Self-adaptive support control system for wind power blade transportation

Abstract

The invention relates to the technical field of wind power equipment transportation control, in particular to a self-adaptive support control system for wind power blade transportation, which comprises a sensor group and a server, and is characterized in that the sensor group collects and uploads blade attitude original data, support contact pressure original data, transportation speed data, road section type data, wind speed original data and rainfall state data; the server comprises a data preprocessing module, a historical data processing and training module, a data fusion module, a control decision module, a cross validation module and a self-adaptive correction module. Collected data are preprocessed and then stored and trained, a posture and supporting force influence coefficient matrix and a dynamic threshold library are generated, a state fusion value is output through multi-source data dynamic weight fusion, control parameters are corrected in combination with a three-dimensional mapping library and a simplified model prediction control algorithm, and the reliability is guaranteed through cross validation. And working conditions are divided according to multi-dimensional indexes to realize self-adaptive adjustment.

Description

Technical Field

[0001] This invention relates to the field of wind power equipment transportation control technology, specifically to an adaptive support control system for wind turbine blade transportation. Background Technology

[0002] As the core component of wind turbine generators, the size and weight of wind turbine blades have continued to increase with the development of wind power technology. Currently, the mainstream blade length has reached 30-60m, and the blade length of some high-power units even exceeds 70m. Moreover, the blades are made of fiber-reinforced composite materials, which are structurally fragile and have weak impact resistance. The attitude stability during transportation directly determines whether the blades will suffer irreversible damage such as cracks or breakage. Therefore, the precise control of the support system has become a core technical requirement for the transportation of wind turbine blades.

[0003] To ensure transportation safety, the technology has evolved from early fixed mechanical supports to support control systems with simple feedback functions. However, existing technologies still have the following shortcomings: First, insufficient scenario adaptability; fixed parameters cannot match diverse needs. The support force parameters, attitude deviation thresholds, and response frequencies of existing systems are mostly fixed settings, failing to fully consider differences in blade specifications and changes in transportation scenarios. This leads to the system being suitable for one scenario but potentially causing problems such as excessive support (generating additional stress) or insufficient support (excessive attitude deviation) in other scenarios. Second, weak control accuracy and robustness; a single data source is prone to decision-making errors. Third, lack of self-optimization capability; existing systems do not utilize a large amount of historical transportation data for model training and parameter optimization. Every time a new blade specification is transported or a transportation route is switched, technicians must debug the control parameters on-site, increasing labor and time costs and making it difficult to develop an optimal control strategy adapted to specific scenarios. In addition, a few attempts to use complex algorithms (such as full-dimensional model predictive control and deep learning), while aiming to improve accuracy, rely on high-performance computing hardware and cannot be deployed on conventional vehicle controllers.

[0004] In summary, existing wind turbine blade transportation support control systems have significant shortcomings in terms of scenario adaptability, control accuracy, and self-optimization capabilities, and cannot meet the safety requirements of long-distance and multi-scenario transportation of large blades. There is an urgent need for an adaptive support control technology that integrates multi-source data verification, historical data driving, and multi-dimensional operating condition adaptation to achieve precise matching of support status and operating conditions under different blade specifications and different transportation scenarios, thereby reducing the risk of blade transportation damage. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention aims to provide an adaptive support control system for wind turbine blade transportation, which can achieve precise matching between transportation conditions and support status, thereby improving support stability and control accuracy under dynamic conditions.

[0006] The basic solution provided by this invention is an adaptive support control system for wind turbine blade transportation, comprising a sensor group and a server; The sensor group includes an inertial measurement unit, a support pressure sensor, a vehicle speed sensor, a road segment identification sensor, and an environmental auxiliary sensor. It is used to collect raw data on wind turbine blade attitude, raw data on support contact pressure, transport vehicle speed, road segment type, raw wind speed, and rainfall status, and upload them to the server. The road segment type data includes highway segments, national highway segments, and rural road segments. The server includes a data preprocessing module, a historical data processing and training module, a data fusion module, a control decision module, a cross-validation module, and an adaptive correction module. The data preprocessing module is used to perform moving average filtering and drift calibration on the raw data of wind turbine blade attitude, support contact pressure, and wind speed. The size of the filtering window is adjusted according to the transport vehicle speed data and road segment type data. After processing, the wind turbine blade attitude data, support contact pressure data, and wind speed data are obtained. The processed data, the transport vehicle speed data collected by the sensors, the road segment type data, the rainfall status data, and the preset blade specification data are stored together in the historical data processing and training module. The blade specification data includes preset parameters adapted to wind turbine blades of different sizes. The historical data processing and training module is used to extract historical data of different blade specifications, road segment types and vehicle speed ranges from the stored data, construct a training set according to the three-dimensional classification of blade specifications, road segment types and vehicle speed ranges, and obtain the attitude and support force influence coefficient matrix and dynamic threshold library including attitude deviation threshold and pressure deviation threshold through gradient boosting algorithm training. During the training process, the optimal control parameter label and fault handling parameters are generated and stored simultaneously. The data fusion module receives wind turbine blade attitude data, support contact pressure data, and wind speed data output by the data preprocessing module, as well as transport vehicle speed data, road segment type data, and rainfall status data collected by sensors. It then fuses the above data using a dynamic weight allocation and Kalman filtering strategy. The dynamic weights are adjusted according to the transport vehicle speed data and road segment type data. After fusion, the output is a blade and support status fusion value that includes real-time attitude deviation, optimal support pressure range, and scenario adaptation coefficient. The control decision module has a pre-set 3D mapping library of attitude deviation support force road segment types, which is used to match the initial control parameters based on the fusion value of the blade and support state and the attitude and support force influence coefficient matrix. It predicts the state changes within a preset time period through a simplified model predictive control algorithm, and iterates the initial control parameters a preset number of times with transport vehicle speed data, real-time attitude deviation and scene adaptation coefficient as the core influencing factors, and outputs the support force adjustment value and response speed command. The cross-validation module is used to perform bidirectional cross-validation between the real-time attitude deviation and optimal support pressure range in the fused values ​​of blade and support status and the theoretical values ​​corresponding to the attitude and support force influence coefficient matrix. When the data deviation exceeds the preset ratio of the dynamic threshold, a data calibration command is triggered to the data preprocessing module. The adaptive correction module is used to classify four operating conditions based on road segment type data, transport vehicle speed change rate, real-time attitude deviation change rate, and rainfall status data: high-speed smooth, national highway bumpy, rural road complex, and rapid turn and acceleration. The high-speed smooth condition corresponds to high-speed road segments, the national highway bumpy condition corresponds to national highway segments, the rural road complex condition corresponds to rural road segments, and the rapid turn and acceleration condition corresponds to cross-road segment dynamic conditions. The module dynamically adjusts the update frequency of control parameters, the adjustment range of support force, and the dynamic threshold for different operating conditions, and outputs the final control signal to the support actuator.

[0007] The principle of this invention is as follows: A sensor array accurately collects data on three types of road segments and multi-dimensional operational data; a data preprocessing module dynamically adjusts filtering strategies based on the characteristics of different road segments to improve the quality of the original data; the processed data is stored together with blade specification data to provide a foundation for model training; a historical data processing and training module performs three-dimensional classification training based on blade specifications, road segment type, and vehicle speed range, generating coefficient matrices and threshold libraries adapted to different road segments to achieve scenario-based parameter adaptation; a data fusion module dynamically allocates data weights according to road segment type and outputs accurate state fusion values; a control decision module combines the fusion values ​​with a three-dimensional mapping library, simplifies prediction algorithms to correct parameters, and balances accuracy and response speed; a cross-validation module eliminates data errors through two-way mutual verification; and an adaptive correction module clarifies the correspondence between four operating conditions and road segment types, dynamically adjusting control strategies for different operating conditions to adapt to both fixed road segment characteristics and dynamic cross-road segment scenarios.

[0008] The beneficial effects of this invention are as follows: It adapts to different blade sizes and various transportation scenarios by using three-dimensional classification training based on blade specifications, road segment type, and vehicle speed range, combined with dynamic adjustment of control parameters according to operating conditions, thus solving the problem of poor adaptability of fixed parameters; it employs a multi-source data fusion and cross-validation mechanism to avoid decision-making errors caused by single data source errors, while improving the quality of raw data through filtering and calibration to ensure the accuracy of supporting control; it continuously trains and updates the coefficient matrix and threshold library using historical data, eliminating the need for repeated manual debugging, reducing operating costs, and gradually improving control accuracy with data accumulation; it uses a simplified model predictive control algorithm, requiring no high-performance computing hardware, can be deployed on conventional vehicle controllers, has a fast response speed, and adapts to dynamic transportation conditions.

[0009] Furthermore, the drift calibration logic of the data preprocessing module includes: when the wind speed data reaches the preset wind speed threshold, the attitude raw data collected by the inertial measurement unit is subjected to a preset compensation amount of deviation compensation, and then the processed wind turbine blade attitude data is output.

[0010] Wind speed is a key environmental disturbance factor affecting attitude measurement. For strong wind scenarios that may occur in different road sections, a threshold-triggered compensation mechanism can be used to offset the attitude data drift caused by wind speed, further improving the data measurement accuracy in multi-road sections and complex environments, and providing reliable input for subsequent control decisions.

[0011] Furthermore, in the attitude and support force influence coefficient matrix output by the historical data processing and training module, the support force coefficients of blades of different specifications under the same road segment type show a linear adaptation ratio according to the blade length; in the dynamic threshold library, the attitude deviation benchmark threshold corresponding to the highway segment is less than that of the national road segment, and the attitude deviation benchmark threshold corresponding to the national road segment is less than that of the rural road segment.

[0012] The support force coefficient can be set linearly according to the blade length, which can quickly adapt to the stiffness differences of different blade specifications without the need for individual training; the attitude deviation threshold can be set according to the characteristics of different road sections, which is in line with the smoothness characteristics of different road sections. Highway sections have higher requirements for attitude stability, while rural road sections allow for moderately larger attitude fluctuations, thereby improving the scenario-specificity of the control strategy.

[0013] Furthermore, the dynamic threshold adaptation rules of the cross-validation module include: as the blade length increases, the attitude deviation threshold increases linearly; as the road bumpiness intensifies, the pressure deviation threshold increases linearly, and the road bumpiness in rural road sections, national road sections, and expressway sections decreases sequentially.

[0014] Longer blades have lower stiffness, so appropriately increasing the attitude deviation threshold is in line with their structural characteristics. Shorter blades have higher stiffness, so the deviation must be strictly controlled. The higher the road bumpiness, the more significant the support pressure fluctuation. Increasing the pressure deviation threshold can avoid control oscillations caused by frequent calibration, thus achieving dual adaptation of the threshold with blade characteristics and road conditions.

[0015] Furthermore, during the parameter correction iteration process of the simplified model predictive control algorithm, each iteration is constrained by the dynamic threshold corresponding to the current operating condition to avoid overload of the support force adjustment range leading to blade damage.

[0016] By adjusting the magnitude of the dynamic threshold constraint parameters, it is possible to effectively avoid a sudden increase in support force caused by road segment switching or sudden changes in operating conditions, thus preventing squeezing or impact damage to the structurally fragile wind turbine blades. It is especially suitable for the smooth adjustment of support force under dynamic operating conditions such as rapid turns and accelerations, ensuring transportation safety. Attached Figure Description

[0017] Figure 1This is a system block diagram of an embodiment of the adaptive support control system for wind turbine blade transportation of the present invention. Detailed Implementation

[0018] The following detailed description illustrates the specific implementation method: The basic implementation examples are as follows: Figure 1 As shown: An adaptive support control system for wind turbine blade transportation, including a sensor array and a server; The sensor group includes an inertial measurement unit, a support pressure sensor, a vehicle speed sensor, a road segment identification sensor, and an environmental auxiliary sensor. It is used to collect raw data on wind turbine blade attitude, raw data on support contact pressure, transport vehicle speed, road segment type, raw wind speed, and rainfall status, and upload them to the server. The road segment type data includes highway segments, national highway segments, and rural road segments. The server includes a data preprocessing module, a historical data processing and training module, a data fusion module, a control decision module, a cross-validation module, and an adaptive correction module. The data preprocessing module is used to perform moving average filtering and drift calibration on the raw data of wind turbine blade attitude, support contact pressure, and wind speed. The size of the filtering window is adjusted according to the transport vehicle speed data and road segment type data. After processing, the wind turbine blade attitude data, support contact pressure data, and wind speed data are obtained. The processed data, the transport vehicle speed data collected by the sensors, the road segment type data, the rainfall status data, and the preset blade specification data are stored together in the historical data processing and training module. The blade specification data includes preset parameters adapted to wind turbine blades of different sizes. The historical data processing and training module is used to extract historical data of different blade specifications, road segment types and vehicle speed ranges from the stored data, construct a training set according to the three-dimensional classification of blade specifications, road segment types and vehicle speed ranges, and obtain the attitude and support force influence coefficient matrix and dynamic threshold library including attitude deviation threshold and pressure deviation threshold through gradient boosting algorithm training. During the training process, the optimal control parameter label and fault handling parameters are generated and stored simultaneously. The data fusion module receives wind turbine blade attitude data, support contact pressure data, and wind speed data output by the data preprocessing module, as well as transport vehicle speed data, road segment type data, and rainfall status data collected by sensors. It then fuses the above data using a dynamic weight allocation and Kalman filtering strategy. The dynamic weights are adjusted according to the transport vehicle speed data and road segment type data. After fusion, the output is a blade and support status fusion value that includes real-time attitude deviation, optimal support pressure range, and scenario adaptation coefficient. The control decision module has a pre-set 3D mapping library of attitude deviation support force road segment types, which is used to match the initial control parameters based on the fusion value of the blade and support state and the attitude and support force influence coefficient matrix. It predicts the state changes within a preset time period through a simplified model predictive control algorithm, and iterates the initial control parameters a preset number of times with transport vehicle speed data, real-time attitude deviation and scene adaptation coefficient as the core influencing factors, and outputs the support force adjustment value and response speed command. The cross-validation module is used to perform bidirectional cross-validation between the real-time attitude deviation and optimal support pressure range in the fused values ​​of blade and support status and the theoretical values ​​corresponding to the attitude and support force influence coefficient matrix. When the data deviation exceeds the preset ratio of the dynamic threshold, a data calibration command is triggered to the data preprocessing module. The adaptive correction module is used to classify four operating conditions based on road segment type data, transport vehicle speed change rate, real-time attitude deviation change rate, and rainfall status data: high-speed smooth, national highway bumpy, rural road complex, and rapid turn and acceleration. The high-speed smooth condition corresponds to high-speed road segments, the national highway bumpy condition corresponds to national highway segments, the rural road complex condition corresponds to rural road segments, and the rapid turn and acceleration condition corresponds to cross-road segment dynamic conditions. The module dynamically adjusts the update frequency of control parameters, the adjustment range of support force, and the dynamic threshold for different operating conditions, and outputs the final control signal to the support actuator.

[0019] Furthermore, the drift calibration logic of the data preprocessing module includes: when the wind speed data reaches the preset wind speed threshold, the attitude raw data collected by the inertial measurement unit is subjected to a preset compensation amount of deviation compensation, and then the processed wind turbine blade attitude data is output.

[0020] Furthermore, in the attitude and support force influence coefficient matrix output by the historical data processing and training module, the support force coefficients of blades of different specifications under the same road segment type show a linear adaptation ratio according to the blade length; in the dynamic threshold library, the attitude deviation benchmark threshold corresponding to the highway segment is less than that of the national road segment, and the attitude deviation benchmark threshold corresponding to the national road segment is less than that of the rural road segment.

[0021] Furthermore, the dynamic threshold adaptation rules of the cross-validation module include: as the blade length increases, the attitude deviation threshold increases linearly; as the road bumpiness intensifies, the pressure deviation threshold increases linearly, and the road bumpiness in rural road sections, national road sections, and expressway sections decreases sequentially.

[0022] Furthermore, during the parameter correction iteration process of the simplified model predictive control algorithm, each iteration is constrained by the dynamic threshold corresponding to the current operating condition to avoid overload of the support force adjustment range leading to blade damage.

[0023] In this embodiment, the inertial measurement unit (IMU) has a sampling frequency of 10 Hz and is used to collect raw blade attitude data in real time; the support pressure sensor has a sampling frequency of 5 Hz, a measurement range of 0-50 kN, and is deployed at the contact point between the blade and the support device to collect raw support contact pressure data; the vehicle speed sensor has a sampling frequency of 1 Hz and collects real-time vehicle speed data; the road segment identification sensor combines GPS positioning and electronic map data, has a sampling frequency of 0.5 Hz, and automatically identifies the road segment type (highway / national highway / rural road); and the environmental auxiliary sensor has a sampling frequency of 0.5 Hz and collects raw wind speed data and rainfall status data.

[0024] Road segment type and working condition judgment criteria: Expressway segment: identified as an expressway by the road segment identification sensor, with a speed limit of 60-120km / h; corresponding to the stable working condition of the expressway, the judgment conditions are that the rate of change of the transport vehicle speed is less than or equal to 5km / h / s and the rate of change of the real-time attitude deviation is less than or equal to 1° / s. National Highway Section: The road section is identified as a national highway by the road section identification sensor. The speed limit is 40-80km / h. It corresponds to the bumpy working conditions of the national highway. The judgment condition is that the real-time attitude deviation change rate is greater than 1° / s and less than or equal to 2° / s. Rural road sections: The road section identification sensor identifies it as a rural road with a speed limit of 0-40km / h. This corresponds to the complex working conditions of rural roads. The judgment condition is that the real-time attitude deviation change rate is greater than 2° / s. Rapid turn and acceleration (crossing road sections): The judgment condition is that the rate of change of the transport vehicle speed is greater than 5km / h / s. Regardless of the current road section, the special control strategy for this condition will be triggered.

[0025] Filter window settings: Window length 5 for highways (vehicle speed greater than or equal to 60km / h), window length 3 for highways (vehicle speed less than 60km / h); window length 4 for national highways (vehicle speed greater than or equal to 40km / h), window length 2 for national highways (vehicle speed less than 40km / h); window length is fixed at 2 for rural roads. Drift calibration: preset wind speed threshold of 8m / s, preset compensation amount of 0.2° (to adapt to deviations caused by actual wind speed interference), automatic compensation when the wind speed is greater than or equal to 8m / s.

[0026] The historical data processing and training module is specifically set up as follows: Three-dimensional classification: blade size (30m / 40m / 50m / 60m / 70m), road segment type (3 types), vehicle speed range (0-40km / h / 40-60km / h / over 60km / h), with a total of 45 training subsets and a total data volume of over 1200 sets; Coefficient matrix: The support force coefficient of the 60m blade in rural road sections is set to be 1.8 times that of the 30m blade, and other specifications are calculated by linear interpolation; Dynamic threshold baseline values ​​(adapted to actual errors): 1.0° attitude deviation and 2kN pressure deviation for highway sections; 1.5° attitude deviation and 3kN pressure deviation for national highway sections; and 2.0° attitude deviation and 4kN pressure deviation for rural road sections.

[0027] Dynamic weights for the data fusion module: Highway section: attitude data 0.6, pressure data 0.3, wind speed data 0.05, rainfall data 0.05; National highway section: attitude data 0.5, pressure data 0.4, wind speed data 0.05, rainfall data 0.05; Rural road sections: attitude data 0.4, pressure data 0.5, wind speed data 0.05, rainfall data 0.05.

[0028] Specific settings for the control decision module: 3D mapping library: containing 45 sets of optimal parameters for typical working conditions; Simplified model prediction control: prediction time 0.5s, 2 iterations, influencing factor weights: vehicle speed 0.3, attitude deviation 0.5, scene adaptation coefficient 0.2.

[0029] Cross-validation and adaptive correction settings: Cross-validation: Calibration is triggered when the deviation exceeds 20% (e.g., when the attitude deviation on a highway exceeds 1.2°). Operating condition control strategy: High-speed and stable operating conditions: update frequency 1Hz, adjustment range 5%, threshold set to 80% of the baseline value (attitude deviation 0.8°); National Highway Bumpy Condition: Update frequency 3Hz, adjustment range 8%, threshold set to 100% of the baseline value (attitude deviation 1.5°); Complex working conditions on rural roads: update frequency 5Hz, adjustment range 12%, threshold set to 110% of the baseline value (attitude deviation 2.2°). Rapid turn and acceleration conditions: core parameters are locked, response speed is improved by 30%, and the threshold is set to 120% of the baseline value (corresponding to the upper limit of attitude deviation on the road section: 1.2° for highways, 1.8° for national highways, and 2.4° for rural areas).

[0030] The above are merely embodiments of the present invention. Commonly known structures and characteristics are not described in detail here. Those skilled in the art are aware of all common technical knowledge in the field prior to the application date or priority date, are aware of all existing technologies in that field, and have the ability to apply conventional experimental methods prior to that date. Those skilled in the art can, under the guidance of this application, improve and implement this solution in combination with their own capabilities. Some typical known structures or methods should not be obstacles for those skilled in the art to implement this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the structure of the present invention. These should also be considered within the scope of protection of the present invention, and will not affect the effectiveness of the implementation of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. An adaptive support control system for wind turbine blade transportation, characterized by: The sensor group and the server are included; The sensor group includes an inertial measurement unit, a support pressure sensor, a vehicle speed sensor, a road section identification sensor, and an environmental auxiliary sensor, which are used to collect wind turbine blade posture original data, support contact pressure original data, transportation vehicle speed data, road section type data, wind speed original data, and rainfall state data and upload them to the server. The road section type data includes highway sections, national road sections, and rural road sections. The server includes a data preprocessing module, a historical data processing and training module, a data fusion module, a control decision module, a cross-validation module, and a self-adaptive correction module. The data preprocessing module is used to perform sliding average filtering and drift calibration on the wind turbine blade posture original data, the support contact pressure original data, and the wind speed original data. The filtering window size is adjusted according to the transportation vehicle speed data and the road section type data. After processing, the wind turbine blade posture data, the support contact pressure data, and the wind speed data are obtained. The processed data, the transportation vehicle speed data, the road section type data, the rainfall state data collected by the sensors, and the preset blade specification data are stored in the historical data processing and training module. The blade specification data includes preset parameters for different sizes of wind turbine blades. The historical data processing and training module is used to extract historical data of different blade specifications, road section types, and vehicle speed intervals from the stored data. A training set is constructed according to the three-dimensional classification of blade specifications, road section types, and vehicle speed intervals. The posture and support force influence coefficient matrix and the dynamic threshold library including the posture deviation threshold and the pressure deviation threshold are obtained through gradient boosting algorithm training. The optimal control parameter label and the fault handling parameter are generated synchronously during the training process and stored. The data fusion module is used to receive the wind turbine blade posture data, the support contact pressure data, and the wind speed data output by the data preprocessing module, as well as the transportation vehicle speed data, the road section type data, and the rainfall state data collected by the sensors. Dynamic weight distribution and Kalman filtering strategy are used to fuse the above data. The dynamic weight is adjusted according to the transportation vehicle speed data and the road section type data. The fusion value of the blade and the support state containing the real-time posture deviation, the optimal support pressure range, and the scene adaptation coefficient is output. The control decision module is preset with a posture deviation support force road section type three-dimensional mapping library. It is used to match the initial control parameter according to the blade and support state fusion value and the posture and support force influence coefficient matrix. The state change within a preset time is predicted through a simplified model predictive control algorithm. The initial control parameter is corrected by iterating a preset number of times with the transportation vehicle speed data, the real-time posture deviation, and the scene adaptation coefficient as the core influencing factors. The support force adjustment value and the response speed instruction are output. The cross-validation module is used to perform bidirectional verification on the real-time posture deviation and the optimal support pressure range in the blade and support state fusion value, and the theoretical value corresponding to the posture and support force influence coefficient matrix. When the data deviation exceeds the preset proportion of the dynamic threshold, a data calibration instruction is triggered to the data preprocessing module. An adaptive correction module is configured to divide four types of working conditions, i.e., high-speed smooth, national road bumpy, rural road complex, and sharp turn and rapid acceleration, according to the road section type data, the transport vehicle speed change rate, the real-time attitude deviation change rate, and the rainfall state data, wherein the high-speed smooth working condition corresponds to a high-speed road section, the national road bumpy working condition corresponds to a national road section, the rural road complex working condition corresponds to a rural road section, and the sharp turn and rapid acceleration working condition corresponds to a cross-road section dynamic working condition; The control parameter update frequency, the support force adjustment amplitude, and the dynamic threshold are dynamically adjusted according to different working conditions, and a final control signal is output to a support execution mechanism.

2. The adaptive support control system for wind blade transportation according to claim 1, wherein: The drift calibration logic of the data preprocessing module includes: when the wind speed data reaches a preset wind speed threshold, the attitude original data collected by the inertial measurement unit is subjected to deviation compensation by a preset compensation amount, and then the processed wind turbine blade attitude data is output.

3. The wind turbine blade transport adaptive support control system according to claim 2, wherein: In the attitude and support force influence coefficient matrix output by the historical data processing and training module, the support force coefficients of different specifications of blades under the same road section type are in a linear adaptive proportional relationship according to the blade length. In the dynamic threshold library, the attitude deviation reference threshold corresponding to the high-speed road section is smaller than that corresponding to the national road section, and the attitude deviation reference threshold corresponding to the national road section is smaller than that corresponding to the rural road section.

4. The adaptive support control system for wind blade transportation according to claim 3, wherein: The dynamic threshold adaptation rule of the cross-validation module includes: the attitude deviation threshold increases linearly with the increase of the blade length; the pressure deviation threshold linearly expands with the increase of the road section bumping degree, and the road section bumping degrees of the rural road section, the national road section, and the high-speed road section decrease in turn.

5. The adaptive support control system for wind blade transportation according to claim 4, wherein: In the parameter correction iteration process of the simplified model predictive control algorithm, the dynamic threshold corresponding to the current working condition is used as a constraint in each iteration to avoid blade damage caused by excessive support force adjustment amplitude.

Images