Wind turbine generator hoisting construction tower drum operation system and construction method thereof

Abstract

The invention relates to the technical field of wind power hoisting construction, and discloses a wind turbine generator hoisting construction tower drum operation system and a construction method thereof. The method comprises the following steps: acquiring initial environment parameters through a multi-source sensor network arranged in a tower hoisting area of the wind turbine generator, and generating a space environment model of the tower hoisting area by adopting a three-dimensional modeling technology; the hoisting area is divided into a plurality of hoisting operation units, tower drum structure state collection operation is executed on each hoisting operation unit, and tower drum displacement characteristics and foundation resonance characteristics are obtained; performing wind load response simulation by using a wind speed measurement value to generate a static wind pressure influence matrix; a planned moving track of the hoisting equipment is obtained, and the advancing sequence of the hoisting equipment during tower drum hoisting on the planned moving track is adjusted; and in the tower drum hoisting process, the dynamic wind pressure influence value is calculated, and the hoisting direction of the hoisting equipment is dynamically corrected based on the dynamic wind pressure influence value and the static wind pressure influence value of the current hoisting operation unit.

Description

Technical Field

[0001] This invention relates to the field of wind power hoisting and construction technology, specifically to a wind turbine hoisting and construction tower operation system and its construction method. Background Technology

[0002] Wind power generation, as an important component of clean energy, has experienced rapid development globally. With the continuous increase in the capacity of individual wind turbine units, the height and weight of their towers have also increased accordingly, making tower hoisting one of the most technically challenging and safety-risk aspects of wind farm construction. Tower hoisting operations are typically conducted at high altitudes in open areas, making them highly susceptible to the significant impact of wind loads. Wind loads are the primary threat to hoisting safety, exhibiting unpredictable and dynamically changing characteristics. Excessive wind speeds not only directly affect the stability and control precision of the hoisting equipment but can also cause swaying, displacement, or even collisions between the already positioned or being hoisted tower sections and the hoisting equipment, leading to major safety accidents.

[0003] Traditional tower hoisting methods rely heavily on the experience and qualitative judgment of construction personnel. Hoisting plans are typically based on historical meteorological data and limited on-site wind speed measurements, lacking a refined and structured analysis of the wind field characteristics in the hoisting area. Decisions during the hoisting process, such as whether to continue work or how to adjust the hoisting posture, are mostly made by on-site supervisors based on anemometer readings and subjective experience, lacking quantitative data support and forward-looking risk assessment. This experience-based approach has significant limitations and is ill-suited to complex and ever-changing wind field environments. Due to a lack of understanding of the spatial distribution differences in wind loads within the hoisting area, hoisting plans often employ a uniform operational procedure, failing to differentiate and sequentialize operations based on the wind pressure influence at different locations within the area.

[0004] While existing technologies monitor wind speed, they typically only focus on whether the average wind speed exceeds a safe threshold, neglecting the gustability and turbulent characteristics of the wind, as well as the potential dynamic interactions between the wind and the tower structure and foundation. These complex wind-induced responses significantly increase the dynamic loads during hoisting, affecting structural stability. Furthermore, the movement of the hoisting equipment itself and the lifting process of tower components alter the local flow field, generating dynamic wind pressure changes that are not adequately considered or compensated for in real time using traditional methods. Therefore, current wind turbine tower hoisting operations suffer from problems such as crude analysis, subjective decision-making, and poor adaptability in addressing the impact of wind loads. There is an urgent need for an intelligent operation method that can precisely assess the spatial distribution of wind loads, predict their impact, and dynamically adjust hoisting strategies according to environmental changes, thereby improving the safety and accuracy of the hoisting process. Summary of the Invention

[0005] The purpose of this invention is to provide a wind turbine tower hoisting operation system and construction method thereof to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides a method for hoisting and constructing the tower of a wind turbine, the method comprising: Initial environmental parameters are collected by a multi-source sensor network deployed in the wind turbine tower hoisting area. These parameters include wind speed, temperature, and humidity measurements. Based on these initial environmental parameters, a spatial environment model of the tower hoisting area is generated using 3D modeling technology. Based on the spatial environment model, the hoisting area is divided into multiple hoisting operation units. The tower structure status acquisition operation is performed on each hoisting operation unit to obtain the tower displacement characteristics and foundation resonance characteristics corresponding to each hoisting operation unit. Wind load response simulation was performed on the tower displacement characteristics and foundation resonance characteristics corresponding to each hoisting operation unit using wind speed measurements, generating a static wind pressure influence matrix. Obtain the planned movement trajectory of the hoisting equipment, and combine each static wind pressure influence value in the static wind pressure influence matrix with its associated hoisting operation unit to adjust the movement sequence of the hoisting equipment when hoisting the tower on the planned movement trajectory; During the tower hoisting process, the instantaneous environmental parameters of the current hoisting operation unit are collected through a real-time monitoring system, the dynamic wind pressure impact value is calculated, and the hoisting direction of the hoisting equipment is dynamically corrected based on the dynamic and static wind pressure impact values ​​of the current hoisting operation unit.

[0007] Preferably, the specific steps of dividing the hoisting area into multiple hoisting operation units according to the spatial environment model, performing tower structure state acquisition operations on each hoisting operation unit, and obtaining the tower displacement characteristics and foundation resonance characteristics corresponding to each hoisting operation unit include: The spatial environment model is processed into a three-dimensional mesh, and the hoisting area is divided into multiple hoisting operation units of equal size based on the elevation gradient change rate. The wind farm monitoring system acquires real-time prevailing wind direction data, determines the position of the initial hoisting operation unit based on the prevailing wind direction data, uses a laser scanner to collect tower displacement characteristics of the initial hoisting operation unit, and collects tower displacement characteristics of subsequent hoisting operation units in sequence according to wind direction. Spatial airflow distribution data is acquired through an airflow sensor array. An airflow motion pattern vector is generated based on the spatial airflow distribution data. The initial hoisting operation unit is reselected according to the airflow motion pattern vector. Ground resonance characteristics of the initial hoisting operation unit are collected using vibration sensors. Ground resonance characteristics of subsequent hoisting operation units are collected sequentially along the direction of the airflow motion pattern vector.

[0008] Preferably, before dividing the hoisting area into multiple hoisting operation units according to the spatial environment model to collect tower structure status, the method further includes: acquiring the prevailing wind direction data and airflow movement pattern vector at the current time point in real time through the data interface of the wind farm central control system, and storing the prevailing wind direction data and airflow movement pattern vector in a temporary buffer for use by the feature acquisition module.

[0009] Preferably, the specific steps for simulating the wind load response of the tower displacement characteristics and foundation resonance characteristics corresponding to each hoisting operation unit using wind speed measurements to generate a static wind pressure influence matrix include: The average wind speed and turbulence intensity are calculated based on the wind speed measurements to determine the calculation benchmark for the reference wind pressure depth. The tower displacement characteristics and foundation resonance characteristics corresponding to the first hoisting operation unit are extracted. Combined with the three-dimensional coordinates and reference wind pressure depth of the first hoisting operation unit, adjacent hoisting operation units are selected as comparison areas. The wind load response is simulated using the finite element analysis method to obtain the static wind pressure influence value corresponding to the first hoisting operation unit. The tower displacement characteristics and foundation resonance characteristics corresponding to the remaining hoisting operation units are extracted sequentially. The same finite element analysis method is used to calculate the static wind pressure influence value of each hoisting operation unit. All static wind pressure influence values ​​are then arranged in a matrix according to the mesh structure of the space environment model to generate a static wind pressure influence matrix.

[0010] Preferably, the specific steps for obtaining the planned movement trajectory of the hoisting equipment, combining each static wind pressure influence value in the static wind pressure influence matrix and its associated hoisting operation unit, and adjusting the movement sequence of the hoisting equipment when hoisting the tower on the planned movement trajectory include: Read each static wind pressure influence value and its associated hoisting operation unit identification information from the static wind pressure influence matrix; Obtain all preset path segments of the hoisting equipment and identify the adjacent hoisting operation units connected to each preset path segment; The initial working position of the hoisting equipment is determined. The static wind pressure impact values ​​of all adjacent hoisting operation units corresponding to each preset path segment are weighted and summed to obtain the comprehensive risk coefficient of the preset path segment. The particle swarm optimization algorithm is used to optimize the movement sequence of the hoisting equipment from the initial working position based on the comprehensive risk coefficient.

[0011] Preferably, the specific steps for dynamically correcting the hoisting direction of the hoisting equipment based on the dynamic and static wind pressure influence values ​​of the current hoisting operation unit include: Calculate the absolute deviation between the dynamic wind pressure influence value and the static wind pressure influence value of the current hoisting operation unit; When the absolute deviation value exceeds the preset safety threshold, the correction radius is determined based on the real-time wind speed data of the current hoisting operation unit, and the dynamic wind pressure influence value and static wind pressure influence value of the surrounding hoisting operation units are obtained based on the correction radius. The dynamic and static wind pressure influence values ​​of the surrounding hoisting operation units are input to the direction correction controller to generate hoisting direction adjustment commands for the hoisting equipment, thereby achieving dynamic correction of the hoisting direction.

[0012] Preferably, when collecting wind speed measurements from initial environmental parameters through a multi-source sensor network deployed in the wind turbine tower hoisting area, a Doppler lidar anemometer is used for wind speed acquisition, and accurate wind speed measurements are obtained through data filtering.

[0013] Preferably, after collecting the instantaneous environmental parameters of the current hoisting operation unit through a real-time monitoring system and calculating the dynamic wind pressure impact value during the tower hoisting process, the method further includes: The inertial measurement unit installed on the hoisting equipment collects the tower's attitude angle change data in real time, fuses the attitude angle change data with the dynamic wind pressure influence value, and uses the Kalman filter algorithm to calibrate the dynamic wind pressure influence value in real time.

[0014] Preferably, after adjusting the movement sequence of the hoisting equipment during tower hoisting on the planned trajectory, the method further includes: Collect wind pressure impact data and environmental parameters from historical hoisting operations, construct a wind pressure impact prediction database, train a wind pressure impact prediction model using time series analysis, and iteratively optimize the generation process of the static wind pressure impact matrix using the wind pressure impact prediction model.

[0015] Preferably, the present invention also includes a wind turbine hoisting construction tower operation system, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of the above-described wind turbine hoisting construction tower operation method.

[0016] Compared with the prior art, the beneficial effects of the present invention are: This invention collects environmental parameters through a multi-source sensor network and constructs a three-dimensional spatial environment model of the hoisting area, achieving a digital and structured representation of the working environment and providing an accurate spatial framework for subsequent refined analysis. By dividing the hoisting area into multiple operational units and collecting structural state data for each unit, obtaining tower displacement characteristics and foundation resonance characteristics, this method decomposes the macroscopic hoisting area into smaller, more easily analyzed units. This allows the assessment of structural response characteristics to be grounded in specific spatial locations, enhancing the targeted nature of the analysis.

[0017] Performing wind load response simulation and generating a static wind pressure influence matrix is ​​one of the core steps of this method. It quantifies the potential impact of wind load on the tower structure and foundation within different hoisting operation units at specific wind speeds. This matrix transforms the abstract effect of wind force into influence values ​​at specific spatial locations, providing crucial data for pre-planning hoisting operation sequences. Adjusting the movement sequence of hoisting equipment along the planned trajectory based on the static wind pressure influence matrix means that the path arrangement for hoisting operations no longer solely considers the shortest geometric path or operational habits, but incorporates consideration of the spatial distribution differences of wind loads. For example, key hoisting steps can be prioritized in areas with less wind pressure influence, or the sequence can be adjusted to avoid operations under unfavorable wind directions, thus improving the safety margin of the operation from the planning stage.

[0018] Introducing real-time monitoring and dynamic correction mechanisms during the hoisting process significantly enhances the method's adaptability and robustness. By collecting instantaneous environmental parameters and calculating dynamic wind pressure impact values, the system can detect deviations between the actual wind field and anticipated conditions. Dynamically correcting the hoisting direction based on pre-calculated static impact values ​​means that the hoisting operation can respond to real-time wind changes. For example, it can fine-tune the tower section's aerial attitude or movement path based on the current wind direction and speed to compensate for offsets or swaying caused by wind loads. This closed-loop control method transforms hoisting operations from a static, predetermined process to an intelligent behavior capable of dynamic interaction with the environment.

[0019] The entire methodology forms a complete technical process from environmental perception, modeling and analysis, impact prediction, strategy planning to dynamic adjustment. By introducing quantitative wind load impact assessment and dynamic control based on real-time feedback, it effectively reduces the over-reliance on personnel experience in hoisting operations and minimizes the uncertainty of subjective judgment. Through spatial unit analysis and temporal dynamic adjustment, the methodology can manage wind load risks more precisely, guiding hoisting operations to be carried out within relatively safer spatial and temporal windows, thereby improving the safety, stability, and positioning accuracy of large tower hoisting processes in complex wind field environments. Attached Figure Description

[0020] Figure 1 This is a schematic diagram illustrating the working principle of the wind turbine tower hoisting operation method described in this invention. Figure 2 A flowchart illustrating the steps for segmenting the hoisting operation unit and collecting the status of the tower structure. Figure 3 This is a flowchart of the steps involved in generating the static wind pressure influence matrix; Figure 4 A comparative analysis chart of industrial protocol latency and data packet loss rate in wind farms; Figure 5 A comparison chart of convergence and paths for the particle swarm optimization algorithm for the hoisting path of wind turbine towers. Detailed Implementation

[0021] 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.

[0022] Please see Figure 1 This invention provides a wind turbine tower hoisting operation system and its construction method. The method includes: a multi-source sensor network deployed in the wind turbine tower hoisting area to collect initial environmental parameters, including wind speed, temperature, and humidity measurements; generating a spatial environment model of the tower hoisting area based on these initial environmental parameters using 3D modeling technology; dividing the hoisting area into multiple hoisting operation units according to the spatial environment model; performing tower structure state acquisition operations on each hoisting operation unit to obtain tower displacement characteristics and foundation resonance characteristics corresponding to each hoisting operation unit; and using wind speed measurements... For each hoisting operation unit, the tower displacement characteristics and foundation resonance characteristics are simulated under wind load to generate a static wind pressure influence matrix. The planned movement trajectory of the hoisting equipment is obtained, and the movement sequence of the hoisting equipment during tower hoisting is adjusted by combining the static wind pressure influence value of each static wind pressure influence value in the static wind pressure influence matrix and its associated hoisting operation unit. During the tower hoisting process, the instantaneous environmental parameters of the current hoisting operation unit are collected through a real-time monitoring system, the dynamic wind pressure influence value is calculated, and the hoisting direction of the hoisting equipment is dynamically corrected based on the dynamic and static wind pressure influence values ​​of the current hoisting operation unit.

[0023] Example 1: See Figure 2In the specific implementation, the spatial environment model is processed into a three-dimensional mesh. Based on the elevation gradient change rate, the hoisting area is divided into multiple hoisting operation units of equal size. The three-dimensional meshing process adopts a regular cube meshing algorithm. The mesh size is determined according to the total area and complexity of the hoisting area. The elevation gradient change rate is calculated through a digital elevation model (DEM), which is generated from terrain data collected by a multi-source sensor network. The boundary coordinates of each hoisting operation unit are stored in a spatial database for subsequent retrieval. In the specific implementation, the three-dimensional meshing process includes importing the three-dimensional point cloud data of the spatial environment model. The point cloud data is collected by a laser scanner and a photogrammetry system. Then, a mesh generation algorithm, such as the traveling cube algorithm or Delaunay triangulation, is applied to convert the point cloud into a continuous mesh surface. The elevation gradient change rate is obtained by calculating the elevation difference of the mesh nodes and dividing it by the horizontal distance. This is used to identify areas with drastic terrain changes, thereby ensuring that the divided hoisting operation units are consistent in the height direction and avoiding uneven unit sizes due to terrain undulations. In some embodiments, the grid size can be set to a fixed value, such as 10 meters × 10 meters × 10 meters, to accommodate the working range of standard hoisting equipment. The threshold for the elevation gradient change rate is set according to the site geological conditions; for example, when the change rate exceeds 5%, the grid density is adjusted to capture subtle terrain features. It can be understood that 3D meshing not only improves the spatial resolution of the hoisting area but also provides a structured framework for feature acquisition.

[0024] The wind farm monitoring system acquires real-time prevailing wind direction data, determines the location of the initial hoisting unit based on this data, and uses a laser scanner to collect tower displacement characteristics of the initial hoisting unit. Subsequent hoisting units are then sequentially acquired according to wind direction. The wind farm monitoring system integrates wind direction sensors and a data processing unit, outputting real-time prevailing wind direction data as angle values. The selection of the initial hoisting unit is based on the direction of the wind vector, typically located on the upwind side of the hoisting area. In practice, the wind farm monitoring system continuously collects wind direction information using ultrasonic wind direction sensors deployed around the tower, with a data sampling frequency of 1 Hz. The prevailing wind direction data, after being processed by moving average filtering, is used to calculate the wind rose diagram, thereby determining the three-dimensional coordinates of the initial unit. The laser scanner uses a phase-type or pulsed lidar, emitting a laser beam and receiving reflected signals. The displacement on the tower surface is calculated by measuring the round-trip time of the light wave. Displacement characteristics include linear and angular displacement. The acquisition sequence strictly follows the wind direction, starting from the initial unit and proceeding unit by unit along the wind vector direction to ensure the logical consistency of data acquisition and environmental uniformity. In some embodiments, the scanning parameters of the laser scanner can be adjusted according to the cell size, such as setting the scanning angle to 90 degrees to cover the entire cell area. During displacement feature acquisition, timestamps and location information are recorded in real time to facilitate subsequent data analysis. Optionally, the wind farm monitoring system can access meteorological satellite data to verify the accuracy of the prevailing wind direction. The laser scanner's data output format is point cloud or grid data, which is directly input into the feature processing software. It is understood that acquiring displacement features in sequence according to wind direction helps to capture the propagation effect of wind loads.

[0025] Spatial airflow distribution data is acquired through an airflow sensor array. Based on the spatial airflow distribution data, an airflow motion pattern vector is generated. The initial hoisting operation unit is reselected according to the airflow motion pattern vector. The ground resonance characteristics of the initial hoisting operation unit are collected using a vibration sensor. The ground resonance characteristics of subsequent hoisting operation units are collected sequentially along the direction of the airflow motion pattern vector. The airflow sensor array consists of multiple hot-film anemometers or pitot tubes, arranged around the hoisting area. The spatial airflow distribution data includes flow velocity, flow direction, and turbulence intensity. The airflow motion pattern vector is generated through a vector synthesis algorithm, representing the overall direction of airflow movement. In practice, the airflow sensor array is deployed in a grid pattern, with the sensor spacing set according to the size of the hoisting area, such as a 20-meter interval. The data acquisition frequency is 10 Hz. Principal component analysis is used to generate the airflow motion pattern vector, extracting the dominant airflow direction from multi-sensor data. The newly selected initial hoisting operation unit is determined based on the vector's starting point, typically located at the airflow injection point. Vibration sensors, using piezoelectric or MEMS accelerometers, are installed on the foundation surface to collect vibration frequency and amplitude data. Foundation resonance characteristics include natural frequencies and damping ratios. The acquisition sequence follows the direction of the airflow motion pattern vector to ensure consistency with airflow dynamics. In practice, the data from the airflow sensor array is transmitted to the central processing unit via a wireless transmission module. The calculation of the airflow motion pattern vector uses linear algebra tools such as vector projection and normalization. The vibration sensor sampling rate is set to 1000 Hz to capture high-frequency vibration signals. Foundation resonance characteristics are obtained by analyzing the vibration data using fast Fourier transform. Optionally, the airflow sensor array can be calibrated for different altitudes to adapt to complex terrain, and the installation depth of the vibration sensor can be adjusted according to the foundation type, such as shallow burial in rock foundations and deep burial in soft soil foundations. It can be understood that the introduction of airflow motion pattern vectors enhances the environmental adaptability of feature acquisition.

[0026] In practical implementation, the tower displacement feature acquisition operation includes scanning each hoisting operation unit using a laser scanner. The laser scanner operates based on a preset program, emitting a laser beam and receiving the echo. High-precision point cloud data is generated by calculating the phase difference or time difference of the light waves. After filtering and registration, the displacement parameters of the tower structure, such as the maximum displacement and displacement gradient, are extracted and stored in a feature database. In practice, the laser scanner's scanning path is optimized according to the geometry of the hoisting operation unit, for example, using a zigzag path to cover the entire unit. The displacement features are calculated using the least squares method to fit the point cloud to a reference model, yielding the displacement deviation. During the acquisition process, environmental factors such as illumination and humidity are monitored in real time to ensure data quality. When acquiring data sequentially by wind direction, the data acquisition time for each unit is controlled within a few minutes to avoid the impact of sudden wind changes. In some embodiments, the laser scanner can integrate an inertial navigation system to compensate for errors caused by equipment movement. The displacement feature data can be output in CSV or JSON format for subsequent simulation. It can be understood that the automated process of displacement feature acquisition improves efficiency.

[0027] The acquisition of foundation resonance characteristics involves the deployment and data analysis of vibration sensors. Vibration sensors are fixed to the foundation surface with bolts or adhesive. After acquiring vibration signals, a signal amplifier amplifies weak signals, and an analog-to-digital converter digitizes the analog signals. Foundation resonance characteristics are identified through spectral analysis, such as using power spectral density estimation methods. In practice, the vibration sensors are arranged based on the center point of the hoisting unit, and the acquisition time lasts for several minutes to capture stable signals. When acquiring data along the airflow motion vector direction, the vector direction is updated in real time using airflow data to ensure that the acquisition path is dynamically synchronized with the airflow. Foundation resonance characteristics include multiple resonant frequencies and corresponding modal shapes, which are used to assess foundation stability. In practice, the vibration sensor data acquisition chain includes a pre-filter to remove noise, spectral analysis using a Hanning window to reduce leakage, and verification of consistency with theoretical models after resonance feature extraction. It can be understood that foundation resonance characteristic acquisition provides structural dynamics input for wind load response.

[0028] In practical implementation, the elevation gradient change rate calculation for 3D meshing employs a differential method. Elevation values ​​of adjacent grid points are extracted from the digital elevation model (DEM), and the gradient vector is calculated. The gradient change rate is defined as the magnitude of the vector. When dividing the hoisting operation units, the equal division of unit size is ensured by a grid index, and each unit's unique identifier is generated based on 3D coordinates. In practice, the resolution of the DEM is set to 1 meter to ensure gradient calculation accuracy. Meshing is performed using commercial software such as GIS tools or custom code. After unit division, the spatial environment model is updated to a mesh structure for easy access by the feature acquisition module. It can be understood that the application of the elevation gradient change rate improves terrain adaptability.

[0029] When acquiring real-time prevailing wind direction data through a wind farm monitoring system, the monitoring system and sensor network are connected via industrial Ethernet, using Modbus or OPCUA as the data protocol. The prevailing wind direction data undergoes quality control, such as outlier removal. The position calculation of the initial hoisting unit uses a geometric projection method to map the wind direction angle to grid coordinates. In practice, the wind farm monitoring system's data interface supports real-time streaming processing, with prevailing wind direction data updated every second. The initial unit selection algorithm is based on nearest neighbor search to ensure rapid response. The laser scanner's trigger signal is synchronized with the wind direction data, achieving automated acquisition. It can be understood that the integration of real-time prevailing wind direction data enhances the rationality of the acquisition timing.

[0030] When collecting tower displacement characteristics of subsequent hoisting units sequentially according to wind direction, the sequential logic is based on the angular increment of the wind direction vector. After each unit is collected, the system automatically indexes the next unit. The motion control of the laser scanner is achieved through servo motors, and path planning avoids collisions. In practice, the wind direction sequence is maintained using a queue data structure, and the collected data is uploaded to the cloud platform in real time. Displacement characteristics include static and dynamic components, which are processed separately to distinguish between permanent deformation and temporary vibration. It can be understood that sequential acquisition optimizes resource utilization.

[0031] Spatial airflow distribution data is acquired through an airflow sensor array. The sensor array is calibrated before deployment, and the velocity-output relationship is calibrated using wind tunnel testing. The spatial airflow distribution data is used to generate a continuous field through interpolation methods such as Kriging interpolation. The calculation of the airflow motion pattern vector is based on vector field integration. In practice, the data acquisition cycle of the airflow sensor array is 1 second, and the generation of the airflow motion pattern vector is updated every minute. When reselecting the initial hoisting operation unit, the vector starting point is aligned with the grid node. The acquisition parameters of the vibration sensor, such as gain, are dynamically adjusted according to the signal strength. It can be understood that the dynamic updating of the airflow motion pattern vector improves the accuracy of feature acquisition.

[0032] When using vibration sensors to collect ground resonance characteristics, the sensor's frequency response range covers 0.1-100 Hz to capture low-frequency ground vibrations. The collected data undergoes bandpass filtering to remove environmental noise, and resonance characteristics are extracted using modal analysis software. In practice, the installation location of the vibration sensors is determined based on the ground geological exploration report. The acquisition time ensures signal stability. When acquiring data along the airflow motion pattern vector direction, the vector direction is corrected using real-time airflow data to avoid path deviation.

[0033] Example 2: See Figure 3In practice, the prevailing wind direction data and airflow motion pattern vector at the current time point are acquired in real time through the data interface of the wind farm central control system. The prevailing wind direction data and airflow motion pattern vector are stored in a temporary buffer for the feature acquisition module to access. The data interface of the wind farm central control system adopts industry standard protocols such as OPCUA or ModbusTCP to establish communication connections with the meteorological stations and sensor networks deployed in the wind farm. The real-time data stream is updated at a specific frequency, such as once per second. The temporary buffer is usually implemented using a high-speed circular buffer or a first-in-first-out queue data structure. Its capacity is configured according to the data update frequency and the historical data retention requirements to prevent data overflow and ensure that the feature acquisition module can access the latest environmental parameters in a timely manner. In practical implementation, the communication parameters of the data interface need to be pre-configured, including IP address, port number, and data type definition. The prevailing wind direction data comes from a high-precision wind direction sensor installed on the anemometer tower, and its data format is angle values ​​(degrees) or unit vectors. The airflow motion pattern vector is generated by the algorithm module built into the central control system after vector synthesis calculation of readings from multiple airflow sensors. The process of storing the data in the temporary buffer includes a data verification step, such as checking whether the data values ​​are within a reasonable physical range; invalid data is marked and discarded. The feature acquisition module reads this data from the temporary buffer through a predefined application programming interface, thereby ensuring that subsequent tower structure status acquisition operations are based on the most accurate current environmental conditions. In some embodiments, the data interface can support redundant communication links, such as using wired Ethernet and wireless cellular networks simultaneously, to enhance communication reliability. The storage period of the temporary buffer can be set to the data from the most recent few minutes to adapt to short-term fluctuations in wind field conditions. Optionally, the data interface can integrate a data timestamp synchronization function to ensure that data received from different sources has a consistent time base. It can be understood that this real-time data acquisition and caching mechanism provides a dynamic environmental information foundation for the entire hoisting operation.

[0034] The specific process of simulating the wind load response of each hoisting operation unit's tower displacement characteristics and foundation resonance characteristics using wind speed measurements to generate a static wind pressure influence matrix involves first calculating the average wind speed and turbulence intensity based on the wind speed measurements to determine the calculation benchmark for the reference wind pressure depth. The average wind speed is calculated by arithmetically averaging wind speed measurements over a continuous time series. The turbulence intensity is calculated by dividing the standard deviation of wind speeds within the same time period by the average wind speed. The reference wind pressure depth is a parameter based on boundary layer wind engineering theory, and its calculation benchmark is usually related to the ground roughness length and observation height. In practice, wind speed measurements are obtained from continuous sampling data from Doppler lidar anemometers or cup anemometers. The calculation time window for the average wind speed is selected based on the stability of the wind field, for example, an average duration of ten minutes. The calculated result of the turbulence intensity is a dimensionless value used to characterize the fluctuation characteristics of the wind. The determination of the reference wind pressure depth requires consulting wind engineering specifications or calculating using empirical formulas. The result serves as an important input parameter for defining the wind pressure load distribution in the finite element analysis model. It is understandable that accurately calculating the average wind speed, turbulence intensity, and reference wind pressure depth is a prerequisite for conducting reliable wind load response simulations.

[0035] In the specific implementation, the tower displacement characteristics and foundation resonance characteristics corresponding to the first hoisting operation unit are extracted. Combining the three-dimensional coordinates of the first hoisting operation unit and the reference wind pressure depth, adjacent hoisting operation units are selected as comparison areas. The finite element analysis method is used to simulate the wind load response and obtain the static wind pressure influence value corresponding to the first hoisting operation unit. The tower displacement characteristics include the displacement of key points of the tower structure extracted from laser scanning data. The foundation resonance characteristics include the natural frequency and damping ratio of the foundation obtained from vibration sensor data. The three-dimensional coordinates of the first hoisting operation unit are obtained from the mesh index of its spatial environment model. The selection rule of adjacent hoisting operation units is based on spatial proximity, such as units sharing surfaces or edges. The finite element analysis model needs to be pre-constructed, including the geometric model of the tower structure, material properties, and boundary conditions. The wind load is applied to the model surface based on the wind pressure converted from the reference wind pressure depth and wind speed measurement value. The stress and strain responses of the tower and foundation are obtained by solving the structural mechanics equations. The static wind pressure influence value can be a comprehensive index, such as the maximum equivalent stress or the maximum displacement. In practice, finite element analysis can be performed using commercial software or a custom program. The analysis process includes mesh generation, load application, solving, and post-processing. The introduction of a comparison region helps to assess the significance of wind pressure influence on the first hoisting unit through relative comparison. The calculated static wind pressure influence value is assigned an identifier to the first hoisting unit and stored. In some embodiments, the complexity of the finite element model can be adjusted according to computational resources, for example, using a simplified beam element model or a more refined shell element model. The size of the comparison region can be set to include a specific number of neighboring elements around the first hoisting unit. Optionally, the static wind pressure influence value can be normalized to facilitate comparison between different elements.

[0036] The tower displacement characteristics and foundation resonance characteristics corresponding to the remaining hoisting operation units are extracted sequentially. The same finite element analysis method is used to calculate the static wind pressure influence value of each hoisting operation unit. All static wind pressure influence values ​​are then arranged into a matrix according to the mesh structure of the space environment model to generate a static wind pressure influence matrix. This process is repeated iteratively, performing wind load response simulations on each hoisting operation unit within the hoisting area. In practice, the extraction of feature data for the remaining units can be achieved by sequentially traversing the entire mesh index list, ensuring that each unit is processed. The finite element analysis method and its parameter settings must be completely consistent with those used when calculating the first hoisting operation unit to guarantee comparability of results. After each unit is calculated, its static wind pressure influence value along with its unique mesh coordinates is recorded. The static wind pressure influence values ​​of all units are then filled into a multidimensional array based on their mesh coordinates. The structure of this array completely corresponds to the three-dimensional mesh of the space environment model, thus generating the static wind pressure influence matrix. In practice, the matrix arrangement operation can be automated using a programming script. The matrix data can be stored as a text file or a binary file for easy reading and retrieval by the subsequent hoisting path planning algorithm. It is understandable that generating the static wind pressure influence matrix is ​​a key step in systematically organizing the discrete unit wind pressure response results, providing a spatialized risk distribution map for global hoisting operation planning.

[0037] In practical implementation, when the data interface of the wind farm central control system acquires data in real time, it needs to handle potential network latency and data packet loss issues. This can be achieved by setting timeout retransmission mechanisms and data retransmission requests. Temporary buffer management strategies include overwriting the oldest data when new data arrives, or marking data as consumed after it is read by the feature acquisition module. Dominant wind direction data and airflow motion pattern vectors may require coordinate transformation before storage to match the coordinate system used by the hoisting operation unit. In some embodiments, the data interface can be configured to simultaneously monitor multiple data sources and use a voting mechanism or weighted average method to handle data differences between different sources. The temporary buffer can be divided into multiple logical partitions to store different types of environmental parameters. It is understood that the robust design of the data interface and buffer ensures high availability of environmental parameters.

[0038] In practical implementation, when using the finite element method to simulate wind load response, the application of wind pressure loads needs to consider the influence of wind direction. Typically, wind pressure is decomposed into components acting on different surfaces of the structure. Tower displacement characteristics and foundation resonance characteristics are input into the finite element model as initial or boundary conditions. For example, measured displacements are used as constraints, or resonant frequencies are used to verify the model's dynamic characteristics. The use of reference wind pressure depth is reflected in defining an exponential or logarithmic law profile of wind pressure variation with height. It is understandable that the accuracy of finite element analysis depends on the realism of the input parameters and the rationality of the model settings.

[0039] In practical implementation, the generation process of the static wind pressure influence matrix can integrate verification steps, such as checking for null or outlier values ​​in the matrix and interpolating or recalculating them. The arrangement order of the matrix must be strictly consistent with the spatial index of the hoisting operation unit to facilitate quick retrieval of static wind pressure influence values ​​at any location using coordinates. In some embodiments, the matrix can be accompanied by metadata, such as the generation time and the version number of the wind speed measurement value used, to facilitate version management and traceability.

[0040] See Figure 4 This paper presents a comparative analysis of the latency characteristics and data packet loss rate of the OPCUA and ModbusTCP protocols used in the data interface of the wind farm central control system. The OPCUA latency (blue curve) and ModbusTCP latency (red curve) are presented in milliseconds (ms), while the data packet loss rate (green dashed line) is a dimensionless indicator. In terms of latency, ModbusTCP latency is generally higher than OPCUA latency and fluctuates more dramatically. For example, at several time points (such as around 30), ModbusTCP latency approaches 30ms, while OPCUA latency mostly remains in the 5-22ms range. Although OPCUA latency also fluctuates, it shows a significant trough (around 5ms) at certain points (such as around 17). Analyzing the data packet loss rate, it is correlated with the latency performance of the two protocols. When the protocol latency fluctuates significantly, the data packet loss rate often peaks (e.g., during periods corresponding to OPCUA latency troughs, the data packet loss rate is also at a low level, while during ModbusTCP latency peaks, the data packet loss rate also increases significantly).

[0041] Example 3: In specific implementation, after obtaining the planned movement trajectory of the hoisting equipment, the movement sequence of the hoisting equipment during tower hoisting is adjusted by combining each static wind pressure influence value in the static wind pressure influence matrix and its associated hoisting operation unit. The specific steps include reading the identification information of each static wind pressure influence value and its associated hoisting operation unit from the static wind pressure influence matrix. The static wind pressure influence matrix is ​​stored in computer memory or database in the form of a multi-dimensional array. Each array element corresponds to the static wind pressure influence value of a hoisting operation unit. The identification information includes the grid index of the hoisting operation unit in the spatial environment model, such as row number, column number, and layer number. The reading operation is implemented through a programming interface such as SQL query or array index access to ensure data integrity and access efficiency. In practice, all preset path segments of the hoisting equipment are acquired, and the adjacent hoisting operation units connected to each preset path segment are identified. The preset path segments are generated by path planning software based on the topology of the hoisting area and are represented as a series of continuous spatial line segments or curve segments. Each preset path segment stores the coordinate information of the start and end points. The identification of adjacent hoisting operation units is completed by calculating the spatial intersection of the preset path segment and the boundary of the hoisting operation unit or by using a nearest neighbor algorithm. For example, the ray casting method is used to detect which units the path segment passes through. The results are stored in the form of a list of the set of hoisting operation units associated with each preset path segment. In practice, the initial operating position of the hoisting equipment is determined. The static wind pressure impact values ​​of all adjacent hoisting operation units corresponding to each preset path segment are weighted and summed to obtain the comprehensive risk coefficient for that preset path segment. The initial operating position is obtained from the hoisting operation plan, typically the entrance point of the hoisting area or the first hoisting point. In the weighted summation operation, the static wind pressure impact value of each adjacent hoisting operation unit is multiplied by a weighting coefficient. This weighting coefficient is determined based on the distance or relative orientation between the preset path segment and the hoisting operation unit; for example, the closer the distance, the greater the weight. The formula for calculating the comprehensive risk coefficient is as follows:

[0042] in: This represents the overall risk coefficient of the j-th preset path segment. This indicates the number of adjacent hoisting operation units connected to the preset path segment. This represents the weight coefficient of the k-th hoisting operation unit. This represents the static wind pressure impact value of the k-th hoisting operation unit. The weighting coefficients are normalized to ensure a sum of 1. After the comprehensive risk coefficient is calculated, it is stored as a vector or list. In specific implementation, the particle swarm optimization algorithm is used to optimize the movement sequence of the hoisting equipment from its initial operating position based on the comprehensive risk coefficient. The particle swarm optimization algorithm initializes a set of particles, each representing a movement sequence scheme, i.e., the access sequence of the hoisting operation units. The particle position is encoded as a discrete sequence of hoisting operation unit indices. The fitness function is defined as the total risk of the entire movement sequence, i.e., the sum of the comprehensive risk coefficients of all preset path segments. The optimization objective is to minimize the total risk. The iterative process of the particle swarm optimization algorithm includes velocity updates and position updates. The velocity update formula is:

[0043] in: This represents the velocity vector of particle i at iteration number t. Indicates the inertia weight parameter. and Represents the acceleration constant. and Represents a random number between 0 and 1. This represents the historical best position of particle i. Indicates the globally optimal position. Let represent the position vector of particle i at iteration number t. Position updates are achieved by adding velocity to the current position and mapping it to a discrete sequence space. The algorithm terminates when the maximum number of iterations is reached or fitness convergence occurs, ultimately outputting the optimal travel order. In some embodiments, the reading operation of the static wind pressure influence matrix can be performed in batches to reduce I / O overhead, and an identification information verification step is added to ensure data consistency. In some embodiments, the generation of preset path segments can consider the mechanical constraints of the hoisting equipment, such as the minimum turning radius, and the identification algorithm of adjacent hoisting operation units can be optimized to use a spatial index structure, such as an R-tree, to improve computation speed. In some embodiments, the weight coefficients of the weighted summation can be learned from historical data using machine learning methods, and the parameters of the particle swarm optimization algorithm, such as inertia weights, can adopt an adaptive adjustment strategy. It can be understood that the introduction of the comprehensive risk coefficient quantifies the wind pressure risk of the hoisting path.

[0044] After adjusting the movement sequence, wind pressure impact data and environmental parameters from historical hoisting operations are collected to construct a wind pressure impact prediction database. Historical wind pressure impact data includes static and dynamic wind pressure impact values ​​recorded in previous hoisting operations. Environmental parameters include wind speed measurements, temperature measurements, humidity measurements, and prevailing wind direction data. Data collection sources include hoisting operation logs, sensor network storage systems, and the central control system database. The wind pressure impact prediction database is implemented using a relational database or a time-series database. The table structure design includes timestamp, hoisting operation unit identifier, wind pressure impact value, and environmental parameter fields. Data cleaning and normalization are performed before data is entered into the database to remove noise and outliers. In practical implementation, a time series analysis method is used to train the wind pressure impact prediction model. The chosen time series analysis method is an autoregressive integral moving average model or a long short-term memory network. Training data is extracted from the wind pressure impact prediction database and arranged chronologically. The model input is a sequence of historical environmental parameters, and the output is a prediction of future wind pressure impact values. The training process includes data splitting into training and testing sets, model parameter initialization, gradient descent optimization, and cross-validation. Model performance is evaluated using root mean square error or mean absolute error. In practical implementation, the generation process of the static wind pressure impact matrix is ​​iteratively optimized using the wind pressure impact prediction model. This iterative optimization process is triggered before each new hoisting operation. The current environmental parameters are input into the wind pressure impact prediction model to obtain the predicted wind pressure impact values. These predicted values ​​are then used to adjust the static wind pressure impact values ​​in the static wind pressure impact matrix. The adjustment method can be weighted averaging or direct replacement. The optimized static wind pressure impact matrix is ​​used for subsequent hoisting path planning, forming a closed-loop feedback. Optionally, the wind pressure impact prediction database can be backed up regularly and have a version control mechanism. It can be understood that the application of time series analysis methods improves the accuracy of wind pressure prediction.

[0045] In practical implementation, the position vector encoding of the particle swarm optimization algorithm needs to handle discrete sequence constraints, such as through permutation encoding or random key representation. When calculating the fitness function, the feasibility of the movement order needs to be verified, such as avoiding repeated visits to cells, and the inertia weight parameter needs to be considered. The speedup constant is typically set between 0.4 and 0.9. and The setting is usually 2.0, for random numbers. and The particle swarm optimization (PSO) algorithm generates pseudo-random numbers. The particle swarm size is chosen based on problem complexity, such as 20 to 50 particles. The number of iterations is set to 100 to 500, and convergence is determined based on a fitness rate threshold. In practice, the construction of the wind pressure impact prediction database includes a data ETL process. Time series analysis training may involve selecting the lag order or optimizing the network structure. The iterative optimization process can use a sliding window mechanism to use only the most recent data to enhance model adaptability. In some embodiments, the PSO algorithm can be combined with local search strategies such as simulated annealing to improve solution quality. The wind pressure impact prediction model can integrate multiple base models and improve robustness through stacking or voting.

[0046] Example 4: In specific implementation, the absolute deviation between the dynamic wind pressure influence value and the static wind pressure influence value of the current hoisting operation unit is calculated. The dynamic wind pressure influence value is calculated from the instantaneous environmental parameters collected by the real-time monitoring system, while the static wind pressure influence value is read from the pre-generated static wind pressure influence matrix. The absolute deviation value is calculated using an absolute value function, i.e., the dynamic wind pressure influence value minus the static wind pressure influence value is taken as the absolute value. The calculation process is executed in real time in the data processing module, and the result is temporarily stored in floating-point form. In specific implementation, when the absolute deviation value exceeds the preset safety threshold, the correction radius is determined based on the real-time wind speed data of the current hoisting operation unit. The preset safety threshold is set based on engineering safety standards and historical hoisting data and is stored in a configuration file. The real-time wind speed data is continuously collected by wind speed sensors installed on the hoisting equipment or tower. The correction radius is calculated using empirical formulas or lookup tables, and its value is positively correlated with the real-time wind speed. For example, the higher the wind speed, the larger the correction radius is to cope with more unstable wind field conditions. In practical implementation, the dynamic and static wind pressure influence values ​​of surrounding hoisting operation units are obtained based on the correction radius. The correction radius defines a circular or spherical search area centered on the current hoisting operation unit. The identification of surrounding hoisting operation units is achieved through spatial querying, such as calculating the three-dimensional spatial distance between the current unit and surrounding units and filtering out all units with a distance smaller than the correction radius. The dynamic wind pressure influence values ​​of these surrounding units are obtained from their respective real-time monitoring data, and the static wind pressure influence values ​​are obtained by indexing from the static wind pressure influence matrix. The obtained data is organized into a list or array for subsequent processing. In practical implementation, the dynamic and static wind pressure influence values ​​of the surrounding hoisting operation units are input to the direction correction controller to generate hoisting direction adjustment commands for the hoisting equipment, thereby realizing dynamic correction of the hoisting direction. The direction correction controller is a software algorithm module or an embedded hardware unit. Its input includes two types of wind pressure influence value sequences from the surrounding units. The control algorithm, such as a proportional-integral-derivative control algorithm, calculates the required hoisting direction adjustment angle and amplitude by comparing the difference trend between the dynamic and static values. The adjustment command is sent to the actuator of the hoisting equipment, such as the steering motor or hydraulic system, through a communication interface.

[0047] Understandably, calculating the absolute deviation value is the core condition for triggering the dynamic correction logic. Setting the preset safety threshold requires consideration of the tower's structural strength, the stability of the hoisting equipment, and the fluctuation characteristics of the current wind conditions. Its value may be adjusted according to different hoisting stages; for example, a stricter threshold may be used at higher hoisting heights. Real-time wind speed data is typically filtered to eliminate instantaneous noise, and the formula for calculating the correction radius may involve the square relationship of wind speed to reflect the physical correlation between wind pressure and wind speed. The search efficiency of the surrounding hoisting operation units can be optimized using spatial index data structures such as quadtrees or octrees. The parameter tuning of the direction correction controller, such as proportional gain, integral time, and derivative time, needs to be adjusted according to the response characteristics of the hoisting system to ensure control stability and speed. Hoisting direction adjustment commands may include angular offsets in the horizontal plane or vector direction changes in three-dimensional space.

[0048] In practical implementation, referring to Table 1, the calculation frequency of the absolute deviation value should be consistent with the data update frequency of the real-time monitoring system, for example, several times per second. The preset safety threshold can be a fixed value or a variable that is dynamically adjusted according to environmental conditions. For example, when the temperature measurement value is low or the humidity measurement value is high, the threshold can be appropriately reduced to cope with the possible increase in structural brittleness. Real-time wind speed data is collected using a calibrated anemometer. The data is input into the processing unit after analog-to-digital conversion. The correction radius can be determined with reference to the following correlation relationship, which is based on wind engineering experience data.

[0049] Table 1: Correspondence between real-time wind speed and correction radius ; In practical implementation, when acquiring the wind pressure influence values ​​of surrounding hoisting operation units, it is necessary to ensure the spatiotemporal consistency of the data. That is, the dynamic wind pressure influence value should be the latest reading, and the static wind pressure influence value should match the version of the static wind pressure influence matrix corresponding to the current hoisting stage. The algorithm implementation of the direction correction controller may adopt an incremental control strategy, that is, calculating only a small adjustment relative to the current hoisting direction each time to avoid swaying of the hoisted object due to abrupt changes in direction. The generated hoisting direction adjustment command typically includes the absolute coordinates of the target direction or the deflection angle relative to the current direction, and the command format must be compatible with the control system protocol of the hoisting equipment. In some embodiments, the output of the direction correction controller may also need to pass through a smoothing filter to eliminate high-frequency jitter in the command, making the movement of the hoisting equipment smoother. Optionally, the calculation of the absolute deviation value can consider using the average value within a sliding window to smooth out false triggers caused by instantaneous fluctuations. Optionally, the determination of the correction radius can also introduce a wind direction stability factor; if the dominant wind direction remains stable in the near future, the correction radius can be appropriately reduced.

[0050] In practical implementation, the acquisition of dynamic wind pressure influence values ​​from surrounding hoisting operation units may face data transmission delays. This can be addressed by setting up a cache locally on the hoisting equipment and employing a timestamp matching mechanism to ensure sufficient freshness of the data used for control decisions. The indexing operation of static wind pressure influence values ​​relies on a mapping table between the unique identifier of the hoisting operation unit and the static wind pressure influence matrix index, which is loaded into memory during the operation initialization phase. The design of the direction correction controller needs to consider the system's hysteresis characteristics, i.e., a certain time is required from the issuance of the command to the actual response of the hoisting equipment; therefore, the control algorithm may include a lead correction component. The execution result of the hoisting direction adjustment command can be fed back through attitude sensors on the hoisting equipment, forming a closed-loop control. However, the description in Example 4 mainly focuses on the open-loop command generation process. It is understood that the real-time performance and accuracy of the dynamic correction process depend on the quality of sensor data, the processing power of the computing unit, and the robustness of the control algorithm.

[0051] In practice, when the absolute deviation value does not exceed the preset safety threshold, the hoisting equipment maintains its current hoisting direction and travel plan without initiating a dynamic correction process. The data processing module continuously monitors the absolute deviation value, and triggers subsequent steps immediately once the threshold is exceeded. The management interface for the preset safety threshold may be integrated into the wind farm's central control system, allowing operators to make fine-tuning adjustments based on site conditions. Real-time wind speed data may come from multiple sensors; in this case, the average or maximum value can be used as input for calculating the correction radius to reduce measurement uncertainty. The boundary of the search area defined by the correction radius may not be strictly circular, but rather appropriately adjusted based on the geometry of the hoisting area and the distribution of obstacles, for example, narrowing the search range near the boundary. The list of wind pressure influence values ​​of surrounding units may need to be normalized before being input into the direction correction controller to eliminate differences in dimensions or orders of magnitude between different units.

[0052] Example 5: In specific implementation, when collecting wind speed measurements from initial environmental parameters through a multi-source sensor network deployed in the wind turbine tower hoisting area, a Doppler lidar anemometer is used for wind speed acquisition. The Doppler lidar anemometer emits a laser beam of a specific wavelength and receives the echo signal reflected from aerosol particles in the air. The frequency shift of the echo signal is calculated using the Doppler effect to derive the wind speed vector. The wind speed acquisition process is performed at a high-frequency sampling rate, for example, tens of times per second. The raw wind speed data contains noise and outliers, which need to be filtered to obtain accurate wind speed measurements. Digital filters such as low-pass filters or Kalman filters are used for data filtering. The cutoff frequency of the low-pass filter is set according to the turbulence characteristics of the wind field to retain effective wind speed fluctuations while suppressing high-frequency noise. The filtered wind speed measurements are stored in a database for subsequent environmental modeling. In practical implementation, the installation location of the Doppler lidar anemometer is optimized to cover the entire hoisting area, typically deployed around the tower base or on a temporary wind measurement tower, ensuring an unobstructed laser beam scanning path. The data acquisition system integrates a time synchronization module to ensure consistent timestamps between wind speed measurements and data from other sensors. The data filtering algorithm runs in real-time on an embedded processor or cloud server, and filtering parameters can be dynamically adjusted according to on-site wind conditions to improve accuracy. In some embodiments, the Doppler lidar anemometer can be configured for multi-mode scanning, such as a planar position display mode or a distance / height indication mode, to obtain wind speed profile data at different heights. Optionally, data filtering can be combined with anomaly detection algorithms to automatically identify and remove interference signals caused by bird flocks or precipitation. It is understood that the high-precision measurement of the Doppler lidar anemometer provides a reliable input for wind load analysis.

[0053] In practical implementation, during the tower hoisting process, a real-time monitoring system collects instantaneous environmental parameters of the current hoisting operation unit. After calculating the dynamic wind pressure impact value, an inertial measurement unit (IMU) installed on the hoisting equipment collects the tower's attitude angle change data in real time. The IMU includes a three-axis gyroscope and a three-axis accelerometer. The gyroscope measures the tower's angular velocity and integrates it to obtain attitude angle changes such as pitch and roll angles. The accelerometer measures linear acceleration to compensate for gravity effects. The attitude angle change data is output at a high frequency, such as 100 Hz. The data acquisition is transmitted to the processing unit via a serial communication interface such as SPI or I2C. In practical implementation, the attitude angle change data and the dynamic wind pressure impact value are fused. Data fusion uses sensor fusion algorithms such as weighted averaging or Kalman filtering. The dynamic wind pressure impact value is calculated by the real-time monitoring system based on instantaneous wind speed and direction. The fusion process first performs coordinate transformation on the attitude angle change data to match the reference frame of the wind pressure impact value, and then calculates the correlation or deviation between the two. The fused data is used to enhance the estimation accuracy of the wind pressure impact. In practical implementation, a Kalman filter algorithm is used to calibrate the dynamic wind pressure impact value in real time. The Kalman filter algorithm includes a prediction step and an update step. The prediction step estimates the current value based on the state equation of the wind pressure impact, and the update step uses attitude angle change data as observations to correct the predicted value. Algorithm parameters, such as process noise covariance and observation noise covariance, are calibrated experimentally. The calibrated dynamic wind pressure impact value is output to the hoisting control system for decision-making. In some embodiments, the inertial measurement unit is preferably installed near the center of gravity of the tower to reduce measurement errors. Optionally, the attitude angle change data can be preprocessed, such as denoising and drift compensation, before data fusion.

[0054] In practical implementation, the wind speed acquisition process of the Doppler lidar anemometer includes equipment calibration, data acquisition, and signal processing. The calibration process uses standard wind tunnel equipment to calibrate the Doppler lidar anemometer to ensure measurement accuracy. During data acquisition, the laser beam scans the suspended area at a specific angle, and the echo signal is received by a photodetector and converted into an electrical signal. The signal processing unit calculates the wind speed by analyzing the frequency shift using a fast Fourier transform. Data filtering uses a moving average filter or an infinite impulse response filter, with the filter order selected based on real-time performance requirements. In practical implementation, the data acquisition chain of the inertial measurement unit involves sensor initialization, data reading, and time synchronization. Initialization configures the range and resolution of the gyroscope and accelerometer. Data reading obtains raw measurement values ​​through interrupts or polling. Time synchronization uses a Global Positioning System clock or a network time protocol. Attitude angle change data is represented using quaternions or Euler angles. The data fusion algorithm is implemented on a microcontroller to reduce latency. Optionally, the implementation of the Kalman filter algorithm can be optimized to a simplified version to reduce computational burden. Optionally, data from the Doppler lidar anemometer can be backed up to local storage in case of network interruption.

[0055] In practical implementation, when the real-time monitoring system collects instantaneous environmental parameters, wind speed data from the Doppler lidar anemometer is collected in parallel with data from temperature and humidity sensors. Timestamp alignment ensures data consistency. The calculation of dynamic wind pressure impact value uses a wind pressure formula combined with wind speed and air density. Attitude angle change data from the inertial measurement unit is streamed to the fusion module in real time. The state vector of the Kalman filter algorithm includes the wind pressure impact value and its rate of change, while the observation vector is the attitude angle change data. The filter gain is dynamically adjusted to adapt to changes in wind conditions. In practical implementation, the calibrated dynamic wind pressure impact value is sent to the hoisting equipment controller via a digital communication bus. The controller adjusts the hoisting strategy according to the calibration value, and the entire process is executed cyclically to achieve real-time response.

[0056] See Figure 5 The diagram contains two subgraphs, which will be analyzed from a professional perspective. Subgraph a shows the convergence process of the particle swarm optimization algorithm: it illustrates the change in the fitness value of the particle swarm optimization algorithm with the number of iterations in optimizing the movement sequence of wind turbine tower hoisting equipment. The fitness value rises rapidly in the early stages, reflecting the algorithm's ability to quickly converge to a better region in the search space; later, it fluctuates within a higher fitness range, indicating that the algorithm performs a fine search within this range to find a better movement sequence of hoisting equipment. This reflects the convergence characteristics of the particle swarm optimization algorithm in handling tower hoisting path optimization problems, providing algorithmic performance support for determining the optimal movement sequence. Subgraph b compares the hoisting path before and after optimization: using the operation unit ID as the x-axis, it compares the situation of each path segment before and after wind turbine tower hoisting path optimization. The different colored bar charts visually present the differences between the optimized and unoptimized paths in each segment, demonstrating the effectiveness of the particle swarm optimization algorithm combined with the static wind pressure influence matrix in adjusting the movement sequence of hoisting equipment. This can be used to evaluate the actual role of path optimization in wind turbine tower hoisting operations and provide data reference for path planning in subsequent hoisting operations.

[0057] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0058] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for hoisting and constructing the tower of a wind turbine generator set, characterized in that, Includes the following steps: Initial environmental parameters are collected by a multi-source sensor network deployed in the wind turbine tower hoisting area. These parameters include wind speed, temperature, and humidity measurements. Based on these initial environmental parameters, a spatial environment model of the tower hoisting area is generated using 3D modeling technology. Based on the spatial environment model, the hoisting area is divided into multiple hoisting operation units. The tower structure status acquisition operation is performed on each hoisting operation unit to obtain the tower displacement characteristics and foundation resonance characteristics corresponding to each hoisting operation unit. Wind load response simulation was performed on the tower displacement characteristics and foundation resonance characteristics corresponding to each hoisting operation unit using wind speed measurements, generating a static wind pressure influence matrix. Obtain the planned movement trajectory of the hoisting equipment, and combine each static wind pressure influence value in the static wind pressure influence matrix with its associated hoisting operation unit to adjust the movement sequence of the hoisting equipment when hoisting the tower on the planned movement trajectory; During the tower hoisting process, the instantaneous environmental parameters of the current hoisting operation unit are collected through a real-time monitoring system, the dynamic wind pressure impact value is calculated, and the hoisting direction of the hoisting equipment is dynamically corrected based on the dynamic and static wind pressure impact values ​​of the current hoisting operation unit.

2. The method for hoisting and constructing the tower of a wind turbine unit according to claim 1, characterized in that, The specific steps for dividing the hoisting area into multiple hoisting operation units based on the spatial environment model, performing tower structure state acquisition operations on each hoisting operation unit, and obtaining the tower displacement characteristics and foundation resonance characteristics corresponding to each hoisting operation unit include: The spatial environment model is processed into a three-dimensional mesh, and the hoisting area is divided into multiple hoisting operation units of equal size based on the elevation gradient change rate. The wind farm monitoring system acquires real-time prevailing wind direction data, determines the position of the initial hoisting operation unit based on the prevailing wind direction data, uses a laser scanner to collect tower displacement characteristics of the initial hoisting operation unit, and collects tower displacement characteristics of subsequent hoisting operation units in sequence according to wind direction. Spatial airflow distribution data is acquired through an airflow sensor array. An airflow motion pattern vector is generated based on the spatial airflow distribution data. The initial hoisting operation unit is reselected according to the airflow motion pattern vector. Ground resonance characteristics of the initial hoisting operation unit are collected using vibration sensors. Ground resonance characteristics of subsequent hoisting operation units are collected sequentially along the direction of the airflow motion pattern vector.

3. The method for hoisting and constructing the tower of a wind turbine unit according to claim 1, characterized in that, Before dividing the hoisting area into multiple hoisting operation units based on the spatial environment model to collect tower structure status, the process also includes: acquiring the prevailing wind direction data and airflow movement pattern vector at the current time point in real time through the data interface of the wind farm central control system, and storing the prevailing wind direction data and airflow movement pattern vector in a temporary buffer for use by the feature acquisition module.

4. The method for hoisting and constructing the tower of a wind turbine unit according to claim 1, characterized in that, The specific steps for simulating the wind load response of the tower displacement characteristics and foundation resonance characteristics corresponding to each hoisting operation unit using wind speed measurements, and generating a static wind pressure influence matrix, include: The average wind speed and turbulence intensity are calculated based on the wind speed measurements to determine the calculation benchmark for the reference wind pressure depth. The tower displacement characteristics and foundation resonance characteristics corresponding to the first hoisting operation unit are extracted. Combined with the three-dimensional coordinates and reference wind pressure depth of the first hoisting operation unit, adjacent hoisting operation units are selected as comparison areas. The wind load response is simulated using the finite element analysis method to obtain the static wind pressure influence value corresponding to the first hoisting operation unit. The tower displacement characteristics and foundation resonance characteristics corresponding to the remaining hoisting operation units are extracted sequentially. The same finite element analysis method is used to calculate the static wind pressure influence value of each hoisting operation unit. All static wind pressure influence values ​​are then arranged in a matrix according to the mesh structure of the space environment model to generate a static wind pressure influence matrix.

5. The method for hoisting and constructing the tower of a wind turbine unit according to claim 1, characterized in that, The specific steps for obtaining the planned movement trajectory of the hoisting equipment, combining each static wind pressure influence value in the static wind pressure influence matrix and its associated hoisting operation unit, and adjusting the movement sequence of the hoisting equipment when hoisting the tower on the planned movement trajectory include: Read each static wind pressure influence value and its associated hoisting operation unit identification information from the static wind pressure influence matrix; Obtain all preset path segments of the hoisting equipment and identify the adjacent hoisting operation units connected to each preset path segment; The initial working position of the hoisting equipment is determined. The static wind pressure impact values ​​of all adjacent hoisting operation units corresponding to each preset path segment are weighted and summed to obtain the comprehensive risk coefficient of the preset path segment. The particle swarm optimization algorithm is used to optimize the movement sequence of the hoisting equipment from the initial working position based on the comprehensive risk coefficient.

6. The method for hoisting and constructing the tower of a wind turbine unit according to claim 1, characterized in that, The specific steps for dynamically correcting the hoisting direction of the hoisting equipment based on the dynamic and static wind pressure influence values ​​of the current hoisting operation unit include: Calculate the absolute deviation between the dynamic wind pressure influence value and the static wind pressure influence value of the current hoisting operation unit; When the absolute deviation value exceeds the preset safety threshold, the correction radius is determined based on the real-time wind speed data of the current hoisting operation unit, and the dynamic wind pressure influence value and static wind pressure influence value of the surrounding hoisting operation units are obtained based on the correction radius. The dynamic and static wind pressure influence values ​​of the surrounding hoisting operation units are input to the direction correction controller to generate hoisting direction adjustment commands for the hoisting equipment, thereby achieving dynamic correction of the hoisting direction.

7. The method for hoisting and constructing the tower of a wind turbine unit according to claim 1, characterized in that, When collecting wind speed measurements from initial environmental parameters through a multi-source sensor network deployed in the wind turbine tower hoisting area, a Doppler lidar anemometer is used for wind speed acquisition, and accurate wind speed measurements are obtained through data filtering.

8. The method for hoisting and constructing the tower of a wind turbine unit according to claim 1, characterized in that, The process of collecting instantaneous environmental parameters of the current hoisting operation unit through a real-time monitoring system and calculating the dynamic wind pressure impact value during tower hoisting also includes: The inertial measurement unit installed on the hoisting equipment collects the tower's attitude angle change data in real time, fuses the attitude angle change data with the dynamic wind pressure influence value, and uses the Kalman filter algorithm to calibrate the dynamic wind pressure influence value in real time.

9. The method for hoisting and constructing the tower of a wind turbine unit according to claim 1, characterized in that, After adjusting the movement sequence of the hoisting equipment during tower hoisting on the planned trajectory, the following is also included: Collect wind pressure impact data and environmental parameters from historical hoisting operations, construct a wind pressure impact prediction database, train a wind pressure impact prediction model using time series analysis, and iteratively optimize the generation process of the static wind pressure impact matrix using the wind pressure impact prediction model.

10. A wind turbine tower hoisting and construction operation system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the wind turbine hoisting construction tower operation method according to any one of claims 1 to 9.

Images