An AI monitoring and early warning system for the transmission chain of wind turbines

Abstract

The present invention discloses an AI monitoring and early warning system for the transmission chain of a wind turbine generator, comprising a vibration sensor connected to a transmission gearbox system, a temperature acquisition module in a cabin, a data collector connected to the vibration sensor and the temperature acquisition module, an industrial controller connected to the data collector, a 5G router connected to the output end of the industrial controller, and a microphone connected to the input end of the industrial controller. The system of the present invention is reasonably configured, and the wind turbine generator is monitored by a microphone and a temperature acquisition module, and data is transmitted by an industrial controller and a 5G router, so that fault features can be extracted, which is convenient for inspection and maintenance, improves the efficiency of inspection and maintenance, reduces the cost of inspection and maintenance, is stable and reliable to use, and has strong applicability and good practicality.

Description

Technical Field

[0001] The present invention belongs to the technical field of wind turbine monitoring, and in particular relates to an AI monitoring and early warning system for a transmission chain of a wind turbine. Background Art

[0002] Wind turbines are mostly installed on mountain tops. They have electrical parts for detecting and judging faults. However, in actual application, they lack monitoring structures for mechanical, transmission chain and gearbox systems. Humans cannot directly judge these faults. They can only judge what the fault is based on the frequency of wind turbine alarms and find and repair it through machine coding. Moreover, the machine needs to be shut down for operation, which greatly reduces the efficiency of inspection and maintenance and increases the cost of inspection and maintenance. Therefore, its applicability and practicality are limited. Summary of the invention

[0003] The purpose of the present invention is to provide an AI monitoring and early warning system for the transmission chain of a wind turbine generator which has a reasonable structure and is conducive to monitoring the transmission chain and gear box system.

[0004] The technical solution for achieving the purpose of the present invention is an AI monitoring and early warning system for the transmission chain of a wind turbine, comprising a vibration sensor connected to a transmission gearbox system, a temperature acquisition module in a cabin, a data collector connected to the vibration sensor and the temperature acquisition module, an industrial controller connected to the data collector, a 5G router connected to the output end of the industrial controller, and a microphone connected to the input end of the industrial controller.

[0005] The microphone comprises a sound collector for monitoring frequency, an FFT converter connected to the sound collector, and a frequency calculator connected to the FFT converter, and an output end of the frequency calculator is connected to the industrial controller.

[0006] The temperature acquisition module includes a temperature sensor, a signal processor connected to the temperature sensor, a large signal processor connected to the signal processor, a value-to-decimal string converter connected to the large signal processor, and a temperature display connected to the value-to-decimal string converter, and the temperature display is connected to the data collector.

[0007] The microphone comprises the following steps:

[0008] A. The sounds of the wind wheel, transmission chain, yaw, hydraulic system, braking system, and generator are superimposed into the initialization audio;

[0009] B. Set the sampling rate, channel and sampling data;

[0010] C. Initialize the audio pair sampling data and enter the FFT spectrum after entering the queue, jumping out of the loop, and entering the loop after exiting the queue;

[0011] D. Compare the sampling rate with the sampling number of the data entering the FFT spectrum to get the resolution and find the maximum value of the index through the index to find the maximum value;

[0012] E. Finally, the various sound frequencies of the wind wheel, transmission chain, yaw, hydraulic system, brake system, and generator are obtained by the maximum value of the index * resolution, and sent to the industrial controller through the data collector;

[0013] F. The data collector sends various sound frequencies of the wind wheel, transmission chain, yaw, hydraulic system, brake system, and generator to the industrial controller, and compares them with the sound frequencies of the wind wheel, transmission chain, yaw, hydraulic system, brake system, and generator stored in the industrial controller when they are working normally;

[0014] G. When there is a deviation in the sound frequencies of the wind wheel, transmission chain, yaw, hydraulic system, braking system, and generator, an alarm will occur and the equipment will be displayed on the monitor.

[0015] The present invention has positive effects: the system of the present invention is reasonably set up, it monitors the wind turbine through a microphone and a temperature acquisition module, and transmits data through an industrial controller and a 5G router, so as to be ready to extract fault characteristics, facilitate overhaul and maintenance, improve the efficiency of overhaul and maintenance, reduce the cost of overhaul and maintenance, and is stable and reliable to use, with strong applicability and good practicality. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to make the content of the present invention more clearly understood, the present invention is further described in detail below based on specific embodiments and in conjunction with the accompanying drawings, wherein:

[0017] Figure 1 It is a structural schematic diagram of the present invention;

[0018] Figure 2 is a step block diagram of the microphone in the present invention;

[0019] Figure 3 It is a structural schematic diagram of the temperature acquisition module in the present invention. DETAILED DESCRIPTION

[0020] The technical solution is now further described in conjunction with the drawings of the above specification, and the specific description is as follows:

[0021] See Figures 1 to 3As shown, an AI monitoring and early warning system for the transmission chain of a wind turbine generator includes a vibration sensor 2 connected to a transmission gearbox system 1, a temperature acquisition module 3 in the cabin, a data collector 4 connected to the vibration sensor and the temperature acquisition module, an industrial controller 5 connected to the data collector, a 5G router 6 connected to the output end of the industrial controller, and a microphone 7 connected to the input end of the industrial controller. The aerodynamic noise and mechanical noise read by the microphone from the cabin are processed by the industrial controller to identify the sound frequency and uploaded to the 5G router, and then sent to the central industrial controller via China Mobile and Telecom Internet, and then displayed on the display. If the frequency is abnormal, an alarm is triggered.

[0022] In this embodiment, the microphone is placed above the tower top, mainly for frequency monitoring of mechanical parts including: wind wheel, transmission system, yaw system, hydraulic system, braking system, generator, control and safety system, cabin, tower, etc.

[0023] The data of the gearbox and transmission chain of the wind turbine collected from analog and digital data is sent to the central industrial controller through the Internet or mobile Internet for judgment and display through the intelligent AI of the industrial controller, so as to obtain the fault of a single wind turbine. After the fault is determined, the location of a single wind turbine can be accurately found through the serial number, shortening the maintenance time and improving the efficiency of the maintenance work.

[0024] The microphone 7 includes a sound collector for monitoring frequency, an FFT converter connected to the sound collector, and a frequency calculator connected to the FFT converter, and an output end of the frequency calculator is connected to the industrial controller.

[0025] The main purpose is to solve the problem of wind turbines being located in remote areas and complex environments. The early "minor problems" can be promptly found through the intelligent diagnosis system of wind turbines with graphical programming, and the faults can be found and repaired according to the provided list to solve the problems in time, which is conducive to improving the wind power generation capacity.

[0026] Content: Use NI controller to simulate and digitally collect and monitor the status of components such as "drive chain, bearing damage, gearbox" of wind turbines. Use graphical programming to monitor the noise level of drive chain and gearbox, monitor the temperature of bearings and generators, and then compare the monitored data with normal values ​​to determine the location of the fault and the problem of a single wind turbine. Through monitoring alarms, spare devices can be brought to a fixed address for repair and maintenance of a specific wind turbine. The specific method is as follows.

[0027] (1) First determine the installation locations of the oscillation sensor and the temperature sensor on the wind turbine generator.

[0028] (2) The collected series of simulation information is processed by graphic programming and transmitted to the upper industrial control computer.

[0029] (3) Finally, all the data of the wind turbine generator set is sent to the industrial controller via 5G, and the alarm is displayed on the display to indicate which wind turbine has failed.

[0030] A microphone is installed above the cabin to collect aerodynamic noise and mechanical noise in the cabin, including: wind turbine, transmission system, yaw system, hydraulic system, brake system, generator, control and safety system, cabin, tower, etc. for frequency monitoring, and the collected sound is transformed by FFT (Fast Fourier Transformation) discrete Fourier transform fast algorithm. Because the microphone collects all the superimposed sounds in the cabin, FFT is used to accelerate polynomial multiplication, such as

[0031] a0 0 +a1X 1 +a2X 2 +...+O n X n

[0032] After transformation:

[0033] f(X)=a0X 0 +a1X 1 +a2X 2 +...+a n X n

[0034] Each frequency is obtained through programming calculation and uploaded to the central industrial control system to send out monitoring frequency data.

[0035] Vibration sensors are installed at the gearbox to monitor the wear of components such as the gearbox speed change gear bearings, and the monitoring data is uploaded to the central industrial control system to provide monitoring data.

[0036] Temperature sensors are installed on the load-bearing bearings of the gearbox and transmission chain to monitor the temperature changes of the load-bearing bearings. The linkage braking system realizes cooling of the load-bearing bearings. The control data is uploaded to the central industrial control system for monitoring and control data.

[0037] The temperature acquisition module 3 includes a temperature sensor 31, a signal processor 32 connected to the temperature sensor, a large signal processor 33 connected to the signal processor, a value-to-decimal string converter 34 connected to the large signal processor, and a temperature display 35 connected to the value-to-decimal string converter, and the temperature display is connected to the data collector.

[0038] The microphone comprises the following steps:

[0039] A. The sounds of the wind wheel, transmission chain, yaw, hydraulic system, braking system, and generator are superimposed into the initialization audio;

[0040] B. Set the sampling rate, channel and sampling data;

[0041] C. Initialize the audio pair sampling data and enter the FFT spectrum after entering the queue, jumping out of the loop, and entering the loop after exiting the queue;

[0042] D. Compare the sampling rate with the sampling number of the data entering the FFT spectrum to get the resolution and find the maximum value of the index through the index to find the maximum value;

[0043] E. Finally, the various sound frequencies of the wind wheel, transmission chain, yaw, hydraulic system, brake system, and generator are obtained by the maximum value of the index * resolution, and sent to the industrial controller through the data collector;

[0044] F. The data collector sends various sound frequencies of the wind wheel, transmission chain, yaw, hydraulic system, brake system, and generator to the industrial controller, and compares them with the sound frequencies of the wind wheel, transmission chain, yaw, hydraulic system, brake system, and generator stored in the industrial controller when they are working normally;

[0045] G. When there is a deviation in the sound frequencies of the wind wheel, transmission chain, yaw, hydraulic system, braking system, and generator, an alarm will occur and the equipment will be displayed on the monitor.

[0046] 1) Configure sound input to initialize the sound channel. In "Configure sound input", set the sound operation acquisition rate to 22050S / s, the total number of samples per channel to 5000, the number of channels to mono, and the bit resolution of each sample to 16 bits. Select the default value 0 for device ID. The sampling mode is continuous sampling.

[0047] 2) Set up the sound input and set the total number of samples per channel to 5000.

[0048] 3) Output the "read sound input" data to the "element queue" and then dequeue it through the VI. The dequeue data jumps out of the {1), 2), 3)} loop and enters the next loop. One path is placed on a virtual oscilloscope to display the original audio, and the other path is used for the FFT spectrum (amplitude-phase) VI time signal input window.

[0049] 4) Set up the FFT (Spectru Mag-phase) spectrum (amplitude-phase) VI, set the 'window' to Hanning, which is also the time domain window value for time signals. Set the 'average parameters' cluster to define how to calculate the average value. The parameter description includes the averaging type as No averaging (default), the weighting type as Exponential (default), the number of averages as 3, and set the 'restart averaging' to 'F'. Amplitude spectrum: A. The amplitude is an array composed of FFT spectra, and the input waveform has FFT spectra. The Fourier series transform formula is obtained.

[0050]

[0051] B. Set the starting frequency, f0=0, f0 returns the starting frequency of the spectrum in Hertz.

[0052] C.df returns the frequency resolution of the spectrum in Hertz. The acquisition rate is 22050S / channel and the total number of samples is 5000 = df4.41.

[0053] D. Amplitude is the amplitude of the averaged FFT spectrum. If the input signal is in volts (V), the units of amplitude are volts rms (Vrms).

[0054] E. Calculate the double-sided amplitude spectrum by the following equation

[0055] FA(i)=x(i) / ni=0.1.2.3....(N-1)

[0056] GA is the double-sided amplitude spectrum, X is the discrete Fourier transform of the signal, and N is the number of points in the signal.

[0057] H. Then, this VI converts the double-sided magnitude spectrum to a single-sided magnitude spectrum using the following equation:

[0058] Amplitude spectrum size = |B|

[0059] I. Array index or channel name to select waveform. Wire data types to the waveform array and index inputs and the data type of the analog waveform Y.

[0060] J. The frequency can be obtained through the above calculation, that is: HZ = df frequency resolution * maximum value (array) index.

[0061] AI monitors the health status of the transmission chain and gearbox equipment, reduces or avoids damage to the transmission chain and gearbox system, reduces direct and indirect maintenance costs, optimizes spare parts accuracy, improves reliability and reduces maintenance costs. AI monitoring is to find out the fault location through the self-diagnosis prompt information analysis provided by the fault diagnosis system in the control system and computer control device of the unit; AI self-diagnosis method is a parameter test comparison method for self-diagnosis, through the sub-industrial controller from the microphone, oscillation sensor and temperature sensor instrument equipment to measure the frequency and pulse oscillation signal and temperature and compare these values ​​and waveforms under normal conditions, so as to find out the fault location; comprehensive troubleshooting. Wind turbines involve many instruments and equipment, and the fault chain may be long in actual operation. In order to shorten the troubleshooting time, maintenance personnel can use AI self-diagnosis method to repair and replace the faults in the fault chain according to the prompts of the central controller and the frequency alarm position prompted by the sub-industrial controller from the microphone, oscillation sensor and temperature sensor.

[0062] Reduce unplanned downtime, improve equipment availability, and extend the service life of the unit; the performance of the wind turbine itself must be regularly maintained by the maintenance station, so that the fault of the wind turbine can be discovered and eliminated in time, thereby improving the overall operating efficiency of the wind turbine [7] The operation and maintenance of wind turbines mainly include regular inspection and daily maintenance of wind turbines:

[0063] Regular maintenance: Regular maintenance of the electrical equipment of wind turbines can keep the equipment in good operating condition. The most important maintenance contents include: bolt torque detection between the connection points of wind turbines, lubrication between transmission parts and testing various functions of the generator. Generally, the maintenance work of bolt torque detection is carried out in the summer when there is no wind or little wind, mainly to avoid high wind force. The lubrication methods of generators are divided into thin oil lubrication and dry oil lubrication. Generally, thin oil lubrication is used for the gearbox and yaw reduction gearbox of wind turbine generators. The parts using dry lubrication are yaw gears and bearings. The maintenance methods are mainly to add lubricating oil and replace expired lubricating oil. Since the generator set will generate heat during operation, excessive temperature will cause the lubricating oil to deteriorate and lose its lubricating effect. Therefore, it is very important to add and replace the lubricating oil in time, but it is important to pay attention to the control of the amount of lubricating oil added to prevent too much or too little lubricating oil from burning the generator. Generally, the functional test of the generator mainly includes the detection of various parameters such as the output voltage and output power of the generator to prevent damage to the generator and other control equipment due to abnormal power output of the generator.

[0064] Daily maintenance: During the operation and maintenance of wind turbines in wind farms, some faults may occur that need to be eliminated immediately on site, and necessary maintenance must be carried out after troubleshooting.

[0065] 1. Observe the looseness of the connecting bolts between the safety platform and the ladder;

[0066] 2. Check whether there is any burning inside the control and monitoring cabinet and whether there is any discharge sound or other noise;

[0067] 3. Check whether the generator cable is offset and whether the clamp is loose;

[0068] 4. Check whether there is any abnormal noise between the wind turbine bearings, gears, hammer plates and brake pads. Finally, after maintaining the wind turbine, the site should be cleaned up and the interfaces between the various components of the hydraulic station and the pipe heads should be wiped clean to facilitate the next maintenance.

[0069] To ensure the normal operation of the equipment and provide a basis for refined equipment management; mainly based on the daily maintenance of wind turbines. After a generator set fails, the maintenance personnel must be able to quickly find the cause of the failure in a short period of time and take effective measures to minimize the loss of the failure. The following is a brief summary of common faults of wind turbines and maintenance methods:

[0070] Common troubleshooting methods for wind turbines: Wind turbines involve a lot of equipment and parts, and each part may fail. Therefore, there are many types of faults in the entire generator set. Simply put, the faults of the unit can be divided into two types: mechanical faults and electrical faults. In general, mechanical faults can be found by listening to the sound of the equipment, testing the surface temperature, observing the vibration, etc. The fault location can be found. Relatively speaking, the maintenance of electrical fault locations is more complicated. The maintenance personnel must be familiar with the working principle of the entire unit to find the cause of the fault and choose the appropriate treatment method. The intelligent diagnosis system of the wind turbine with graphic programming is the main way to find faults. During the work process, the maintenance personnel must be able to find the cause of the fault through the fault information provided by the intelligent diagnosis system of the wind turbine with graphic programming, and analyze the accurate fault point. Generally speaking, the analysis of the fault phenomenon can be carried out in the following steps: first, analyze the symptoms of the equipment in detail, then carry out equipment inspection work, confirm the fault location, then check the damaged devices, circuits, etc. in turn, and finally replace the damaged parts. After the fault handling is completed, the working condition and performance of the unit after repair should be checked to ensure that all instruments and equipment are in normal working condition.

[0071] Fault inspection of wind turbines: Fault inspection of wind turbines is an important part of the fault inspection process. Common inspection methods include routine inspection, self-diagnosis, comprehensive analysis of mechanical, electrical and hydraulic systems, circuit parameter test comparison, isolation, and segmented elimination. Appropriate methods should be flexibly selected during the inspection process. Routine inspection is to rely on the maintenance personnel to observe whether the instrument has abnormal temperature, vibration, odor, and sound, and find out the cause of the fault by looking, smelling, touching, and listening. Keep records of consumable parts of wind turbines: During the maintenance of wind turbines, it is also necessary to keep maintenance records, summarize the maintenance content and results of the day, fill in relevant maintenance reports and schedules, and promptly feedback and quickly handle problems that arise during the maintenance process to improve the efficiency of maintenance work. In addition, components that often fail and parts that are easily damaged should be summarized and recorded. The summary report of consumable parts should include system classification, unit model, name of consumable components, such as yaw system, etc. The vulnerable components in GE1500sle, AW1500 and WD750 units mainly include wind vanes, unwinding switches, row brakes, yaw motor brakes, etc. In the later maintenance, maintenance personnel can identify the inspection focus based on these data, and the cause of the same type of faults can be quickly and roughly determined, which can shorten the maintenance time to a certain extent and improve the efficiency of maintenance work. Strengthening the daily maintenance and management of wind power equipment is an important and effective means for wind farms to reduce maintenance costs and reduce the cost of wind power generation. It is also an important means to increase corporate competitiveness, accelerate the localization of wind power equipment and promote the development of wind power industry.

[0072] The system of the present invention is reasonably configured. It monitors the wind turbine through a microphone and a temperature acquisition module, and transmits data through an industrial controller and a 5G router. It can be prepared to extract fault characteristics, facilitate inspection and maintenance, improve the efficiency of inspection and maintenance, reduce the cost of inspection and maintenance, and is stable and reliable to use, with strong applicability and good practicality.

[0073] The standard parts used in this embodiment can be purchased directly from the market, and the non-standard structural components recorded in the specification and the drawings can also be directly processed according to the existing technical common sense. At the same time, the connection method of each component adopts the mature conventional means in the prior art, and the machinery, parts and equipment all adopt the conventional models in the prior art, so no specific description will be given here.

[0074] Obviously, the above embodiments of the present invention are merely examples for clearly explaining the present invention, and are not intended to limit the implementation methods of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. However, these obvious changes or modifications derived from the essential spirit of the present invention still fall within the protection scope of the present invention.

Claims

1. An AI monitoring and early warning system for the transmission chain of a wind turbine generator, characterized in that: It includes a vibration sensor connected to the transmission gearbox system, a temperature acquisition module in the cabin, a data collector connected to the vibration sensor and the temperature acquisition module, an industrial controller connected to the data collector, a 5G router connected to the output end of the industrial controller, and a microphone connected to the input end of the industrial controller; The microphone includes a sound collector for monitoring frequency, an FFT converter connected to the sound collector, and a frequency calculator connected to the FFT converter, wherein the output end of the frequency calculator is connected to the industrial controller; The temperature acquisition module includes a temperature sensor, a signal processor connected to the temperature sensor, a large signal processor connected to the signal processor, a value-to-decimal string converter connected to the large signal processor, and a temperature display connected to the value-to-decimal string converter, and the temperature display is connected to the data collector; The microphone comprises the following steps: A. The sounds of the wind wheel, transmission chain, yaw, hydraulic system, braking system, and generator are superimposed into the initialization audio; B. Set the sampling rate, channel and sampling data; C. Initialize the audio pair sampling data and enter the FFT spectrum after entering the queue, jumping out of the loop, and entering the loop after exiting the queue; D. Compare the sampling rate with the sampling number of the data entering the FFT spectrum to get the resolution and find the maximum value of the index through the index to find the maximum value; E. Finally, the various sound frequencies of the wind wheel, transmission chain, yaw, hydraulic system, brake system, and generator are obtained by the maximum value of the index * resolution, and sent to the industrial controller through the data collector; F. The data collector sends various sound frequencies of the wind wheel, transmission chain, yaw, hydraulic system, brake system, and generator to the industrial controller, and compares them with the sound frequencies of the wind wheel, transmission chain, yaw, hydraulic system, brake system, and generator stored in the industrial controller when they are working normally; G. When there is a deviation in the sound frequencies of the wind wheel, transmission chain, yaw, hydraulic system, braking system, and generator, an alarm will occur and the equipment will be displayed on the monitor.

Images