force amplified by the gearbox 123) into electricity. The main shaft 121, the main bearing 122, the gearbox 123, and the generator 124 correspond to the "mechanism installed in the nacelle" in this disclosure.

[0022] The nacelle 105 is further provided with a vibration sensor 125, a rotation sensor 126, a vibration sensor 127, and a vibration sensor 128. The vibration sensor 125, the rotation sensor 126, the vibration sensor 127, and the vibration sensor 128 detect the vibration value of the main bearing 122, the rotational speed (rotational position) of the main shaft 121, the vibration value of the gearbox 123, and the vibration value of the generator 124, respectively. The vibration sensor 125, the rotation sensor 126, the vibration sensor 127, and the vibration sensor 128 correspond to the "nacelle sensor" in this disclosure. The vibration value of the main bearing 122, the rotational speed (rotational position) of the main shaft 121, the vibration value of the gearbox 123, and the vibration value of the generator 124 correspond to the "parameter" in this disclosure. The vibration sensor 125, the rotation sensor 126, the vibration sensor 127, and the vibration sensor 128 transmit their detection values ​​to the nacelle monitoring device 300 as time-series data.

[0023] The nacelle monitoring device 300 includes a determination device 310. The determination device 310 determines whether or not there is an abnormality in the mechanisms installed in the nacelle 105 (i.e., the main shaft 121, the main bearing 122, the gearbox 123, and the generator 124) based on time-series data from the vibration sensor 125, the rotation sensor 126, the vibration sensor 127, and the vibration sensor 128. This determination process is also referred to as a nacelle abnormality determination process (see step S114 in FIG. 12 ).

[0024] In the following description, the first blade 101, the second blade 102, and the third blade 103 may be collectively referred to as the "three blades." Furthermore, the first vibration sensor 111, the second vibration sensor 112, and the third vibration sensor 113 may be collectively referred to as the "three vibration sensors."

[0025] [Hardware Configuration of State Monitoring Device and Nacelle Monitoring Device] Figure 3 is a diagram showing an example of the hardware configuration of state monitoring device 200 and nacelle monitoring device 300. State monitoring device 200 has a CPU (Central Processing Unit) 201, memory 202, and a communication interface 203. Note that in Figure 3, the communication interface is written as "I / F." Communication interface 203 corresponds to the "signal input unit" and "external interface" of the present disclosure. Nacelle monitoring device 300 has a CPU 301, memory 302, and communication interface 303. CPU 201 and CPU 301 execute various processes.

[0026] The memory 202 and the memory 302 include a read-only memory (ROM) and a random access memory (RAM). The ROM is a non-volatile memory that cannot be rewritten, and the RAM is a volatile memory. The communication interface 203 and the communication interface 303 communicate with external devices. For example, the status monitoring device 200 transmits a lightning strike signal (described below) to the communication interface 303 via the communication interface 203 of the status monitoring device 200. The lightning strike signal corresponds to the "detection signal" of this disclosure.

[0027] The CPU 201 and at least a portion of the storage area of ​​the memory 202 constitute a control device 240. The control device 240 may also be referred to as a "control circuit" or "at least one controller." The CPU 301 and at least a portion of the storage area of ​​the memory 302 constitute a determination device 310 (see FIG. 2).

[0028] 4 is a diagram showing waveforms of time-series data of vibration values ​​measured by the state monitoring device 200 from the first vibration sensor 111, the second vibration sensor 112, and the third vibration sensor 113. Hereinafter, the vibration value measured by the first vibration sensor 111, the vibration value measured by the second vibration sensor 112, and the vibration value measured by the third vibration sensor 113 will also be referred to as the first vibration value, the second vibration value, and the third vibration value, respectively. Furthermore, the time-series data of the first vibration value, the second vibration value, and the third vibration value will also be referred to as the first time-series data, the second time-series data, and the third time-series data, respectively.

[0029] 4 and the following Figures 5 to 8, the vertical axis represents vibration value (acceleration) and the horizontal axis represents time. In Figure 4, the first vibration value, the second vibration value, and the third vibration value are represented by a solid line, a dashed line, and a dashed line, respectively.

[0030] The first vibration value, the second vibration value, and the third vibration value are values ​​that reflect the gravitational acceleration that accompanies the rotation of the first blade 101, the second blade 102, and the third blade 103, respectively. Therefore, the waveforms of the time-series data of the first vibration value, the second vibration value, and the third vibration value are sine curves. Furthermore, the first blade 101, the second blade 102, and the third blade 103 are arranged at intervals of 120 degrees in the rotation direction of the hub 104. Therefore, the phases of the first vibration value, the second vibration value, and the third vibration value at the same time are shifted by 120 degrees. If the rotation speed of the hub 104 is, for example, 20 min －１ In this case, the first vibration value, the second vibration value, and the third vibration value increase and decrease at a cycle of 20 / 60≈0.33 Hz.

[0031] 5 is a diagram showing the relationship between the rotational position (rotational angle) of the first blade 101 and the first vibration value. In the example of FIG. 5, when the first blade 101 is located at the uppermost position, the rotational angle of the first blade 101 is set to 0 degrees. In the example of FIG. 5, when the rotational angle of the first blade 101 is 0 degrees, the first vibration value is 0, and when the rotational angle of the first blade 101 is 90 degrees, the first vibration value is 9.8 m / s ２Furthermore, when the rotation angle of the first blade 101 is 180 degrees, the first vibration value is 0, and when the rotation angle of the first blade 101 is 270 degrees, the first vibration value is −9.8 m / s ２ This becomes:

[0032] The gravity g acting on the vibration sensor (first vibration sensor 111) is 9.8 m / s ２ Therefore, as shown in FIG. 5, the maximum value of the first vibration is 9.8 m / s ２ The minimum value of the first vibration is −9.8 m / s ２ However, in reality, the vibration value data may have a phase lead due to the influence of the electronic circuitry of the condition monitoring device 200, which may cause a discrepancy with the actual change in the gravitational acceleration component. Also, depending on the mounting orientation and frequency characteristics of the vibration sensor, the maximum vibration value may be 9.8 m / s ２ The minimum vibration value is -9.8 m / s ２ Furthermore, if noise (high frequency noise, etc.) is mixed in the time series data output by the vibration sensor, the state monitoring device 200 may detect an actual vibration value of 9.8 m / s ２ Vibration values ​​greater than or equal to -9.8 m / s ２ The noise is generated by the hub 104 or the creaking of the three blades, etc.

[0033] [Lightning Strike Detection Process] Next, the lightning strike detection process by the status monitoring device 200 will be described. As explained in Fig. 2, when lightning strikes the wind turbine generator 100, a lightning current flows through the down conductor. Also, as explained in Fig. 2, for example, inside the hub 104, the wiring 141 from the vibration sensor is arranged near the down conductor (see point a in Fig. 2). Therefore, an induced lightning surge caused by the lightning strike enters the wiring 141 of the vibration sensor via the down conductor.

[0034] 6 is a diagram showing waveforms of the first vibration value, the second vibration value, and the third vibration value when lightning strikes the wind turbine generator 100. In the waveforms of the vibration values, an induced lightning surge that has entered the wiring 141 from the vibration sensor appears as noise with an impulse waveform (see point A in FIG. 6).

[0035] Fig. 7 is an enlarged view of a portion of the waveform in Fig. 6 where the vibration value (acceleration) is 0. Fig. 7 also shows waveforms of the absolute values ​​of the first vibration value, the second vibration value, and the third vibration value. As shown in Fig. 7, when lightning strikes the wind turbine generator 100, the absolute values ​​of the first vibration value, the second vibration value, and the third vibration value all fall below the first threshold value of 9.8 m / s ２ Exceeds.

[0036] Therefore, the state monitoring device 200 of this embodiment acquires the first vibration value, the second vibration value, and the third vibration value (first time-series data, second time-series data, and third time-series data) in a predetermined first collection period, and calculates whether the absolute values ​​of the first vibration value, the second vibration value, and the third vibration value are all equal to or less than the first threshold value of 9.8 m / s ２ If the absolute values ​​of the first vibration value, the second vibration value, and the third vibration value all equal the first threshold value, it is detected that lightning has struck the wind turbine generator 100 (see "Lightning Strike Detection" in FIG. 7). Note that if the absolute values ​​of the first vibration value, the second vibration value, and the third vibration value all equal the first threshold value, it may be detected that lightning has struck the wind turbine generator 100, or it may be detected that lightning has not struck the wind turbine generator 100.

[0037] For example, as shown in FIG. 2, when the first down conductor 161 and the wiring 141 of any of the first vibration sensors 111 are close to each other inside the hub 104, the absolute values ​​of the first vibration value, the second vibration value, and the third vibration value may be increased to 9.8 m / s due to noise caused by an induced lightning surge. ２ Therefore, the state monitoring device 200 determines whether the absolute values ​​of the first vibration value, the second vibration value, and the third vibration value all exceed the first threshold value of 9.8 m / s at the same timing. ２ If the value of the lightning strike exceeds the threshold, a lightning strike to the wind turbine generator 100 is detected.

[0038] The "same timing" in this embodiment does not necessarily mean that the timings at which the absolute values ​​of the first vibration value, the second vibration value, and the third vibration value all exceed the first threshold are exactly the same, but may also mean that the timings are approximately the same. As long as a lightning strike to the wind turbine generator 100 can be detected, the "approximately the same timing" may mean that there is a time difference (for example, 0.1 seconds) between at least two of the timings at which the absolute value of the first vibration value exceeds the first threshold, the timing at which the absolute value of the second vibration value exceeds the first threshold, and the timing at which the absolute value of the third vibration value exceeds the first threshold are detected.

[0039] In this embodiment, the first collection period is, for example, 20 seconds, but is determined appropriately depending on the specifications of the state monitoring device 200, such as the data capacity and filtering process described below.

[0040] Furthermore, when only two vibration values ​​exceed the first threshold value, or when only one vibration value exceeds the first threshold value, the condition monitoring device 200 of this embodiment determines that lightning has not struck the wind turbine generator 100. This is because when only two vibration values ​​exceed the first threshold value, or when only one vibration value exceeds the first threshold value, it is highly likely that the vibration value exceeded the first threshold value due to a factor other than a lightning strike.

[0041] [Blade Abnormality Determination Process] When lightning strikes wind turbine generator 100, an abnormality (e.g., damage) may occur in at least one of the three blades (first blade 101, second blade 102, and third blade 103). Therefore, when state monitoring device 200 detects that lightning has struck wind turbine generator 100, it executes blade abnormality determination process to detect whether or not there is an abnormality in the three blades.

[0042] 8A, 8B, and 8C are diagrams showing exemplary waveforms of the first vibration value, the second vibration value, and the third vibration value acquired during a predetermined second collection period by the state monitoring device 200. Figures 8A, 8B, and 8C show the waveforms of the first vibration value, the second vibration value, and the third vibration value, respectively.

[0043] The condition monitoring device 200 extracts time series data for a predetermined number of rotations L of the three blades from the time series data for the first collection period output by each of the three vibration sensors. L is an integer equal to or greater than 1, and in this embodiment, L=1, i.e., the predetermined number of rotations is one rotation.

[0044] Fig. 9 shows three pieces of vibration data (the extracted data) for one rotation of the three blades when no abnormality occurs in the three blades. The vertical axis of Fig. 9 represents the vibration value, and the horizontal axis represents the blade rotation angle. Figs. 9(A), 9(B), and 9(C) show waveforms of the first vibration value, the second vibration value, and the third vibration value, respectively. In Figs. 9(A), 9(B), and 9(C), the rotation angles are shown all at once.

[0045] After a lightning strike is detected, if no abnormality occurs in any of first blade 101, second blade 102, and third blade 103, the correlation between the vibration values ​​in time series tends to be strong, as shown in Figure 9. On the other hand, if an abnormality occurs in any of first blade 101, second blade 102, and third blade 103, the correlation between the vibration value of the blade in which the abnormality occurred and the vibration values ​​of the other blades tends to be weak.

[0046] In consideration of the above-described tendency, the condition monitoring device 200 of this embodiment detects the presence or absence of an abnormality in the three blades by using a correlation parameter that indicates the correlation between the time-series data of the first vibration value, the second vibration value, and the third vibration value. In this embodiment, the correlation parameter is a correlation coefficient.

[0047] 10 is a diagram illustrating an example of application of the correlation coefficient. As shown in FIG. 10, the condition monitoring device 200 calculates a correlation coefficient indicating the correlation between two pieces of time series data for all possible combinations of the two pieces of time series data for the three blades. Hereinafter, the two pieces of time series data will also be referred to as "one piece of time series data" and "the other piece of time series data."

[0048] Specifically, the status monitoring device 200 extracts N (N is an integer equal to or greater than 2) vibration values ​​for the same rotation angle from each of the three pieces of time-series data during a predetermined second collection period. Then, the status monitoring device 200 calculates correlation coefficients σ12, σ23, and σ31 between one piece of time-series data and the other piece of time-series data.

[0049] Furthermore, all of the above combinations are "the first time series data and the second time series data," "the second time series data and the third time series data," and "the third time series data and the first time series data." That is, the correlation coefficients σ12, σ23, and σ31 are the correlation coefficient between the first time series data and the second time series data, the correlation coefficient between the second time series data and the third time series data, and the correlation coefficient between the third time series data and the first time series data, respectively. The status monitoring device 200 calculates the correlation coefficients, for example, using the following equation (1):

[0050] σ=(A1) / (B1×B2) (1) where A1 in equation (1) represents the covariance between “one time series data” and “the other time series data.” Also, B1 represents the standard deviation of N vibration values ​​in one time series data, and B2 represents the standard deviation of N vibration values ​​in the other time series data.

[0051] 10A shows an example where the correlation coefficient σ12, the correlation coefficient σ23, and the correlation coefficient σ31 are "0.73," "0.80," and "0.23," respectively. 10B shows an example where the correlation coefficient σ12, the correlation coefficient σ23, and the correlation coefficient σ31 are "0.70," "0.48," and "0.35," respectively. 10C shows an example where the correlation coefficient σ12, the correlation coefficient σ23, and the correlation coefficient σ31 are "0.51," "0.34," and "0.28," respectively.

[0052] The state monitoring device 200 determines whether each of the three calculated correlation coefficients (correlation coefficient σ12, correlation coefficient σ23, and correlation coefficient σ31) is less than a second threshold value. In the present embodiment, the second threshold value is 0.60.

[0053] Although not shown, if all three correlation coefficients are greater than the second threshold, the condition monitoring device 200 determines that all three blades are normal. Also, as shown in Figure 10(A), if one of the three correlation coefficients is less than the second threshold and two of the correlation coefficients are equal to or greater than the second threshold, the condition monitoring device 200 determines that all three blades are normal.

[0054] Next, as shown in Figure 10(B), if two of the three correlation coefficients are less than the second threshold and one correlation coefficient is equal to or greater than the second threshold, it is highly likely that an abnormality has occurred in one of the three blades. In the case of Figure 10(B), it is highly likely that an abnormality has occurred in the third blade 103. Therefore, in the case of Figure 10(B), the condition monitoring device 200 determines that an abnormality has occurred in one blade.

[0055] Next, as shown in Fig. 10(C), if all three correlation coefficients are less than the second threshold, there is a high possibility that an abnormality has occurred in at least two of the three blades. Therefore, in the case of Fig. 10(C), the condition monitoring device 200 determines that an abnormality has occurred in two or three blades.

[0056] As described above, the condition monitoring device 200 detects that an abnormality has occurred in one of the three blades when the number of correlation coefficients of the two time series data in all of the above combinations that is less than the second threshold is two or more (i.e., in the cases of Figures 10(B) and 10(C)).

[0057] As a modified example, the above-mentioned definitions "less than the second threshold value" and "greater than or equal to the second threshold value" may be replaced with the definitions "less than or equal to the second threshold value" and "greater than the second threshold value", respectively.

[0058] [Functional Block Diagram of State Monitoring Device] Fig. 11 is a functional block diagram of the state monitoring device 200. The state monitoring device 200 has a signal input unit 250, a first LPF (Low pass filter) 251, a second LPF 252, a third LPF 253, a first absolute value unit 261, a second absolute value unit 262, a third absolute value unit 263, a comparison unit 266, a correlation unit 268, a determination unit 270, a generation unit 272, a communication unit 274, a first storage unit 281, and a second storage unit 282. The first storage unit 281 stores a first threshold value (in this embodiment, 9.8 m / s ２ The second storage unit 282 also stores a second threshold value (0.6 in this embodiment).

[0059] 3 has the functions of a signal input unit 250, a first LPF (low pass filter) 251, a second LPF 252, a third LPF 253, a first absolute value unit 261, a second absolute value unit 262, a third absolute value unit 263, a comparison unit 266, a correlation unit 268, a determination unit 270, a generation unit 272, and a communication unit 274. The memory 202 has the functions of a first storage unit 281 and a second storage unit 282.

[0060] The signal input unit 250 acquires the first time series data (first vibration value) transmitted from the first vibration sensor 111, the second time series data (second vibration value) transmitted from the second vibration sensor 112, and the third time series data (third vibration value) transmitted from the third vibration sensor 113 over the above-mentioned first collection period.

[0061] Next, the first LPF 251, the second LPF 252, and the third LPF 253 perform filtering to suppress noise in the first time-series data, the second time-series data, and the third time-series data, respectively. In this embodiment, the cutoff frequencies of the first LPF 251, the second LPF 252, and the third LPF 253 are 3 Hz.

[0062] Next, the first absolute value section 261, the second absolute value section 262, and the third absolute value section 263 calculate the absolute values ​​of the output value from the first LPF 251, the output value from the second LPF 252, and the output value from the third LPF 253, respectively. Through the processing by the first absolute value section 261, the second absolute value section 262, and the third absolute value section 263, waveform data such as that shown in FIG.

[0063] Next, the comparison unit 266 compares the output values ​​(absolute values ​​of the vibration values) from the first absolute value unit 261 to the third absolute value unit 263 with the first threshold value (9.8 m / s ２ ) and the output values ​​(absolute values ​​of the vibration values) from the first absolute value unit 261 to the third absolute value unit 263. Then, when all of the output values ​​(absolute values ​​of the vibration values) from the first absolute value unit 261 to the third absolute value unit 263 exceed the first threshold, the comparison unit 266 detects that lightning has struck the wind turbine generator 100. Furthermore, when at least one of the output values ​​from the first absolute value unit 261 to the third absolute value unit 263 is equal to or less than the first threshold, the comparison unit 266 determines that lightning has not struck the wind turbine generator 100.

[0064] When the comparison unit 266 detects that lightning has struck the wind turbine generator 100, it transmits a lightning signal to the correlation unit 268 and the communication unit 274. The lightning signal is a signal indicating that lightning has been detected striking the wind turbine generator 100. The communication unit 274 transmits the lightning signal to the nacelle monitoring device 300. When the nacelle monitoring device 300 receives the lightning signal, the determination unit 310 of the nacelle monitoring device 300 executes the above-described nacelle abnormality determination process.

[0065] Furthermore, upon receiving a lightning strike signal, the correlation unit 268 calculates correlation coefficients σ12, σ23, and σ31 using the first time-series data, the second time-series data, and the third time-series data from the signal input unit 250 (see the above formula (1) and the description of FIG. 10 ). The determination unit 270 then determines whether the correlation coefficients σ12, σ23, and σ31 exceed the second threshold. If the number of correlation coefficients σ12, σ23, and σ31 that are less than the second threshold is two or more (as in FIG. 10(B) or FIG. 10(C)), an abnormality is detected among the three blades (at least one of the three blades is determined to be abnormal). On the other hand, if the number of correlation coefficients σ12, σ23, and σ31 that are less than the second threshold is less than one (as in FIG. 10(A)), the determination unit 270 detects that the three blades are normal (all three blades are determined to be normal).

[0066] When the determination unit 270 determines that the three blades are abnormal, the generation unit 272 generates an abnormality signal. Then, the communication unit 274 outputs the signal to the external device 400. The external device 400 is, for example, a management device of the manager of the wind turbine generator 100, or an owner device of the owner of the wind turbine generator 100. Furthermore, when the determination unit 270 determines that the three blades are normal, the generation unit 272 generates a normality signal and outputs it to the external device 400. When the external device 400 receives the abnormality signal, it notifies the owner or manager that the three blades are abnormal. When the external device 400 receives the normality signal, it notifies the owner or manager that the three blades are normal. The external device 400 and the nacelle monitoring device 300 correspond to the "external equipment" in this disclosure.

[0067] [Flowchart] Figure 12 is a flowchart showing the flow of main processes of the state monitoring device 200 and the like. The state monitoring device 200 executes the processes of steps S102, S104, and S106 in parallel. Step S102 is a process in which the state monitoring device 200 acquires first time series data for the above-mentioned first collection period from the first vibration sensor 111. Step S104 is a process in which the state monitoring device 200 acquires second time series data for the above-mentioned first collection period from the second vibration sensor 112. Step S106 is a process in which the state monitoring device 200 acquires third time series data for the above-mentioned first collection period from the third vibration sensor 113.

[0068] Next, in step S108, the status monitoring device 200 executes the above-described lightning strike detection process based on the first time-series data, the second time-series data, and the third time-series data. Next, in step S110, the status monitoring device 200 determines whether or not a lightning strike to the wind turbine generator 100 has been detected in the lightning strike detection process of step S108.

[0069] If no lightning strike is detected in step S110 (NO in step S110), the process in Fig. 12 ends. On the other hand, if a lightning strike is detected in step S110 (YES in step S110), the process proceeds to step S111.

[0070] In step S111, the state monitoring device 200 outputs a lightning strike signal to an external device (for example, the nacelle monitoring device 300). The process then proceeds to steps S112 and S114. The lightning strike signal in FIG. 11 is sent to the correlation unit 268, thereby executing the process of step S112, and the lightning strike signal is sent to the nacelle monitoring device 300, thereby executing the process of step S114.

[0071] In step S112, the state monitoring device 200 executes the blade abnormality determination process described above. In step S114, the nacelle monitoring device 300 executes the nacelle abnormality determination process described above. The nacelle monitoring device 300 transmits the result of the nacelle abnormality determination process to the external device 400 described above.

[0072] Next, in step S116, the state monitoring device 200 determines whether or not an abnormality has been detected in the blade abnormality determination process in step S112 and the nacelle abnormality determination process in step S114.

[0073] If an abnormality is detected (YES in step S116), in step S118, the status monitoring device 200 outputs an abnormality signal to the external device 400. The abnormality signal includes information indicating the location where the abnormality occurred (the blade and the mechanism inside the nacelle 105). If no abnormality is detected (NO in step S116), in step S120, the status monitoring device 200 outputs a normality signal to the external device 400. As a variant, in step S120, the status monitoring device 200 may not output the normality signal.

[0074] [Summary] (1) As described above, the condition monitoring device 200 detects a lightning strike to the wind turbine generator 100 when all of the absolute values ​​of the vibration values ​​of the M blades exceed the first threshold. Therefore, the condition monitoring device 200 can detect a lightning strike to the wind turbine generator 100 without requiring a special device such as the lightning strike detection device described above. Therefore, the condition monitoring device 200 of this embodiment can detect a lightning strike while keeping manufacturing costs down.

[0075] (2) As described in FIG. 11 , the first LPF 251, the second LPF 252, and the third LPF 253 perform filtering to suppress noise in the first time-series data, the second time-series data, and the third time-series data, respectively. Therefore, even if noise is mixed in the time-series data output by the three vibration sensors, the noise can be suppressed. Therefore, the condition monitoring device 200 of this embodiment can improve the accuracy of lightning strike detection.

[0076] (3) As explained with reference to FIG. 7, the first threshold is 9.8 m / s ２ According to this configuration, the manufacturer of the state monitoring device 200 does not need to determine the first threshold value through experiments or the like, and therefore the burden of determining the first threshold value can be reduced.

[0077] 11, the first absolute value unit 261 to the third absolute value unit 263 calculate the absolute values ​​of the first time-series data to the third time-series data, respectively. Then, when all of the absolute values ​​exceed the first threshold, the status monitoring device 200 detects a lightning strike to the wind turbine generator.

[0078] When lightning strikes the wind turbine generator 100, the vibration values ​​detected by the three vibration sensors may become excessively negative. To detect such a case, the status monitoring device 200 may be configured to store both a positive first threshold value and a negative first threshold value. However, such a configuration may increase the storage capacity used to store the first threshold values. However, the status monitoring device 200 can calculate the absolute value, thereby making the vibration value a positive value. Therefore, the status monitoring device 200 does not need to store the negative first threshold value.

[0079] (5) Furthermore, when the absolute values ​​of the vibration values ​​of the three blades exceed the first threshold value at the same time, the condition monitoring device 200 detects a lightning strike to the wind turbine generator 100. When lightning strikes the wind turbine generator 100, the absolute values ​​of the vibration values ​​of the three blades exceed the first threshold value at the same time. Therefore, with this configuration, the condition monitoring device 200 can accurately detect a lightning strike to the wind turbine generator 100.

[0080] (6) Furthermore, the condition monitoring device 200 calculates three correlation parameters (correlation coefficient σ12, correlation coefficient σ23, and correlation coefficient σ31) when detecting a lightning strike to the wind turbine generator 100. Then, the condition monitoring device 200 detects an abnormality in the three blades when two or more of the three correlation parameters are less than the second threshold value.

[0081] For example, when a lightning strike on the wind turbine generator 100 is detected, a confirmation process may be executed to confirm whether or not the three blades have been affected by the lightning strike (such as damage). Such confirmation process may involve a configuration in which data before the lightning strike was detected (such as blade vibration data) is compared with data after the lightning strike was detected. However, in such a configuration, the control device must store the data before the lightning strike was detected. In contrast, the above-described configuration detects an abnormality in the three blades using the correlation between the vibration values ​​of the three blades. Therefore, the condition monitoring device 200 can detect an abnormality in the three blades without storing data before the lightning strike was detected.

[0082] (7) In this embodiment, the correlation parameter is a correlation coefficient. Therefore, an abnormality in multiple blades can be detected using a known correlation coefficient.

[0083] (8) Furthermore, the nacelle monitoring device 300 executes the above-described nacelle abnormality determination process when it receives a lightning strike signal from the state monitoring device 200. When lightning strikes the wind turbine generator 100, the mechanisms installed in the nacelle 105 may be damaged. Therefore, the state monitoring device 200 can cooperate with the nacelle monitoring device 300 to execute abnormality determination process not only for the blades but also for the mechanisms installed in the nacelle 105.

[0084] [Modification] (A) In the above example, the first threshold is 9.8 m / s ２ However, the first threshold may be another value. For example, time-series data in a state where there is no impulse waveform due to an induced lightning surge may be collected in advance, and the first threshold may be determined from the collected time-series data. For example, the previously collected time-series data may be filtered using an LPF, and the maximum value after the filtering process may be set as the first threshold.

[0085] (B) In the above example, the second collection period is a predetermined period. However, the status monitoring device 200 may change the second collection period, for example, in accordance with the rotation period of the hub 104. For example, the minimum value of the second collection period may be determined by the following equation (2):

[0086] Minimum value of second collection period=rotation period×(1+((M−1) / M)) (2) In this embodiment, the number of blades is three, so the minimum value of the second collection period is rotation period×(5 / 3). The maximum value of the second collection period is determined appropriately depending on the specifications of the condition monitoring device 200, such as the data capacity.

[0087] Next, an example of a method for calculating the rotation period of equation (2) will be described. For example, the state monitoring device 200 performs a fast Fourier transform (FFT) on the output values ​​of the first LPF 251 to the third LPF 253 in FIG. 11 . The state monitoring device 200 then identifies the peak value obtained by the fast Fourier transform as the frequency corresponding to the rotation period. The state monitoring device 200 determines the reciprocal of the identified frequency as the rotation period.

[0088] (C) The three correlation coefficients described above may be calculated as follows. For example, the condition monitoring device 200 may use a frequency filter on the three pieces of time-series data to determine the correlation coefficients of each blade in a specific frequency band. Alternatively, the condition monitoring device 200 may determine the correlation coefficients of each blade in multiple frequency bands. Furthermore, when such a configuration is employed, the condition monitoring device 200 may determine an abnormality using a second threshold for each of the multiple frequency bands.

[0089] (D) The condition monitoring device 200 may store the calculated correlation coefficients as statistical data. This storage location may be the condition monitoring device 200 or the above-mentioned management device. The second threshold may be determined from the stored statistical data. For example, the second threshold may be set to ⅓ of the median absolute deviation of the multiple correlation coefficients indicated by the statistical data. The second threshold may be set from a statistical quantity such as the average value or standard deviation of the multiple correlation coefficients, or may be set as an absolute evaluation value based on past performance. Although the second thresholds for the correlation coefficients σ12, σ23, and σ31 are set to a common value (0.6) in the above-described configuration, the second thresholds for the correlation coefficients σ12, σ23, and σ31 may be set individually.

[0090] 11 , the condition monitoring device 200 is configured to include the first absolute value unit 261 to the third absolute value unit 263. However, the first absolute value unit 261 to the third absolute value unit 263 may not be included (the absolute values ​​of the blade vibration values ​​may not be calculated). Even with this configuration, the condition monitoring device 200 detects a lightning strike to the wind turbine generator 100 when all of the absolute values ​​of the vibration values ​​of the multiple blades exceed the first threshold value.

[0091] (F) The above correlation parameter may be other parameters instead of the correlation coefficient. For example, the correlation parameter may be the degree of similarity of the waveforms in FIG.

[0092] (G) In the above embodiment, the configuration has been described in which the abnormality of the wind turbine generator 100 is a lightning strike to the wind turbine generator. However, the abnormality of the wind turbine generator 100 may be another abnormality, for example, an abnormality caused by a gust of wind hitting the wind turbine generator 100. An abnormality caused by a gust of wind hitting the wind turbine generator 100 may be, for example, damage to the blades due to the gust of wind.

[0093] [Notes] It will be understood by those skilled in the art that the exemplary embodiments described above are specific examples of the following aspects.

[0094] (Item 1) A condition monitoring device disclosed herein monitors the condition of a wind turbine generator having multiple blades. The condition monitoring device includes a signal input unit, a control device, and an external interface. The signal input unit acquires vibration values ​​of the multiple blades from multiple vibration sensors that detect vibration values ​​of the multiple blades. The control device detects an abnormality in the wind turbine generator when all absolute values ​​of the vibration values ​​of the multiple blades exceed a first threshold. The external interface transmits a detection signal indicating that an abnormality in the wind turbine generator has been detected to an external device.

[0095] With this configuration, it is possible to detect abnormalities in the wind turbine generator without requiring a special device such as the above-mentioned detection device. Therefore, the condition monitoring device can detect abnormalities in the wind turbine generator while keeping manufacturing costs down.

[0096] (2) In the condition monitoring device according to the first aspect, the plurality of vibration sensors output time-series data of vibration values ​​detected by the vibration sensors, and the control device performs filtering on the time-series data to suppress noise.

[0097] With this configuration, even if noise is mixed in the time-series data output by the vibration sensor, it is possible to suppress the noise.

[0098] (Item 3) In the condition monitoring device according to item 2, the plurality of vibration sensors are acceleration sensors. The first threshold is 9.8 m / s ２ is.

[0099] With this configuration, the manufacturer of the condition monitoring device does not need to determine the first threshold value through experiments or the like, thereby reducing the burden of determining the first threshold value.

[0100] (4) In the condition monitoring device according to any one of the first to third aspects, the control device calculates absolute values ​​of vibration values ​​of a plurality of blades, and when all of the absolute values ​​of the vibration values ​​of the plurality of blades exceed a first threshold, the control device detects a lightning strike to the wind turbine generator.

[0101] When lightning strikes a wind turbine generator, the vibration value detected by the vibration sensor may become an excessively negative value. To detect this case, the control device may be configured to store both a positive first threshold value and a negative first threshold value. However, such a configuration may increase the storage capacity used to store the first threshold values. In contrast, with the above configuration, the control device can calculate the absolute value of the vibration value, thereby making the vibration value a positive value. Therefore, the control device does not need to store the negative first threshold value.

[0102] (5) In the condition monitoring device described in any one of paragraphs 1 to 4, the control device detects a lightning strike to the wind turbine generator when the absolute values ​​of the vibration values ​​of multiple blades exceed a first threshold value at the same time.

[0103] When a lightning strike occurs on the wind turbine generator, the absolute values ​​of the vibration values ​​of the multiple blades exceed the first threshold value at the same time. Therefore, with this configuration, the condition monitoring device can accurately detect a lightning strike on the wind turbine generator.

[0104] (Item 6) In the condition monitoring device according to any one of Items 1 to 5, the plurality of blades are three or more blades. The plurality of vibration sensors are three or more vibration sensors. The three or more vibration sensors output time series data of vibration values ​​detected by the vibration sensors. The control device extracts time series data for a predetermined number of rotations of the plurality of blades from the time series data output by the three or more vibration sensors. When the control device detects a lightning strike to the wind turbine generator, it calculates a correlation parameter indicating the correlation between two possible time series data of the extracted time series data of the three or more blades for all combinations of the two time series data. The control device detects an abnormality in one of the plurality of blades when two or more correlation parameters of the correlation parameters of the two time series data for all combinations are less than a second threshold. The external interface transmits an abnormality signal indicating the detection of the abnormality to an external device.

[0105] For example, when a lightning strike on a wind turbine generator is detected, a confirmation process may be performed to check whether or not the lightning strike has caused abnormalities (such as damage) in multiple blades. Such confirmation process may involve a configuration in which data before the lightning strike (e.g., blade vibration data) is compared with data after the lightning strike is detected. However, in such a configuration, the control device must store the data before the lightning strike is detected. In contrast, the above-described configuration detects abnormalities in multiple blades using the correlation between the vibration values ​​of three or more blades. Therefore, a condition monitoring device having the above configuration can detect abnormalities in multiple blades without storing data before the lightning strike is detected.

[0106] (Item 7) In the condition monitoring device according to item 6, the correlation parameter is a correlation coefficient.

[0107] With this configuration, it is possible to detect abnormalities in multiple blades using known correlation coefficients.

[0108] (Item 8) In the condition monitoring device according to any one of items 1 to 7, the abnormality in the wind turbine generator is a lightning strike to the wind turbine generator.

[0109] With this configuration, it is possible to detect lightning strikes on the wind turbine generator. (Item 9) The wind turbine generator further comprises a nacelle, a mechanism installed in the nacelle, and a nacelle sensor that detects parameters for determining whether or not an abnormality exists in the mechanism. The nacelle monitoring device comprises an interface that communicates with the condition monitoring device described in any one of Items 1 to 8. The nacelle monitoring device comprises a determination device that, when receiving a detection signal from the condition monitoring device, determines whether or not an abnormality exists in the mechanism based on the parameters.

[0110] When a lightning strike is detected in a wind turbine generator, the mechanisms installed inside the nacelle may be damaged. Therefore, with the above configuration, when the condition monitoring device detects a lightning strike, it is possible to determine whether or not there is an abnormality in the mechanisms.

[0111] (10) A condition monitoring system includes the condition monitoring device according to any one of the first to eighth paragraphs and a plurality of vibration sensors.

[0112] (Item 11) A condition monitoring method disclosed herein is a method for monitoring the condition of a wind turbine generator having a plurality of blades. The condition monitoring method includes acquiring vibration values ​​of the plurality of blades. The condition monitoring method also includes detecting that lightning has struck the wind turbine generator when absolute values ​​of the vibration values ​​of the plurality of blades exceed a first threshold.

[0113] The embodiments disclosed herein should be considered to be illustrative in all respects and not restrictive. The scope of the present disclosure is defined by the claims, not by the description of the above embodiments, and is intended to include all modifications within the meaning and scope of the claims.

[0114] 100 Wind turbine generator, 101 First blade, 102 Second blade, 103 Third blade, 104 Hub, 105 Nacelle, 111 First vibration sensor, 112 Second vibration sensor, 113 Third vibration sensor, 121 Main shaft, 122 Main bearing, 123 Gearbox, 124 Generator, 126 Rotation sensor, 161 First down conductor, 162 Second down conductor, 163 Third down conductor, 180 Tower, 200 Status monitoring device, 240 Control device, 250 Acquisition unit, 261 First absolute value unit, 262 Second absolute value unit, 263 Third absolute value unit, 266 Comparison unit, 268 Correlation unit, 270 Determination unit, 272 Generation unit, 281 First memory unit, 282 Second memory unit, 300 Nacelle monitoring device, 310 Determination device, 500 condition monitoring system.

Claims

1. A condition monitoring device for monitoring the condition of a wind power generation device having a plurality of blades, comprising: a signal input unit that acquires vibration values of the plurality of blades from a plurality of vibration sensors that detect vibration values of the plurality of blades; a control device that detects an abnormality of the wind power generation device when all of the absolute values of the vibration values of the plurality of blades exceed a first threshold; and an external interface that transmits a detection signal indicating that an abnormality of the wind power generation device has been detected to an external device.

2. The plurality of vibration sensors output time-series data of the vibration values detected by the vibration sensors, and the control device performs a filtering process for suppressing noise on the time-series data. The condition monitoring device according to claim 1.

3. The plurality of vibration sensors are acceleration sensors, and the first threshold value is 9.8 m / s ２ The state monitoring device according to claim 2, wherein the state monitoring device is in the state monitoring device.

4. The control device calculates the absolute values of the vibration values of the plurality of blades, and detects a lightning strike on the wind power generation device when all of the absolute values of the vibration values of the plurality of blades exceed the first threshold. The condition monitoring device according to any one of claims 1 to 3.

5. The control device detects a lightning strike on the wind power generation device when the absolute values of the vibration values of the plurality of blades exceed the first threshold at the same timing. The condition monitoring device according to any one of claims 1 to 3.

6. The plurality of blades are three or more blades, the plurality of vibration sensors are three or more vibration sensors, the three or more vibration sensors output time-series data of the vibration values detected by the vibration sensors, and the control device extracts time-series data for a predetermined number of rotations of the plurality of blades from the time-series data output by the three or more vibration sensors. When a lightning strike on the wind power generation device is detected, a correlation parameter indicating the correlation between two time-series data in all combinations of two time-series data that can be obtained from the extracted time-series data of the three or more blades is calculated. When the number of correlation parameters less than a second threshold among the correlation parameters of the two time-series data in all combinations is two or more, an abnormality of any one of the plurality of blades is detected. The external interface transmits an abnormality signal indicating that an abnormality of the blade has been detected to an external device. The condition monitoring device according to any one of claims 1 to 3.

7. The state monitoring device according to claim 6, wherein the correlation parameter is a correlation coefficient.

8. The state monitoring device according to any one of claims 1 to 3, wherein the abnormality of the wind power generation device is a lightning strike on the wind power generation device.

9. The wind power generation device further includes a nacelle, a mechanism installed in the nacelle, and a nacelle sensor that detects a parameter for determining the presence or absence of an abnormality in the mechanism, an interface that communicates with the state monitoring device according to any one of claims 1 to 3, and a determination device that determines the presence or absence of an abnormality in the mechanism based on the parameter when receiving the detection signal from the state monitoring device. A nacelle monitoring device.

10. A state monitoring system comprising the state monitoring device according to any one of claims 1 to 3 and the plurality of vibration sensors.

11. A state monitoring method for monitoring the state of a wind power generation device having a plurality of blades, the state monitoring method comprising: obtaining vibration values of the plurality of blades; and detecting that there is an abnormality in the wind power generation device when an absolute value of the vibration values of the plurality of blades exceeds a first threshold value. And transmitting a detection signal indicating that an abnormality of the wind power generation device has been detected to an external device.

Images