Offshore wind power structure multi-source monitoring data real-time synchronization system and method based on heterogeneous architecture
By using a multi-source monitoring system for offshore wind power structures based on a heterogeneous architecture, the shortcomings of the offshore wind power monitoring system in the evaluation of the whole equipment are solved, and real-time data synchronization and risk identification are realized, thereby improving the real-time performance and reliability of the monitoring system.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHANDONG ACAD OF MARINE SCI (QINGDAO NAT MARINE SCI RES CENT)
- Filing Date
- 2026-01-15
- Publication Date
- 2026-05-01
AI Technical Summary
Current offshore wind power monitoring systems cannot achieve comprehensive evaluation of the entire equipment under the same time and environment. They suffer from problems such as limited sensor connection ports, lack of real-time data processing capabilities, inability to synchronize data without network access, and data synchronization issues among multiple generating devices.
A real-time synchronization system for multi-source monitoring data of offshore wind power structures based on a heterogeneous architecture is adopted, including monitoring systems for the middle and bottom of the pile foundation, blade structure, and environmental monitoring system. Combined with GPS, LoRa, and 4G modules, the system enables real-time data synchronization and processing.
It achieves the ability to synchronize data from multiple power generation piles, has powerful interface expansion capabilities, and has data timing function under network-free conditions, supporting real-time risk identification and control.
Smart Images
Figure CN121968041A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of industrial internet technology and relates to an offshore wind power structure monitoring system. Background Technology
[0002] Over the past decade, global offshore wind power capacity has expanded at an average annual rate of 25%, with single-unit power increasing from 3 MW to over 20 MW. Wind farm site selection has also rapidly shifted from nearshore shallow waters (≤30 m) to deep-sea areas (≥60 m). China's offshore wind power industry has experienced rapid growth in scale and technology in recent years, ranking first in the world in both total installed capacity and new additions in 2021. Offshore wind power development is of great significance for China's energy structure adjustment, carbon peaking, and carbon neutrality goals.
[0003] The marine environment is complex and variable, and wind turbine structures are directly exposed to the harsh marine environment, making them highly susceptible to damage due to corrosion, structural fatigue, and other factors, which can even lead to the loss of power generation capabilities. Real-time online monitoring of wind turbine structures helps to detect potential problems in components such as foundations and blades early, reducing economic losses caused by component damage and improving power generation efficiency. Furthermore, by combining monitoring data of the operating environment of offshore wind turbines, the safety of wind turbines under different operating conditions can be assessed and tested, providing data support for the construction of smart wind farms such as digital twins.
[0004] Currently, researchers have proposed various monitoring technologies, including the acquisition and analysis of parameters such as stress, vibration, and tilt based on sensor networks. However, most current offshore wind power monitoring research focuses on single structural units, failing to provide a comprehensive assessment of the entire equipment under the same conditions and at the same time. The distributed data acquisition instruments used in existing multi-source data synchronization schemes for offshore wind power monitoring suffer from prominent problems such as limited sensor connection ports, lack of real-time data processing capabilities, inability to synchronize data without a network, and lack of data synchronization capabilities across multiple generating devices. Summary of the Invention
[0005] To address the current challenges in full-structure monitoring of offshore wind power, this invention proposes a real-time synchronization system and method for multi-source monitoring data of offshore wind power structures based on a heterogeneous architecture.
[0006] The technical solution adopted in this invention is: a real-time synchronization system for multi-source monitoring data of offshore wind power structures based on a heterogeneous architecture, including a pile foundation mid-structure monitoring system, a pile foundation bottom structure monitoring system, a blade structure monitoring system, an environmental monitoring system, a heterogeneous architecture data processing module, a GPS module, a 4G processing module, a LoRa module, and an industrial control board; the pile foundation mid-structure monitoring system, the pile foundation bottom structure monitoring system, the blade structure monitoring system, the environmental monitoring system, and the GPS module are respectively connected to the input ports of the heterogeneous architecture data processing module, and the output of the heterogeneous architecture data processing module is connected to the industrial control module; the LoRa module and the 4G module can receive wireless signals sent by the central control room and connect to the heterogeneous architecture data processing module; the heterogeneous architecture data processing module includes data input interfaces, implementation logic, and synchronous real-time data processing algorithms that match each system and the industrial control board; it includes a GPS module data input interface and time extraction logic; and it includes LoRa module and 4G module data receiving interfaces and command extraction logic.
[0007] Preferably, the pile foundation mid-structure monitoring system includes an acceleration sensor, an inclination sensor, a fiber optic strain sensor, and a demodulator.
[0008] Preferably, the pile foundation bottom structure monitoring system includes an acceleration sensor, an inclination sensor, a fiber optic strain sensor, and a demodulator.
[0009] Preferably, the blade structure monitoring system includes an acceleration sensor, an tilt sensor, a fiber optic strain sensor, and a demodulator.
[0010] Preferably, the environmental monitoring system includes a water level and wave radar, a wind speed and direction meter, and a camera monitoring device.
[0011] Preferably, the GPS module is a low-power navigation and positioning module that supports multiple satellite navigation systems.
[0012] Preferably, the LoRa module is a LoRaWAN node module with a pin-type connection and can be equipped with an antenna.
[0013] This invention also provides a method for real-time synchronization of multi-source monitoring data of offshore wind power structures based on heterogeneous architecture, which includes the following steps using the system described above: (1) The main control room sends a data read command via LoRa protocol or 4G network broadcast; (2) The heterogeneous architecture module receives and distributes the parsed instructions from the LoRa / 4G module; (3) Each sensor receives the corresponding data reading command; (4) Each sensor returns monitoring data; (5) The heterogeneous architecture of the single wind power synchronous system receives and processes data; (6) The heterogeneous architecture frames the data from each sensor and adds timestamps; (7) The heterogeneous architecture end sends framed data to the industrial control board; (8) The industrial control board sends the monitoring data of a single wind turbine pile to the main control room.
[0014] Preferably, the parallel data processing method within the heterogeneous architecture includes: the heterogeneous architecture simultaneously or rapidly receives data packets from various sensors through its hardware interface controller and temporarily stores them in an external high-speed buffer (such as DDR); after preliminary processing, the real-time data processing method for the accelerometer, tilt sensor, and fiber optic strain sensor data within the heterogeneous architecture is as follows: the accelerometer data undergoes low-pass filtering in the microprocessor chip within the heterogeneous architecture, followed by wavelet transform processing, using a 5 / 3 discrete wavelet transform and an inertial splitting method. The tilt sensor data undergoes weighted moving average processing in the microprocessor chip within the heterogeneous architecture, followed by low-pass filtering; the fiber optic strain data undergoes band-pass filtering in the microprocessor chip within the heterogeneous architecture. After noise reduction processing, the above three types of data are transmitted into the logic chip for accelerated neural network risk feature identification, and the pre-trained neural network parameters are stored in the external memory chip of the heterogeneous architecture logic chip. Based on the identified risk information, the heterogeneous architecture unit can perform risk control such as blade retraction at the "edge" end. The three data processing methods differ, and the processing times vary. Therefore, each data processing step requires a different delay to ensure alignment of the three data time scales. Subsequently, the three data sets are combined within the heterogeneous architecture according to the accelerometer, tilt sensor, and fiber optic strain data. Each data set has a fixed bit width of 16 bits. If the actual bit width is insufficient, sign bit extension is used to pad the data. The combined data forms a 48-bit data packet for subsequent data transmission.
[0015] This invention not only has the ability to synchronize data from multiple power generation piles, but also has strong interface expansion capabilities and can achieve data timing under network-free conditions. Attached Figure Description
[0016] Figure 1 This is a structural diagram of a real-time synchronization system for multi-source monitoring data of offshore wind power structures based on a heterogeneous architecture, as described in an embodiment of the present invention. Figure 2 This is a flowchart of a method for real-time synchronization of multi-source monitoring data of offshore wind power structures based on heterogeneous architecture, as described in an embodiment of the present invention. Figure 3 This is a flowchart of data processing methods within a heterogeneous architecture. Detailed Implementation
[0017] The principles and features of the present invention are described below with reference to specific embodiments. The described embodiments are only some embodiments of this disclosure, and not all embodiments. The examples are only used to explain the present invention and are not intended to limit the scope of the present invention.
[0018] Example 1, such as Figure 1 As shown, this embodiment provides a real-time synchronization system for multi-source monitoring data of offshore wind power structures based on a heterogeneous architecture. The system includes a pile foundation mid-structure monitoring system, a pile foundation bottom structure monitoring system, a blade structure monitoring system, an environmental monitoring system, a heterogeneous architecture data processing module, a GPS module, a 4G processing module, a LoRa module, and an industrial control board. The pile foundation mid-structure monitoring system, the pile foundation bottom structure monitoring system, and the blade structure monitoring system each contain an accelerometer, a tilt sensor, a fiber optic strain sensor, and a corresponding demodulator. The accelerometer and tilt sensor data output interfaces are RS485 interfaces, while the fiber optic strain demodulator data output interface is a network port. The environmental monitoring system includes a water level and wave radar, a wind speed and direction indicator, and a video monitoring device. The water level and wave radar and the wind speed and direction indicator data output interfaces are also RS485 interfaces, while the video monitoring device data output interface is a network port. The heterogeneous architecture data processing module uses an automatic transceiver isolation chip to handle the RS485 interface processing. For the network port portion, a physical layer chip plus a MAC layer IP core is used to achieve network communication. The GPS module in the system is model ATGM336H, and its data output supports multiple modes including UART, I2C, and SPI. The heterogeneous architecture data processing module uses the UART interface to communicate with the GPS module. The LoRa module in the system is mainly model Ra-08-Kit, which connects to the heterogeneous architecture data processing module using a UART interface. The 4G module in the system is model E840-TTL, which includes a SIM card slot and connects to the heterogeneous architecture data processing module using a TTL pin header, with UART as the data communication protocol.
[0019] Example 2: This example provides a method for real-time synchronization of multi-source monitoring data of offshore wind power structures based on heterogeneous architecture. This method is based on the system in Example 1, and its processing flow is as follows: Figure 2 As shown, the specific steps include: (1) The main control room sends a data read command via LoRa protocol or 4G network broadcast. The host computer software in the central control room generates a standard "data read request" command frame based on a preset monitoring plan (e.g., every 5 minutes) or manual instructions. This command frame is encapsulated into a specific network data packet. The command is then sent to the LoRa base station for broadcast via radio waves. Alternatively, the command can be sent to a 4G base station and transmitted to each node via the network.
[0020] (2) The heterogeneous architecture module receives and distributes the parsed instructions from the LoRa / 4G module. The communication module (LoRa or 4G DTU) installed on the wind turbine pile receives the radio frequency signal and demodulates it into a digital data stream. This data stream is then transmitted to the heterogeneous architecture via a hardware interface such as a serial port (e.g., UART) or SPI.
[0021] The heterogeneous architecture internally implements a lightweight communication protocol stack. It parses data packets, checks if the target address matches the local machine, and verifies the data integrity (using a checksum). If a match is found, the core "read data" instruction is extracted. Based on a pre-defined sensor address mapping table, the heterogeneous architecture sends the "read data" sub-instruction in parallel or rapidly polling mode to each sensor connected to the bus (such as strain gauges, accelerometers, tilt sensors, etc.) via its I / O pins (e.g., SPI, I2C bus) or serial port.
[0022] (3) Each sensor receives the corresponding data reading command. Each sensor has a unique address or ID. They listen to the communication bus but only respond to commands that match their address. Upon receiving a command, the microcontroller inside the sensor initiates a high-precision analog-to-digital converter (ADC) to acquire analog signals such as vibration, deformation, and temperature at the current moment and convert them into digital values.
[0023] (4) Each sensor returns monitoring data The sensors package the acquired digital values according to an agreed-upon communication protocol format, typically including the sensor ID, sensor type, acquired data value, and status code. The packaged data frames are then sent back to the heterogeneous architecture via bus protocols such as SPI / I2C, either sequentially or according to the scheduling of the heterogeneous architecture. Because all sensors are triggered almost simultaneously, the data they return naturally exhibits temporal synchronization.
[0024] (5) The heterogeneous architecture of the single wind power synchronous system receives and processes data. like Figure 3As shown, the heterogeneous architecture receives data packets from various sensors simultaneously or rapidly through its hardware interface controller and temporarily stores them in an external high-speed buffer (such as DDR). The real-time data processing method for accelerometer, tilt sensor, and fiber optic strain sensor data within the heterogeneous architecture is as follows: Accelerometer data undergoes low-pass filtering within the microprocessor chip of the heterogeneous architecture, followed by wavelet transform processing. The wavelet transform uses a 5 / 3 discrete wavelet transform with inertial splitting. Tilt sensor data undergoes weighted moving average processing within the microprocessor chip of the heterogeneous architecture, followed by low-pass filtering. Fiber optic strain data undergoes band-pass filtering within the microprocessor chip of the heterogeneous architecture. After noise reduction, these three types of data are transmitted into the logic chip for accelerated neural network risk feature identification. The pre-trained neural network parameters are stored in an external memory chip within the heterogeneous architecture logic chip. Based on the identified risk information, the heterogeneous architecture unit can perform risk control measures such as blade retraction at the "edge." Since the three data processing methods differ and their processing times differ, different delays are required after each data processing step to ensure alignment of the three data time scales.
[0025] Subsequently, various monitoring data are packaged within the heterogeneous architecture in the following order: monitoring data of the pile foundation mid-section structure, monitoring data of the pile foundation bottom structure, monitoring data of the blade structure, and environmental monitoring data. The first three types of monitoring data consist of accelerometer, tilt sensor, and fiber optic strain data arranged in sequence. Environmental monitoring data consists of wave radar data, anemometer data, and video data arranged in sequence. Before uploading, timestamp data is added to the end of the packaged data to complete the construction of a time-identified data packet.
[0026] (6) The heterogeneous architecture frames the data from each sensor and adds timestamps. This is the core step in achieving multi-source data synchronization. The heterogeneous architecture utilizes its high-precision internal clock or external GPS / BeiDou module to assign the same, millisecond- or even microsecond-level timestamp to the processed monitoring data packets received from step (4) that belong to the same acquisition cycle. This timestamp marks the unified moment when the data was acquired. Data from different sensors but with the same timestamp are packaged together according to a predefined order and structure to form a complete "wind turbine pile status snapshot" data frame. This frame includes a frame header, timestamp, data volumes from each sensor, and a frame tail checksum.
[0027] (7) The heterogeneous architecture end sends framed data to the industrial control board. After framing is completed, the heterogeneous architecture transmits the entire frame data quickly and stably to the industrial control board (host computer) through a high-speed parallel bus or high-speed serial interface (such as PCIe, Ethernet RGMII).
[0028] (8) The industrial control board sends the monitoring data of a single wind turbine pile to the central control room. The data receiving program on the industrial control board reads complete, pre-processed, and synchronized data frames from the heterogeneous architecture. The industrial control board temporarily stores the data on its local hard drive and displays it in real-time on the human-machine interface for on-site personnel to view. Using standard network protocols such as TCP / IP, the industrial control board encapsulates the data frames into data packets that can be transmitted via 4G / LoRa networks. These data packets are then sent back to the server in the central control room via a connected 4G / LoRa module. This completes one full data acquisition and transmission cycle.
Claims
1. A real-time synchronization system for multi-source monitoring data of offshore wind power structures based on heterogeneous architecture, comprising a pile foundation mid-structure monitoring system, a pile foundation bottom structure monitoring system, a blade structure monitoring system, an environmental monitoring system, a heterogeneous architecture data processing module, a GPS module, a 4G processing module, a LoRa module, and an industrial control board; characterized in that: The pile foundation mid-structure monitoring system, pile foundation bottom structure monitoring system, blade structure monitoring system, environmental monitoring system, and GPS module are respectively connected to the input ports of the heterogeneous architecture data processing module. The output of the heterogeneous architecture data processing module is connected to the industrial control module. The LoRa module and 4G module can receive wireless signals sent by the central control room and connect to the heterogeneous architecture data processing module. The heterogeneous architecture data processing module includes a programmable logic chip and a microprocessor chip, and internally runs data input interfaces, interface implementation logic, and data processing algorithms that are compatible with each system and industrial control board. It includes a GPS module data input interface and time extraction logic; and LoRa module and 4G module data receiving interfaces and command extraction logic.
2. The real-time synchronization system for multi-source monitoring data of offshore wind power structures based on heterogeneous architecture according to claim 1, characterized in that: The pile foundation mid-structure monitoring system includes an acceleration sensor, an inclination sensor, a fiber optic strain sensor, and a demodulator.
3. The real-time synchronization system for multi-source monitoring data of offshore wind power structures based on heterogeneous architecture according to claim 1, characterized in that: The pile foundation bottom structure monitoring system includes an acceleration sensor, an inclination sensor, a fiber optic strain sensor, and a demodulator.
4. The real-time synchronization system for multi-source monitoring data of offshore wind power structures based on heterogeneous architecture according to claim 1, characterized in that: The blade structure monitoring system includes an accelerometer, an inclination sensor, a fiber optic strain sensor, and a demodulator.
5. The real-time synchronization system for multi-source monitoring data of offshore wind power structures based on heterogeneous architecture according to claim 1, characterized in that: The environmental monitoring system includes water level and wave radar, wind speed and direction sensor, and video monitoring device.
6. The real-time synchronization system for multi-source monitoring data of offshore wind power structures based on heterogeneous architecture according to claim 1, characterized in that: The GPS module is a low-power navigation and positioning module that supports multiple satellite navigation systems.
7. The real-time synchronization system for multi-source monitoring data of offshore wind power structures based on heterogeneous architecture according to claim 1, characterized in that: LoRa modules are LoRaWAN node modules with a pin-type connection method, allowing for the installation of antennas.
8. A method for real-time synchronization of multi-source monitoring data of offshore wind power structures based on heterogeneous architecture, characterized in that, The system according to any one of claims 1-7 includes the following steps: (1) The main control room sends a data read command via LoRa protocol or 4G network broadcast; (2) The heterogeneous architecture module receives and distributes the parsed instructions from the LoRa / 4G module; (3) Each sensor receives the corresponding data reading command; (4) Each sensor returns monitoring data; (5) The heterogeneous architecture of the single wind power synchronous system receives data and performs real-time algorithm processing on each; (6) In the heterogeneous architecture, the logic unit frames the data from each sensor and adds a timestamp; (7) In a heterogeneous architecture, the logic unit sends framed data to the industrial control board; (8) The industrial control board sends the monitoring data of a single wind turbine pile to the main control room.
9. The method for real-time synchronization of multi-source monitoring data of offshore wind power structures based on heterogeneous architecture according to claim 8, characterized in that, The parallel data processing method within the heterogeneous architecture includes: the heterogeneous architecture simultaneously or rapidly receives data packets from various sensors through its hardware interface controller and temporarily stores them in an external high-speed buffer; after preliminary processing, the real-time data processing method for accelerometer, tilt sensor, and fiber optic strain sensor data within the heterogeneous architecture is as follows: accelerometer data undergoes low-pass filtering on the microprocessor chip within the heterogeneous architecture, followed by wavelet transform processing. The wavelet transform uses a 5 / 3 discrete wavelet transform with inertial splitting; tilt sensor data undergoes weighted moving average processing on the microprocessor chip within the heterogeneous architecture, followed by low-pass filtering; fiber optic strain data undergoes band-pass filtering on the microprocessor chip within the heterogeneous architecture; after noise reduction processing, the three types of data are transmitted into the logic chip for neural network acceleration. Risk characteristics are assessed in real time, and pre-trained neural network parameters are stored in the external memory chip of the heterogeneous architecture logic chip. For the identified risk information, the heterogeneous architecture unit completes risk control at the edge. After the three types of data processing are completed, they need to undergo different delays to ensure that the three data time scales are aligned. Subsequently, multiple monitoring data are packaged in the heterogeneous architecture in the following order: pile foundation mid-structure monitoring data, pile foundation bottom structure monitoring data, blade structure monitoring data, and environmental monitoring data. The first three types of monitoring data are composed of acceleration sensor, tilt sensor, and fiber optic strain data arranged in sequence. The environmental monitoring data is composed of wave radar data, anemometer data, and video data arranged in sequence. Before uploading, timestamp data is added to the end of the packaged data to complete the construction of a time-identified data packet.