Methods, devices, equipment, media, and aircraft for estimating aircraft disturbance parameters

Abstract

This application provides a method, apparatus, device, medium, and aircraft for estimating aircraft disturbance parameters, relating to the field of aircraft control technology. The method includes: acquiring flight state data of the aircraft in the current flight cycle; the flight state data includes flight speed, flight acceleration, and thrust; constructing a residual measurement model based on the flight speed, flight acceleration, and thrust; the residual measurement model is constructed from residual measurements, wind speed, and load changes; and estimating the disturbance in the current cycle using the residual measurement model to obtain estimated wind speed and load change values. This application, by constructing a residual measurement model, transforms the complex coupling relationship between aerodynamic drag and mass change into a linear measurement model concerning wind speed and load changes, avoiding the cumbersome and unstable nature of nonlinear estimation. The estimation of wind speed and load changes is achieved by solving the residual measurement model.

Description

Technical Field

[0001] This application relates to the field of aircraft control technology, and more specifically, to a method, apparatus, device, medium, and aircraft for estimating aircraft disturbance parameters. Background Technology

[0002] When an aircraft flies in a windy field, the relative velocity of the air causes aerodynamic drag, resulting in deviations in the aircraft's speed and acceleration during flight. Additionally, due to factors such as cargo dropping or icing, the aircraft's payload often differs from its nominal mass.

[0003] When factors such as wind speed and load changes cause an aircraft to output a certain thrust, its theoretical acceleration may differ from its actual acceleration, leading to instability during flight. Currently, there is a lack of methods to estimate wind speed and load changes. Summary of the Invention

[0004] The purpose of this application is to provide a method, apparatus, device, medium, and aircraft for estimating aircraft disturbance parameters, so as to estimate the disturbances received by the aircraft during flight and provide a data basis for the stable flight of the aircraft.

[0005] In a first aspect, embodiments of this application provide a method for estimating aircraft disturbance parameters, including: Acquire flight status data of the aircraft during the current flight cycle; the flight status data includes flight speed, flight acceleration, and thrust; Based on flight speed, flight acceleration, and thrust, a residual measurement model is constructed; the residual measurement model is constructed from residual measurements, wind speed, and load changes. The disturbance of the aircraft in the current period is estimated using a residual measurement model to obtain the corresponding disturbance parameter estimates, which include wind speed estimates and load change estimates.

[0006] This application embodiment cleverly transforms the complex coupling relationship between aerodynamic drag and mass change into a linear measurement model of wind speed and load change by constructing a residual measurement model, thereby avoiding the tediousness and instability of nonlinear estimation. Then, by solving the residual measurement model, the estimation of wind speed and load change is realized.

[0007] In one possible implementation of the first aspect, the flight state data further includes air drag parameters and nominal mass; based on flight speed, flight acceleration, and thrust, a residual measurement model is constructed, including: The translational dynamics equations are constructed based on the parameters of flight speed, flight acceleration, thrust, air resistance, and nominal mass. A residual measurement model is constructed based on the translational dynamics equation.

[0008] This application embodiment constructs a translational dynamics equation, and then derives and transforms the translational dynamics equation to obtain a residual measurement model. Since the residual measurement model is a linear combination of wind speed and load changes, it simplifies the solution for wind speed and load changes.

[0009] In one possible implementation of the first aspect, a residual measurement model is constructed based on the translational dynamics equations, including: Substituting the actual mass of the aircraft into the translational dynamics equations yields the true dynamics equations; the actual mass is the sum of the nominal mass and the change in payload. The combination of all known quantities in the true dynamic equation is used to define the residual measurement formula. Based on the real dynamic equations and the definition of residual measurement, a residual measurement model is obtained.

[0010] This application's embodiments introduce load variation into the translational dynamics equation, and combine the known and unknown quantities in the translational dynamics equation separately. The combination of unknown quantities is defined as the residual. Then, the translational dynamics equation is substituted into the definition of the residual to obtain the final residual measurement model. Through the derivation of the above formula, a residual measurement model for the linear combination of wind speed and load variation is obtained, which provides convenience for subsequent solutions to wind speed and load variation.

[0011] In one possible implementation of the first aspect, the disturbance of the aircraft in the current period is estimated using a residual measurement model to obtain the corresponding disturbance parameter estimates, including: Construct the state vector to be estimated based on changes in wind speed and load; Construct a measurement model based on the state vector and the residual measurement model; The measurement model is recursively estimated to obtain wind speed and load change estimates.

[0012] This application embodiment uses the Kalman filter algorithm to recursively estimate the constructed linear time-varying measurement model, which can calculate the estimated values ​​of wind speed and load changes in real time and efficiently, providing a reliable quantitative basis for applications such as flight control compensation and load change alarm.

[0013] In one possible implementation of the first aspect, the measurement model is constructed based on the state vector and the residual measurement model, including: Use the horizontal component of the residual measurement as the observed value; A measurement model representing the relationship between observed values ​​and state vectors is constructed based on the residual measurement model; The measurement model includes an observation matrix, which is a linear time-varying matrix containing horizontal acceleration.

[0014] The observation matrix constructed in this embodiment is time-varying with horizontal acceleration but linear with respect to the state. Therefore, the standard Kalman filter algorithm can be used directly for recursive estimation, thereby stably outputting the estimated values ​​of wind speed and load changes.

[0015] In one possible implementation of the first aspect, a recursive estimation of the measurement model is performed to obtain wind speed estimates and load change estimates, including: A state prediction equation is constructed based on the random walk model. The state prediction equation is used to characterize the relationship between the prior estimate of the current flight cycle and the optimal estimate of the previous flight cycle. Based on the state prediction equation, process noise, and covariance of the previous flight cycle, predict the estimated state vector and estimation error covariance of the current flight cycle. The estimated state vector and the estimated error covariance are updated based on the actual observations at the current moment, the given process model, and the measurement model to obtain the estimated wind speed and the estimated load change.

[0016] In one possible implementation of the first aspect, the estimated state vector and the estimation error covariance are updated based on the actual observations at the current moment, the given process model, and the measurement model to obtain the wind speed estimate and the load change estimate, including: The gain is determined based on the estimated error covariance, the measurement noise covariance matrix, and the residual measurement model. The deviation between the actual and predicted observations is calculated, and the prior estimates in the state prediction equation are corrected using the deviation and gain to obtain the posterior estimate of the current flight cycle. The wind speed estimate and the load change estimate are determined based on the posterior estimation.

[0017] Based on the characteristic that wind speed and load changes do not change significantly instantaneously during flight, the state prediction equation constructed according to the random walk model conforms to the actual flight conditions of the aircraft, thereby improving the accuracy of wind speed and load change estimates.

[0018] In one possible implementation of the first aspect, the method further includes: The observation matrix, gain, and measurement noise covariance matrix are used to update the estimated error covariance.

[0019] This application embodiment uses Kalman filtering to recursively estimate the constructed linear time-varying measurement model, which can not only calculate the estimated values ​​of wind speed and load changes in real time and efficiently, but also output the estimated covariance as a reliability index, providing a reliable quantitative basis for applications such as flight control compensation and load change alarm.

[0020] In one possible implementation of the first aspect, the state vector to be estimated is constructed based on the changes in wind speed and load, including: When the aircraft is found to meet the vertical excitation condition, the state vector to be estimated is constructed based on the horizontal component of the wind speed, the vertical component of the wind speed, and the change in load. Vertical excitation conditions include at least one of the following: The absolute value of the vertical acceleration of the aircraft in the current flight cycle is greater than the first threshold; The estimated covariance of the payload change of the aircraft in the current flight cycle is greater than the second threshold, and the absolute value of the vertical acceleration is greater than the first threshold; The aircraft is in a preset flight mode or flight phase.

[0021] This application embodiment introduces the vertical component of wind speed under the condition of satisfying the vertical excitation condition to accelerate the convergence of load change.

[0022] In one possible implementation of the first aspect, if the horizontal acceleration of the aircraft is less than a first threshold, the step of correcting the prior estimate in the state prediction equation using bias and gain is skipped, and the posterior estimate of the current flight cycle is determined to be equal to the predicted value of the current flight cycle; or, the gain is reduced, and the prior estimate in the state prediction equation is corrected using the reduced gain and bias.

[0023] This application embodiment skips updating or reduces the Kalman gain when the horizontal acceleration of the aircraft is less than a first threshold, in order to avoid mistaking noise for load changes or wind under weak excitation.

[0024] In one possible implementation of the first aspect, the measurement noise covariance matrix is ​​determined based on the aircraft's throttle position, attitude angle, or acceleration amplitude.

[0025] The embodiments of this application adaptively determine the measurement noise covariance matrix based on throttle size, attitude angle, or acceleration amplitude. This enables the Kalman filter to adaptively optimize the trust weights of the measurement data under different flight conditions, thereby improving the estimation accuracy of wind speed and load changes.

[0026] Secondly, embodiments of this application provide an apparatus for estimating aircraft disturbance parameters, comprising: The data acquisition module is used to acquire the flight status data of the aircraft in the current flight cycle; the flight status data includes flight speed, flight acceleration and thrust; The residual measurement model construction module is used to construct a residual measurement model based on flight speed, flight acceleration, and thrust; wherein, the residual measurement model is constructed from residual measurements, wind speed, and load changes; The estimation module is used to estimate the disturbance of the aircraft in the current period using the residual measurement model, and obtain the corresponding disturbance parameter estimates; the disturbance parameter estimates include wind speed estimates and load change estimates.

[0027] Thirdly, embodiments of this application provide an aircraft, including: a processor, a memory, and a bus, wherein: The processor and memory communicate with each other via a bus; The memory stores program instructions that can be executed by the processor, and the processor can execute the method of the first aspect by calling the program instructions.

[0028] Fourthly, embodiments of this application provide a non-transitory computer-readable storage medium, comprising: A non-transitory computer-readable storage medium stores computer instructions that cause the computer to perform the methods in the various possible implementations of the first aspect.

[0029] Fifthly, embodiments of this application provide a computer program product, including computer program instructions, which, when read and executed by a processor, perform the methods in various possible implementations of the first aspect.

[0030] Other features and advantages of this application will be set forth in the following description and will be apparent in part from the description or may be learned by practicing embodiments of this application. The objectives and other advantages of this application may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings. Attached Figure Description

[0031] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 A schematic flowchart of a method for joint estimation of wind speed and load provided in an embodiment of this application; Figure 2 This is a schematic flowchart of a Kalman filter algorithm solution method provided in an embodiment of this application; Figure 3 This application provides a schematic diagram of vertical enhancement enabling logic for its embodiments; Figure 4 A schematic diagram of a wind speed and load joint estimation device provided in this application embodiment; Figure 5This is a schematic diagram of the physical structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0033] The embodiments of the technical solution of this application will now be described in detail with reference to the accompanying drawings. These embodiments are only used to more clearly illustrate the technical solution of this application and are therefore merely examples, and should not be used to limit the scope of protection of this application.

[0034] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit this application; the terms “comprising” and “having”, and any variations thereof, in the specification and the foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0035] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0036] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0037] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0038] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0039] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0040] When an aircraft flies outdoors, its flight status is inevitably affected by both environmental wind field changes and its own load variations. On the one hand, ambient wind speed causes a relative velocity between the aircraft and the air, which in turn introduces aerodynamic drag, causing the aircraft's speed and acceleration responses to deviate from the expected values ​​under windless conditions. On the other hand, when performing certain tasks, such as pesticide spraying, cargo dropping, and material handling, the aircraft's mass can change abruptly, leading to a mismatch in the mapping relationship between thrust and acceleration. If changes in wind speed and load cannot be accurately estimated, the accuracy and stability of flight control will be directly affected.

[0041] Currently, the main methods for estimating changes in wind speed or load include the following: 1. Wind speed can be directly measured by installing sensors such as pitot tubes, ultrasonic anemometers, or pitot tubes on the aircraft, or by detecting load changes using load cells. However, this approach increases additional hardware costs and system integration complexity. Furthermore, for multi-rotor aircraft, measurement accuracy is often difficult to guarantee due to factors such as rotor downwash and installation location, and the sensors themselves are also subject to failure risks.

[0042] 2. Estimating wind speed or load changes in isolation, for example, by establishing a drag model and using the deviation between flight speed and thrust to infer wind speed; or by analyzing the ratio of thrust to acceleration to estimate mass change. However, when only wind speed is estimated, the thrust and acceleration deviations caused by load changes are incorrectly attributed to wind resistance, leading to a drift in the wind speed estimate; when only load is estimated, the velocity response deviation caused by wind resistance is misjudged as a mass change. Because the effects of wind and load are coupled in the dynamic equations, single-variable estimation methods cannot fundamentally solve the problem of distinguishing between the two.

[0043] To address the aforementioned issues, this application provides a method for joint estimation of wind speed and load. This method constructs a measurement equation that decouples the effects of wind and load without adding additional hardware. By solving this measurement equation, the estimated values ​​of wind speed and load change can be obtained.

[0044] It should be noted that the aircraft described in the embodiments of this application may be a multi-rotor drone (such as a quadcopter or hexacopter drone), a fixed-wing drone, a vertical take-off and landing compound wing drone, and a small manned aircraft / helicopter, etc.

[0045] Furthermore, the entity executing the disturbance estimation method for aircraft provided in this application can be the aircraft's flight control system, or it can be a separate computational unit for disturbance estimation, which is communicatively connected to the flight control system. For ease of description, this application uses a flight control system as an example.

[0046] It should be noted that the embodiments in this application use the Earth-fixed coordinate system. The aircraft's flight speed, acceleration, and thrust are all within... express.

[0047] Figure 1 This is a schematic flowchart illustrating a method for joint estimation of wind speed and load, provided in an embodiment of this application. The method includes: Step 101: Obtain the flight status data of the aircraft in the current flight cycle; wherein, the flight status data includes flight speed, flight acceleration and thrust.

[0048] Since the flight control system is a real-time discrete system, it cannot process data continuously. It can only perform calculations once every fixed time interval, which is called the flight cycle. The current flight cycle refers to the flight cycle at the current moment.

[0049] Flight speed refers to the three-dimensional linear velocity vector of an aircraft in a ground-fixed coordinate system, which can be expressed as: ,in, Let be the northbound velocity during the k-th flight cycle. Let be the eastward velocity during the k-th flight cycle. This represents the ground velocity (positive downwards) during the k-th flight cycle. Flight velocity describes the speed and direction of the aircraft's motion relative to the ground. It can be obtained using GNSS, visual odometry, or a combined navigation system that integrates data from multiple sources such as IMU, GNSS, and magnetic data.

[0050] Thrust refers to the total thrust vector of an aircraft rotor in the geodetic coordinate system, which can be expressed as: , For the northward thrust during the k-th flight cycle, For the eastward thrust during the k-th flight cycle, Let be the vertical thrust during the k-th flight cycle. This thrust is the primary power source for the aircraft to overcome gravity and air resistance and generate acceleration. In this embodiment, the thrust can be calculated using a thrust model, and the calculation method is as follows: Based on the current throttle command, motor speed or power, the total thrust is calculated using a pre-calibrated thrust model. Then, using the aircraft's current attitude (including roll angle, pitch angle, and yaw angle), the thrust in the airframe coordinate system is rotated to the ground-fixed coordinate system to obtain the thrust required for the implementation of this application.

[0051] Flight acceleration refers to the linear acceleration of an aircraft in a ground-fixed coordinate system, specifically the pure motion acceleration after removing the gravitational component. It represents the rate of change of the aircraft's velocity, i.e., the change in velocity per unit time. Flight acceleration can be measured and processed by an inertial measurement unit (IMU) within the aircraft. For example, the accelerometer in the IMU can measure the vector sum of motion acceleration and gravitational acceleration, based on the aircraft's coordinate system. Therefore, after acquiring the IMU data, the aircraft's current attitude can be used to rotate the IMU data to a ground-fixed coordinate system, and then the gravity vector can be subtracted to obtain the flight acceleration.

[0052] Step 102: Construct a residual measurement model based on flight speed, flight acceleration, and thrust.

[0053] In the specific implementation process, based on the translational dynamics of the aircraft in the Earth-fixed coordinate system, a force balance equation including thrust, inertial force, and air resistance is established. The actual mass is expressed as the sum of the nominal mass and the load deviation to be estimated. Among them, the thrust, the inertial force generated by the nominal mass, and the drag calculated based on the current speed are known quantities. All known quantities are moved to one side of the equation to define a residual measurement vector that can be calculated in real time.

[0054] By substituting and simplifying, it can be found that the constructed residual measurement model is physically equivalent to a linear combination of the unknown quantities to be estimated (i.e., load deviation and wind speed) and the known quantities (i.e., flight acceleration and air drag coefficient). This ingenious construction successfully isolates the wind speed and load information that were originally coupled to the thrust and acceleration response, forming a simple and directly measurable linear equation, laying the theoretical foundation for subsequent joint estimation.

[0055] The following describes the process of deriving the residual measurement model based on the translational dynamics equations: Flight status data can also include air resistance parameters and nominal mass. Based on translational dynamics, the net force received by the aircraft is equal to its mass multiplied by its acceleration. That is: .

[0056] The resultant force here includes the aircraft's thrust F and air resistance. , For the mass of the aircraft, The acceleration of the aircraft.

[0057] Air drag is typically related to the relative speed between the aircraft and the air. Assuming air drag is linear (first-order model) and the wind speed is... The flight speed of the aircraft is Then the speed of the aircraft relative to the air is Therefore, air resistance .in, These are air resistance parameters, describing the proportional relationship between the drag experienced by the UAV and its flight speed. Essentially, they are parameters of a first-order linear drag model. , This is the northbound drag coefficient. This is the eastward drag coefficient. This is the vertical drag coefficient. In this embodiment, since the wind speed is assumed to be horizontal, it can be... Simplification involves assuming the aircraft's drag characteristics are the same in the north and east directions, using a scalar. Description, that is, This can be understood as the drag coefficients in the north and east directions. It can be obtained through wind tunnel testing, CFD simulation, or flight data identification. It is a 2×2 identity matrix. Air resistance parameters can also be presented as a diagonal matrix, in the form of: .

[0058] Substituting air resistance and the actual mass of the aircraft into the translational dynamics equations above, we can obtain the actual dynamics equations, as follows: .

[0059] in, The nominal mass of an aircraft is a pre-defined engineering parameter that represents the reference mass of the aircraft during flight. It is not necessarily equal to the actual mass of the aircraft. The methods for obtaining nominal mass include: (1) directly taking a fixed value, such as the empty mass of an aircraft; (2) estimating the current mass using a payload estimation algorithm in the flight control system and updating it dynamically. For example, it can be determined based on Kalman filtering, using accelerometer data during takeoff or the relationship between thrust and acceleration during flight. Nominal mass can be obtained using... express. Let r be the change in load. Moving all terms containing known quantities to one side of the equation above, we define the combination of known quantities as the residual measurement vector, denoted by r, i.e.: .

[0060] The above formula is the definition of residual measurement. Since the residual measurement vector is a combination of known quantities, r is a quantity that can be calculated in real time.

[0061] Next, the dynamic equations Substituting into the definition of residual measurement, we can obtain: .

[0062] After simplification, and The results are canceled out, thus yielding the residual measurement model: .

[0063] As can be seen from the residual measurement model, residual measurement involves measuring the thrust F and nominal mass of the spacecraft in the geodetic coordinate system. Corresponding inertial force and known air resistance By combining these parameters, a vector capable of real-time computation is constructed. That is, the residual measurement is exactly equal to a linear combination of the unknowns (load change and wind speed) and the knowns (flight acceleration and air resistance parameters). Here, the load change is the actual total mass of the aircraft. With nominal quality The difference between them is a quantity to be estimated in this application embodiment. Faced with this residual measurement model, this application embodiment transforms the problem from a difficult-to-describe, coupled nonlinear estimation problem into a clear, solvable linear problem.

[0064] This application's embodiments introduce load variation into the translational dynamics equation, and combine the known and unknown quantities in the translational dynamics equation separately. The combination of unknown quantities is defined as the residual. Then, the translational dynamics equation is substituted into the definition of the residual to obtain the final residual measurement model. Through the derivation of the above formula, a residual measurement model for the linear combination of wind speed and load variation is obtained, which provides convenience for subsequent solutions to wind speed and load variation.

[0065] Step 103: Use the residual measurement model to jointly estimate the wind speed and load changes of the aircraft in the current cycle to obtain the corresponding disturbance parameter estimates.

[0066] After obtaining the residual measurement model, it can be solved. For example, the Kalman filter algorithm can be used to jointly estimate the wind speed and load changes in the current cycle, or the least squares method, machine learning, and other methods can be used for joint estimation to obtain disturbance parameter estimates. These disturbance parameter estimates can include wind speed estimates and load change estimates. The wind speed estimates include horizontal wind speed estimates and may also include vertical wind speed estimates.

[0067] Taking the Kalman filter algorithm as an example, the algorithm can be divided into a prediction stage and an update stage. In the prediction stage, a state prediction equation is constructed based on a random walk model. The random walk model can be used to describe the changes in disturbance parameters. In this embodiment, the aircraft is considered to have slow changes in disturbance parameters (wind speed and load) during flight. Therefore, in the constructed state prediction equation, the state at the next moment is equal to the state at the current moment plus a random disturbance. The magnitude of the random disturbance is controlled by the covariance matrix Q. The larger Q is, the faster the state changes.

[0068] The estimation error covariance of the current flight cycle is predicted based on the process noise and the covariance of the previous flight cycle, where the estimation error covariance of the current flight cycle is equal to the sum of the covariance of the previous flight cycle and the process noise.

[0069] During the update phase, residuals and Kalman gains are calculated based on the actual observations at the current moment, the given process model, and the measurement model. These residuals and Kalman gains are then used for state and covariance updates. The residuals are calculated by subtracting the predicted observations from the actual observations, representing the difference between the prediction and the actual data. The Kalman gain is a weight used to determine which is more reliable, the prediction model or the observation data. The prediction model is used to predict the estimation error covariance of the current flight cycle, and the observation data is the measurement noise covariance matrix. After obtaining the residuals and Kalman gains, the product of the residuals and the Kalman gain can be used to correct the prior estimate, yielding the posterior estimate, which is the optimal estimate of the disturbance parameters at the current moment.

[0070] Next, the residual measurement model is solved based on the Kalman filter algorithm. Figure 2 A schematic flowchart of a Kalman filter algorithm solution method provided in this application embodiment is shown below: Step 201: Construct the state vector to be estimated based on the changes in wind speed and load.

[0071] The horizontal component of wind speed includes the first directional component. Second direction component The change in load is Where k represents the k-th flight cycle, which is the current flight cycle. Therefore, the constructed state vector to be estimated can be expressed as: .

[0072] Step 202: Construct the measurement model based on the state vector and the residual measurement model.

[0073] The horizontal component of the residual measurement is taken as the observed value of the filter, and the horizontal drag coefficient is assumed to be isotropic. , It is a 2×2 identity matrix. We can obtain: .

[0074] according to This allows us to obtain the observation value during the k-th flight cycle. The state vector expected to be estimated in the k-th flight cycle The relationship between them:

[0075] That is, the measurement model is: .

[0076] in, The horizontal component of the residual measurement of the aircraft during the k-th flight cycle. Let be the horizontal thrust of the aircraft during the k-th flight cycle. Let be the acceleration of the aircraft in the first direction of the horizontal component during the k-th flight cycle. Let be the acceleration of the aircraft in the second direction of the horizontal component during the k-th flight cycle. Let be the measurement noise during the k-th flight cycle, and assume... It follows a Gaussian distribution with zero mean, i.e. , To measure the noise covariance matrix, These are the drag coefficients for northward and eastward directions. Let be the horizontal velocity of the aircraft during the k-th flight cycle. This is the observation matrix for the k-th flight cycle. Since this matrix includes the horizontal acceleration for the current flight cycle, although... The measurement model varies over time, but it is linear with respect to the state vector; therefore, a standard Kalman filter algorithm can be used for recursive estimation. It should be noted that the subscript k in the above formula indicates the k-th flight cycle.

[0077] Step 203: Use the Kalman filter algorithm to recursively estimate the measurement model and obtain the estimated values ​​of wind speed and load change.

[0078] In the filtering timescale, wind speed and load are considered to change slowly. Therefore, this application embodiment uses a random walk model to describe the variation of wind speed and load, which can be expressed as follows: .

[0079] in, Let this be the state vector for the (k+1)th flight cycle. Let be the state vector for the k-th flight cycle. For process noise, This indicates that the distribution follows a zero-mean Gaussian distribution. The formula above shows that the state at the next time step is equal to the state at the current time step plus a random perturbation. The covariance matrix Q of this perturbation describes the severity of the state change, and is typically assumed to be... ,in, To control for the rate of wind change, To control the rate of load change, both parameters are assumed to be adjustable according to actual conditions.

[0080] The Kalman filter algorithm can be divided into a prediction step and an update step, where the prediction step is as follows: Based on the given process model and measurement model , Given the identity matrix, the recursive process for each flight cycle k is as follows: According to the random walk model, the predicted value for the current flight cycle is equal to the optimal estimate for the previous flight cycle. Therefore, the state prediction equation can be expressed as: . This is a predicted value for the current flight cycle. This is the optimal estimate for the previous flight cycle. It should be noted that, in this embodiment, the k-th flight cycle is taken as the current flight cycle.

[0081] Error covariance is used to characterize the uncertainty of the prediction. It can be calculated based on the process noise and the covariance of the previous flight cycle. The specific formula is as follows: .in, This is the estimated error covariance for the current flight cycle, which is the predicted value. The error covariance of the previous flight cycle, This is the process noise matrix.

[0082] After completing the above prediction steps, proceed to the update step, as follows: The gain of the Kalman filter algorithm is a dynamic weight that determines whether to place more weight on predictions or measurements during fusion. The gain can be determined based on the estimation error covariance, the measurement noise covariance matrix, and the residual measurement model. The formula for calculating the gain is as follows: .in, The gain for the k-th flight cycle, Let T be the observation matrix for the k-th flight cycle, and let T denote the matrix transpose. This is the measurement noise covariance matrix.

[0083] After determining the gain, the actual observations were used. Compared with predicted observations The deviation between the two values ​​is multiplied by the gain to correct the predicted value for the current flight cycle, thus obtaining the posterior estimate of the current flight cycle. That is, the optimal estimate, which is calculated using the following formula: .

[0084] After obtaining the posterior estimate After that, you can... Extract wind speed estimates and load change estimates. The wind speed estimate is... The estimated value of the load change is .

[0085] Based on the above embodiments, the estimated error covariance can also be updated according to the observation matrix, gain, and measurement noise covariance matrix. The specific implementation formula is as follows: ,in, Let T be the gain for the k-th flight cycle, and let T denote the matrix transpose.

[0086] Because the observation matrix is ​​incorporated into the above update formula, the updated error covariance is smaller than that before the update. The embodiments in this application use the Josephus form, which exhibits good numerical stability.

[0087] This application embodiment uses Kalman filtering to recursively estimate the constructed linear time-varying measurement model, which can not only calculate the estimated values ​​of wind speed and load changes in real time and efficiently, but also output the estimated covariance as a reliability index, providing a reliable quantitative basis for applications such as flight control compensation and load change alarm.

[0088] In the above embodiments, the state vector includes a horizontal wind speed component and a load change, and horizontal residual measurement is used. This method of updating the data yields good estimation results in most flight scenarios (e.g., horizontal cruise, turning, etc.). However, when the UAV performs vertical maneuvers (e.g., rapid climb, rapid descent, recovery after dropping, etc.), the vertical acceleration increases significantly, which is related to physical relationships. ,in, For vertical residual measurement, For vertical acceleration, This is the vertical drag coefficient. Vertical wind speed, Let be the vertical velocity. It can be seen that the vertical residual contains a large amount of information about load changes. If this information can be fully utilized, it can not only accelerate the convergence of load estimation, but also estimate the vertical wind speed under certain conditions, providing more comprehensive wind field information for flight control.

[0089] However, since the vertical channel is more sensitive to attitude errors, gravity compensation errors, and thrust model errors, forcibly introducing vertical estimation when vertical excitation is insufficient may introduce noise and model errors into the filtering process, leading to a decrease in estimation accuracy. Therefore, in this embodiment, the flight control system monitors the flight status of the aircraft in each flight cycle and switches to enhanced estimation mode when the vertical excitation conditions are met. For example, a third measurement can be introduced, namely: . The vertical observation value for the k-th flight cycle. For the vertical residual measurement of the k-th flight cycle, Let be the vertical component of the thrust during the k-th flight cycle. Let be the vertical acceleration during the k-th flight cycle. This is the vertical drag coefficient. Let be the vertical velocity in the k-th flight cycle. Therefore, the state vector to be estimated can be constructed based on the horizontal component of the wind speed, the vertical component of the wind speed, and the load change. The constructed state vector to be estimated in the k-th flight cycle is: .

[0090] in, Let be the northward acceleration during the k-th flight cycle. Let be the eastward acceleration during the k-th flight cycle. The vertical excitation conditions include at least one of the following: (1) The absolute value of the vertical acceleration of the aircraft in the current flight cycle is greater than the first threshold. When the UAV has significant vertical acceleration, the inertial effect in the vertical direction is significantly enhanced. At this time, the contribution of load change to the vertical residual is... Dominant in its position with a high signal-to-noise ratio, vertical estimation can effectively extract information. Conversely, when the aircraft is hovering or in a constant-speed ascent / descent state, i.e. If the load information in the vertical residual is relatively small, the enhancement mode can be left unused.

[0091] (2) The estimated covariance of the payload change of the aircraft in the current flight cycle is greater than the second threshold, and the absolute value of the vertical acceleration is greater than the first threshold. When the filter lacks confidence in the payload estimation (large covariance), it indicates that more observation information is needed to compress the uncertainty. At this time, if the presence of vertical excitation is detected at the same time, vertical residuals are actively introduced to accelerate the convergence of payload estimation by using the information brought by vertical maneuvering. It can automatically use available vertical excitations for calibration when the uncertainty increases.

[0092] (3) The aircraft is in a preset flight mode or flight phase. If it belongs to a preset high vertical information scenario, the enhanced mode is activated. The preset flight modes or flight phases include: take-off / landing phase, drop / release mission in progress, vertical climb / descent mode, etc.

[0093] This application embodiment introduces the vertical component of wind speed under the condition of satisfying the vertical excitation condition to accelerate the convergence of load change.

[0094] Figure 3 This application provides a schematic diagram of vertical enhancement enabling logic, including: Step 301: Obtain the flight status of the aircraft; the flight status includes vertical acceleration, flight mode, flight phase, etc.

[0095] Step 302: Determine whether the vertical excitation condition is met; the determination of whether the vertical excitation condition is met can be found in the above embodiments, and will not be repeated here.

[0096] Step 303: Generate a state vector containing the vertical component of wind speed; the state vector contains the horizontal component of wind speed, the vertical component of wind speed, and the load change.

[0097] Step 304: Use the state vector containing the vertical component of wind speed for subsequent solutions; Step 305: Disable the vertical component and construct a state vector using only the horizontal component of the wind speed and the load change; the state vector includes the horizontal component of the wind speed and the load change.

[0098] Step 306: Construct a state vector using the horizontal component of wind speed and the load change for subsequent solutions.

[0099] Based on the above embodiments, when the horizontal acceleration amplitude of the aircraft is less than a preset threshold, it indicates that the UAV is in a weak maneuvering state (such as hovering, constant speed cruise, or low-speed flight) during the current flight cycle. In this state, the contribution of load changes to the observations is close to zero. At this time, the observations almost only reflect wind speed information and cannot effectively distinguish load changes. If Kalman filtering updates continue, the filter will attempt to correct the load estimate using these observations that contain almost no load information, which may easily misjudge the observation noise as load changes, leading to unnecessary fluctuations or even divergence in the estimated load change. Therefore, one strategy is that the flight control system can control the update step according to the horizontal acceleration amplitude. For example, when the horizontal acceleration of the aircraft is less than a first threshold, the step of correcting the prior estimate in the state prediction equation using bias and gain is skipped, and the posterior estimate of the current flight cycle is equal to the predicted value of the current flight cycle. In this mode, the flight control system only performs the prediction step of Kalman filtering, while skipping steps such as gain calculation, state correction, and covariance update.

[0100] Another strategy is to reduce the gain of the Kalman filter. In this mode, the flight control system still performs the complete prediction-update process, but the Kalman gain is forced through dynamic adjustment. This can be significantly reduced, for example, by temporarily increasing the measurement noise covariance R or by directly multiplying the calculated gain by an attenuation factor less than 1. The reduced gain and bias are then used to correct the prior estimates in the state prediction equation.

[0101] Building upon the above embodiments, in Kalman filtering, the measurement noise covariance matrix R is used to quantify the reliability of the observations. A larger R indicates greater noise and lower reliability in the observations, resulting in the filter assigning less weight to the observations during updates. Conversely, a smaller R indicates more accurate observations, allowing the filter to correct state estimates. During actual flight, sensor noise characteristics and model errors are not constant but dynamically change with the flight state. Therefore, dynamically adjusting the R matrix based on real-time flight state parameters such as throttle position, attitude angle, or acceleration amplitude allows the filter to maintain optimal estimation performance throughout the entire flight envelope.

[0102] Throttle position directly reflects the rotor thrust output level, thus affecting the accuracy of the thrust model. At low throttle, the thrust value is small, and the nonlinear characteristics in the thrust model (such as the dead zone at low throttle and motor response delay) and model errors are relatively significant, leading to lower thrust values ​​calculated by the thrust model. The accuracy decreases, which in turn affects the residual measurement. Noise increases. In this case, the corresponding elements of the R matrix can be increased to reduce the filter's confidence in the current observations and avoid model errors contaminating the state estimation. At high throttle, the thrust output is large, the signal-to-noise ratio is high, and the relative proportion of model error decreases. At this point, R should be appropriately reduced to allow the filter to quickly respond to real changes in wind speed and load. In practice, a mapping relationship between throttle and thrust model error can be established beforehand through calibration experiments. The throttle value can be divided into several intervals, each interval corresponding to a set of R values, or a continuous R-throttle curve can be fitted to achieve smooth adjustment.

[0103] The magnitude of the attitude angles (roll angle, pitch angle) determines the accuracy of the thrust vector projection in the geodetic coordinate system, directly affecting... and acceleration The accuracy of the calculation is crucial. When the UAV is maneuvering at large attitude angles, even small errors in attitude estimation can lead to significant deviations in thrust projection and gravity compensation. For example, when the roll angle is large, the calculation of the horizontal thrust component is more sensitive to attitude errors. In this case, the measurement noise covariance should be increased. This reflects the coupled attitude error uncertainty in the observations. Conversely, at small attitude angles (near level flight), the attitude error has a smaller impact on the projection, the reliability of the observations is higher, and a smaller [value] can be used. In addition, the rate of change of attitude angles can also be used as an auxiliary indicator: when the attitude changes rapidly, the delay and error of attitude estimation usually increase, and should be increased accordingly. .

[0104] The magnitude of acceleration directly determines the load variation in residual measurements. The amount of information. From the measurement equation. It can be seen that when the acceleration amplitude is small, the contribution of load change to the observed values ​​is weak, and the noise proportion in the observed values ​​is relatively high. In this case, the load should be increased. To avoid the filter misinterpreting noise as a change in load. When the acceleration amplitude is large, the signal energy is strong, the load information is significant, and the signal-to-noise ratio of the observed value is high; therefore, the speed should be reduced. To accelerate the filter's response speed and enable rapid convergence of load estimation, an acceleration threshold can be set. Expressed as a piecewise function or exponentially decaying function of acceleration amplitude: when hour, Take the larger value; when hour, Take the smaller value; use linear interpolation or smooth transition in the middle area. and Set specific values ​​as needed.

[0105] The embodiments of this application adaptively determine the measurement noise covariance matrix based on throttle size, attitude angle, or acceleration amplitude. This enables the Kalman filter to adaptively optimize the trust weights of the measurement data under different flight conditions, thereby improving the estimation accuracy of wind speed and load changes.

[0106] The following embodiment of this application provides an estimation of wind speed and load changes using the sliding window least squares method: A historical data window of fixed length (N) is pre-defined. and the corresponding observation matrix At each time step, a standard least squares solution is performed on all data within this window. The calculation formula is as follows: .

[0107] in, Stack the observation matrices for all times within the window; that is, stack the observation matrices for N times within the sliding window. The matrix is ​​formed by stacking rows and describes the linear mapping between all observations and states within the window, where T is the matrix transpose. This is a stack of all observations within a window, i.e., residual measurements over N time steps within a sliding window. A vector formed by stacking rows.

[0108] In another embodiment, machine learning methods can also be used to estimate changes in wind speed and load, as follows: The problem of estimating wind speed and load changes can be viewed as a supervised learning task. A large amount of input data and corresponding real output data are collected. The input data includes parameters such as thrust, flight acceleration, flight speed, nominal mass, and air resistance. The real output data includes specific wind speed and load changes, which can be obtained through high-precision sensors or simulation. A neural network is trained using the input data and real output data to learn the mapping from input to output. The network structure can be a fully connected network (MLP) or a recurrent neural network (LSTM), etc.

[0109] Figure 4 This is a schematic diagram of an aircraft disturbance parameter estimation device provided in an embodiment of this application. The device can be a module, program segment, or code on an electronic device. It should be understood that this device is similar to the one described above. Figure 1 The method implementation corresponds to this and can be executed. Figure 1 The specific functions of the device involved in the various steps of the method embodiment can be found in the description above; to avoid repetition, detailed descriptions are omitted here. The device includes: a data acquisition module 401, a residual measurement model construction module 402, and an estimation module 403, wherein: The data acquisition module is used to acquire the flight status data of the aircraft in the current flight cycle; the flight status data includes flight speed, flight acceleration and thrust; The residual measurement model construction module is used to construct a residual measurement model based on flight speed, flight acceleration, and thrust; wherein, the residual measurement model is constructed from residual measurements, wind speed, and load changes; The estimation module is used to estimate the disturbance of the aircraft in the current period using the residual measurement model, and obtain the corresponding disturbance parameter estimates; the disturbance parameter estimates include wind speed estimates and load change estimates.

[0110] Based on the above embodiments, the flight status data also includes air resistance parameters and nominal mass; the residual measurement model construction module is specifically used for: The translational dynamics equations are constructed based on the flight speed, flight acceleration, thrust, air resistance parameters, and nominal mass. The residual measurement model is constructed based on the translational dynamics equation.

[0111] Based on the above embodiments, the residual measurement model construction module is specifically used for: Substituting the actual mass of the aircraft into the translational dynamics equation, the true dynamics equation is obtained; the actual mass is the sum of the nominal mass and the load change. The combination of all known quantities in the actual dynamic equation is determined as the definition of residual measurement; Based on the true dynamic equation and the definition of residual measurement, the residual measurement model is obtained.

[0112] Based on the above embodiments, the estimation module is specifically used for: Construct a state vector to be estimated based on the wind speed and the load change; Construct a measurement model based on the state vector and the residual measurement model; The measurement model is recursively estimated to obtain the wind speed estimate and the load change estimate.

[0113] Based on the above embodiments, the estimation module is specifically used for: The horizontal component of the residual measurement is used as the observed value; Based on the residual measurement model, a measurement model characterizing the relationship between the observed values ​​and the state vector is constructed; The measurement model includes an observation matrix, which is a linear time-varying matrix containing horizontal acceleration.

[0114] Based on the above embodiments, the estimation module is specifically used for: A state prediction equation is constructed based on a random walk model; the state prediction equation is used to characterize the relationship between the prior estimate of the current flight cycle and the optimal estimate of the previous flight cycle. Based on the state prediction equation, process noise, and covariance of the previous flight cycle, predict the estimated state vector and estimation error covariance of the current flight cycle. The estimated state vector and the estimated error covariance are updated based on the actual observations at the current moment, the given process model, and the measurement model to obtain the estimated wind speed and the estimated load change.

[0115] Based on the above embodiments, the estimation module is specifically used for: The gain is determined based on the estimated error covariance, the measurement noise covariance matrix, and the residual measurement model. The deviation between the actual observation and the predicted observation is calculated, and the prior estimate in the state prediction equation is corrected using the deviation and the gain to obtain the posterior estimate of the current flight cycle. The wind speed estimate and the load change estimate are determined based on the posterior estimate.

[0116] Based on the above embodiments, the device further includes an update module, used for: The estimated error covariance is updated based on the observation matrix, the gain, and the measurement noise covariance matrix.

[0117] Based on the above embodiments, the estimation module is specifically used for: When the aircraft is detected to meet the vertical excitation condition, a state vector to be estimated is constructed based on the horizontal component of the wind speed, the vertical component of the wind speed, and the load change. The vertical excitation condition includes at least one of the following: The absolute value of the vertical acceleration of the aircraft in the current flight cycle is greater than a first threshold. The estimated covariance of the payload change of the aircraft in the current flight cycle is greater than the second threshold, and the absolute value of the vertical acceleration is greater than the first threshold. The aircraft is in a preset flight mode or flight phase.

[0118] Based on the above embodiments, if the horizontal acceleration of the aircraft is less than a first threshold, the step of correcting the prior estimate in the state prediction equation using the deviation and the gain is skipped, and the posterior estimate of the current flight cycle is determined to be equal to the predicted value of the current flight cycle; or, the gain is reduced, and the prior estimate in the state prediction equation is corrected using the reduced gain and the deviation.

[0119] Based on the above embodiments, the measurement noise covariance matrix is ​​determined according to the throttle size, attitude angle, or acceleration amplitude of the aircraft.

[0120] Figure 5 This is a schematic diagram of the physical structure of the electronic device provided in the embodiments of this application, such as... Figure 5 As shown, the electronic device includes: a processor 501, a memory 502, and a bus 503; wherein: The processor 501 and the memory 502 communicate with each other through the bus 503; The processor 501 is used to call program instructions in the memory 502 to execute the methods provided in the above-described method embodiments, such as: acquiring flight status data of the aircraft in the current flight cycle; wherein the flight status data includes flight speed, flight acceleration, and thrust; constructing a residual measurement model based on the flight speed, flight acceleration, and thrust; wherein the residual measurement model is constructed from residual measurements, wind speed, and load changes; and using the residual measurement model to estimate the disturbance of the aircraft in the current cycle to obtain the corresponding disturbance parameter estimates.

[0121] Processor 501 can be an integrated circuit chip with signal processing capabilities. The processor 501 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor.

[0122] The memory 502 may include, but is not limited to, random access memory (RAM), read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), etc.

[0123] This embodiment discloses a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions, and when the program instructions are executed by the computer, the computer can execute the methods provided in the above-described method embodiments, such as: acquiring flight status data of the aircraft in the current flight cycle; wherein the flight status data includes flight speed, flight acceleration, and thrust; constructing a residual measurement model based on the flight speed, flight acceleration, and thrust; wherein the residual measurement model is constructed from residual measurements, wind speed, and load changes; and estimating the disturbance of the aircraft in the current cycle using the residual measurement model to obtain the corresponding disturbance parameter estimates.

[0124] This embodiment provides a non-transitory computer-readable storage medium storing computer instructions that cause the computer to execute the methods provided in the above-described method embodiments. For example, the methods include: acquiring flight status data of an aircraft in the current flight cycle; wherein the flight status data includes flight speed, flight acceleration, and thrust; constructing a residual measurement model based on the flight speed, flight acceleration, and thrust; wherein the residual measurement model is constructed from residual measurements, wind speed, and load changes; and estimating the disturbance of the aircraft in the current cycle using the residual measurement model to obtain corresponding disturbance parameter estimates.

[0125] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0126] Furthermore, the units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0127] Furthermore, the functional modules in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0128] In this document, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, without necessarily requiring or implying any such actual relationship or order between these entities or operations.

[0129] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for estimating aircraft disturbance parameters, characterized in that, include: Acquire flight status data of the aircraft during the current flight cycle; the flight status data includes flight speed, flight acceleration, and thrust; Based on the flight speed, the flight acceleration, and the thrust, a residual measurement model is constructed; wherein, the residual measurement model is constructed from residual measurements, wind speed, and load changes; The disturbance of the aircraft in the current period is estimated using the residual measurement model to obtain the corresponding disturbance parameter estimates; the disturbance parameter estimates include wind speed estimates and load change estimates.

2. The method according to claim 1, characterized in that, The flight status data also includes air resistance parameters and nominal mass; the construction of the residual measurement model based on the flight speed, flight acceleration, and thrust includes: The translational dynamics equations are constructed based on the flight speed, flight acceleration, thrust, air resistance parameters, and nominal mass. The residual measurement model is constructed based on the translational dynamics equation.

3. The method according to claim 2, characterized in that, The construction of the residual measurement model based on the translational dynamics equation includes: Substituting the actual mass of the aircraft into the translational dynamics equation, the true dynamics equation is obtained, where the actual mass is the sum of the nominal mass and the load change. The combination of all known quantities in the actual dynamic equation is determined as the definition of residual measurement; Based on the true dynamic equation and the definition of residual measurement, the residual measurement model is obtained.

4. The method according to claim 1, characterized in that, The step of estimating the disturbance of the aircraft in the current period using the residual measurement model to obtain the corresponding disturbance parameter estimates includes: Construct a state vector to be estimated based on the wind speed and the load change; Construct a measurement model based on the state vector and the residual measurement model; The measurement model is recursively estimated to obtain the wind speed estimate and the load change estimate.

5. The method according to claim 4, characterized in that, Constructing a measurement model based on the state vector and the residual measurement model includes: The horizontal component of the residual measurement is used as the observed value; Based on the residual measurement model, a measurement model characterizing the relationship between the observed values ​​and the state vector is constructed; The measurement model includes an observation matrix, which is a linear time-varying matrix containing horizontal acceleration.

6. The method according to claim 4, characterized in that, The step of recursively estimating the measurement model to obtain the estimated wind speed and the estimated load change includes: A state prediction equation is constructed based on a random walk model; the state prediction equation is used to characterize the relationship between the prior estimate of the current flight cycle and the optimal estimate of the previous flight cycle. Based on the state prediction equation, process noise, and covariance of the previous flight cycle, predict the estimated state vector and estimation error covariance of the current flight cycle. The estimated state vector and the estimated error covariance are updated based on the actual observations at the current moment, the given process model, and the measurement model to obtain the estimated wind speed and the estimated load change.

7. The method according to claim 6, characterized in that, The step of updating the estimated state vector and the estimated error covariance based on the actual observed values ​​at the current moment, the given process model, and the measurement model to obtain the estimated wind speed and the estimated load change includes: The gain is determined based on the estimated error covariance, the measurement noise covariance matrix, and the residual measurement model. The deviation between the actual observation and the predicted observation is calculated, and the prior estimate in the state prediction equation is corrected using the deviation and the gain to obtain the posterior estimate of the current flight cycle. The wind speed estimate and the load change estimate are determined based on the posterior estimate.

8. The method according to any one of claims 4-7, characterized in that, The process of constructing the state vector to be estimated based on the wind speed and the load change includes: When the aircraft is detected to meet the vertical excitation condition, a state vector to be estimated is constructed based on the horizontal component of the wind speed, the vertical component of the wind speed, and the load change. The vertical excitation condition includes at least one of the following: The absolute value of the vertical acceleration of the aircraft in the current flight cycle is greater than a first threshold. The estimated covariance of the payload change of the aircraft in the current flight cycle is greater than the second threshold, and the absolute value of the vertical acceleration is greater than the first threshold. The aircraft is in a preset flight mode or flight phase.

9. The method according to claim 7, characterized in that, If the horizontal acceleration of the aircraft is less than a first threshold, the step of correcting the prior estimate in the state prediction equation using the bias and the gain is skipped, and the posterior estimate of the current flight cycle is determined to be equal to the predicted value of the current flight cycle; or, the gain is reduced, and the prior estimate in the state prediction equation is corrected using the reduced gain and the bias.

10. A device for estimating aircraft disturbance parameters, characterized in that, include: The data acquisition module is used to acquire the flight status data of the aircraft in the current flight cycle; the flight status data includes flight speed, flight acceleration and thrust; The residual measurement model construction module is used to construct a residual measurement model based on the flight speed, the flight acceleration, and the thrust; wherein the residual measurement model is constructed from residual measurements, wind speed, and load changes. The estimation module is used to estimate the disturbance of the aircraft in the current period using the residual measurement model, and obtain the corresponding disturbance parameter estimates; the disturbance parameter estimates include wind speed estimates and load change estimates.

11. An aircraft, characterized in that, include: Processor, memory, and bus, among which: The processor and the memory communicate with each other via the bus; The memory stores program instructions that can be executed by the processor, and the processor can execute the method as described in any one of claims 1-9 by calling the program instructions.

Images