An Adaptive Disturbance Resistant Control Method and System for Offshore Wind Power Operation and Maintenance

Abstract

This application relates to an adaptive disturbance rejection control method and system for aircraft used in offshore wind power operation and maintenance. The method includes receiving wind field data from a lidar terminal mounted on the aircraft in real time and acquiring wake field data of the wind turbine to be repaired / tested; fusing the wind field data and wake field data and projecting them onto a three-dimensional grid space in the aircraft's body coordinate system to generate a three-dimensional disturbed wind field grid map; inputting the three-dimensional disturbed wind field grid map into a preset nonlinear model predictive controller and acquiring the aircraft's flight state data in real time; the nonlinear model predictive controller predicts, based on the aircraft's current flight state data, that the aircraft will pass through grid points in the three-dimensional disturbed wind field grid map within the next n seconds; and calculating the feedforward compensation amount of the aircraft's power system based on the disturbance data of the grid points to be passed. This application has the effect of improving the aircraft's disturbance rejection kinetic energy in complex airflow environments around offshore wind turbines.

Description

Technical Field

[0001] This application relates to the technical field of aircraft control, and in particular to an adaptive disturbance rejection control method and system for offshore wind power operation and maintenance. Background Technology

[0002] Traditional offshore wind power operation and maintenance relies on ships, which is inefficient and poses high safety risks. With the development of drone technology, using aircraft for material transportation and inspection has become a trend.

[0003] However, the marine environment is complex, with strong turbulence, wakes, and sudden gusts near wind turbines. Existing aircraft control systems mostly rely on preset parameters and suffer from control lag, making it difficult to respond in real time to such rapidly changing and nonlinear airflow disturbances. This can easily lead to loss of attitude control, positional deviation, or even collisions with wind turbines or crashes into the sea. Therefore, improvements are needed. Summary of the Invention

[0004] To address the problem of insufficient anti-disturbance capability of aircraft in complex airflow environments around offshore wind turbines in existing technologies, this application provides an adaptive anti-disturbance control method and system for offshore wind power operation and maintenance.

[0005] The above-mentioned objective of this application is achieved through the following technical solution:

[0006] An adaptive disturbance rejection control method for aircraft used in offshore wind power operation and maintenance includes the following steps:

[0007] It receives wind field data from the lidar terminal on board the aircraft in real time and acquires wake field data of the wind turbine to be repaired and inspected.

[0008] The wind field data and wake field data are fused and projected into a three-dimensional grid space under the aircraft's body coordinate system to generate a three-dimensional perturbation wind field grid map. The grid points in the three-dimensional perturbation wind field grid map are bound to the perturbation data.

[0009] The three-dimensional perturbation wind field grid map is input into the preset nonlinear model prediction controller and the flight status data of the aircraft is acquired in real time. Based on the current flight status data of the aircraft, the nonlinear model prediction controller predicts that the aircraft will pass through the grid points in the three-dimensional perturbation wind field grid map within the next n seconds.

[0010] Based on the disturbance data of the grid points that are about to be passed, the feedforward compensation of the aircraft's propulsion system is calculated.

[0011] By adopting the above technical solution, real-time scanning data of the wind turbine wake field based on an airborne environmental perception system, such as multi-beam lidar, is introduced. This enables real-time perception of offshore wind field data and also considers the complex environment of the wind turbine wake's impact on the wind field during wind turbine maintenance and inspection. Wind field data and wake field data are fused to form a three-dimensional disturbed wind field grid map. Combined with the overlay of flight status data, a four-dimensional predictive model is formed, capable of predicting the aircraft's route in three-dimensional space and time dimensions. This model can detect airflow disturbances that the aircraft will encounter 1-3 seconds in advance, such as wind turbine wakes and gust fronts. The nonlinear model predictive controller makes pre-emptive adjustments based on the predicted aircraft trajectory and wind field disturbance data along the trajectory, calculating the feedforward control compensation required to counteract the disturbance in advance. This ensures precise time alignment between the aircraft's control commands and future disturbances, achieving a proactive anti-disturbance effect of "commands issued before disturbances arrive," solving the lag problem of traditional feedback control, and taking into account the complex scenarios affected by wind turbine wakes.

[0012] Optionally, the step of receiving wind field data from the lidar terminal mounted on the aircraft in real time and obtaining wake field data of the wind turbine to be repaired and inspected includes:

[0013] The lidar terminal emits a laser beam at a preset frequency into a preset fan-shaped area in front to obtain the echo reflected by aerosol particles and obtain point cloud data.

[0014] By analyzing the displacement vector of the same aerosol in two consecutive frames of point cloud data, wind field data in the vicinity of the aircraft is obtained. The wind field data includes the instantaneous three-dimensional wind field observation field.

[0015] When the position of the wind turbine tower and blades is identified through point cloud data, the relative position of the aircraft and the wind turbine is determined. When the aircraft is within the preset area, the model parameters of the wind turbine to be repaired and inspected are identified.

[0016] The wake effect parameters of the fan are matched based on the model parameters.

[0017] By employing the above technical solution, a clearly structured wind field is obtained by acquiring the echoes reflected from aerosol particles. Since a single frame scan by the lidar requires a certain time (e.g., 100ms), the aircraft's position and attitude may have changed significantly between the start and end of this frame, leading to motion distortion in the point cloud data. Therefore, the point cloud data needs distortion correction. The three-dimensional wind field observation field can then be obtained through aerosol motion field inversion using the three-dimensional optical flow method. Utilizing continuous frame motion estimation, this wind field model inherently possesses predictive capabilities—combining the aircraft's current speed and heading, it can calculate which grid nodes the aircraft will pass through in the next 1-3 seconds. By identifying the positions of the wind turbine tower and blades, it can be determined that the aircraft is located within a certain range downstream of the wind turbine, i.e., when the aircraft is affected by the wind turbine wake, providing an accurate data source for the subsequent three-dimensional disturbed wind field grid map.

[0018] Optionally, the step of fusing wind field data and wake field data and projecting them into a three-dimensional grid space in the aircraft's body coordinate system to generate a three-dimensional disturbed wind field grid map, wherein the grid points in the three-dimensional disturbed wind field grid map are bound to the disturbed data, includes:

[0019] A three-dimensional regular grid is preset, and the weather forecast wind field is used as the background field;

[0020] The three-dimensional wind field observation field is used as the observation value, and the confidence weight of the three-dimensional wind field observation field is used to update the wind field in the background field.

[0021] When the aircraft is located within the preset area, the wake effect parameters are used as a strong constraint to make secondary corrections to the wind field within the preset area.

[0022] Generate a three-dimensional perturbation wind field grid with the current position of the aircraft as the origin and a fixed size. Each grid point stores a three-dimensional wind vector.

[0023] By adopting the above technical solution and combining the three-dimensional wind field observation field with the parameterized model fitting of the wind turbine wake influence, multi-dimensional perception of complex wind field environments is achieved. When the lidar echo signal is weak or the aerosol distribution is uneven, the wake influence parameters can fill the observation gaps; when the wind turbine wake morphology is complex, optical flow observation can correct the model parameters. The fusion of the two enables the wind field perception system to still output reliable disturbance estimates even if any single information source fails or is inaccurate.

[0024] Optionally, the step of inputting the three-dimensional perturbation wind field grid map into a preset nonlinear model prediction controller and acquiring the aircraft's flight status data in real time, wherein the nonlinear model prediction controller predicts that the aircraft will pass through the grid points in the three-dimensional perturbation wind field grid map within the next n seconds based on the aircraft's current flight status data, includes the following steps in constructing the nonlinear model prediction controller:

[0025] Acquire the current flight status data of the aircraft, including the current state vector and the aircraft's dynamic parameters;

[0026] The six-degree-of-freedom rigid body dynamics equations of the aircraft are established as follows:

[0027] ẋ=f(x, u, w),

[0028] Where u is the control input vector, including the square of the rotational speed of each rotor or the thrust, w is the external disturbance, including the wind speed, and x is the state vector;

[0029] Decouple airspeed, which is the speed of an aircraft relative to the air, from ground speed, which is the speed relative to the ground. The relationship is: v_air = v_ground - v_wind, where v_ground is the ground speed and v_wind is the airspeed.

[0030] In each round of prediction, for each future time t+k*Δt and prediction location p_pred(t+kΔt) in the prediction time domain, the wind speed value at that location is queried from the three-dimensional perturbation wind field grid map;

[0031] Discretize the six-degree-of-freedom rigid body dynamics equation ẋ= f(x, u, w) into a difference equation:

[0032] x(k+1) = F[x(k), u(k), w(k)], where K is the acquisition time. The difference equation is used to predict the state trajectory of the spacecraft at a series of future time points.

[0033] By adopting the above technical solution, the aerodynamic forces on the aircraft, such as lift and drag, depend on airspeed rather than ground speed. Therefore, in the dynamic equations, all aerodynamic-related items, such as rotor thrust model and airframe drag model, must be calculated using airspeed. Since the wind field model is gridded, three-dimensional interpolation algorithms, such as trilinear interpolation or cubic spline interpolation, are required to obtain the wind speed at any location. This allows the aircraft's trajectory at a series of future time points to be predicted based on the current state and future control sequence.

[0034] Furthermore, it does not employ complex online learning algorithms to adjust the internal model of the nonlinear model predictive controller. However, its model-predictive nature inherently gives it parameter adaptive capabilities. As long as the aircraft's dynamic equations f(x, u, w) are sufficiently accurate, the nonlinear model predictive controller can automatically calculate the optimal control sequence to match different flight weights and environmental conditions. When changing to a different aircraft configuration, such as from a quadcopter to a hexacopter, only the internal dynamic equations need to be changed.

[0035] Optionally: the step of calculating the feedforward compensation amount of the aircraft propulsion system based on the disturbance data of the grid points to be passed further includes:

[0036] The nonlinear model predictive controller calculates the deviation between the predicted trajectory and the desired trajectory of the aircraft at future moments based on a preset cost function. The cost function includes the calculation of trajectory tracking cost, control cost, control change rate cost, and terminal cost. The desired trajectory is obtained by collecting pre-planned route points.

[0037] Based on the current state, construct a nonlinear programming problem, namely, the optimal control sequence that minimizes the cost function under the premise of satisfying the difference equation and the preset constraints.

[0038] Execute the first control input of the optimal control sequence and output the feedforward compensation input.

[0039] By adopting the above technical solution, the nonlinear model predictive controller not only pursues the minimization of output attitude error, but also comprehensively balances multiple objectives such as trajectory tracking accuracy, control energy consumption, and control smoothness in the cost function. After executing the first control variable, the above steps are repeated in the next control cycle, for example, after 20ms, using the latest state and the updated wind field model to perform optimization again; this process is repeated cyclically, forming a real-time, four-dimensional feedforward control compensation variable based on future wind field prediction calculations for rolling time-domain control.

[0040] Optionally: After the step of calculating the feedforward compensation amount of the aircraft propulsion system based on the disturbance data of the grid points to be passed, the following steps are also performed:

[0041] Acquire the position and attitude data of the aircraft after feedforward control compensation, and perform real-time error calculation with the expected position and attitude data in the expected trajectory;

[0042] The error calculation results are input into the parallel PID controller. The PID controller calculates the feedback control quantity based on the proportional, integral and derivative coefficients and sends it to the aircraft power system to eliminate real-time errors.

[0043] By adopting the above technical solution, the PID controller is responsible for correcting the prediction error of the feedforward, the unmodeled random disturbance, and the inaccuracy of the model itself, so as to reduce the error between the aircraft and the desired trajectory and ensure that the system eventually converges to the desired state.

[0044] Optional: Simultaneously with sending the feedforward compensation and feedback control quantities to the aircraft propulsion system, the following steps are also included:

[0045] Obtain the current maximum available power of each motor in the power system;

[0046] The upper limit of available power is used as a constraint on the current feedforward compensation or feedback control. When the feedforward compensation or feedback control of the motor exceeds the current upper limit of available power, the excess feedforward compensation or feedback control is transferred to the other available motors.

[0047] If the power of each motor in the current system is close to saturation and still cannot meet the requirements of feedforward compensation or feedback control, the requirements for attitude control accuracy will be automatically reduced.

[0048] By adopting the above technical solutions, the energy management coordination mechanism incorporates the real-time power limits of each motor as hard constraints into the control allocation process, ensuring that the power system will not collapse due to overload of a single motor under any circumstances. For example, the weight of attitude error in the objective function is relaxed to ensure that the aircraft can first maintain the most basic flight safety. At this time, the aircraft may not be able to hover precisely, but it can perform a safe, power-limited return or forced landing maneuver.

[0049] Traditional control systems often treat attitude control and energy management separately. This invention achieves deep coupling between the two by directly embedding motor power constraints into the optimization solution of the control distributor; thus improving the fault tolerance and survivability of the aircraft from a systems engineering perspective.

[0050] The second objective of this invention is achieved through the following technical solution:

[0051] An adaptive disturbance rejection control system for aircraft used in offshore wind power operation and maintenance includes:

[0052] The data acquisition module is used to receive wind field data emitted by the lidar terminal on board the aircraft in real time, and to acquire wake field data of the wind turbine to be repaired and inspected.

[0053] The wind field grid map module is used to fuse wind field data and wake field data and project them into a three-dimensional grid space under the aircraft's body coordinate system to generate a three-dimensional perturbation wind field grid map. The grid points in the three-dimensional perturbation wind field grid map are bound to perturbation data.

[0054] The prediction module is used to input the three-dimensional disturbance wind field grid map into the preset nonlinear model prediction controller and acquire the flight status data of the aircraft in real time. Based on the current flight status data of the aircraft, the nonlinear model prediction controller predicts that the aircraft will pass through the grid points in the three-dimensional disturbance wind field grid map within the next n seconds.

[0055] The feedforward compensation module is used to calculate the feedforward compensation amount of the aircraft's propulsion system based on the disturbance data of the grid points that are about to be passed.

[0056] The above-mentioned objective three of this application is achieved through the following technical solution:

[0057] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the aforementioned adaptive anti-disturbance control method for aircraft used in offshore wind power operation and maintenance.

[0058] The fourth objective of this application is achieved through the following technical solution:

[0059] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the aforementioned adaptive disturbance rejection control method for offshore wind power operation and maintenance.

[0060] In summary, this application includes at least one of the following beneficial technical effects:

[0061] 1. Wind field data and wake field data are fused to form a three-dimensional disturbance wind field grid map. Combined with the overlay of flight status data, this creates a four-dimensional predictive model capable of predicting aircraft routes in both three-dimensional space and time dimensions. This model can detect impending airflow disturbances, such as wind turbine wakes and gust fronts, 1-3 seconds in advance. The nonlinear model predictive controller, based on the predicted aircraft trajectory and wind field disturbance data along that trajectory, makes pre-calculation adjustments, determining the feedforward control compensation required to counteract the disturbance. This ensures precise time alignment between the aircraft's control commands and future disturbances, achieving a proactive disturbance rejection effect of "commands issued before disturbances arrive," solving the lag problem of traditional feedback control, and taking into account the complex scenarios of wind turbine wake effects.

[0062] 2. Utilizing continuous frame motion estimation, this wind field model inherently possesses predictive capabilities—by combining the aircraft's current speed and heading, it can calculate which grid nodes the aircraft will pass through in the next 1-3 seconds. By identifying the positions of the wind turbine tower and blades, it can determine that the aircraft is located within a certain range downstream of the wind turbine, i.e., determine when the aircraft is affected by the wind turbine wake, providing an accurate data source for subsequent three-dimensional perturbation wind field grid maps;

[0063] 3. The aerodynamic forces acting on an aircraft, such as lift and drag, depend on airspeed, not ground speed. Therefore, in the dynamic equations, all aerodynamic-related items, such as rotor thrust models and airframe drag models, must be calculated using airspeed. Since the wind field model is gridded, three-dimensional interpolation algorithms, such as trilinear interpolation or cubic spline interpolation, are required to obtain the wind speed at any location. This allows the aircraft's trajectory at a series of future time points to be predicted based on the current state and future control sequences.

[0064] 4. The energy management coordination mechanism incorporates the real-time power limits of each motor as hard constraints into the control allocation process, ensuring that the power system will not collapse under any circumstances due to the overload of a single motor. Attached Figure Description

[0065] Figure 1 This is a flowchart of an embodiment of an adaptive disturbance rejection control method for aircraft used in offshore wind power operation and maintenance according to this application;

[0066] Figure 2 This is a flowchart of step 20 in an embodiment of an adaptive disturbance rejection control method for aircraft used in offshore wind power operation and maintenance according to this application;

[0067] Figure 3 This is a schematic block diagram of a computer device according to this application. Detailed Implementation

[0068] The following is in conjunction with the appendix Figure 1-3 This application will be described in further detail.

[0069] In one embodiment, such as Figure 1 As shown, this application discloses an adaptive disturbance rejection control method for aircraft used in offshore wind power operation and maintenance, which specifically includes the following steps:

[0070] S10: Receive wind field data from the lidar terminal on board the aircraft in real time and obtain wake field data of the wind turbine to be repaired and inspected.

[0071] In this embodiment, the aircraft is also equipped with an IMU (Inertial Measurement Unit) and a GPS positioning module to acquire flight status data. Specifically, it collects triaxial acceleration and angular velocity at a frequency of 200Hz, and GPS position, velocity, and heading angle at a frequency of 10Hz. Then, it performs tightly coupled calculations using extended Kalman filtering to obtain the aircraft's current high-precision attitude, position, and velocity information. The wind field data is a point cloud map of aerosol reflection, and the wake field data is a parameterized model of the wind turbine wake intensity and its influence range.

[0072] Specifically, step S10 includes:

[0073] S11: The lidar terminal emits a laser beam at a preset frequency into a preset fan-shaped area in front to obtain the echo reflected by aerosol particles and obtain point cloud data.

[0074] S12: Analyze the displacement vector of the same aerosol in two consecutive frames of point cloud data to obtain wind field data in the vicinity of the aircraft. The wind field data includes the instantaneous three-dimensional wind field observation field.

[0075] S13: When the position of the wind turbine tower and blades is identified through point cloud data, the relative position of the aircraft and the wind turbine is determined. When the aircraft is within the preset area, the model parameters of the wind turbine to be repaired and inspected are identified.

[0076] S14: Wake effect parameters of matching fans based on model parameters.

[0077] In this embodiment, a multi-beam lidar with a wavelength of 905nm or 1550nm is used to scan a fan-shaped area in front at a frequency of 10Hz to obtain point cloud data. Further motion distortion correction is performed on the original point cloud data, including a distortion correction algorithm. Specifically, for each point P_i in a frame of point cloud, based on its corresponding precise acquisition time t_i, the relative attitude transformation matrix ΔT obtained from time t_i to the end of the frame t_end, obtained by IMU integration, is used to inversely compensate the point to a unified coordinate system at time t_end. Finally, statistical filtering is performed on the corrected point cloud to remove isolated outliers; then, voxel filtering downsampling is performed, dividing the three-dimensional space into a fixed-size (e.g., 0.2m x 0.2m x 0.2m) voxel grid. The centroid of all points within each voxel is used to replace all points within that voxel to reduce the data volume and improve subsequent processing speed, ultimately outputting a clear and sparse three-dimensional point cloud.

[0078] The construction of the three-dimensional wind field observation field is achieved through aerosol motion field inversion using the three-dimensional optical flow method. For each point P_t in PointCloud(t), several nearest neighbors (e.g., K=10) within its spatial neighborhood are searched in PointCloud(t-1). Assuming these nearest neighbors undergo approximately rigid body motion with the airflow within a short time interval Δt=0.1 seconds, the motion vector of P_t relative to these nearest neighbors can be obtained by solving a weighted least squares problem, which is essentially solving a three-dimensional optical flow constraint equation. A variant of the Iterative Closest Point (ICP) algorithm can be used, but the goal is not to achieve global registration between two point clouds, but rather to find the small displacement vector of each local region.

[0079] By traversing all points or key points, a set of sparse, spatially distributed three-dimensional motion vectors V_i=(vx_i, vy_i, vz_i) is obtained, where each vector represents the average velocity of the aerosol cluster at that location over the past 0.1 seconds. This vector field is a direct observation of the instantaneous wind field at the current moment.

[0080] A confidence weight is calculated for each motion vector, which is related to the density of the local point cloud, its geometry, and the fitting residuals from optical flow calculations. The denser the point cloud and the clearer its structure, the higher the weight.

[0081] Regarding the extraction of wake influence parameters, the wake region is first segmented. Based on the known location and orientation of the wind turbine, the fan-shaped region downstream of the wind turbine is automatically segmented in the point cloud as the region of interest for the wake. Then, the classic Jensen wake model or the more complex Ainslie eddy viscosity model is used to fit the point cloud density or motion characteristics in the three-dimensional wind field observation field. Through nonlinear optimization algorithms, the key parameters in the model, such as the wake centerline, the location of maximum velocity deficit, and the turbulence intensity enhancement coefficient, are estimated using the motion characteristics or the sparsity of the point cloud distribution in the three-dimensional wind field observation field.

[0082] S20: The wind field data and wake field data are fused and projected into a three-dimensional grid space under the aircraft's body coordinate system to generate a three-dimensional perturbation wind field grid map. The grid points in the three-dimensional perturbation wind field grid map are bound to the perturbation data.

[0083] In this embodiment, the disturbance data is the three-dimensional wind vector data of the grid point.

[0084] Specifically, refer to Figure 2 Step S20 includes:

[0085] S21: Preset a three-dimensional regular grid and use the weather forecast wind field as the background field;

[0086] S22: Use the three-dimensional wind field observation field as the observation value, and use the confidence weight of the three-dimensional wind field observation field to update the wind field in the background field;

[0087] S23: When the aircraft is located within the preset area, the wake influence parameter is used as a strong constraint to perform secondary correction on the wind field within the preset area;

[0088] S24: Generate a three-dimensional perturbation wind field grid map with the current position of the aircraft as the origin and a fixed size. Each grid point stores a three-dimensional wind vector.

[0089] In this embodiment, three-dimensional Kalman filtering or variational data assimilation techniques are used to fuse wind field information from different sources and with different levels of precision onto a unified three-dimensional regular grid; the weather forecast wind field is used as the background field. If no background field is available, the wind speed across the entire field is assumed to be zero or constant.

[0090] The sparse three-dimensional wind field observation field is used as the observation value, and its confidence weight is used to update the background field. The Kalman filter calculates the optimal gain of each grid point based on the prediction error covariance and the observation error covariance, thus obtaining an analysis field that incorporates the latest observations.

[0091] In the downstream region of the wind turbine, wake influence parameters are incorporated as a strong constraint into the fusion process. Specifically, in the Kalman filter prediction step, the error covariance of this region is modified according to the wake model to make it conform to the structure of the wake model. Alternatively, after the analysis step, the wake model is used to perform a secondary correction on the wind field in the corresponding region.

[0092] S30: Input the three-dimensional disturbance wind field grid map into the preset nonlinear model prediction controller and obtain the flight status data of the aircraft in real time. Based on the current flight status data of the aircraft, the nonlinear model prediction controller predicts that the aircraft will pass through the grid points in the three-dimensional disturbance wind field grid map within the next n seconds.

[0093] In this embodiment, n is typically set to 1 second. Flight status data is acquired through the IMU (Inertial Measurement Unit) and GPS positioning module, including the aircraft's current state vector x(t): containing position (px, py, pz), velocity (vx, vy, vz), attitude angles (roll φ, pitch θ, yaw ψ), and angular velocity (p, q, r). This results in a total of 12-dimensional states.

[0094] The dynamic parameters of an aircraft include: mass, moment of inertia, rotor configuration, and motor response delay.

[0095] Specifically, step S30 includes:

[0096] S31: Obtain the current flight status data of the aircraft, including the current state vector and the dynamic parameters of the aircraft;

[0097] S32: Establish the six-degree-of-freedom rigid body dynamics equations for the aircraft, expressed as follows:

[0098] ẋ=f(x, u, w),

[0099] Where u is the control input vector, including the square of the rotational speed of each rotor or the thrust, w is the external disturbance, including the wind speed, and x is the state vector;

[0100] S33: Decouple airspeed, i.e. the speed of the aircraft relative to the air, from ground speed, i.e. the speed relative to the ground, with the relationship: v_air = v_ground - v_wind, where v_ground is the ground speed and v_wind is the airspeed;

[0101] S34: In each round of prediction, for each future time t+k*Δt and prediction location p_pred(t+kΔt) in the prediction time domain, query the wind speed value at that location from the three-dimensional perturbation wind field grid map;

[0102] S35: Discretize the six-degree-of-freedom rigid body dynamics equation ẋ= f(x, u, w) into a difference equation:

[0103] x(k+1) = F[x(k), u(k), w(k)], where K is the acquisition time. The difference equation is used to predict the state trajectory of the spacecraft at a series of future time points.

[0104] S40: Calculate the feedforward compensation amount of the aircraft's propulsion system based on the disturbance data of the grid points that are about to be passed.

[0105] In this embodiment, the feedforward compensation amount is a control amount used to compensate for the power system parameters of the aircraft in the next second.

[0106] Specifically, step S40 includes:

[0107] S41: The nonlinear model predictive controller calculates the deviation between the predicted trajectory and the desired trajectory of the aircraft at future moments based on a preset cost function. The cost function includes the calculation of trajectory tracking cost, control cost, control change rate cost, and terminal cost. The desired trajectory is obtained by collecting pre-planned route points.

[0108] S42: Based on the current state, construct a nonlinear programming problem, that is, the optimal control sequence with the minimum cost function under the premise of satisfying the difference equation and the preset constraints;

[0109] S43: Execute the first control input of the optimal control sequence and output the feedforward compensation input.

[0110] In this embodiment, the cost function includes four terms:

[0111] The first term represents the trajectory tracking cost, where ||x(k) - x_ref(k)||_Q^2 is the square of the weighted Euclidean norm, measuring the deviation between the predicted and desired states at each prediction step k. Q is a diagonal weight matrix, where the size of each element on the diagonal represents the importance we place on the tracking accuracy of the corresponding state variable (such as position or attitude). For example, if hovering accuracy is critical, the weight of the position error can be set very large.

[0112] The second term is the control cost. ||u(k)||_R^2 is used to penalize excessive control inputs in order to save energy and prevent actuator saturation. R is the weight matrix of the control input.

[0113] The third term is the cost of the rate of change of the control input. ||Δu(k)||_S^2 is used to penalize drastic changes in the control input, ensuring the smoothness of the control command and reducing the impact and vibration on the mechanical structure; S is the weight matrix of the rate of change of the control input.

[0114] The fourth term is the terminal cost, ||x(N) - x_ref(N)||_P^2, which is used to ensure the stability of the predicted terminal state in the time domain. The terminal weight matrix P is obtained by solving a Riccati equation.

[0115] Sequential quadratic programming is used: the original problem is decomposed into a series of quadratic programming subproblems for solution, which has a fast convergence speed.

[0116] In one embodiment, after step S40, the method further includes:

[0117] S50: Acquire the position and attitude data of the aircraft after feedforward control compensation, and perform real-time error calculation with the expected position and attitude data in the expected trajectory;

[0118] S60: The error calculation result is input into the parallel PID controller. The PID controller calculates the feedback control quantity based on the proportional, integral and derivative coefficients and sends it to the aircraft power system to eliminate real-time error.

[0119] In one embodiment, while the feedforward compensation and feedback control quantities are sent to the aircraft propulsion system, the following is also included:

[0120] S70: Obtain the current available power limit of each motor in the power system;

[0121] S80: Use the upper limit of available power as a constraint on the current feedforward compensation or feedback control. When the feedforward compensation or feedback control of the motor exceeds the current upper limit of available power, the excess feedforward compensation or feedback control will be transferred to the remaining available motors.

[0122] S90: If the power of each motor in the current system is close to saturation and still cannot meet the requirements of feedforward compensation or feedback control, the requirements for attitude control accuracy will be automatically reduced.

[0123] In this embodiment, an allocatable motor refers to a motor whose current available power limit is greater than the current excess feedforward compensation amount.

[0124] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0125] In one embodiment, an adaptive disturbance rejection control system for aircraft used in offshore wind power operation and maintenance is provided. This adaptive disturbance rejection control system for aircraft used in offshore wind power operation and maintenance corresponds to the adaptive disturbance rejection control method for aircraft used in offshore wind power operation and maintenance described in the previous embodiment. The adaptive disturbance rejection control system for aircraft used in offshore wind power operation and maintenance includes:

[0126] The data acquisition module is used to receive wind field data emitted by the lidar terminal on board the aircraft in real time, and to acquire wake field data of the wind turbine to be repaired and inspected.

[0127] The wind field grid map module is used to fuse wind field data and wake field data and project them into a three-dimensional grid space under the aircraft's body coordinate system to generate a three-dimensional perturbation wind field grid map. The grid points in the three-dimensional perturbation wind field grid map are bound to perturbation data.

[0128] The prediction module is used to input the three-dimensional disturbance wind field grid map into the preset nonlinear model prediction controller and acquire the flight status data of the aircraft in real time. Based on the current flight status data of the aircraft, the nonlinear model prediction controller predicts that the aircraft will pass through the grid points in the three-dimensional disturbance wind field grid map within the next n seconds.

[0129] The feedforward compensation module is used to calculate the feedforward compensation amount of the aircraft's propulsion system based on the disturbance data of the grid points that are about to be passed.

[0130] Specific limitations regarding the adaptive disturbance rejection control system for aircraft used in offshore wind power operation and maintenance can be found in the above-described limitations of the adaptive disturbance rejection control method for aircraft used in offshore wind power operation and maintenance, and will not be repeated here. The various modules in the aforementioned adaptive disturbance rejection control system for aircraft used in offshore wind power operation and maintenance can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0131] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 3As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements an adaptive disturbance rejection control method for aircraft used in offshore wind power operation and maintenance.

[0132] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements an adaptive anti-disturbance control method for aircraft used in offshore wind power operation and maintenance.

[0133] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements an adaptive disturbance rejection control method for aircraft used in offshore wind power operation and maintenance.

[0134] 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 a computer program instructing related hardware. The 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 a variety of 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.

[0135] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0136] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. An adaptive disturbance rejection control method for aircraft used in offshore wind power operation and maintenance, characterized in that, Including the following steps: It receives wind field data from the lidar terminal on board the aircraft in real time and acquires wake field data of the wind turbine to be repaired and inspected. The wind field data and wake field data are fused and projected into a three-dimensional grid space under the aircraft's body coordinate system to generate a three-dimensional perturbation wind field grid map. The grid points in the three-dimensional perturbation wind field grid map are bound to the perturbation data. The three-dimensional perturbation wind field grid map is input into the preset nonlinear model prediction controller and the flight status data of the aircraft is acquired in real time. Based on the current flight status data of the aircraft, the nonlinear model prediction controller predicts that the aircraft will pass through the grid points in the three-dimensional perturbation wind field grid map within the next n seconds. Based on the disturbance data of the grid points that are about to be passed, the feedforward compensation of the aircraft's propulsion system is calculated.

2. The adaptive disturbance rejection control method for aircraft used in offshore wind power operation and maintenance according to claim 1, characterized in that, The step of receiving wind field data from the lidar terminal mounted on the aircraft in real time and acquiring wake field data of the wind turbine to be repaired and inspected includes: The lidar terminal emits a laser beam at a preset frequency into a preset fan-shaped area in front to obtain the echo reflected by aerosol particles and obtain point cloud data. By analyzing the displacement vector of the same aerosol in two consecutive frames of point cloud data, wind field data in the vicinity of the aircraft is obtained. The wind field data includes the instantaneous three-dimensional wind field observation field. When the position of the wind turbine tower and blades is identified through point cloud data, the relative position of the aircraft and the wind turbine is determined. When the aircraft is within the preset area, the model parameters of the wind turbine to be repaired and inspected are identified. The wake effect parameters of the fan are matched based on the model parameters.

3. The adaptive disturbance rejection control method for aircraft used in offshore wind power operation and maintenance according to claim 2, characterized in that, The step of fusing wind field data and wake field data and projecting them into a three-dimensional grid space in the aircraft's body coordinate system to generate a three-dimensional perturbation wind field grid map, wherein the grid points in the three-dimensional perturbation wind field grid map are bound to perturbation data, includes: A three-dimensional regular grid is preset, and the weather forecast wind field is used as the background field; The three-dimensional wind field observation field is used as the observation value, and the confidence weight of the three-dimensional wind field observation field is used to update the wind field in the background field. When the aircraft is located within the preset area, the wake effect parameters are used as a strong constraint to make secondary corrections to the wind field within the preset area. Generate a three-dimensional perturbation wind field grid with the current position of the aircraft as the origin and a fixed size. Each grid point stores a three-dimensional wind vector.

4. The adaptive disturbance rejection control method for aircraft used in offshore wind power operation and maintenance according to claim 1, characterized in that, The step of inputting the three-dimensional perturbation wind field grid map into a preset nonlinear model prediction controller and acquiring the aircraft's flight status data in real time, and the nonlinear model prediction controller predicting that the aircraft will pass through the grid points in the three-dimensional perturbation wind field grid map within the next n seconds based on the aircraft's current flight status data, includes the following steps: Acquire the current flight status data of the aircraft, including the current state vector and the aircraft's dynamic parameters; The six-degree-of-freedom rigid body dynamics equations of the aircraft are established as follows: ẋ=f(x, u, w), Where u is the control input vector, including the square of the rotational speed of each rotor or the thrust, w is the external disturbance, including the wind speed, and x is the state vector; Decouple airspeed, which is the speed of an aircraft relative to the air, from ground speed, which is the speed relative to the ground. The relationship is: v_air = v_ground - v_wind, where v_ground is the ground speed and v_wind is the airspeed. In each round of prediction, for each future time t+k*Δt and prediction location p_pred(t+kΔt) in the prediction time domain, the wind speed value at that location is queried from the three-dimensional perturbation wind field grid map; Discretize the six-degree-of-freedom rigid body dynamics equation ẋ= f(x, u, w) into a difference equation: x(k+1) = F[x(k), u(k), w(k)], where K is the acquisition time. The difference equation is used to predict the state trajectory of the spacecraft at a series of future time points.

5. The adaptive disturbance rejection control method for aircraft used in offshore wind power operation and maintenance according to claim 4, characterized in that, The step of calculating the feedforward compensation amount of the aircraft propulsion system based on the disturbance data of the grid points to be passed also includes: The nonlinear model predictive controller calculates the deviation between the predicted trajectory and the desired trajectory of the aircraft at future moments based on a preset cost function. The cost function includes the calculation of trajectory tracking cost, control cost, control change rate cost, and terminal cost. The desired trajectory is obtained by collecting pre-planned route points. Based on the current state, construct a nonlinear programming problem, namely, the optimal control sequence that minimizes the cost function under the premise of satisfying the difference equation and the preset constraints. Execute the first control input of the optimal control sequence and output the feedforward compensation input.

6. The adaptive disturbance rejection control method for aircraft used in offshore wind power operation and maintenance according to claim 1, characterized in that, After the step of calculating the feedforward compensation amount of the aircraft propulsion system based on the disturbance data of the grid points to be passed, the following steps are also performed: Acquire the position and attitude data of the aircraft after feedforward control compensation, and perform real-time error calculation with the expected position and attitude data in the expected trajectory; The error calculation results are input into the parallel PID controller. The PID controller calculates the feedback control quantity based on the proportional, integral and derivative coefficients and sends it to the aircraft power system to eliminate real-time errors.

7. The adaptive disturbance rejection control method and system for aircraft used in offshore wind power operation and maintenance according to claim 1, characterized in that, The process of sending feedforward compensation and feedback control quantities to the aircraft's propulsion system also includes the following steps: Obtain the current maximum available power of each motor in the power system; The upper limit of available power is used as a constraint on the current feedforward compensation or feedback control. When the feedforward compensation or feedback control of the motor exceeds the current upper limit of available power, the excess feedforward compensation or feedback control is transferred to the other available motors. If the power of each motor in the current system is close to saturation and still cannot meet the requirements of feedforward compensation or feedback control, the requirements for attitude control accuracy will be automatically reduced.

8. An adaptive disturbance rejection control system for aircraft used in offshore wind power operation and maintenance, characterized in that, include: The data acquisition module is used to receive wind field data emitted by the lidar terminal on board the aircraft in real time, and to acquire wake field data of the wind turbine to be repaired and inspected. The wind field grid map module is used to fuse wind field data and wake field data and project them into a three-dimensional grid space under the aircraft's body coordinate system to generate a three-dimensional perturbation wind field grid map. The grid points in the three-dimensional perturbation wind field grid map are bound to perturbation data. The prediction module is used to input the three-dimensional disturbance wind field grid map into the preset nonlinear model prediction controller and acquire the flight status data of the aircraft in real time. Based on the current flight status data of the aircraft, the nonlinear model prediction controller predicts that the aircraft will pass through the grid points in the three-dimensional disturbance wind field grid map within the next n seconds. The feedforward compensation module is used to calculate the feedforward compensation amount of the aircraft's propulsion system based on the disturbance data of the grid points that are about to be passed.

9. 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 steps of the adaptive anti-disturbance control method for aircraft used in offshore wind power operation and maintenance as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the adaptive anti-disturbance control method for aircraft used in offshore wind power operation and maintenance as described in any one of claims 1 to 7.

Images