Real-time health monitoring method and system for mixed tower conversion section based on optical fiber sensing

Abstract

The invention relates to a mixed tower conversion section real-time health monitoring method based on optical fiber sensing, and the method comprises the steps: obtaining multi-source sensing data of a mixed tower conversion section, carrying out the feature extraction and compensation operation based on the multi-source sensing data through a pre-constructed temperature-slippage two-stage decoupling model, and outputting pure mechanical strain data after the temperature interference is eliminated; on the basis of the virtual work principle, full-domain slip distribution inversion is carried out along a sensor layout path by utilizing pure mechanical strain data, and the circumferential slip physical quantity of the steel-concrete conversion section contact surface is obtained; and a health assessment result is generated based on the circumferential slip physical quantity, and when the health assessment result meets a preset triggering condition, a hierarchical control instruction is sent to a wind turbine generator master control system. The method has the characteristics of high accuracy, high reliability and high real-time performance in evaluation of the health state of the steel-concrete joint part, and can prevent serious accidents such as breakage of a steel-concrete section bolt or a steel strand and the like.

Description

Technical Field

[0001] This application relates to the field of wind turbine health monitoring, and in particular to a method, system, computer equipment, and computer-readable storage medium for real-time health monitoring of the hybrid tower conversion section based on fiber optic sensing. Background Technology

[0002] This application relates to the field of wind power technology, and in particular to a method and system for real-time health monitoring of hybrid tower conversion sections based on fiber optic sensing.

[0003] Wind turbine hybrid tower structures face the challenge of health monitoring at the steel-concrete joint section during operation. Existing technologies mainly focus on monitoring wind turbine hybrid tower structures, often employing distributed fiber optic sensors, such as fiber Bragg gratings (FBGs), deployed at the steel-concrete joint section. These sensors achieve static health assessment of the structure through strain and temperature parameters.

[0004] However, existing technologies have the following problems: First, existing monitoring schemes use discrete point-type FBG sensors, for example, deployed at 12-meter intervals. This deployment method can only acquire strain data at local points and cannot cover the slip gradient changes over the entire steel-concrete interface, resulting in missed detections of circumferential shear displacement distribution.

[0005] Second, the existing system does not integrate a temperature compensation algorithm. Expansion or contraction deformation of concrete caused by changes in ambient temperature, such as deformation of ±0.1 mm per degree Celsius, is misjudged as mechanical slippage. This misjudgment leads to insufficient reliability of the monitoring data.

[0006] Existing technologies are insufficient for real-time monitoring of the circumferential slip physical quantity of the steel-concrete conversion section, and are even less able to connect the corresponding indicators to the wind turbine data monitoring system to achieve control functions such as early warning and shutdown. Summary of the Invention

[0007] This application provides a method, system, computer device, and computer-readable storage medium for real-time health monitoring of hybrid tower switching sections based on fiber optic sensing, in order to at least solve the problem of insufficient reliability of monitoring data in related technologies.

[0008] In a first aspect, embodiments of this application provide a method for real-time health monitoring of a hybrid tower switching section based on fiber optic sensing, the method comprising: Acquire multi-source sensor data for the hybrid tower conversion section, wherein the multi-source sensor data includes interface strain data acquired through a preset spiral circumferential sensor array, temperature reference data acquired through a temperature reference sensor group, and unit operating parameters acquired from the wind turbine main control system. By using a pre-constructed temperature-slip dual-level decoupling model, feature extraction and compensation calculations are performed based on the multi-source sensing data to output pure mechanical strain data after removing temperature interference. Based on the principle of virtual work, the circumferential slip physical quantity of the contact surface of the steel-concrete transition section is obtained by inverting the global slip distribution along the sensor deployment path using the pure mechanical strain data. A health assessment result is generated based on the circumferential slip physical quantity, and a hierarchical control command is sent to the wind turbine main control system when the health assessment result meets the preset triggering conditions.

[0009] In some embodiments, acquiring multi-source sensor data of the mixing tower switching section includes: The interface strain data is collected using a hybrid array of micro-strain optical fibers and fiber grating sensors arranged in the interface sensing channel of the steel-concrete joint section. The hybrid array is arranged in a spiral circumferential manner with a helical angle of 30°, and a circumferential sensing unit containing axial, circumferential and shear direction optical fibers is set at a preset distance in the circumferential direction. The temperature reference data is collected using a group of temperature sensors arranged in the non-stressed areas of the concrete. While collecting the interface strain data and the temperature reference data, the unit operating parameters are collected through the data communication interface, including the main shaft speed, generator speed and impeller load data.

[0010] In some embodiments, the method further includes, prior to computation via a pre-built temperature-slip two-stage decoupling model: The interface strain data and the temperature reference data are subjected to wavelet threshold denoising processing, wherein the db4 wavelet basis function is used for multi-level decomposition, retaining the effective signal of the preset frequency band and filtering out high-frequency electromagnetic and wind vibration interference.

[0011] In some embodiments, a pre-built temperature-slip two-stage decoupling model is used to perform feature extraction and compensation calculations based on the multi-source sensing data, outputting pure mechanical strain data after removing temperature interference, including: Based on the temperature reference data, the pure temperature strain components are calculated using a preset concrete thermal expansion model. The interface strain data, the temperature reference data, and the unit operating parameters are constructed into a multi-dimensional feature vector, which is then input into the trained LSTM network model. Using the gating unit mechanism of the LSTM network model, non-mechanical slip deformation under the coupling effect of temperature and load is predicted and eliminated, and the pure mechanical strain data is output; wherein, the LSTM network model is obtained by supervised training using experimentally calibrated pure mechanical strain as label data.

[0012] In some embodiments, the LSTM network model adopts a dual-branch input structure, which includes a thermal environment branch for processing temperature reference data and a mechanical condition branch for processing unit operating parameters. The feature branch weighting module calculates the weights of the working condition perception weight vector based on the processing results of the thermal environment branch and the mechanical working condition branch. The condition-aware weight vector is introduced into the forget gate and input gate of the LSTM network model. The forget gate and input gate dynamically adjust the contribution of the thermal environment branch features to the model state and output according to the condition-aware weight vector, so as to adaptively predict the non-mechanical slip deformation in the interface strain data.

[0013] In some embodiments, based on the principle of virtual work, the global slip distribution inversion using the purely mechanical strain data along the sensor deployment path includes: Based on the aforementioned purely mechanical strain data, an integral relationship model is established between the fiber optic deployment angle, the sensor unit length, and the slippage:

[0014] in, This is the circumferential slip. For the purely mechanical strain data, For the angle of fiber optic cable deployment, The length of the sensing unit is given; the integral relationship model is solved using a numerical integration algorithm to obtain the circumferential slip physical quantity of the steel-concrete transition section contact surface.

[0015] In some embodiments, generating health assessment results based on the circumferential slip physical quantity includes: Determine the maximum slip amplitude, slip rate, and slip gradient of the circumferential slip physical quantities; The maximum slip amplitude, slip rate, and slip gradient are compared with preset warning thresholds and shutdown thresholds to generate graded evaluation results. The graded assessment results include normal status, early warning status, and dangerous shutdown status.

[0016] In some embodiments, when the health assessment result meets a preset trigger condition, sending a tiered control command to the wind turbine main control system includes: When any single indicator in the health assessment results reaches the warning threshold, a warning instruction is generated and sent to the remote monitoring center; A shutdown command is generated when any single indicator in the health assessment results reaches the shutdown threshold, or when at least two indicators simultaneously reach the warning threshold. The shutdown command is sent to the wind turbine main control PLC in real time via the Modbus-TCP protocol, triggering the unit to perform an emergency shutdown action until all indicators remain below the warning threshold for a preset time, after which the shutdown is lifted.

[0017] Secondly, embodiments of this application provide a real-time health monitoring system for a hybrid tower switching section based on fiber optic sensing, the system comprising: The acquisition module is used to acquire multi-source sensor data of the hybrid tower conversion section. The multi-source sensor data includes interface strain data acquired through a preset spiral circumferential sensor array, temperature reference data acquired through a temperature reference sensor group, and unit operating parameters acquired from the wind turbine main control system. The computation module is used to perform feature extraction and compensation calculations based on the multi-source sensor data through a pre-constructed temperature-slip dual-level decoupling model, and output pure mechanical strain data after removing temperature interference. Furthermore, based on the principle of virtual work, the circumferential slip physical quantity of the contact surface of the steel-concrete transition section is obtained by using the pure mechanical strain data to perform global slip distribution inversion along the sensor deployment path. The control module generates a health assessment result based on the circumferential slip physical quantity, and sends a hierarchical control command to the wind turbine main control system when the health assessment result meets the preset trigger conditions.

[0018] Thirdly, embodiments of this application provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method described in the first aspect above.

[0019] Fourthly, embodiments of this application provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described in the first aspect above.

[0020] Compared to related technologies, the real-time health monitoring method for the hybrid tower transition section based on fiber optic sensing provided in this application can monitor the circumferential slip physical quantity of the steel-concrete transition section in real time, while effectively eliminating interference from environmental parameters such as temperature. This method acquires multi-source sensor data and utilizes a temperature-slip dual-level decoupling model to accurately separate the purely mechanical strain components. Furthermore, it obtains the global slip distribution of the steel-concrete transition section contact surface based on the principle of virtual work. It generates graded health assessment results through the circumferential slip physical quantity and can send graded control commands, such as early warnings or emergency shutdown commands, to the wind turbine main control system. This method features high accuracy, high reliability, and high real-time performance in assessing the health status of the steel-concrete joint, preventing serious accidents such as bolt or strand breakage in the steel-concrete section, and directly guiding hybrid tower operation and maintenance decisions, thus extending the structural service life. Attached Figure Description

[0021] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart illustrating a real-time health monitoring method for a hybrid tower switching section based on fiber optic sensing, according to an embodiment of this application. Figure 2 This is a structural block diagram of a real-time health monitoring system for a hybrid tower switching section based on fiber optic sensing, according to an embodiment of this application.

[0022] Figure 3 This is a schematic diagram of the internal structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this application clearer, the application is described and illustrated below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. All other embodiments obtained by those skilled in the art based on the embodiments provided in this application without inventive effort are within the scope of protection of this application.

[0024] Obviously, the accompanying drawings described below are merely some examples or embodiments of this application. Those skilled in the art can apply this application to other similar scenarios based on these drawings without any inventive effort. Furthermore, it is understood that although the efforts made in this development process may be complex and lengthy, for those skilled in the art related to the content disclosed in this application, any changes to design, manufacturing, or production based on the technical content disclosed in this application are merely conventional technical means and should not be construed as insufficient disclosure of the content of this application.

[0025] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment that is mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0026] Unless otherwise defined, the technical or scientific terms used in this application shall have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms “a,” “an,” “an,” “the,” and similar words used in this application do not indicate quantity limitation and may indicate singular or plural. The terms “comprising,” “including,” “having,” and any variations thereof used in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units not listed, or may include other steps or units inherent to these processes, methods, products, or devices. The terms “connected,” “linked,” “coupled,” and similar words used in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. “Multiple” used in this application refers to two or more. “And / or” describes the relationship between related objects, indicating that three relationships may exist; for example, “A and / or B” can represent: A alone, A and B simultaneously, and B alone. The character " / " generally indicates that the preceding and following objects are in an "or" relationship. The terms "first," "second," and "third" used in this application are merely to distinguish similar objects and do not represent a specific ordering of the objects.

[0027] This application provides a real-time health monitoring method for hybrid tower switching sections based on fiber optic sensing. This method is implemented through four main functional steps: data acquisition, data decoupling operation, slip inversion, and health assessment.

[0028] Figure 1 This is a flowchart illustrating a real-time health monitoring method for hybrid tower switching sections based on fiber optic sensing, according to an embodiment of this application. Figure 1 As shown, the method includes the following steps: S101: Acquire multi-source sensor data of the mixing tower conversion section.

[0029] The hybrid tower transition section is the middle section of the generator tower, used to connect the upper part (steel tower) and the lower part (hybrid tower, i.e., concrete tower). Its function is to achieve effective connection and stress transfer between the tower sections made of different materials (steel and concrete). Due to the significant differences in the thermal expansion coefficients, elastic moduli, and other physical properties of steel and concrete, this part is the most complex area in the entire tower structure, most prone to stress concentration and interface slippage, and therefore a key area for health monitoring.

[0030] The multi-source sensor data includes: interface strain data acquired through a preset spiral circumferential sensor array; temperature reference data acquired through a temperature reference sensor group; and unit operating parameters acquired from the wind turbine main control system.

[0031] Specifically, the interface strain data is acquired using a hybrid array of micro-strain optical fibers and fiber optic grating sensors arranged within the interface sensing channel of the steel-concrete joint section. The radial depth of the sensing channel is set to 5-8 mm to ensure that the sensor is in direct contact with the interface slip generation layer.

[0032] The hybrid array is arranged in a helical circumferential configuration with a helical angle of 30°. This configuration allows the sensors to acquire shear strain information along the helical path on the steel-concrete contact surface, effectively covering the global slip gradient changes that traditional point sensors cannot capture. A circumferential sensing unit is set at preset intervals in the circumferential direction, and this unit contains optical fibers in the axial, circumferential, and shear directions.

[0033] Furthermore, the preset distance of the circumferential sensing unit can be determined based on the circumferential structural characteristics and expected defect size of the hybrid tower (e.g., 50 mm) to ensure sufficient sampling of potential slippage. The micro-strain fiber is preferably a Φ1550μm large-core fiber (accuracy ±1με), and the sensor encapsulation uses an epoxy-based nanocomposite material with a thermal conductivity ≤0.2W / (m·K) to provide fatigue resistance (≥10^7 cycles) and corrosion resistance.

[0034] In addition, temperature reference data is acquired using a group of temperature sensors deployed in the non-stressed areas of the concrete. Simultaneously with acquiring interface strain data and temperature reference data, unit operating parameters are collected via a data communication interface, including spindle speed, generator speed, and impeller load data.

[0035] In step S101, an innovative spiral array and three-dimensional sensing unit are used, along with a high-density spiral layout and a dedicated packaging process, to construct a blind-zone-free three-dimensional strain monitoring network. This solves the problem of missed detection of slip gradient caused by traditional discrete layout. At the same time, the low thermal conductivity packaging significantly reduces temperature conduction interference, ensuring the full range and high fidelity of the original data from the source.

[0036] S102: Preprocess the multi-source sensor data.

[0037] Before performing calculations using a pre-constructed temperature-slip two-stage decoupled model, wavelet threshold denoising is applied to the interface strain data and temperature reference data. Optionally, the original data acquisition frequency for the interface sensing group and temperature reference group is set to 100Hz; and multi-level decomposition (e.g., 5 levels) is performed using the db4 wavelet basis function. Valid signals within a preset frequency band (0.1-10Hz) are retained, while high-frequency electromagnetic and wind-induced vibration interference are filtered out.

[0038] Step S102 uses high-frequency sampling and a targeted wavelet threshold denoising strategy to accurately remove common electromagnetic radiation and high-frequency wind vibration noise in the wind farm environment, while retaining effective frequency band signals that reflect structural slip characteristics, significantly improving the signal-to-noise ratio and laying a high-quality data foundation for subsequent small strain feature extraction and decoupling operations.

[0039] S103: Outputs pure mechanical strain data after eliminating temperature interference through a temperature-slip dual-level decoupling model.

[0040] This two-stage decoupling model performs feature extraction and compensation operations, specifically including: First-level decoupling: Based on temperature reference data, a pre-defined concrete thermal expansion model is used to calculate the pure temperature strain components. This model establishes a mapping relationship between temperature change and thermally induced strain based on the properties of concrete materials and temperature variation patterns, for primary compensation. The linear expansion coefficient α in the model is approximately 1.2 × 10⁻⁶. -5 / ℃, the secondary temperature coefficient β is approximately 2.5×10 -8 / ℃².

[0041] Second-level decoupling: The interfacial strain data, temperature reference data, and unit operating parameters are used to construct a multidimensional feature vector.

[0042] The feature vector is input into the trained LSTM network model. This LSTM network model uses the interface strain data ε_total, temperature reference data T, and unit operating parameters to construct a multi-dimensional feature vector as input features, and is trained using experimentally calibrated pure mechanical strain as the label. The model learns the strain variation law under temperature-load coupling through gating units and outputs the compensated mechanical strain: ε_mech = ε_total - LSTM(ε_total, T, operating parameters).

[0043] Furthermore, the LSTM network model in this embodiment adopts a dual-branch input structure. This structure includes a thermal environment branch for processing temperature reference data and a mechanical condition branch for processing unit operating parameters.

[0044] The feature branch weighting module calculates the condition-aware weight vector based on the processing results of the thermal environment branch and the mechanical condition branch. The calculation of the condition-aware weight vector can employ an attention mechanism, dynamically assigning higher or lower weights to thermal environment features according to the state of the real-time input data, to adapt to different operating conditions where temperature changes or mechanical load changes dominate. Specifically, the weights of the load's influence on the total strain (W_{L}) and the temperature's influence on the total strain (W_{T}) are calculated.

[0045] The load condition-aware weight vector is introduced into the forget gate and input gate of the LSTM network model. When W_{L} is high (load-dominant), the weight of the mechanical load condition branch is increased and the temperature branch is suppressed; when W_{T} is high (temperature-dominant), the weight of the temperature branch is increased.

[0046] The forget gate and input gate dynamically adjust the contribution of thermal environment branch features to the model state and output based on the working condition-aware weight vector, so as to adaptively predict non-mechanical slip deformation in interface strain data.

[0047] Finally, the output of the LSTM network model is the pure mechanical strain data after predicting and removing non-mechanical slip deformation under the coupling effect of temperature and load.

[0048] It should also be noted that the LSTM network model uses experimentally calibrated pure mechanical strain as label data for supervised training. In a preferred embodiment, the training process includes: independently and coupledly applying temperature and mechanical loads to the concrete-slab transition section in a bench test; simultaneously, using high-precision extensometers and other physical quantity measurement equipment, calibrating the pure mechanical slip deformation of the steel-concrete interface under different working conditions; this calibration data serves as the ground truth label; and the LSTM network is trained based on this ground truth label to minimize the loss function, for example, using the mean squared error (MSE) loss function, and the network weights are adjusted through the backpropagation algorithm.

[0049] This step S103 introduces an LSTM model with a dual-branch feature weighted attention mechanism, enabling a leap from passive learning to active judgment of strain sources. It can adaptively adjust the weights of temperature and load under nonlinear coupling conditions, reducing temperature interference errors by more than 80% and solving the problem of tiny mechanical slip features being masked by temperature deformation in complex environments.

[0050] S104: Based on the principle of virtual work, the physical quantity of circumferential slip at the contact surface of the steel-concrete transition section is obtained by inversion.

[0051] This step utilizes purely mechanical strain data to perform a global slip distribution inversion along the sensor deployment path. Specifically, it includes: Based on purely mechanical strain data, an integral relationship model is established between the fiber optic deployment angle, the length of the sensing unit, and the slippage.

[0052] The integral relationship model is as follows:

[0053] in, This is the circumferential slip. For the purely mechanical strain data, For the angle of fiber optic cable deployment, The length of the sensing unit is given. Furthermore, a numerical integration algorithm (such as Simpson's rule) is used to solve the integral relationship model to obtain the circumferential slip physical quantity of the steel-concrete transition section contact surface. This physical quantity characterizes the global slip distribution.

[0054] This step S104 utilizes the principle of virtual work and the Simpson numerical integration algorithm to transform discrete micro-strain data into continuous circumferential slip physical quantities, realizing the transformation from point monitoring to field inversion, accurately reconstructing the slip distribution map of the steel-concrete interface, and ensuring that the slip peak and gradient changes are captured without omission.

[0055] S105: Generate health assessment results based on circumferential slip physical quantities, and send hierarchical control commands to the wind turbine main control system when the health assessment results meet the preset trigger conditions.

[0056] Specifically, the generation of health assessment results includes: Sensing data, SCADA data, and environmental data are fused using edge computing nodes, and the Kalman filter algorithm is used to achieve spatiotemporal alignment and redundancy verification of the data. The maximum slip amplitude, slip rate, and slip gradient of the circumferential slip physical quantity are determined.

[0057] Among them, the maximum slip amplitude can reflect the degree of local damage, the slip rate reflects the rate of damage deterioration, and the slip gradient reflects the non-uniformity of slip distribution.

[0058] Furthermore, the maximum slip amplitude, slip rate, and slip gradient are compared with preset warning thresholds and shutdown thresholds, respectively, to generate a graded evaluation result. As a preferred embodiment, the warning / shutdown thresholds for the first-level indicator, maximum circumferential slip, are set to 0.5 mm / 1.0 mm; the warning / shutdown thresholds for the second-level indicator, slip rate, are set to 0.05 mm / h / 0.1 mm / h; and the warning / shutdown thresholds for the third-level indicator, slip gradient, are set to 0.02 mm / m / 0.05 mm / m. In addition, the graded evaluation result includes a normal state, a warning state, and a dangerous shutdown state.

[0059] By constructing a multi-dimensional health indicator system that includes amplitude, rate, and gradient, and combining Kalman filtering to fuse multi-source data, real-time assessment is achieved through edge computing. This hierarchical early warning mechanism can sensitively identify early, minor slippage and distinguish between danger levels, providing a quantitative, intuitive, and reliable scientific basis for operation and maintenance decisions.

[0060] Sending hierarchical control commands to the wind turbine main control system includes: When any single indicator in the health assessment results reaches the warning threshold, a warning instruction is generated and sent to the remote monitoring center.

[0061] A shutdown command is generated when any single indicator in the health assessment results reaches the shutdown threshold, or when at least two indicators simultaneously reach the warning threshold.

[0062] The shutdown command is sent to the wind turbine's main control PLC in real time via the Modbus-TCP protocol. The Modbus-TCP protocol ensures real-time and reliable data transmission with the wind turbine's main control PLC. The wind turbine's main control PLC triggers the unit to execute emergency shutdown actions, which may include placing the rotor in the shutdown position to quickly eliminate external loads.

[0063] The unit will lift the lockdown after all indicators have remained below the warning threshold for a preset time (e.g., 30 minutes). The preset duration for lifting the lockdown is intended to confirm that environmental and load conditions have stabilized.

[0064] Step S105 establishes a closed-loop control link between the monitoring system and the unit's main control system, enabling millisecond-level command issuance via the Modbus-TCP protocol. Upon detecting dangerous slippage, it automatically triggers a shutdown, effectively preventing catastrophic accidents such as bolt breakage in the reinforced concrete section or steel strand failure, achieving a qualitative leap from passive monitoring to proactive safety defense.

[0065] Through the above steps, multi-source sensor data is acquired, and a temperature-slip dual-level decoupling model is used to accurately separate the purely mechanical strain components. Furthermore, the global slip distribution of the steel-concrete transition section contact surface is obtained based on the principle of virtual work. Graded health assessment results are generated using circumferential slip physical quantities, and graded control commands, such as early warnings or emergency shutdown commands, can be sent to the wind turbine main control system. This method features high accuracy, high reliability, and high real-time performance in assessing the health status of the steel-concrete joint, preventing serious accidents such as bolt or strand breakage in the steel-concrete section, and directly guiding tower operation and maintenance decisions, thus extending the structural service life.

[0066] This application also provides a real-time health monitoring system for hybrid tower switching sections based on fiber optic sensing. Figure 2 This is a structural block diagram of a real-time health monitoring system for a hybrid tower switching section based on fiber optic sensing, according to an embodiment of this application. Figure 2 As shown, the system includes: The acquisition module 20 is used to acquire multi-source sensor data of the hybrid tower conversion section. The multi-source sensor data includes interface strain data acquired through a preset spiral circumferential sensor array, temperature reference data acquired through a temperature reference sensor group, and unit operating parameters acquired from the wind turbine main control system. The computation module 21 is used to perform feature extraction and compensation calculations based on multi-source sensor data through a pre-constructed temperature-slip dual-level decoupling model, and output pure mechanical strain data after removing temperature interference; and, based on the principle of virtual work, to perform global slip distribution inversion using pure mechanical strain data along the sensor deployment path to obtain the circumferential slip physical quantity of the steel-concrete conversion section contact surface. The control module 22 generates a health assessment result based on the circumferential sliding physical quantity, and sends a hierarchical control command to the wind turbine main control system when the health assessment result meets the preset trigger conditions.

[0067] The above system enables the assessment of the health status of steel-concrete joints to be highly accurate, reliable, and real-time. It can prevent serious accidents such as bolt or strand breakage in steel-concrete sections and directly guide the operation and maintenance decisions of steel-concrete towers, thus extending the service life of the structure.

[0068] In one embodiment, Figure 3 This is a schematic diagram of the internal structure of an electronic device according to an embodiment of this application, such as... Figure 3 As shown, an electronic device is provided, which can be a server, and its internal structure diagram can be as follows. Figure 3 As shown, the electronic device includes a processor, a network interface, internal memory, and non-volatile memory connected via an internal bus. The non-volatile memory stores the operating system, computer programs, and a database. The processor provides computing and control capabilities, the network interface communicates with external terminals via a network connection, the internal memory provides the environment for the operating system to run, the computer programs are executed by the processor to implement a real-time health monitoring method for hybrid tower switching sections based on fiber optic sensing, and the database stores data.

[0069] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the electronic device to which the present application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements.

[0070] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by hardware related to computer program instructions. This computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0071] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0072] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for real-time health monitoring of a hybrid tower switching section based on fiber optic sensing, characterized in that, The method includes: Acquire multi-source sensor data for the hybrid tower conversion section, wherein the multi-source sensor data includes interface strain data acquired through a preset spiral circumferential sensor array, temperature reference data acquired through a temperature reference sensor group, and unit operating parameters acquired from the wind turbine main control system. By using a pre-constructed temperature-slip dual-level decoupling model, feature extraction and compensation calculations are performed based on the multi-source sensing data to output pure mechanical strain data after removing temperature interference. Based on the principle of virtual work, the circumferential slip physical quantity of the contact surface of the steel-concrete transition section is obtained by inverting the global slip distribution along the sensor deployment path using the pure mechanical strain data. A health assessment result is generated based on the circumferential slip physical quantity, and a hierarchical control command is sent to the wind turbine main control system when the health assessment result meets the preset triggering conditions.

2. The method according to claim 1, characterized in that, Acquiring multi-source sensor data for the mixing tower conversion section includes: The interface strain data is collected using a hybrid array of micro-strain optical fibers and fiber grating sensors arranged in the interface sensing channel of the steel-concrete joint section. The hybrid array is arranged in a spiral circumferential manner with a helical angle of 30°, and a circumferential sensing unit containing axial, circumferential and shear direction optical fibers is set at a preset distance in the circumferential direction. The temperature reference data is collected using a group of temperature sensors arranged in the non-stressed areas of the concrete. While collecting the interface strain data and the temperature reference data, the unit operating parameters are collected through the data communication interface, including the main shaft speed, generator speed and impeller load data.

3. The method according to claim 1, characterized in that, The method further includes, prior to computation using a pre-built temperature-slip two-stage decoupling model: The interface strain data and the temperature reference data are subjected to wavelet threshold denoising processing, wherein the db4 wavelet basis function is used for multi-level decomposition, retaining the effective signal of the preset frequency band and filtering out high-frequency electromagnetic and wind vibration interference.

4. The method according to any one of claims 1 to 3, characterized in that, Using a pre-constructed temperature-slip dual-stage decoupling model, feature extraction and compensation calculations are performed based on the multi-source sensing data to output pure mechanical strain data after removing temperature interference, including: Based on the temperature reference data, the pure temperature strain components are calculated using a preset concrete thermal expansion model. The interface strain data, the temperature reference data, and the unit operating parameters are constructed into a multi-dimensional feature vector, which is then input into the trained LSTM network model. Using the gating unit mechanism of the LSTM network model, non-mechanical slip deformation under the coupling effect of temperature and load is predicted and eliminated, and the pure mechanical strain data is output; wherein, the LSTM network model is obtained by supervised training using experimentally calibrated pure mechanical strain as label data.

5. The method according to claim 4, characterized in that, The LSTM network model adopts a dual-branch input structure, which includes a thermal environment branch for processing temperature reference data and a mechanical condition branch for processing unit operating parameters. The feature branch weighting module calculates the weights of the working condition perception weight vector based on the processing results of the thermal environment branch and the mechanical working condition branch. The condition-aware weight vector is introduced into the forget gate and input gate of the LSTM network model. The forget gate and input gate dynamically adjust the contribution of the thermal environment branch features to the model state and output according to the condition-aware weight vector, so as to adaptively predict the non-mechanical slip deformation in the interface strain data.

6. The method according to claim 1, characterized in that, Based on the principle of virtual work, the global slip distribution inversion is performed using the purely mechanical strain data along the sensor deployment path, including: Based on the aforementioned purely mechanical strain data, an integral relationship model is established between the fiber optic deployment angle, the sensor unit length, and the slippage: in, This is the circumferential slip. For the purely mechanical strain data, For the angle of fiber optic cable deployment, The length of the sensing unit is given; the integral relationship model is solved using a numerical integration algorithm to obtain the circumferential slip physical quantity of the steel-concrete transition section contact surface.

7. The real-time health monitoring method according to claim 1, characterized in that, The health assessment results generated based on the circumferential slip physical quantities include: Determine the maximum slip amplitude, slip rate, and slip gradient of the circumferential slip physical quantities; The maximum slip amplitude, slip rate, and slip gradient are compared with preset warning thresholds and shutdown thresholds to generate graded evaluation results. The graded assessment results include normal status, early warning status, and dangerous shutdown status.

8. The method according to claim 7, characterized in that, When the health assessment result meets the preset triggering conditions, the hierarchical control command sent to the wind turbine main control system includes: When any single indicator in the health assessment results reaches the warning threshold, a warning instruction is generated and sent to the remote monitoring center; A shutdown command is generated when any single indicator in the health assessment results reaches the shutdown threshold, or when at least two indicators simultaneously reach the warning threshold. The shutdown command is sent to the wind turbine main control PLC in real time via the Modbus-TCP protocol, triggering the unit to perform an emergency shutdown action until all indicators remain below the warning threshold for a preset time, after which the shutdown is lifted.

9. A real-time health monitoring system for a hybrid tower switching section based on fiber optic sensing, characterized in that, The system includes: The acquisition module is used to acquire multi-source sensor data of the hybrid tower conversion section. The multi-source sensor data includes interface strain data acquired through a preset spiral circumferential sensor array, temperature reference data acquired through a temperature reference sensor group, and unit operating parameters acquired from the wind turbine main control system. The computation module is used to perform feature extraction and compensation calculations based on the multi-source sensing data through a pre-constructed temperature-slip dual-level decoupling model, and output pure mechanical strain data after removing temperature interference. Furthermore, based on the principle of virtual work, the circumferential slip physical quantity of the contact surface of the steel-concrete transition section is obtained by using the pure mechanical strain data to perform global slip distribution inversion along the sensor deployment path. The control module generates a health assessment result based on the circumferential slip physical quantity, and sends a hierarchical control command to the wind turbine main control system when the health assessment result meets the preset trigger conditions.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1 to 8.

Images