Fan tower inclination online monitoring and early warning method and system based on AIoT

Abstract

The invention relates to an AIoT-based fan tower inclination on-line monitoring and early warning method and system, and belongs to the technical field of intelligent operation and maintenance of wind power equipment, and the system comprises an end side sensing layer which is used for collecting inclination, settlement, vibration and environment data in real time through multi-point arrangement of inclination angle, vibration and wind speed sensor groups; the edge calculation layer is used for executing space-time consistency processing, dynamic weight fusion and vibration data compression sensing on the multi-source heterogeneous data; the cloud analysis layer is used for establishing a mapping relation among wind speed, vibration and inclination amount based on an LSTM neural network, inputting a time serialization wind speed mean value, vibration spectrum energy and an inclination angle difference value, outputting an inclination trend prediction value, adaptively adjusting a dynamic threshold according to the environment wind speed intensity, and generating an early warning instruction when the threshold is exceeded; and the IoT transmission network is used for converging edge data through ZigBee-Mesh and uploading the edge data to the cloud through 4G / 5G. Through an end-edge-cloud cooperative intelligent architecture, core technical support is provided for safe and efficient operation of the wind power plant.

Description

Technical Field

[0001] This application relates to the field of intelligent operation and maintenance technology for wind power equipment, and more specifically, to a method and system for online monitoring and early warning of wind turbine tower tilt based on AIoT. Background Technology

[0002] As a key supporting structure of wind turbine generators, the stability of the wind turbine tower directly affects equipment safety and power generation efficiency. Traditional monitoring methods rely on periodic manual inspections, which suffer from drawbacks such as large measurement errors, poor continuity, and delayed response, making it impossible to capture real-time changes in tower tilt and settlement. Especially under harsh operating conditions (such as strong winds and vibrations), the tower is prone to collapse due to cumulative deformation, causing significant economic losses and even casualties. While existing technologies include online monitoring solutions, they generally suffer from difficulties in fusing multi-source heterogeneous data, low data transmission efficiency, and rigid warning thresholds, resulting in insufficient accuracy and timeliness of warnings. Therefore, there is an urgent need for a comprehensive solution that can achieve high-precision real-time monitoring, intelligent analysis, and dynamic early warning. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention proposes an online monitoring and early warning method and system for wind turbine tower tilt based on AIoT.

[0004] The technical solution of this invention is as follows:

[0005] On one hand, this invention proposes an online monitoring and early warning system for wind turbine tower tilt based on AIoT, characterized in that it includes:

[0006] End-side sensing layer: Configured with multiple types of sensor groups to monitor the status data of the wind turbine tower in real time. The sensor groups include: tilt sensor group, vibration sensor group and wind speed sensor group; the tilt sensor group is used to collect tilt and settlement data; the vibration sensor group is used to collect vibration signals; the wind speed sensor group is used to collect ambient wind speed data.

[0007] Edge computing layer: Embedded data processing module to perform spatiotemporal consistency processing and quality optimization on multi-source heterogeneous data acquired by the edge perception layer, and generate standardized data with time synchronization and error suppression;

[0008] Cloud-based analytics layer: Based on standardized data generated by the edge computing layer and a deep learning-based early warning model, it establishes a mapping relationship between wind speed, vibration, and tilt / settlement, generates early warning commands through dynamic threshold comparison, and displays tower status and early warning information in real time through a visualization platform;

[0009] IoT transmission network: Data is aggregated from the edge perception layer to the edge computing layer via the ZigBee protocol, and the processed data is transmitted to the cloud analysis layer via the mobile network.

[0010] Preferably, the tilt sensor array is arranged at the top, middle, halfway between the top and middle of the tower, and bottom; the settlement is calculated based on the offset using the tilt sensor located at the bottom.

[0011] Preferably, the multi-source heterogeneous data acquired by the end-side sensing layer is subjected to spatiotemporal consistency processing and quality optimization. Specifically, error suppression is performed on the multi-source heterogeneous data by weighted average fusion to eliminate sensor noise and environmental interference; and the spatiotemporal reference of the asynchronous sampling data of the wind speed sensor and the tilt sensor is unified by cubic spline interpolation.

[0012] Preferably, the edge computing layer also performs data compression sensing:

[0013] A sparse basis matrix is ​​constructed from the vibration data collected by the vibration sensor group, and the data is reconstructed by minimizing the L1 norm.

[0014] The wind speed and tilt angle data sampled asynchronously by the wind speed sensor group and the tilt angle sensor group are unified to a unified time reference.

[0015] Preferably, the early warning model of the cloud analysis layer is an LSTM neural network. The inputs are time-series average wind speed, vibration spectrum energy, and multi-location tilt angle difference values. The output is a tilt trend prediction value, and a dynamic threshold is set. The dynamic threshold is adaptively adjusted according to the ambient wind speed intensity.

[0016] Preferably, the IoT transmission network adopts a hierarchical topology:

[0017] The sensors in the edge-side sensing layer are connected to the routing node;

[0018] Routing nodes forward data to the gateway in the edge computing layer via the Mesh network;

[0019] The gateway of the edge computing layer connects to the server of the cloud analytics layer via 4G / 5G.

[0020] Preferably, the visualization platform pushes early warning information and maintenance work orders to the mobile terminals of management personnel in real time through an API interface.

[0021] On the other hand, the present invention also provides an AIoT-based online monitoring and early warning method for wind turbine tower tilt, comprising the following steps:

[0022] End-side data acquisition: Real-time acquisition of wind turbine tower status data through tilt sensor group, vibration sensor group and wind speed sensor group;

[0023] Perform the following on the collected data at the edge computing layer:

[0024] Time synchronization of asynchronously sampled tilt angle, vibration, and wind speed data is achieved through cubic spline interpolation.

[0025] The weighted average fusion adopts a dynamic weight allocation strategy, and the weight influence factors include location weight factor, time effect weight factor and environmental interference correction factor.

[0026] Pre-processed data is aggregated to the edge gateway via the ZigBee protocol and then transmitted to the cloud via the mobile network.

[0027] At the cloud analytics layer:

[0028] The time-series average wind speed, vibration spectrum energy, and multi-location tilt angle difference values ​​are input into the LSTM neural network model, and the tilt trend prediction value is output.

[0029] The dynamic threshold is adaptively adjusted based on the ambient wind speed intensity, and an early warning command is generated when the predicted value exceeds the threshold.

[0030] The tower status and early warning information are displayed in real time through a visualization platform and pushed to the mobile terminals of managers via API interface.

[0031] In another aspect, the present invention also provides an electronic device having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements an AIoT-based online monitoring and early warning method for wind turbine tower tilt as described in any embodiment of the present invention.

[0032] In another aspect, the present invention also provides a computer-readable medium for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement an AIoT-based online monitoring and early warning method for wind turbine tower tilt as described in any embodiment of the present invention.

[0033] The present invention has the following beneficial effects:

[0034] By deploying tilt sensors at multiple key locations such as the top, middle, and bottom of the tower, and combining the measurement data from the bottom sensor, millimeter-level high-precision monitoring of tilt settlement is achieved, completely solving the problems of large errors and discontinuity in traditional manual detection.

[0035] The edge computing layer uses cubic spline interpolation technology to unify the temporal reference of wind speed, vibration, and tilt angle data, eliminating asynchronous sampling errors. It also innovates a dynamic weighted fusion strategy, which comprehensively considers the importance of sensor location, data timeliness, and environmental interference, significantly suppressing noise and improving data reliability.

[0036] By implementing compressed sensing technology on massive vibration data, the amount of transmitted data can be compressed by 50% while ensuring high fidelity of the reconstructed data, which greatly reduces the network bandwidth requirements and cloud storage pressure.

[0037] Compared to a fixed threshold scheme, the wind speed adaptive dynamic threshold setting significantly reduces the false alarm rate and the missed alarm rate. Detailed Implementation

[0038] The technical solutions in the embodiments of the present invention will be clearly and completely described below. 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.

[0039] It should be understood that the step numbers used in the text are for ease of description only and are not intended to limit the order in which the steps are performed.

[0040] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0041] The terms “comprising” and “including” indicate the presence of the described feature, whole, step, operation, element and / or component, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or collections thereof.

[0042] The term “and / or” refers to any combination of one or more of the associated listed items, as well as all possible combinations, and includes these combinations.

[0043] Example 1:

[0044] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with specific embodiments of this application.

[0045] To address the problems of existing technologies, this invention provides an AIoT-based online monitoring and early warning system for wind turbine tower tilt, comprising:

[0046] End-side sensing layer: Configured with multiple types of sensor groups to monitor the status data of the wind turbine tower in real time. The sensor groups include: tilt sensor group, vibration sensor group and wind speed sensor group; the tilt sensor group is used to collect tilt and settlement data; the vibration sensor group is used to collect vibration signals; the wind speed sensor group is used to collect ambient wind speed data.

[0047] In a preferred embodiment of this invention, the tilt sensor group uses SCA100T tilt sensors, which are installed at the top, middle, halfway between the top and middle of the tower, and bottom, with four sensors installed at each location and adjacent sensors spaced 90° apart. The settlement is calculated based on the offset using the tilt sensor located at the bottom. The specific formula for calculating the settlement is as follows:

[0048]

[0049] In the formula: S is the settlement; Δx and Δy are the offsets of the tower bottom center along the x-axis and y-axis in the horizontal plane, respectively; θ is the change in tilt angle.

[0050] The vibration sensor group is installed on the frame at the rear of the generator in the top nacelle of the tower. It adopts an FBA force balance accelerometer and the vibration sensor collects the vibration spectrum energy in the 0.5–10Hz frequency band.

[0051] The wind speed sensor group adopts a three-cup wind speed sensor, which is installed on the windward side of the outside of the tower. The wind speed sensor collects the ambient wind speed every 0.5 seconds.

[0052] Edge computing layer: Embedded data processing module to perform spatiotemporal consistency processing and quality optimization on multi-source heterogeneous data acquired by the edge perception layer, and generate standardized data with time synchronization and error suppression;

[0053] In a preferred embodiment of this invention, the spatiotemporal consistency processing and quality optimization of the multi-source heterogeneous data acquired by the edge-side sensing layer are performed. Specifically, weighted average fusion is used to suppress errors in the multi-source heterogeneous data, eliminating sensor noise and environmental interference; cubic spline interpolation is used to unify the spatiotemporal reference of asynchronous sampling data from wind speed and tilt sensors.

[0054] Weighted average fusion uses data from multiple tilt sensors at the same time point, employing a dynamic weight allocation strategy. Weight calculation includes:

[0055] Location weighting factors: top 0.4, middle 0.3, halfway point 0.2, bottom 0.1 (because top deformation is more sensitive to collapse risk).

[0056] Timeliness weighting factor: w t =e -0.5Δt ; where Δt is the difference between the current time and the acquisition time, in seconds.

[0057] Environmental disturbance correction factor: When the wind speed is greater than 15 m / s and the energy of the dominant vibration frequency band is greater than 80% of the historical maximum energy, the weight is multiplied by a reliability coefficient of 0.7, and the final tilt angle fusion value is as follows:

[0058]

[0059] In the formula: θ f This is the final tilt angle blending value; w loc For location weighting factors; w t Timeliness weighting factor; w env θ is the environmental disturbance correction factor. i This is the original tilt angle value measured by the i-th tilt sensor.

[0060] In a preferred embodiment of this invention, the edge computing layer also performs data compression sensing:

[0061] A sparse basis matrix is ​​constructed from the vibration data collected by the vibration sensor array, and the data is reconstructed by minimizing the L1 norm; a data compression rate of 50% is achieved, reducing the amount of data transmitted; the specific steps are as follows:

[0062] Constructing a sparse basis matrix: Obtain raw vibration data from the vibration sensor group, select Discrete Cosine Transform (DCT) as the sparse representation method, and create a 1000×1000-dimensional DCT orthogonal basis matrix as the sparse representation space of the vibration signal;

[0063] Compressed observations: A random Gaussian observation matrix with dimensions of 500×1000 is generated. The original vibration data is projected into a low-dimensional space through matrix multiplication to obtain a 500-dimensional compressed observation vector, which reduces the amount of data by 50% compared to the original data.

[0064] L1 norm minimization reconstruction: Receive compressed observation vectors at the edge computing layer and establish an optimization problem: data reconstruction under L1 norm constraints; solve the optimization problem using an iterative threshold algorithm to recover the original vibration signal and verify the reconstruction error: ensure signal-to-noise ratio ≥ 40dB.

[0065] To unify the wind speed and tilt angle data asynchronously sampled by the wind speed sensor group and the tilt angle sensor group to a unified time reference, the specific steps are as follows:

[0066] Timestamp alignment: Add a high-precision timestamp to each sensor data packet to establish a unified time reference: use the edge gateway clock as a reference to calibrate the time delay deviation (GPS synchronization or NTP protocol);

[0067] Cubic spline interpolation: Construct time series from wind speed and tilt data; construct a cubic spline interpolation function, defining a piecewise cubic polynomial that satisfies continuity and smoothness conditions, specifically as follows:

[0068] S(t) = a i (tt i ) 3 +b i (tt i ) 2 +ci (tt i )+d i ;

[0069] In the formula: a i b is the coefficient of the cubic term, representing the rate of change of curvature of the control function within that interval; i c is the coefficient of the quadratic term, which controls the curvature of the function in that interval; i d is the coefficient of the linear term, controlling the slope of the function in that interval; i The constant term is equal to the function at node t. i The value of t; i Let be the timestamp of the i-th data point.

[0070] Data quality verification: Check the continuity of the interpolated data and verify that the boundary conditions and interpolation error values ​​are less than 0.1.

[0071] Cloud-based analytics layer: Based on standardized data generated by the edge computing layer and a deep learning-based early warning model, it establishes a mapping relationship between wind speed, vibration, and tilt / settlement, generates early warning commands through dynamic threshold comparison, and displays tower status and early warning information in real time through a visualization platform;

[0072] As a preferred embodiment of this example, the early warning model of the cloud analysis layer is an LSTM neural network. The inputs are time-series average wind speed, vibration spectrum energy, and multi-location tilt angle difference values. The output is a tilt trend prediction value, and a dynamic threshold is set. The dynamic threshold is adaptively adjusted according to the ambient wind speed intensity.

[0073] In this embodiment, an LSTM neural network is used. The inputs are the average wind speed (1 sampling point per second), vibration spectrum energy (0.5–10Hz frequency band integral value) and multi-location tilt angle difference value (top and bottom tilt angle difference) within a time window (10 minutes); the output is the tilt trend prediction value for the next 5 minutes.

[0074] In this embodiment, the base value of the dynamic threshold is 5°; when the wind speed is greater than 20m / s and less than 70m / s, the adaptive adjustment formula for the threshold is:

[0075] θ th =θ base ×(1-0.02×(v-20));

[0076] In the formula: θ th The threshold is dynamically adjusted; θ base The base threshold is denoted by v, where v is the wind speed. An early warning command is generated when the predicted tilt trend output by the LSTM exceeds the dynamically adjusted threshold.

[0077] As a preferred embodiment of this example, the visualization platform pushes early warning information and maintenance work orders to the mobile terminals of management personnel in real time through the API interface.

[0078] IoT transmission network: Data is aggregated from the edge perception layer to the edge computing layer via the ZigBee protocol, and the processed data is transmitted to the cloud analysis layer via the mobile network.

[0079] In a preferred embodiment of this invention, the IoT transmission network adopts a hierarchical topology:

[0080] The sensors in the edge-side sensing layer are connected to the routing node;

[0081] Routing nodes forward data to the gateway in the edge computing layer via the Mesh network;

[0082] The gateway of the edge computing layer connects to the server of the cloud analytics layer via 4G / 5G.

[0083] Example 2:

[0084] This embodiment provides an AIoT-based online monitoring and early warning method for wind turbine tower tilt, including the following steps:

[0085] End-side data acquisition: Real-time acquisition of wind turbine tower status data through tilt sensor group, vibration sensor group and wind speed sensor group;

[0086] Perform the following on the collected data at the edge computing layer:

[0087] Time synchronization of asynchronously sampled tilt angle, vibration, and wind speed data is achieved through cubic spline interpolation.

[0088] The weighted average fusion adopts a dynamic weight allocation strategy, and the weight influence factors include location weight factor, time effect weight factor and environmental interference correction factor.

[0089] Pre-processed data is aggregated to the edge gateway via the ZigBee protocol and then transmitted to the cloud via the mobile network.

[0090] At the cloud analytics layer:

[0091] The time-series average wind speed, vibration spectrum energy, and multi-location tilt angle difference values ​​are input into the LSTM neural network model, and the tilt trend prediction value is output.

[0092] The dynamic threshold is adaptively adjusted based on the ambient wind speed intensity, and an early warning command is generated when the predicted value exceeds the threshold.

[0093] The tower status and early warning information are displayed in real time through a visualization platform and pushed to the mobile terminals of managers via API interface.

[0094] Example 3:

[0095] This embodiment provides an electronic device that stores a computer program. When the computer program is executed by a processor, it implements an AIoT-based online monitoring and early warning method for wind turbine tower tilt as described in any embodiment of the present invention.

[0096] Example 4:

[0097] This embodiment provides a computer-readable medium for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement an AIoT-based online monitoring and early warning method for wind turbine tower tilt as described in any embodiment of the present invention.

[0098] In this application embodiment, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent the existence of A alone, A and B simultaneously, or B alone. A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of singular or plural items. For example, at least one of a, b, and c can represent: a, b, c, a and b, a and c, b and c, or a and b and c, where a, b, and c can be single or multiple.

[0099] Those skilled in the art will recognize that the units and algorithm steps described in the embodiments disclosed herein can be implemented using electronic hardware, computer software, or a combination of electronic hardware and software. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0100] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0101] In the several embodiments provided in this application, any function, if implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes: USB flash drive, mobile hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk, and other media capable of storing program code. The above descriptions are merely embodiments of the present invention and do not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made using the present invention specification, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. An AIoT-based online monitoring and early warning system for wind turbine tower tilt, characterized in that, include: End-side sensing layer: Configured with multiple types of sensor groups to monitor the status data of the wind turbine tower in real time. The sensor groups include: tilt sensor group, vibration sensor group and wind speed sensor group; the tilt sensor group is used to collect tilt and settlement data; the vibration sensor group is used to collect vibration signals; the wind speed sensor group is used to collect ambient wind speed data. Edge computing layer: Embedded data processing module to perform spatiotemporal consistency processing and quality optimization on multi-source heterogeneous data acquired by the edge perception layer, and generate standardized data with time synchronization and error suppression; Cloud-based analytics layer: Based on standardized data generated by the edge computing layer and a deep learning-based early warning model, it establishes a mapping relationship between wind speed, vibration, and tilt / settlement, generates early warning commands through dynamic threshold comparison, and displays tower status and early warning information in real time through a visualization platform; IoT transmission network: Data is aggregated from the edge perception layer to the edge computing layer via the ZigBee protocol, and the processed data is transmitted to the cloud analysis layer via the mobile network.

2. The wind turbine tower tilt online monitoring and early warning system based on AIoT according to claim 1, characterized in that: The tilt sensor array is deployed at the top, middle, halfway between the top and middle, and bottom of the tower; the settlement is calculated based on the offset using the tilt sensor located at the bottom.

3. The wind turbine tower tilt online monitoring and early warning system based on AIoT according to claim 1, characterized in that: The multi-source heterogeneous data acquired by the end-side sensing layer undergoes spatiotemporal consistency processing and quality optimization. Specifically, weighted average fusion is used to suppress errors in the multi-source heterogeneous data and eliminate sensor noise and environmental interference. Cubic spline interpolation is used to unify the spatiotemporal reference of asynchronous sampling data from wind speed and tilt sensors.

4. The wind turbine tower tilt online monitoring and early warning system based on AIoT according to claim 1, characterized in that: The edge computing layer also performs data compression awareness: A sparse basis matrix is ​​constructed from the vibration data collected by the vibration sensor group, and the data is reconstructed by minimizing the L1 norm. The wind speed and tilt angle data sampled asynchronously by the wind speed sensor group and the tilt angle sensor group are unified to a unified time reference.

5. The wind turbine tower tilt online monitoring and early warning system based on AIoT according to claim 1, characterized in that: The early warning model of the cloud analysis layer is an LSTM neural network. The inputs are time-series average wind speed, vibration spectrum energy, and multi-location tilt angle difference values. The output is a tilt trend prediction value, and a dynamic threshold is set. The dynamic threshold is adaptively adjusted according to the ambient wind speed intensity.

6. The wind turbine tower tilt online monitoring and early warning system based on AIoT according to claim 1, characterized in that: The IoT transmission network adopts a hierarchical topology: The sensors in the edge-side sensing layer are connected to the routing node; Routing nodes forward data to the gateway in the edge computing layer via the Mesh network; The gateway of the edge computing layer connects to the server of the cloud analytics layer via 4G / 5G.

7. The wind turbine tower tilt online monitoring and early warning system based on AIoT according to claim 1, characterized in that: The visualization platform pushes early warning information and maintenance work orders to the mobile terminals of management personnel in real time through API interfaces.

8. A method for online monitoring and early warning of wind turbine tower tilt based on AIoT, characterized in that, Includes the following steps: End-side data acquisition: Real-time acquisition of wind turbine tower status data through tilt sensor group, vibration sensor group and wind speed sensor group; Perform the following on the collected data at the edge computing layer: Time synchronization of asynchronously sampled tilt angle, vibration, and wind speed data is achieved through cubic spline interpolation. The weighted average fusion adopts a dynamic weight allocation strategy, and the weight influence factors include location weight factor, time effect weight factor and environmental interference correction factor. Pre-processed data is aggregated to the edge gateway via the ZigBee protocol and then transmitted to the cloud via the mobile network. At the cloud analytics layer: The time-series average wind speed, vibration spectrum energy, and multi-location tilt angle difference values ​​are input into the LSTM neural network model, and the tilt trend prediction value is output. The dynamic threshold is adaptively adjusted based on the ambient wind speed intensity, and an early warning command is generated when the predicted value exceeds the threshold. The tower status and early warning information are displayed in real time through a visualization platform and pushed to the mobile terminals of managers via API interface.

9. An electronic 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 program, it implements the AIoT-based online monitoring and early warning method for wind turbine tower tilt as described in claim 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the AIoT-based online monitoring and early warning method for wind turbine tower tilt as described in claim 8.

Images