An online monitoring and diagnostic system and method for wind turbine drivetrain faults

Abstract

This invention discloses an online monitoring and diagnosis system and method for wind turbine drivetrain faults. The system mainly includes multiple accelerometers, a digital signal processing unit, a wind turbine speed acquisition module, a data communication unit, and a remote server. The digital signal processing unit performs monitoring and diagnosis based on an online monitoring and diagnosis method for wind turbine drivetrain faults. Specifically, multiple accelerometers acquire acceleration signals from the wind turbine drivetrain; the acceleration signals are converted into vibration velocity signals using frequency domain integration; the vibration velocity signals are processed to obtain vibration trend values, and combined with a speed-adaptive vibration threshold, the system monitors wind turbine drivetrain faults in real time; when a fault occurs, the system calculates the energy characteristics of each frequency band at the location of the faulty wind turbine drivetrain based on the order bandwidth and the real-time speed of the gearbox, further confirming the fault; the method also periodically performs statistical analysis on the characteristic data to optimize the initial value of the vibration threshold. This invention improves the accuracy of wind turbine drivetrain fault monitoring and diagnosis.

Description

Technical Field

[0001] This invention belongs to the field of intelligent operation and maintenance of wind turbines, and specifically relates to an online monitoring and diagnosis system and method for wind turbine drive chain faults. Background Technology

[0002] Against the backdrop of both "dual carbon" goals and the construction of a new power system, my country's wind power industry has entered a period of rapid development, reaching approximately 330 million kilowatts by the end of 2021. This rapid growth has placed higher demands on the safety and reliability of wind power equipment. Wind power equipment is typically installed in high-wind-energy areas such as mountains, deserts, and wastelands, with a wide distribution, poor road conditions, and significant maintenance challenges. As a key component of wind power equipment, the wind turbine drivetrain operates under complex alternating loads and is susceptible to failure due to manufacturing and installation processes, leading to downtime and severely impacting the economic benefits of wind farms. Furthermore, wind farms operate on a "minimal staffing, unmanned operation" model, with streamlined on-site personnel and long inspection cycles, making it difficult to monitor the real-time operation of the wind turbine drivetrain. This can result in some faults going undetected and escalating. Therefore, fault monitoring and diagnosis of the wind turbine drivetrain, along with alarm activation when faults occur, are crucial for improving its service performance.

[0003] Currently, condition monitoring of wind turbine drivetrains primarily relies on Supervisory Control and Data Acquisition (SCADA) and Condition Monitoring Systems (CMS). These systems collect and analyze physical information such as temperature, current, and vibration in real time to determine the health status of the wind turbine drivetrain. This approach places high demands on the expertise of on-site wind farm personnel, requiring them to possess both theoretical knowledge and extensive engineering experience. Ordinary personnel often struggle to accurately diagnose drivetrain faults based on this physical information, frequently resulting in delayed fault detection and inaccurate fault assessments. Furthermore, the complex operating conditions of wind turbine drivetrains, including variations in speed and load, introduce uncertainty into fault monitoring thresholds, further complicating the accurate assessment of drivetrain health and hindering efficient and organized maintenance by on-site personnel. Moreover, the limited number of on-site personnel at wind farms increases the risk of delayed fault detection, leading to escalation of faults and triggering chain reactions that result in severe drivetrain failures, significant economic losses, and reduced economic benefits for the wind farm. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention proposes an online monitoring and diagnosis system and method for wind turbine drivetrain faults.

[0005] The technical solution of this invention is an online monitoring and diagnosis system for wind turbine drivetrain faults, comprising:

[0006] Multiple accelerometers, digital signal processing unit, wind turbine speed acquisition module, data communication unit, and remote server;

[0007] The plurality of accelerometers are connected in sequence to the digital signal processing unit; the wind turbine speed acquisition module is connected to the digital signal processing unit; the digital signal processing unit, the data communication unit, and the remote server are connected in sequence.

[0008] The multiple accelerometers are arranged sequentially at multiple positions in the wind turbine drive chain;

[0009] The multiple accelerometers are used to collect acceleration signals at multiple moments at each wind turbine drive chain position, and transmit the acceleration signals at multiple moments at each wind turbine drive chain position to the digital signal processing unit.

[0010] The wind turbine speed acquisition module is connected to the wind turbine nacelle control system and is used to acquire the speed of the gearbox input at multiple times and transmit the speed of the gearbox input at multiple times to the digital signal processing unit.

[0011] The digital signal processing unit converts the conditioned acceleration signals at multiple moments at each wind turbine drive chain position into multiple acceleration signals at each moment at each wind turbine drive chain position through analog-to-digital conversion. Combining the multiple acceleration signals at multiple moments at each wind turbine drive chain position with the speed increaser input speed at multiple moments, the unit calculates the real-time health status and real-time fault information using the online monitoring and diagnosis method for wind turbine drive chain faults. The digital signal processing unit then transmits the fault status of the wind turbine drive chain position to the remote server through the data communication unit.

[0012] The remote server provides fault warnings based on the fault status of the wind turbine drive chain position.

[0013] The technical solution of this invention is an online monitoring and diagnosis method for wind turbine drivetrain faults, comprising the following steps:

[0014] Step 1: The multiple accelerometers collect acceleration signals at multiple moments at each position of the wind turbine drive chain and transmit these acceleration signals to the digital signal processing unit; the wind turbine speed acquisition module collects the speed increaser input at multiple moments and transmits these speed increasers to the digital signal processing unit.

[0015] Step 2: The digital signal processing unit calculates the vibration velocity signal at multiple moments at each wind turbine drive chain position using frequency domain integration based on the acceleration signal at multiple moments at each wind turbine drive chain position.

[0016] Step 3: Calculate the vibration trend value of each fan drive chain position based on the vibration velocity signals at multiple moments at each fan drive chain position, and calculate the average input speed of the speed increaser based on the speed of the speed increaser at multiple moments;

[0017] Step 4: Repeat steps 1-3 multiple times to obtain the vibration trend value of each fan drive chain position and the average input speed of the speed increaser in multiple executions. Manually define the minimum and maximum values ​​of the average input speed of the speed increaser. Divide the average speed range into N evenly distributed average speed intervals. Combine the value range of each average speed interval to divide the average input speed of the speed increaser into the corresponding average speed interval, and obtain the average input speed of the speed increaser for all executions in each average speed interval. Combine the execution number to obtain the vibration trend value of each fan drive chain position in each average speed interval. Calculate the normal distribution probability of the vibration trend values ​​of each fan drive chain position in each average speed interval to obtain the mean and standard deviation of the vibration trend of each fan drive chain position in each average speed interval.

[0018] Step 5: The speed of the speed increaser input at multiple moments is collected in real time and transmitted to the digital signal processing unit. The digital signal processing unit calculates the real-time average speed of the speed increaser according to Step 3, and combines the value range of each average speed interval described in Step 4 to divide the real-time average speed of the speed increaser into the average speed interval corresponding to each position of the wind turbine drive chain. The vibration trend value threshold of each wind turbine drive chain position is calculated by combining the mean of the vibration trend of the corresponding average speed interval of each wind turbine drive chain position in Step 4 and the standard deviation of the vibration trend of the corresponding average speed interval of each wind turbine drive chain position.

[0019] Step 6: The multiple accelerometers collect acceleration signals at multiple moments at each wind turbine drive chain position in real time, and transmit the acceleration signals at multiple moments at each wind turbine drive chain position to the digital signal processing unit; the digital signal processing unit calculates the vibration trend value of the wind turbine drive chain position according to steps 2 and 3. If the vibration trend value of the wind turbine drive chain position exceeds the vibration trend value threshold of the wind turbine drive chain position, the corresponding wind turbine drive chain position is determined to be in a potential hazard state;

[0020] Step 7: If the position of the wind turbine drive chain is in a potential hazard state, use Fast Fourier Transform to calculate the instantaneous spectrum of the vibration velocity signal at the position of the wind turbine drive chain at multiple moments.

[0021] Step 8: Define the bandwidth of each order of the fan drive train position, and calculate the frequency range corresponding to each order bandwidth of the fan drive train position based on the real-time average input speed of the speed increaser and the bandwidth of each order of the fan drive train position.

[0022] Step 9: Within the frequency range corresponding to each order bandwidth of the wind turbine drive chain position, calculate the root mean square value of the instantaneous spectrum of the vibration velocity signal corresponding to each order bandwidth of the wind turbine drive chain position.

[0023] Step 10: Calculate the frequency band alarm threshold corresponding to each order bandwidth of the wind turbine drive chain position. If the root mean square value of the instantaneous spectrum of the vibration velocity signal corresponding to the order bandwidth of the wind turbine drive chain position exceeds the frequency band alarm threshold corresponding to the order bandwidth of the wind turbine drive chain position, then the wind turbine drive chain position status is determined to be a fault state.

[0024] Preferably, step 2 involves calculating the vibration velocity signals at multiple moments for each wind turbine drive chain position, specifically as follows:

[0025] A i ={a i,1 ,a i,2 ,a i,3 ,…a i,T}, i∈[1,M]

[0026] Among them, A i Let M be the set of vibration velocity signals at multiple moments for the i-th position of the wind turbine drive train, where M is the number of positions in the wind turbine drive train, and a i,t Let be the vibration velocity signal of the i-th wind turbine drive chain position at time t, where t∈[1,T] and T is the number of time sequences;

[0027] Preferably, step 3, calculating the vibration trend value at each position of the wind turbine drive train, specifically involves:

[0028]

[0029] Among them, B i Let be the vibration trend value at the i-th position of the fan drive train;

[0030] Step 3, calculating the average input speed of the speed increaser, specifically involves:

[0031]

[0032] Where S represents the average input speed of the speed increaser, st Input the rotational speed of the speed increaser at time t, where t∈[1,T];

[0033] Preferably, step 4 involves uniformly dividing the area into N average rotational speed intervals, specifically as follows:

[0034] {[n min ,n min +2],[n min +2,n min +4],[n min +4,n min +6],…,[n max -2,n max ]}

[0035] Where, n min n max Let n represent the minimum and maximum average input speeds of the gearbox, respectively. min +2(j-1),n min +2j] represents the j-th average speed interval, j∈[0,N], N=floor((n max -n min ) / 2, floor() rounds down;

[0036] Step 4 involves calculating the mean value of the vibration trend for each average speed range at each position of the wind turbine drive train based on the normal distribution probability. Specifically:

[0037]

[0038] Where, μ i,j S represents the mean of the vibration trend values ​​within the j-th average speed range of the i-th fan drive train position; j,e B represents the average input speed of the speed increaser during the e-th execution within the j-th average speed range. i,j,e L represents the vibration trend value of the e-th execution within the j-th average speed range of the i-th fan drive train position. i,j The number of vibration trend values ​​within the j-th average speed range of the i-th fan drive train position, where st represents the constraint condition;

[0039] Step 4 describes calculating the standard deviation of the vibration trend for each average speed range at each position of the wind turbine drive train based on the normal distribution probability. Specifically:

[0040]

[0041] Where, σ i,j The standard deviation of the vibration trend value in the j-th average speed range at the i-th fan drive train position is represented.

[0042] Preferably, step 5, which involves calculating the vibration trend threshold value for each position in the wind turbine drive train, specifically involves:

[0043] BL i =μ i +ησ i , i∈[1,M], η∈{2,3}

[0044] Among them, BL i The threshold value representing the vibration trend at the i-th position of the wind turbine drive train; μ i , σ i These are the mean and standard deviation of the vibration trend of the i-th fan drive chain position described in step 4, respectively, within the average speed range to which the real-time average input speed of the speed increaser belongs. η is the 3σ principle coefficient in the Gaussian distribution.

[0045] Preferably, the instantaneous spectrum of the vibration velocity signal at the position of the wind turbine drive chain calculated in step 7 is as follows:

[0046] {f y,1 ,f y,2 ,……f y,H}, H = floor(K / 2), y ∈ [1, Y]

[0047] Among them, f y,h The frequency h of the vibration velocity signal at the position of the fan drive chain in the y-th hidden danger state is represented by the spectral characteristics, where h∈[1,H], K is the sampling frequency, Y is the number of fan drive chain positions in the hidden danger state, and H is the number of spectral sequences;

[0048] Preferably, the bandwidth of each order of the wind turbine drive train position described in step 8 is defined as:

[0049] {[o y,1 ,o y,2 ],……,[o y,Q-1 ,o y,Q ]},y∈[1,Y]

[0050] Among them, [o y,q ,o y,q+1 ] represents the q-th order bandwidth of the wind turbine drive train position in the y-th hidden danger state described in step 6, q∈[1,Q-1], where Q is the number of the order bandwidths;

[0051] The frequency range mentioned in step 8 is specifically as follows:

[0052] {[w y,1 ,w y,2 ],……,[w y,q-1 ,wy,Q ]},y∈[1,Y],w y,q =z·θ·o y,q

[0053] Among them, [w y,q ,w y,q+1 ] represents the bandwidth of the q-th order [o y,q ,o y,q+1 The corresponding frequency range, θ represents the real-time average input speed of the speed increaser, and z represents the transmission ratio of each gear transmission in the speed increaser;

[0054] Preferably, the instantaneous root mean square value of the vibration velocity signal corresponding to each order bandwidth of the wind turbine drive train position in step 9 is defined as:

[0055]

[0056] Among them, FB y,q f represents the root mean square value of the instantaneous spectrum of the vibration velocity signal with the q-th order bandwidth representing the position of the wind turbine drive train in the y-th potential hazard state. y,v This represents the spectral characteristics of frequency v in the instantaneous spectrum of the wind turbine drivetrain position at the y-th potential hazard state, where v∈[1,V] and V represents the frequency range [w y,q-1 ,w y,q The number of corresponding spectral features within ];

[0057] Preferably, the frequency band alarm threshold corresponding to each order bandwidth of the wind turbine drive train position in step 10 is defined as:

[0058] AL y,q =λ y,q BL y ,y∈[1,Y]

[0059] Among them, AL y,q [o] represents the q-th order bandwidth of the wind turbine drive train position for the y-th potential hazard state. y,q-1 ,o y,q The corresponding frequency band alarm threshold, λ y,q The coefficient representing the proportion of the bandwidth of the qth order of the wind turbine drive train position at the yth potential hazard state.

[0060] The advantages of this invention are: it uses a speed-adaptive vibration threshold for online monitoring and diagnosis of wind turbine drivetrain faults; it periodically collects statistical data on the characteristic data of each wind turbine drivetrain position under normal operating conditions to optimize the initial value of the speed-adaptive vibration threshold; it adopts a multi-level fault diagnosis mode of "on-site-remote" to detect wind turbine drivetrain faults in a timely manner; it reduces the dependence on manual labor for wind turbine drivetrain fault monitoring and diagnosis, improves the accuracy of wind turbine drivetrain fault monitoring and diagnosis, and ensures the safe and reliable operation of the wind turbine drivetrain during its service life. Attached Figure Description

[0061] Figure 1 System structure block diagram of an embodiment of the present invention;

[0062] Figure 2 : Arrangement diagram of multiple accelerometers according to an embodiment of the present invention;

[0063] Figure 3 : Flowchart of the method according to an embodiment of the present invention;

[0064] Figure 4 Flowchart of the frequency domain integration method according to an embodiment of the present invention;

[0065] Figure 5 Topology diagram of the remote server solution in this embodiment of the invention. Detailed Implementation

[0066] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0067] In specific implementation, the method proposed in the technical solution of this invention can be automatically executed by those skilled in the art using computer software technology. System devices for implementing the method, such as computer-readable storage media storing the corresponding computer program of the technical solution of this invention and computer equipment including the computer program running the corresponding computer program, should also be within the protection scope of this invention.

[0068] like Figure 1 The diagram shown is a schematic representation of the system structure according to an embodiment of the present invention. The specific technical solution of the system according to an embodiment of the present invention is as follows:

[0069] An online monitoring and diagnostic system for wind turbine drivetrain faults includes:

[0070] Multiple accelerometers, digital signal processing unit, wind turbine speed acquisition module, data communication unit, and remote server;

[0071] The plurality of accelerometers are connected in sequence to the digital signal processing unit; the wind turbine speed acquisition module is connected to the digital signal processing unit; the digital signal processing unit, the data communication unit, and the remote server are connected in sequence.

[0072] like Figure 2 The diagram shows multiple accelerometers arranged sequentially at multiple positions in the wind turbine drive chain, according to an embodiment of the present invention.

[0073] The accelerometers are arranged in key components of the wind turbine drive chain, such as the main shaft, speed increaser, generator, and bearings. These include the front main bearing (radial), rear main bearing (radial and axial), gearbox input shaft bearing (radial), gearbox internal gear ring (radial), gearbox intermediate shaft generator side bearing (radial), gearbox high-speed shaft generator side bearing (radial and axial), generator drive end bearing (radial), and generator non-drive end bearing (radial). Low-frequency accelerometers are used for the radial, radial, and axial sections of the front and rear main shaft support bearings.

[0074] All of the accelerometers selected are IEPE type accelerometers;

[0075] The digital signal processing unit is selected as a DSP digital signal processing module, an AD sampling module, and a PLC;

[0076] The fan speed acquisition module is selected as a DSP digital signal processing module;

[0077] The data communication unit is selected as a gigabit managed industrial Ethernet switch.

[0078] The remote server is selected as a rack-mount server;

[0079] The multiple accelerometers are used to collect acceleration signals at multiple moments at each wind turbine drive chain position, and transmit the acceleration signals at multiple moments at each wind turbine drive chain position to the digital signal processing unit.

[0080] The wind turbine speed acquisition module is connected to the wind turbine nacelle control system and is used to acquire the speed of the gearbox input at multiple times and transmit the speed of the gearbox input at multiple times to the digital signal processing unit.

[0081] The digital signal processing unit converts the conditioned acceleration signals at multiple moments at each wind turbine drive chain position into multiple acceleration signals at each moment at each wind turbine drive chain position through analog-to-digital conversion. Combining the multiple acceleration signals at multiple moments at each wind turbine drive chain position with the speed of the gearbox at multiple moments, the unit calculates the real-time health status and real-time fault information using the online monitoring and diagnosis method for wind turbine drive chain faults. The digital signal processing unit then transmits the fault status of the wind turbine drive chain position to the remote server through the data communication unit.

[0082] The remote server provides fault warnings based on the fault status of the wind turbine drive chain position.

[0083] The following is combined with Figure 3 This invention introduces an online monitoring and diagnosis method for wind turbine drivetrain faults, as detailed below:

[0084] Step 1: The multiple accelerometers collect acceleration signals at multiple moments at each position of the wind turbine drive chain and transmit these acceleration signals to the digital signal processing unit; the wind turbine speed acquisition module collects the speed increaser input at multiple moments and transmits these speed increasers to the digital signal processing unit.

[0085] Step 2: The digital signal processing unit processes the acceleration signals at multiple moments from each wind turbine drivetrain position using methods such as... Figure 4 The frequency domain integration method shown is used to calculate the vibration velocity signals at multiple moments for each position of the wind turbine drive chain;

[0086] Step 2 involves calculating the vibration velocity signals at multiple moments for each position in the wind turbine drive train, specifically as follows:

[0087] A i ={a i,1 ,a i,2 ,a i,3 ,…a i,T}, i∈[1,M]

[0088] Among them, A i Let M = 10 be the set of vibration velocity signals at multiple moments for the i-th wind turbine drive train position, where a is the number of wind turbine drive train positions. i,t Let be the vibration velocity signal of the i-th wind turbine drive chain position at time t, where t∈[1,T] and T=1024 is the number of time sequences;

[0089] Step 3: Calculate the vibration trend value of each fan drive chain position based on the vibration velocity signals at multiple moments at each fan drive chain position, and calculate the average input speed of the speed increaser based on the speed of the speed increaser at multiple moments;

[0090] Step 3, calculating the vibration trend value at each position of the wind turbine drive chain, specifically involves:

[0091]

[0092] Among them, B i Let be the vibration trend value at the i-th position of the fan drive train;

[0093] Step 3, calculating the average input speed of the speed increaser, specifically involves:

[0094]

[0095] Where S represents the average input speed of the speed increaser, s t Input the rotational speed of the speed increaser at time t, where t∈[1,T];

[0096] Step 4: Repeat steps 1-3 multiple times to obtain the vibration trend value of each fan drive chain position and the average input speed of the speed increaser in multiple executions. Manually define the minimum and maximum values ​​of the average input speed of the speed increaser. Divide the average speed range into N evenly distributed average speed intervals. Combine the value range of each average speed interval to divide the average input speed of the speed increaser into the corresponding average speed interval, and obtain the average input speed of the speed increaser for all executions in each average speed interval. Combine the execution number to obtain the vibration trend value of each fan drive chain position in each average speed interval. Calculate the normal distribution probability of the vibration trend values ​​of each fan drive chain position in each average speed interval to obtain the mean and standard deviation of the vibration trend of each fan drive chain position in each average speed interval.

[0097] Step 4, which involves uniformly dividing the area into N average speed intervals, specifically involves:

[0098] {[n min ,n min +2],[n min +2,n min +4],[n min +4,n min +6],…,[n max -2,n max ]}

[0099] Where, n min =3, n max =20 represent the minimum and maximum average input speeds of the gearbox, respectively, [n min +2(j-1),nmin +2j] represents the j-th average speed interval, j∈[0,N], N=floor((n max -n min ) / 2 = 8, floor() rounds down;

[0100] Step 4 involves calculating the mean value of the vibration trend for each average speed range at each position of the wind turbine drive train based on the normal distribution probability. Specifically:

[0101]

[0102] Where, μ i,j S represents the mean of the vibration trend values ​​within the j-th average speed range of the i-th fan drive train position; j,e B represents the average input speed of the speed increaser during the e-th execution within the j-th average speed range. i,j,e L represents the vibration trend value of the e-th execution within the j-th average speed range of the i-th fan drive train position. i,j The number of vibration trend values ​​within the j-th average speed range of the i-th fan drive train position;

[0103] Step 4 describes calculating the standard deviation of the vibration trend for each average speed range at each position of the wind turbine drive train based on the normal distribution probability. Specifically:

[0104]

[0105] Where, σ i,j The standard deviation of the vibration trend value in the j-th average speed range at the i-th fan drive train position is represented.

[0106] Step 5: The speed of the gearbox is collected in real time at multiple moments. The collected speed of the gearbox is transmitted to the digital signal processing unit. The digital signal processing unit calculates the real-time average input speed of the gearbox according to Step 3. Combining the value range of each average speed interval mentioned in Step 4, the real-time average input speed of the gearbox is divided into the average speed interval corresponding to each position of the fan drive chain. Combining the mean of the vibration trend of the corresponding average speed interval of each position of the fan drive chain in Step 4 and the standard deviation of the vibration trend of the corresponding average speed interval of each position of the fan drive chain, the vibration trend value threshold of each position of the fan drive chain is calculated. st represents the constraint condition.

[0107] Step 5, which involves calculating the vibration trend threshold value for each position in the wind turbine drive train, specifically involves:

[0108] BL i =μ i +ησ i, i∈[1,M], η∈{2,3}

[0109] Among them, BL i The threshold value representing the vibration trend at the i-th position of the wind turbine drive train; μ i , σ i These are the mean and standard deviation of the vibration trend of the i-th fan drive chain position described in step 4, respectively, within the average speed range to which the real-time average input speed of the speed increaser belongs. η is the 3σ principle coefficient in the Gaussian distribution.

[0110] Step 6: The multiple accelerometers collect acceleration signals at multiple moments at each wind turbine drive chain position in real time, and transmit the acceleration signals at multiple moments at each wind turbine drive chain position to the digital signal processing unit; the digital signal processing unit calculates the vibration trend value of the wind turbine drive chain position according to steps 2 and 3. If the vibration trend value of the wind turbine drive chain position exceeds the vibration trend value threshold of the wind turbine drive chain position, the corresponding wind turbine drive chain position is determined to be in a potential hazard state;

[0111] Step 7: If the position of the wind turbine drive chain is in a potential hazard state, use Fast Fourier Transform to calculate the instantaneous spectrum of the vibration velocity signal at the position of the wind turbine drive chain at multiple moments.

[0112] Step 7 describes the calculation of the instantaneous spectrum of the vibration velocity signal at the position of the wind turbine drive train, specifically as follows:

[0113] {f y,1 ,f y,2 ,……f y,H}, H = floor(K / 2), y ∈ [1, Y]

[0114] Among them, f y,h The frequency h of the vibration velocity signal at the position of the fan drive chain in the y-th hidden danger state represents the spectral characteristics of the frequency h, where K = 8192 is the sampling frequency, h ∈ [1, H], H = 4096, and Y is the number of fan drive chain positions in the hidden danger state.

[0115] Step 8: Define the bandwidth of each order of the fan drive train position, and calculate the frequency range corresponding to each order bandwidth of the fan drive train position based on the real-time average input speed of the speed increaser and the bandwidth of each order of the fan drive train position.

[0116] The bandwidth of each order of the wind turbine drive train position described in step 8 is defined as follows:

[0117] {[o y,1 ,o y,2 ],……,[o y,Q-1 ,o y,Q ]},y∈[1,Y]

[0118] Among them, [o y,q ,o y,q+1 ] represents the q-th order bandwidth of the wind turbine drive train position in the y-th hidden danger state described in step 6, q∈[1,Q-1], Q=8 is the number of the order bandwidths;

[0119] The frequency range mentioned in step 8 is specifically as follows:

[0120] {[w y,1 ,w y,2 ],……,[w y,q-1 ,w y,Q ]},y∈[1,Y],w y,q =z·θ·o y,q

[0121] Among them, [w y,q ,w y,q+1 ] represents the bandwidth of the q-th order [o y,q ,o y,q+1 The corresponding frequency range, θ represents the real-time average input speed of the speed increaser, and z represents the transmission ratio of each gear transmission in the speed increaser;

[0122] For example, to monitor faults in components such as the main shaft and bearings of a wind turbine drivetrain, the order bandwidth can be defined as:

[0123] Order bandwidth Starting Boundary Termination Limit Band0 0.2x 0.8x Band1 0.8x 1.2x Band2 1.2x 2.2x Band3 2.2x 3.2x Band4 0.8BPFO 1.2BPFO Band5 0.8BPFI 1.2BPFI Band6 0.8BSF 1.2BSF Band7 BNF-FTF BNF+FTF

[0124] To monitor faults in the wind turbine drive chain gear components, the order bandwidth of the gear component is defined as:

[0125] Order bandwidth Starting Boundary Termination Limit Band0 0.2x 0.8x Band1 <![CDATA[GMF-3O ca ]]> <![CDATA[GMF-0.5O ca ]]> Band2 <![CDATA[GMF-0.5O ca ]]> <![CDATA[GMF+0.5O ca ]]> Band3 <![CDATA[GMF+0.5O ca ]]> <![CDATA[GMF+3O ca ]]> Band4 <![CDATA[2GMF-3O ca ]]> <![CDATA[2GMF-0.5O ca ]]> Band5 <![CDATA[2GMF-0.5O ca ]]> <![CDATA[2GMF+0.5O ca ]]> Band6 <![CDATA[2GMF+0.5O ca ]]> <![CDATA[2GMF+3O ca <!-- 8 -->]]> Band7 0.8GNF <![CDATA[0.5f s ]]>

[0126] f in the table s Indicates the sampling frequency, O ca The order of rotation of the shaft is indicated; BPFO, BPFI, BSF, and FTF represent the order of the outer ring of the bearing, the order of the inner ring, the order of the rolling element spin, and the order of the cage rotation, respectively; GMF and GNF represent the order of gear meshing and the order of the natural frequency, respectively; Dx represents the D-fold frequency, such as 2GMF and 2x, which represent the meshing order and the order of rotation of the shaft, respectively.

[0127] Step 9: Within the frequency range corresponding to each order bandwidth of the wind turbine drive chain position, calculate the root mean square value of the instantaneous spectrum of the vibration velocity signal corresponding to each order bandwidth of the wind turbine drive chain position.

[0128] Step 9 describes calculating the root mean square value of the instantaneous spectrum of the vibration velocity signal corresponding to each order bandwidth of the wind turbine drive train position, defined as:

[0129]

[0130] Among them, FB y,q f represents the root mean square value of the instantaneous spectrum of the vibration velocity signal with the q-th order bandwidth representing the position of the wind turbine drive train in the y-th potential hazard state. y,v This represents the spectral characteristics of frequency v in the instantaneous spectrum of the wind turbine drivetrain position at the y-th potential hazard state, where v∈[1,V] and V represents the frequency range [w y,q-1 ,w y,q The number of corresponding spectral features within ];

[0131] Step 10: Calculate the frequency band alarm threshold corresponding to each order bandwidth of the wind turbine drive chain position. If the root mean square value of the instantaneous spectrum of the vibration velocity signal corresponding to the order bandwidth of the wind turbine drive chain position exceeds the frequency band alarm threshold corresponding to the order bandwidth of the wind turbine drive chain position, then the wind turbine drive chain position status is determined to be a fault state.

[0132] Step 10, which calculates the frequency band alarm threshold corresponding to each order bandwidth of the wind turbine drive train position, is defined as:

[0133] AL y,q =λ y,q BL y ,y∈[1,Y]

[0134] Among them, AL y,q [o] represents the q-th order bandwidth of the wind turbine drive train position for the y-th potential hazard state. y,q-1 ,o y,q The corresponding frequency band alarm threshold, λ y,q The coefficient representing the proportion of the bandwidth of the qth order of the wind turbine drive train position at the yth potential hazard state.

[0135] This invention provides an online monitoring and diagnosis system and method for wind turbine drivetrain faults. It employs a speed-adaptive vibration threshold method to monitor and diagnose wind turbine drivetrain faults in real time. Taking into full account the complex alternating operating conditions of the wind turbine drivetrain, it monitors the vibration acceleration and speed information during operation, preprocessing to obtain characteristic data characterizing the health status of the wind turbine drivetrain: vibration trend values ​​and frequency band energy characteristics. Simultaneously, it combines these with an initial value of the speed-adaptive vibration threshold to monitor wind turbine drivetrain faults. When potential faults are detected, it further diagnoses the faults using frequency band energy characteristics, thereby achieving real-time graded alarms. Furthermore, as... Figure 5The diagram shows the topology of the remote server solution. The remote server is connected to each wind turbine in the wind farm via a fiber optic ring network, enabling centralized management and control of drivetrain faults in each turbine. After firewall isolation, fault monitoring data is synchronized via a dedicated line, allowing internal and remote experts to analyze and process the data for precise diagnosis and analysis of the wind turbine drivetrain health status, generating diagnostic and maintenance recommendations, and enabling remote expert diagnosis of wind turbine drivetrain faults. This invention improves the reliability of wind turbine drivetrain fault monitoring and the accuracy of fault diagnosis, reduces the professional maintenance requirements for on-site production personnel, helps to promptly eliminate wind turbine drivetrain faults and defects, improves the effectiveness and scientific nature of wind turbine drivetrain maintenance, and contributes to the realization of "digital" and "intelligent" management of wind farms.

[0136] It should be understood that any parts not described in detail in this specification belong to the prior art.

[0137] Although this document uses numerous terms such as accelerometer, digital signal processing unit, wind turbine speed acquisition module, data communication unit, and remote server, the possibility of using other terms is not excluded. These terms are used merely for the convenience of describing the essence of this invention, and interpreting them as any additional limitation would contradict the spirit of this invention.

[0138] It should be understood that the above description of the preferred embodiments is quite detailed, but it should not be considered as a limitation on the scope of protection of this invention. Those skilled in the art, under the guidance of this invention, can make substitutions or modifications without departing from the scope of protection of the claims of this invention, and all such substitutions or modifications fall within the scope of protection of this invention. The scope of protection of this invention should be determined by the appended claims.

Claims

1. A method for online monitoring and diagnosis of wind turbine drivetrain faults based on a monitoring system, characterized in that: The monitoring system includes: multiple accelerometers, a digital signal processing unit, a wind turbine speed acquisition module, a data communication unit, and a remote server; The plurality of accelerometers are connected in sequence to the digital signal processing unit; the wind turbine speed acquisition module is connected to the digital signal processing unit; the digital signal processing unit, the data communication unit, and the remote server are connected in sequence. The multiple accelerometers are arranged sequentially at multiple positions in the wind turbine drive chain; The multiple accelerometers are used to collect acceleration signals at multiple moments at each wind turbine drive chain position, and transmit the acceleration signals at multiple moments at each wind turbine drive chain position to the digital signal processing unit. The wind turbine speed acquisition module is connected to the wind turbine nacelle control system and is used to acquire the speed of the gearbox input at multiple times and transmit the speed of the gearbox input at multiple times to the digital signal processing unit. The digital signal processing unit converts the conditioned acceleration signals at multiple moments at each wind turbine drive chain position into multiple acceleration signals at each moment at each wind turbine drive chain position through analog-to-digital conversion. Combining the multiple acceleration signals at multiple moments at each wind turbine drive chain position with the speed increaser input speed at multiple moments, the unit calculates the real-time health status and real-time fault information using the online monitoring and diagnosis method for wind turbine drive chain faults. The digital signal processing unit then transmits the fault information of the wind turbine drive chain position to the remote server through the data communication unit. The remote server provides fault warnings based on fault information in the wind turbine drive chain. The online monitoring and diagnosis method for wind turbine drivetrain faults includes: Step 1: The multiple accelerometers collect acceleration signals at multiple moments at each position of the wind turbine drive chain and transmit these acceleration signals to the digital signal processing unit; the wind turbine speed acquisition module collects the speed increaser input at multiple moments and transmits these speed increasers to the digital signal processing unit. Step 2: The digital signal processing unit calculates the vibration velocity signal at multiple moments at each wind turbine drive chain position using frequency domain integration based on the acceleration signal at multiple moments at each wind turbine drive chain position. Step 3: Calculate the vibration trend value of each fan drive chain position based on the vibration velocity signals at multiple moments at each fan drive chain position, and calculate the average input speed of the speed increaser based on the speed increaser input speed at multiple moments; Step 4: Repeat steps 1-3 multiple times to obtain the vibration trend value of each fan drive chain position and the average input speed of the speed increaser in multiple executions. Manually define the minimum and maximum values ​​of the average input speed of the speed increaser. Divide the average speed range into N evenly distributed average speed intervals. Combine the value range of each average speed interval to divide the average input speed of the speed increaser into the corresponding average speed interval, and obtain the average input speed of the speed increaser for all executions in each average speed interval. Combine the execution number to obtain the vibration trend value of each fan drive chain position in each average speed interval. Calculate the normal distribution probability of the vibration trend values ​​of each fan drive chain position in each average speed interval to obtain the mean and standard deviation of the vibration trend of each fan drive chain position in each average speed interval. Step 5: The wind turbine speed acquisition module acquires the speed of the gearbox input at multiple moments in real time and transmits the acquired speed of the gearbox input at multiple moments to the digital signal processing unit. The digital signal processing unit calculates the real-time average input speed of the gearbox according to Step 3, and combines the value range of each average speed interval described in Step 4 to divide the real-time average input speed of the gearbox into the average speed interval corresponding to each wind turbine drive chain position. Combining the mean of the vibration trend of the corresponding average speed interval of each wind turbine drive chain position in Step 4 and the standard deviation of the vibration trend of the corresponding average speed interval of each wind turbine drive chain position, the vibration trend value threshold of each wind turbine drive chain position is calculated. Step 6: The multiple accelerometers collect acceleration signals at multiple moments at each wind turbine drive chain position in real time, and transmit the acceleration signals at multiple moments at each wind turbine drive chain position to the digital signal processing unit; the digital signal processing unit calculates the vibration trend value of the wind turbine drive chain position according to steps 2 and 3. If the vibration trend value of the wind turbine drive chain position exceeds the vibration trend value threshold of the wind turbine drive chain position, the corresponding wind turbine drive chain position is determined to be in a potential hazard state; Step 7: If the position of the wind turbine drive chain is in a potential hazard state, use Fast Fourier Transform to calculate the instantaneous spectrum of the vibration velocity signal at the position of the wind turbine drive chain at multiple moments. Step 8: Define the bandwidth of each order of the fan drive train position, and calculate the frequency range corresponding to each order bandwidth of the fan drive train position based on the real-time average input speed of the speed increaser and the bandwidth of each order of the fan drive train position. Step 9: Within the frequency range corresponding to each order bandwidth of the wind turbine drive chain position, calculate the root mean square value of the instantaneous spectrum of the vibration velocity signal corresponding to each order bandwidth of the wind turbine drive chain position. Step 10: Calculate the frequency band alarm threshold corresponding to each order bandwidth of the wind turbine drive chain position. If the root mean square value of the instantaneous spectrum of the vibration velocity signal corresponding to the order bandwidth of the wind turbine drive chain position exceeds the frequency band alarm threshold corresponding to the order bandwidth of the wind turbine drive chain position, then the wind turbine drive chain position status is determined to be a fault state.

2. The online monitoring and diagnosis method for wind turbine drivetrain faults according to claim 1, characterized in that: Step 2 involves calculating the vibration velocity signals at multiple moments for each position in the wind turbine drive train, specifically as follows: A i = { a i,1 , a i,2 , a i,3 ,… a i,T }, i ∈[1, M] Among them, A i For the first i The set of vibration velocity signals at multiple moments for each position in the wind turbine drive chain, where M is the number of positions in the wind turbine drive chain. a i,t For the i-th wind turbine drive train position at the th t The vibration velocity signal at each moment. t ∈[1,T], where T is the number of time sequences.

3. The online monitoring and diagnosis method for wind turbine drivetrain faults according to claim 2, characterized in that: Step 3, calculating the vibration trend value at each position of the wind turbine drive chain, specifically involves: , i ∈[1, M] Among them, B i For the first i Vibration trend values ​​at each position of the fan drive chain.

4. The online monitoring and diagnosis method for wind turbine drivetrain faults according to claim 3, characterized in that: Step 5, which involves calculating the vibration trend threshold value for each position in the wind turbine drive train, specifically involves: BL i = μ i + ησ i ， i ∈[1, M]， η ∈{2,3} in, BL i Indicates the first i Vibration trend threshold values ​​at individual fan drive chain locations; μ i , σ i The real-time average input speed of the speed increaser belongs to the average speed range described in step 4. i The mean of the vibration trend at each position of the fan drive train, the first i The standard deviation of the vibration trend at each position of the fan drive chain η 3 in Gaussian distribution σ Principle coefficient.

5. The online monitoring and diagnosis method for wind turbine drivetrain faults according to claim 4, characterized in that: Step 7 describes the calculation of the instantaneous spectrum of the vibration velocity signal at the position of the wind turbine drive train, specifically as follows: { f y,1 , f y,2 ,…… f y,H }, H = floor(K / 2), y ∈[1, Y] in, f y,h The frequency of the vibration velocity signal representing the position of the wind turbine drive chain in the y-th potential hazard state. h Spectral characteristics, h ∈[1, H], K is the number of sampling points, Y is the number of positions of the wind turbine drive chain in the hidden danger state, and H is the number of spectrum sequences.

Images