A monitoring method for the drive system of an offshore wind turbine

Abstract

The present invention discloses a monitoring method for a transmission system of an offshore wind turbine, comprising the following steps: S1. According to the characteristics of the offshore wind turbine, select multiple components of the transmission system of the offshore wind turbine as monitoring points; S2. Arrange sensors at the monitoring points for monitoring; S3. Establish a real-time data acquisition system; S4. Establish a satellite communication system, and transmit the monitored parameters collected in real time to a ground station through the satellite communication system; S5. Establish a cloud data center, which is communicatively connected to the ground station and receives the monitored parameters transmitted by the ground station; S6. Establish a central data processing platform to receive the monitored parameters transmitted by the cloud data center and realize the processing and analysis of the monitored parameters. The present invention realizes the low-cost operation and maintenance of the offshore wind turbine, realizes the real-time monitoring of the parameters of each component of the offshore wind turbine, can timely know the operation situation, perform maintenance treatment according to the operation situation, and avoid greater economic losses caused by greater damage.

Description

Technical Field

[0001] The present invention relates to the technical field of health monitoring of offshore wind turbines, and in particular to a method for monitoring a transmission system of an offshore wind turbine. Background Art

[0002] The wind turbine transmission system is a key equipment in wind farms, whether on land or at sea. For a long time, it has been using a planned maintenance method, that is, routine maintenance is carried out after the wind turbine components have been running for a certain period of time. This maintenance method cannot fully and timely understand the operating status of the equipment, and because the maintenance status of the equipment is unknown, it is easy to cause inadequate preparation, long maintenance time, and serious losses and waste. In particular, the maintenance and repair of offshore wind turbine transmission systems require special mechanical resources and weather conditions, which greatly increase the maintenance cost of offshore wind turbine transmission systems. For example, replacing wind turbine transmission system components requires special offshore ships for transportation, and expensive machinery such as cranes are required. Due to factors such as wind and waves, these ships and machinery may not be able to enter the offshore wind farm for maintenance in time, which greatly increases the cost of machinery rental. In addition, the maintenance downtime plus the time for machinery to wait to enter the wind farm prolongs the total downtime of the wind turbine transmission system, which is also one of the aspects of the increase in offshore wind farm maintenance costs. Therefore, it is crucial to improve the stability and reliability of wind turbine monitoring systems, predict and monitor wind turbine transmission system failures, detect wind turbine transmission system failures early and take corresponding protective measures, and reduce operating and maintenance costs in order to promote the development of the offshore wind power industry.

[0003] At present, the development of wind turbine transmission system monitoring is in its infancy. The existing research on wind turbine transmission system failure modes and monitoring methods has created theoretical conditions for the development of monitoring systems. Because of the harsh environment at sea and signal coverage problems, the practical application of these works in engineering is basically based on onshore wind turbine transmission systems, and rarely involves offshore wind turbines. Due to the special application environment of offshore wind turbines, the onshore wind turbine transmission system monitoring technology cannot be completely transplanted to the offshore wind turbine system. For this reason, it is necessary to conduct targeted analysis and summary of the prone failures of large offshore wind turbines and their important components, and lay the foundation for the subsequent establishment of offshore wind turbine transmission system health monitoring and diagnosis. At the same time, the amount of data transmitted by the monitoring data of the offshore wind turbine transmission system is relatively large, and there is no 4G / 5G signal. The traditional Beidou short message rate is too limited. How to solve the problem of large data transmission is also an important problem that needs to be solved.

[0004] At present, manufacturers at home and abroad have proposed corresponding monitoring solutions for wind turbines. For example, the SKF WindCon monitoring system in Sweden monitors various data points of wind turbines. The monitoring points include: unbalance state of the rotor shaft, misalignment, shaft offset, machine looseness, bearing state, gear damage, generator rotor / stator, resonance, detection of poor lubrication state, etc. Or the Gram&Juhl SCADA in Denmark supports the comprehensive monitoring and display of the basic parameters of the drive systems of wind turbines in the entire wind farm, including wind speed, power, rotational speed, etc.

[0005] Looking at the current monitoring solutions for wind turbines, there are some missing details in monitoring, and the offshore working conditions are not considered. For example, the monitoring of the drive system of wind turbines, etc. In particular, the offshore data transmission solutions do not involve the transmission of high-speed big data. Summary of the Invention

[0006] The purpose of the present invention is to provide a monitoring method for the drive system of an offshore wind turbine to solve the deficiencies in the prior art. Sensors are installed at the common failure points to collect data in real time, and then interconnected with the central data processing platform through a satellite data channel to achieve real-time monitoring and early warning of the monitoring data. That is, by monitoring the conventional failure points, the faulty parts of the offshore wind turbine can be replaced or processed in advance, reducing the operation and maintenance risks.

[0007] The purpose of the present invention is achieved through the following technical solutions: A monitoring method for the drive system of an offshore wind turbine includes the following steps:

[0008] S1. Select multiple components of the drive system of the offshore wind turbine as monitoring points according to the characteristics of the offshore wind turbine;

[0009] S2. Arrange sensors at the monitoring points for monitoring to obtain monitoring parameters;

[0010] S3. Establish a real-time data acquisition system to collect the monitoring parameters in real time, and perform preprocessing for noise elimination and compression storage of the local monitoring parameters;

[0011] S4. Establish a satellite communication system to transmit the real-time collected monitoring parameters to the ground station through this satellite communication system, and at the same time, the ground station can also send control commands through the satellite communication system;

[0012] S5. Establish a cloud data center, which is communicatively connected to the ground station and receives the monitoring parameters transmitted by the ground station;

[0013] S6. Establish a central data processing platform to receive the monitoring parameters transmitted by the cloud data center and finally achieve the processing and analysis of the monitoring parameters.

[0014] Further, step S1 includes the following steps:

[0015] Based on the common faults of offshore wind turbines being gear faults and bearing faults, select the main shaft bearing, planetary gear, low-speed shaft bearing, high-speed shaft bearing, and generator bearing of the offshore wind turbine for monitoring as the monitoring points.

[0016] Further, establishing the real-time data acquisition system includes the following steps:

[0017] Adopt an integrated embedded design to establish a real-time data acquisition system. The real-time data acquisition system integrates a high-performance Linux core board, FPGA, multi-channel A / D converter, and data storage unit. The external signal interface is connected to the terminals of the real-time data acquisition system, where the high-performance Linux core board and FPGA are the cores of the data acquisition system;

[0018] The multi-channel A / D converter includes high-speed A / D sampling channels and low-speed A / D sampling channels. The high-speed A / D sampling channels achieve synchronous sampling of all high-speed data of the offshore wind turbine, providing real-time, reliable, and high-precision on-site data for subsequent health status analysis; at the same time, the real-time data acquisition system provides low-speed A / D sampling channels for sampling on-site process parameter data; on this basis, the real-time data acquisition system will provide at least 3 RS485 and 1 CAN bus and other serial communication channels for data acquisition of on-site digital sensors and can establish an algorithm model based on the Linux system.

[0019] Further, in step S3, the preprocessing of eliminating noise and compression storage of the localized monitoring parameters includes the following steps:

[0020] Build a set of on-board high-speed data acquisition and storage system, which integrates a DDR3 dynamic data memory and an on-board solid-state data memory for real-time data processing of the real-time data acquisition system and saving the acquired data;

[0021] In terms of data compression storage, introduce AI technology based on the high computing power performance of the linux core board and adopt the methods of machine learning, timed segmented storage, and fault data storage, including the following steps:

[0022] S3.1. Obtain various raw monitoring parameter signals through the acquisition model under the Linux system via the front-end drive circuit;

[0023] S3.2. Perform preprocessing of eliminating noise on the raw monitoring parameter signals to obtain the preprocessed signals;

[0024] S3.3. Feature extraction is performed on the preprocessed signal through time-domain and frequency-domain analysis to construct a state feature vector. After adding a timestamp to this state feature vector, it is stored in the database as a feature parameter;

[0025] S3.4. Based on machine learning, incremental learning is carried out on the accumulated feature parameters to adaptively train the threshold of the feature parameters;

[0026] S3.5. Based on the obtained threshold, it is judged whether the currently acquired state feature vector is abnormal. If it is abnormal, the original monitoring parameters are stored in the database; if not, go to step S3.6;

[0027] S3.6. Judge whether the current timing segmented storage condition is met. If it is met, the currently collected original monitoring parameters are stored in the database; if not, the data is discarded.

[0028] Furthermore, the real-time data acquisition system further includes:

[0029] The real-time data acquisition system is integrated with at least 2 gigabit Ethernet interfaces and a spare gigabit network interface. One of the gigabit Ethernet interfaces is communicatively connected to the ground station to periodically receive the instructions sent by the ground and the system maintenance and upgrade data, and at the same time upload the compressed data collected on-site to the ground cloud service center; the other gigabit Ethernet interface is used for communication with the on-site control terminal or system wind farm networking, and both transmissions are based on a preset data transmission protocol to ensure the validity of the communication data. The spare gigabit network interface is used for testing or handling emergency requirements.

[0030] Furthermore, the establishment of the satellite communication system includes the following steps:

[0031] Based on the AsiaSat 6D high-throughput satellite communication technology, big data transmission is realized, breaking through the previous satellite responsive communication mode, and adopting the instant communication mode. The satellite relied on is the AsiaSat 6D satellite, which can achieve an uplink data rate of 5 Mbit / s and a downlink data rate of 10 Mbit / s;

[0032] At the same time, the OTM45 satellite communication terminal is used as the remote on-site intelligent terminal to realize the up and down transmission of data. This terminal is fixedly installed by a bracket. Its enhanced signal module can automatically point to and connect without adjusting the satellite pointing angle when the platform deflects by 30°, to cope with the offshore strong wind environment.

[0033] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0034] 1. By adopting the integrated technology, the embedded computer, FPGA, A / D converter, on-board big data storage, and network communication are designed together, making the wind power on-site data acquisition system more robust, reliable, and compact, suitable for the offshore environment;

[0035] 2. Support offline warning, introduce edge computing and AI data learning. Even when the system is offline, it can still monitor the health status of the unit and save data, effectively avoiding the fatal drawback of the network cloud monitoring platform, that is, support for offline warning.

[0036] 3. Satellite Internet of Things communication. Based on the Asia-Pacific 6D satellite, it realizes the transmission of big data on the health status monitoring of offshore wind turbines and the real-time sending of land control instructions, making the operation of the transmission system of the wind turbine generator more safe and reliable.

[0037] 4. Achieve low-cost operation and maintenance of offshore wind turbines, realize real-time monitoring of the parameters of each component of offshore wind turbines, be able to know the operation situation in a timely manner, carry out maintenance and treatment according to the operation situation, and avoid greater economic losses caused by greater damage. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is the architecture diagram of the offshore wind turbine monitoring.

[0039] Figure 2 It is the architecture diagram of the real-time data acquisition system.

[0040] Figure 3 It is the logical block diagram of data compression and storage of the real-time data acquisition system. DETAILED DESCRIPTION OF THE INVENTION

[0041] The present invention will be further described below in conjunction with specific embodiments.

[0042] See Figures 1 to 3 As shown, it is the monitoring method for the transmission system of the offshore wind turbine provided by this embodiment, which is characterized by including the following steps:

[0043] S1. For the characteristics of the offshore wind turbine, select multiple components of the transmission system of the offshore wind turbine as monitoring points, including the following steps:

[0044] According to the common faults of the offshore wind turbine being gear faults and bearing faults, select the main shaft bearing, planetary gear, low-speed shaft bearing, high-speed shaft bearing and generator bearing of the offshore wind turbine for monitoring as monitoring points. The distribution of its main monitoring points is shown in Table 1 below:

[0045] Measuring point Measurement object Test direction Sensor position 1 Main shaft bearing Radial Directly above the bearing housing 2 Main shaft bearing Axial Directly above the bearing housing 3 Planet gear Radial Above the planet gear 4 Low-speed shaft bearing Radial Directly above the bearing housing 5 Low-speed shaft bearing Axial Directly above the bearing housing 6 High-speed shaft bearing Radial Directly above the front bearing housing of the high-speed shaft 7 High-speed shaft bearing Axial Directly above the bearing housing at the output end of the gearbox 8 Generator bearing Radial Directly above the input bearing 9 Generator bearing Axial Directly above the input bearing

[0046] Table 1 Distribution of monitoring points for offshore wind turbines

[0047] S2. Arrange sensors at the monitoring points for monitoring to obtain monitoring parameters.

[0048] S3. Establish a real-time data acquisition system to collect monitoring parameters in real time, and perform preprocessing for noise elimination and compressed storage of local monitoring parameters, including the following steps:

[0049] Adopt an integrated embedded design to establish a real-time data acquisition system. The real-time data acquisition system integrates a high-performance Linux core board, an FPGA, multi-channel A / D converters, and a data storage unit. The external signal interface is connected to the terminals of the real-time data acquisition system. Among them, the high-performance Linux core board and the FPGA serve as the core of the data acquisition system;

[0050] The multi-channel A / D converters include 16-channel 16-bit or 24-bit high-speed A / D sampling channels and 8-channel 16-bit low-speed A / D sampling channels. The 16-channel 16-bit or 24-bit high-speed A / D sampling channels achieve synchronous sampling of all high-speed data of the offshore wind turbine, providing real-time, reliable, and high-precision on-site data for subsequent health status analysis. At the same time, the real-time data acquisition system provides 8-channel 16-bit low-speed A / D sampling channels for sampling on-site process parameter data. On this basis, the real-time data acquisition system will provide at least 3 RS485 and 1 CAN bus and other serial communication channels for data acquisition of on-site digital sensors, and can establish an algorithm model based on the Linux system. The main technical parameters of the real-time data acquisition system are shown in Table 2:

[0051] Serial number Hardware interface Parameter Model selection Remarks 1 CPU Dual-core with a main frequency of 800M 2 Memory 1G 3 Solid-state drive 64G eMMC 5 Ethernet RJ45 1000Mbps (×3) 6 RS485 4 channels 7 CAN 1 channel 9 Analog signal AD acquisition 16 / 32 channels 24bit / 100kHz Vibration acceleration sensor that can be collected

[0052] Table 2 Main technical parameters of the real-time data acquisition system

[0053] The real-time data acquisition system integrates at least 2 gigabit Ethernet interfaces and a spare gigabit network interface. One gigabit Ethernet interface is communicatively connected to the ground station to receive instructions and system maintenance and upgrade data sent from the ground regularly, and at the same time upload the compressed data collected on-site to the ground cloud service center. The other gigabit Ethernet interface is used for communication with the on-site control terminal or system wind farm networking. Both transmissions are based on a preset data transmission protocol to ensure the validity of communication data. The spare gigabit network interface is used for testing or handling emergency requirements.

[0054] The preprocessing for noise elimination and compressed storage of local monitoring parameters includes the following steps:

[0055] Build a set of on-board high-speed data acquisition and storage system, which integrates 1G of DDR3 dynamic data memory and 64G of on-board solid-state data memory for real-time data processing and storage of the real-time data acquisition system;

[0056] In terms of data compression storage, the high computing power performance based on the Linux core board introduces AI technology and adopts machine learning, timed segmented storage, and fault data storage methods, including the following steps:

[0057] S3.1. Obtain various types of original monitoring parameter signals through the acquisition model under the Linux system via the front-end drive circuit;

[0058] S3.2. Perform preprocessing on the original monitoring parameter signals to eliminate noise, such as smoothing, filtering, conversion, etc., to eliminate noise, improve the signal-to-noise ratio, and obtain the preprocessed signals;

[0059] S3.3. Extract features from the preprocessed signals using time-domain and frequency-domain analysis, construct a state feature vector, and store the state feature vector with a timestamp as a feature parameter in the database;

[0060] S3.4. Based on machine learning, perform incremental learning on the accumulated feature parameters and adaptively train the threshold of the feature parameters;

[0061] S3.5. Based on the obtained threshold, determine whether the currently acquired state feature vector is abnormal. If it is abnormal, store the original monitoring parameter in the database; if it is not abnormal, proceed to step S3.6;

[0062] S3.6. Determine whether the current timed segmented storage condition is met. If it is met, store the currently acquired original monitoring parameter in the database; if it is not met, discard the data.

[0063] S4. Establish a satellite communication system. Transmit the real-time acquired monitoring parameters to the ground station through this satellite communication system. At the same time, the ground station can also send control instructions through the satellite communication system, including the following steps:

[0064] Based on the AsiaSat 6D high-throughput satellite communication technology, achieve big data transmission, break through the previous satellite responsive communication mode, and adopt the instant communication mode. The satellite relied on is the AsiaSat 6D satellite, which can achieve an uplink data rate of 5 Mbit / s and a downlink data rate of 10 Mbit / s;

[0065] At the same time, use the OTM45 satellite communication terminal as the remote field intelligent terminal to achieve data up and downlink. This terminal is fixed and installed using a bracket. Its enhanced signal module can automatically point and connect without adjusting the satellite pointing angle when the platform deflects by 30°, to cope with the offshore high-wind environment.

[0066] S5. Establish a cloud data center. This cloud data center is communicatively connected to the ground station and receives the monitoring parameters transmitted by the ground station;

[0067] S6. Establish a central data processing platform to receive the monitoring parameters transmitted by the cloud data center and finally achieve the processing and analysis of the monitoring parameters.

[0068] The following is the application scenario of the monitoring method for offshore wind turbines provided by this embodiment:

[0069] I. Main application scenario

[0070] Operate, analyze, edit, and store the information on the status of the wind turbine drive system, the status of the unit, and the environmental status in real time. Transmit relevant sensing data, over-limit information, control instructions, audio, and video to the cloud data center through satellite communication. Managers can monitor the working status of the unit at any time through mobile phones and computers according to their permissions, eliminate potential accident hazards in a timely manner, reduce the failure rate, and effectively improve the working efficiency of the unit. Automatically solve and correct small problems through the emergency application software built into the intelligent terminal of the monitoring system, and professional engineers can remotely analyze data to achieve remote troubleshooting and handling of small faults, reducing the maintenance cost and the risk of offshore operations. Through the practical accumulation of the solutions to the faults of the offshore wind turbine drive system, a mathematical model can be established to continuously improve the application software of the intelligent terminal and enhance the ability of the offshore wind turbine drive system to automatically solve problems, gradually improving the artificial intelligence level of the operation of the offshore wind turbine drive system.

[0071] II. Auxiliary application scenario

[0072] 1. Provide Internet WIFI for the staff working on offshore wind turbines to improve work efficiency;

[0073] 2. More Internet application terminals can be added to offshore wind turbines, such as cameras, environmental and marine monitoring and acquisition sensing terminals, etc., to achieve global perception of the offshore wind turbine drive system and provide communication services for other fields;

[0074] 3. It can be used as an offshore emergency communication station to provide emergency network services for passenger ships and fishing boats in distress and special situations at sea.

[0075] The above embodiments are only the preferred embodiments of the present invention and do not limit the scope of implementation of the present invention. Therefore, all changes made according to the shape and principle of the present invention should be covered within the protection scope of the present invention.

Claims

1. A monitoring method for a drive system of an offshore wind turbine, characterized in that, Including the following steps: S1. Select multiple components of the offshore wind turbine drive system as monitoring points according to the characteristics of the offshore wind turbine; S2. Arrange sensors at the monitoring points for monitoring to obtain monitoring parameters; S3. Establish a real-time data acquisition system to collect the monitoring parameters in real time, and perform preprocessing for noise elimination and compression storage of the localized monitoring parameters; Adopt an integrated embedded design to establish a real-time data acquisition system. The real-time data acquisition system integrates a high-performance Linux core board, an FPGA, a multi-channel A / D converter, and a data storage unit. The external signal interface is connected to the terminals of the real-time data acquisition system, where the high-performance Linux core board and the FPGA are the cores of the data acquisition system; The multi-channel A / D converter includes a high-speed A / D sampling channel and a low-speed A / D sampling channel. The high-speed A / D sampling channel realizes synchronous sampling of all high-speed data of the offshore wind turbine, providing real-time, reliable, and high-precision on-site data for subsequent health status analysis; meanwhile, the real-time data acquisition system provides a low-speed A / D sampling channel for sampling on-site process parameter data; on this basis, the real-time data acquisition system will provide at least 3 RS485 and 1 CAN bus serial communication channels for data acquisition of on-site digital sensors, and can establish an algorithm model based on the Linux system; Build a board-mounted high-speed data acquisition and storage system, which integrates a DDR3 dynamic data memory and a board-mounted solid-state data memory for real-time data processing and storage of the acquired data of the real-time data acquisition system; In terms of data compression storage, introduce AI technology based on the high computing power performance of the linux core board and adopt the methods of machine learning, timed segmented storage, and fault data storage, including the following steps: S3.

1. Obtain various types of original monitoring parameter signals through the acquisition model and the front-end drive circuit under the Linux system; S3.

2. Perform preprocessing for noise elimination on the original monitoring parameter signals to obtain the preprocessed signals; S3.

3. Extract features from the preprocessed signals through time-domain and frequency-domain analysis, construct a state feature vector, and store the state feature vector with a timestamp as a feature parameter in the database; S3.

4. Based on machine learning, perform incremental learning on the accumulated feature parameters and adaptively train the threshold of the feature parameters; S3.

5. Based on the obtained threshold, judge whether the currently obtained state feature vector is abnormal. If it is abnormal, store the original monitoring parameters in the database; if not, go to step S3.6; S3.

6. Judge whether the current condition for timed segmented storage is met. If it is met, store the currently acquired original monitoring parameters in the database; if not, discard the data; S4. Establish a satellite communication system, and transmit the real-time acquired monitoring parameters to the ground station through the satellite communication system. At the same time, the ground station can also send control instructions through the satellite communication system; S5. Establish a cloud data center, which is communicatively connected to the ground station and receives the monitoring parameters transmitted by the ground station; S6. Establish a central data processing platform to receive the monitoring parameters transmitted by the cloud data center and finally realize the processing and analysis of the monitoring parameters.

2. The monitoring method for a drive system of an offshore wind turbine according to claim 1, characterized in that, The step S1 includes the following steps: According to the gear faults and bearing faults of the offshore wind turbine, select the main shaft bearing, planetary gear, low-speed shaft bearing, high-speed shaft bearing and generator bearing of the offshore wind turbine for monitoring as the monitoring points.

3. A monitoring method for a transmission system of an offshore wind turbine according to claim 1, characterized in that, In step S3, the real-time data acquisition system further includes: The real-time data acquisition system is integrated with at least two Gigabit Ethernet interfaces and a spare Gigabit Ethernet port. One of the Gigabit Ethernet interfaces is communicatively connected to the ground station to regularly receive the instructions sent by the ground and the system maintenance and upgrade data, and at the same time upload the compressed data collected on-site to the cloud data center; the other Gigabit Ethernet interface is used for communication with the on-site control terminal or system wind farm networking, and the transmission of both is based on a preset data transmission protocol to ensure the validity of the communication data. The spare Gigabit Ethernet port is used for testing or handling emergency requirements.

4. A monitoring method for a transmission system of an offshore wind turbine according to claim 1, characterized in that In step S4, the establishment of the satellite communication system includes the following steps: Based on the AsiaSat 6D high-throughput satellite communication technology, realize big data transmission, break through the previous satellite responsive communication mode, and adopt the instant communication mode. The satellite relied on is the AsiaSat 6D satellite, which can achieve an uplink data rate of 5 Mbit / s and a downlink data rate of 10 Mbit / s; At the same time, adopt the OTM45 satellite communication terminal as the remote on-site intelligent terminal to realize the up and down transmission of data. The terminal is fixedly installed in a bracket type. Its enhanced signal module can automatically point and connect without adjusting the satellite pointing angle when the platform deflects by 30°, so as to cope with the offshore strong wind environment.

Images