Autonomous energy-saving soaring route planning method for small low-cost aircraft

Abstract

The invention relates to an autonomous energy-saving soaring flight path planning method for a small-sized low-cost aircraft, belongs to the technical field of aircraft trajectory planning, solves the problem of low-cost wind field energy acquisition of the small-sized low-cost aircraft in the prior art, and comprises the following steps: S1, configuring a sensor for the aircraft, and measuring through the sensor to obtain observation parameters; s2, establishing a state vector of the aircraft; s3, establishing an aerodynamic force model, introducing a dynamic equation and a state transition equation, and performing accurate modeling on aerodynamic force; s4, performing multi-source data fusion by adopting extended Kalman filtering, establishing an extended Kalman filter of a nonlinear system, and executing real-time wind vector high-precision sensing; and S5, performing global wind field modeling, estimating a wind field environment, and performing energy-obtaining flight path planning to obtain an optimal energy-obtaining soaring flight path planning scheme.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of aircraft trajectory planning, in particular to an autonomous energy-saving soaring flight path planning method for small and low-cost aircraft. BACKGROUND

[0002] Compared with large and medium-sized unmanned aerial vehicles, small fixed-wing unmanned aerial vehicles have unique advantages such as low cost, strong flexibility and convenient deployment. When deployed in clusters, they can break through the task boundaries of single machines, have high system robustness and diverse task capabilities. These characteristics enable small fixed-wing unmanned aerial vehicles to be applied to advanced air defense systems.

[0003] However, due to the size limitation of unmanned aerial vehicles, the weight of the batteries and fuel that can be carried by unmanned aerial vehicles is limited, which in turn leads to insufficient endurance, becoming the main technical bottleneck that limits the improvement of the task efficiency of small fixed-wing unmanned aerial vehicles. For small unmanned aerial vehicles, techniques such as in-flight refueling have high costs and high task difficulty, and are difficult to implement at the level of small unmanned aerial vehicles; solar energy technology is greatly affected by weather and has low energy conversion efficiency; wind field energy technology relies on the widely existing wind field conditions in nature, fully extracts energy from the wind field, and has high application prospects.

[0004] The research on wind field energy technology originated from the observation of scholars on birds such as albatrosses in nature. The existing technical route mainly focuses on two aspects: one is the wind field energy flight path planning technology based on optimal control theory, including dynamic soaring, static soaring and other trajectory optimization algorithms; the other is the maneuvering flight control technology for energy flight path, which uses PID, LQR and other classical control methods to realize accurate tracking of the preset trajectory. However, it should be pointed out that the above technical research is generally based on the prior assumption of known wind field distribution.

[0005] In actual engineering applications, the efficiency and capability of wind field energy flight technology are highly dependent on the real-time and accurate perception effect of wind field information. Among the existing wind field measurement methods, the observation data of ground meteorological stations or wind measurement towers have low spatial resolution, which cannot meet the needs of local wind field reconstruction of aircraft; although airborne laser radar can achieve high-precision measurement and prediction, there is a fundamental contradiction between its high cost, weight and power consumption characteristics and the small and low-cost unmanned aerial vehicle platform; the direct measurement method based on angle of attack and sideslip angle sensors has significant limitations - the original data is severely disturbed by airflow disturbance and high-frequency noise, and small unmanned aerial vehicles cannot carry professional incoming flow angle sensors due to size and cost limitations; and indirect estimation methods based on unscented Kalman filter (UKF), moving horizon estimation (MHE) and other nonlinear estimation and optimal control have exponential computational complexity, which cannot meet the real-time requirements on small computing platforms, and the additional cost of custom hardware acceleration modules is contrary to the low-cost design goal.

[0006] Chinese invention patent application, publication number CN119596994A, invention name "Stratosphere aircraft cross-day and night task trajectory planning method and system in dynamic wind field environment", discloses a combination framework of level set algorithm and particle swarm algorithm to consider the speed constraints, energy consumption constraints, task feasibility and other conditions of aircraft cross-day and night arrival or residence task in dynamic wind field environment, to realize trajectory optimization. However, this method involves trajectory planning of airships in the stratosphere, and fails to provide an effective flight perception and planning method for small unmanned aerial vehicles in low-altitude wind field environment.

[0007] Chinese invention patent application, publication number CN117193370A, invention name "Four-rotor unmanned aerial vehicle safe and fast flight method based on risk assessment", discloses that the environmental risk is assessed by the angle change of the direction vector of the local planning start and end point and the direction vector of the original trajectory. However, this method does not consider the influence of wind field on unmanned aerial vehicles.

[0008] Therefore, there is a need in the current technical field for an autonomous energy-saving soaring flight path planning method applicable to low-cost aircraft. SUMMARY

[0009] In view of the above problems, the present application provides an autonomous energy-saving soaring flight path planning method for small low-cost aircraft.

[0010] The autonomous energy-saving soaring flight path planning method for small low-cost aircraft provided by the embodiment of the present application comprises the following steps: Step S1, configuring a sensor for the aircraft, and obtaining observation parameters through the sensor; Step S2, establishing a state vector of the aircraft, including airspeed components, three-axis angular velocity components and attitude angles of the aircraft in the body coordinate system, and three-dimensional components of wind speed in the inertial coordinate system; Step S3, establishing an aerodynamic force model, and introducing the aerodynamic force model into the dynamic equation and the state transition equation to accurately model the aerodynamic force; Step S4, performing multi-source data fusion by using extended Kalman filtering, establishing an extended Kalman filter of the nonlinear system, and performing real-time wind vector high-precision perception to realize high-precision perception of the state quantity; Step S5, performing global wind field modeling, estimating the wind field environment to obtain autonomously perceived wind field information, and performing energy-gaining flight path planning to obtain an optimal energy-gaining soaring flight path planning scheme for guiding the actual flight of the aircraft.

[0011] Optionally, step S1 specifically comprises: Step S1.1, configuring a sensor for the aircraft comprises setting a global satellite positioning system, an inertial navigation system and an airspeed tube on the aircraft; Step S1.2, measuring the ground speed component of the aircraft in the inertial coordinate system by the global satellite positioning system, measuring the three-axis angular velocity component of the aircraft in the body coordinate system by the inertial navigation system, and measuring the airspeed scalar value of the aircraft under the airflow coordinate system by the airspeed tube, as the observation parameters obtained by the sensor measurement.

[0012] Optionally, step S3 specifically comprises: Step S3.1, obtaining the inertia constant of the aircraft in the body coordinate system based on the inertia moment of the aircraft in the body coordinate system obtained by simulation or experiment and the inertia product of the aircraft in the body coordinate system; Step S3.2, obtaining the engine thrust based on the throttle position; Step S3.3, establishing an aerodynamic force model, and fusing the aerodynamic force model to establish a dynamic equation and a state transition equation.

[0013] Optionally, step S3.3 specifically comprises: Step S3.3.1, calculating and obtaining the airspeed by using the estimated airspeed component of the aircraft in the body coordinate system; Step S3.3.2, obtaining the coefficients in the aerodynamic force model to establish the aerodynamic force model, the coefficients in the aerodynamic force model including the drag coefficient, the side force coefficient, the lift coefficient, the roll moment coefficient, the pitch moment coefficient and the yaw moment coefficient; Step S3.3.3, obtaining the aerodynamic force and the aerodynamic moment received by the aircraft based on the aerodynamic force model, the aerodynamic force received by the aircraft including the drag, the side force and the lift received by the aircraft, and the aerodynamic moment received by the aircraft including the roll moment, the pitch moment and the yaw moment received by the aircraft; Step S3.3.4, establishing the aircraft dynamic equation fused with the aerodynamic force model, for calculating the state vector of the aircraft; Step S3.3.5, discretizing the aircraft dynamic equation by using the forward Euler method to establish the state transition equation, for calculating the discretized state vector.

[0014] Optionally, step S4 specifically comprises: Step S4.1, establishing the observation equation of the aircraft based on the obtained state vector of the aircraft and the state transition equation, for calculating the observation vector and the discretized observation vector; Step S4.2, establishing the nonlinear system of the discretized state vector and the observation vector based on the Gaussian distribution, and presetting the initial state of the aircraft and the covariance matrix of the initial state quantity; Step S4.3, performing state estimation of the nonlinear system by using the extended Kalman filter to perform real-time high-precision perception of the wind vector, and obtaining the real-time state estimation value of the aircraft.

[0015] Optionally, step S4.3 specifically comprises: a state prediction is performed on the state estimation value of the previous round by using a state transition equation, to obtain a state prediction value of the current round, wherein for the first iteration, a preset initial state of the aircraft is used as the state estimation value of the previous round; a Jacobian matrix of the state transition equation with respect to the state quantity is obtained based on the state estimation value of the previous round; an error covariance matrix is obtained based on the Jacobian matrix of the state transition equation with respect to the state quantity and the covariance matrix of the state quantity obtained in the previous round, wherein for the first iteration, a preset covariance matrix of the initial state quantity is used as the covariance matrix of the state quantity of the previous round; a Jacobian matrix of the observation equation with respect to the state quantity is obtained based on the state prediction value of the current round and the control vector input at the time of the current round; an observation update is performed based on the obtained error covariance matrix and the Jacobian matrix of the observation equation with respect to the state quantity, to calculate a Kalman gain value at the time of the current round; an observation update is performed based on the state prediction value of the current round and the Kalman gain value, to obtain a state estimation value of the current round as a real-time state estimation value of the aircraft; an observation update is performed based on the error covariance matrix of the current round, the Jacobian matrix of the observation equation with respect to the state quantity and the Kalman gain value, to obtain a covariance matrix of the state quantity of the current round; the updated covariance matrix of the state quantity of the current round and the state estimation value are substituted into the next round of iterative estimation, to realize real-time high-precision perception of the wind vector.

[0016] Optionally, step S5 specifically comprises: step S5.1, global wind field modeling is performed, and a wind gradient estimation of the unknown wind field profile is obtained based on the real-time state estimation value of the aircraft as an estimated wind field environment; step S5.2, energy-gaining flight path planning is performed in the estimated wind field environment, to obtain an optimal energy-gaining soaring flight path planning scheme.

[0017] Optionally, step S5.2 specifically comprises: a three-degree-of-freedom center-of-mass dynamics equation of the aircraft is established; flight performance constraints are added for energy-saving soaring flight of the aircraft, including establishing constraints for lift coefficients, airspeeds, path inclinations, path declinations, roll angles and engine thrusts of the aircraft; periodic flight boundary constraint conditions are added in the hovering mode, including establishing constraints for positions, path inclinations, path declinations and estimated airspeeds of the aircraft in an inertial coordinate system in a flight period of time; Add periodic flight boundary constraints in forward mode, including the position of the aircraft in the inertial coordinate system during a flight cycle, the track inclination angle, the track deviation angle, and the estimated airspeed to establish constraints; The cost function of the trajectory optimization problem is set to the minimum average power consumed in a single cycle; The three-degree-of-freedom center-of-mass dynamic equations of the aircraft are solved by nonlinear programming problems to obtain the optimal energy-generating soaring flight trajectory information planned under the estimated wind field environment, including the position of the periodic flight waypoints and the corresponding speed and track angle information at the waypoints, as well as the average power consumed in a single cycle and the total cycle time.

[0018] Compared with the prior art, the autonomous energy-saving soaring trajectory planning method for a small, low-cost aircraft provided in accordance with an embodiment of the present invention has at least the following beneficial effects.

[0019] 1) The autonomous energy-saving soaring trajectory planning method proposed in this invention relies on the aircraft to autonomously sense wind vector information and integrates an aerodynamic model. This method can achieve rapid tracking and accurate estimation of the aircraft's aerodynamic characteristics, thereby achieving accurate estimation of steady and time-varying wind fields with low latency.

[0020] 2) The autonomous energy-saving soaring trajectory planning method proposed in this invention uses extended Kalman filtering in the high-precision wind vector sensing step, which can effectively eliminate the interference of sensor noise and obtain high-precision wind speed sensing results by fusing model prediction data with observation update data.

[0021] 3) The proposed autonomous, energy-saving soaring trajectory planning method can be applied to small, low-cost aircraft, enabling autonomous, high-precision wind vector sensing. By incorporating an aerodynamic model, the aircraft's wind sensing does not rely on high-precision wind measurement equipment or specialized wind sensors, thus reducing costs.

[0022] 4) The autonomous energy-saving soaring trajectory planning method proposed in this invention uses least squares estimation to enable the aircraft to discretize and sample horizontal wind speed information at different altitudes, accurately fit the estimated wind profile information, and extract the wind gradient value, enabling small, low-cost aircraft to perceive unknown wind fields.

[0023] 5) The autonomous energy-saving soaring trajectory planning method proposed in this invention establishes an optimal control problem and discretizes it into a nonlinear programming problem using the pseudo-spectral method. The problem is directly solved by the built-in function of Matlab, which has high solution efficiency. It can obtain the optimal reference trajectory for the aircraft's autonomous energy-saving soaring flight using wind energy, and can provide a reference for improving the aircraft's endurance performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. By referring to the drawings, the features and advantages of the present invention can be more clearly understood. The drawings are schematic and should not be understood as limiting the present invention in any way. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0025] Figure 1 This is a flowchart of an autonomous energy-saving soaring trajectory planning method for a small, low-cost aircraft provided according to an embodiment of the present invention.

[0026] Figure 2 Schematic diagram of sensor arrangement in an autonomous energy-saving soaring trajectory planning method for a small, low-cost aircraft provided according to an embodiment of the present invention.

[0027] Figure 3 The present invention provides a multi-source data fusion wind speed estimation flow chart for an autonomous energy-saving soaring trajectory planning method for a small, low-cost aircraft according to an embodiment of the present invention.

[0028] Figure 4 Schematic diagram of the velocity triangle in the autonomous energy-saving soaring trajectory planning method for small low-cost aircraft provided according to an embodiment of the present invention.

[0029] Figure 5 This is a flow chart of an extended Kalman filter estimation algorithm in an autonomous energy-saving soaring trajectory planning method for a small, low-cost aircraft provided according to an embodiment of the present invention.

[0030] Figure 6 This is a schematic diagram of global wind field modeling in an autonomous energy-saving soaring trajectory planning method for small, low-cost aircraft provided according to an embodiment of the present invention.

[0031] Description of reference numerals: 1- Global Positioning System (GNSS); 2- Inertial Navigation System (INS); 3- Pitot tube. DETAILED DESCRIPTION

[0032] In order to more clearly understand the above-mentioned objects, features and advantages of the present invention, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present invention and the features therein can be combined with each other without conflict.

[0033] Many specific details are set forth in the following description in order to provide a thorough understanding of the application. However, the application can be practiced according to other embodiments that can not be described in detail herein, and the scope of the present application is not limited to the specific embodiments described in this description.

[0034] An autonomous energy-saving soaring path planning method for small and low-cost aircraft is provided in the following description of an embodiment according to the present application. In this embodiment, multi-source data fusion is carried out by collecting sensor information and combining aerodynamic force models.

[0035] The principle of the method is to fuse state prediction information and sensor observation update information based on extended Kalman filter (EKF) estimation theory, rely only on the on-board sensors of the unmanned aerial vehicle without introducing high-cost on-board equipment, reduce the computational load by approximate linearization, optimize the covariance matrix update strategy in combination with the statistical characteristics of sensor noise, and achieve low-cost, low-latency, and high-precision wind field estimation. Further, real-time wind field estimation results are used for global wind field modeling, and Chebyshev pseudospectral method is used for optimal energy-gaining path planning based on the wind field information perceived by the aircraft, to obtain a reference path for energy-saving flight using wind energy, thereby providing a technical foundation for improving the endurance performance of small and low-cost aircraft.

[0036] First, the aircraft body motion parameters and environmental perception data are obtained. The ground speed vector information of the aircraft is read by the on-board global navigation satellite system (GNSS) module, the angular velocity, attitude angle, and acceleration data output by the inertial navigation system (INS) module are synchronously collected, and the airspeed scalar value is measured by the airspeed tube, to complete the data synchronous collection and preprocessing of multi-source heterogeneous sensors.

[0037] Second, a state space model for wind field estimation is constructed. The body axis coordinate system components of the aircraft, angular velocity, attitude angle, and three-dimensional wind speed are taken as state variables, the measured ground speed, angular velocity, attitude angle, acceleration, and airspeed value are taken as observation quantities, a nonlinear state transition equation is established based on the six-degree-of-freedom dynamics model and the aerodynamic model of the aircraft, and a measurement equation is derived according to the vector relationship (velocity triangle) between airspeed and wind speed, to form a complete system state estimation model framework.

[0038] Third, an extended Kalman filter estimator is designed. The Jacobian matrix of the nonlinear state transition equation and the measurement equation is solved to obtain a linearized system, and through iterative calculation in two stages of state prediction and observation update, the fusion of real-time sensor data and model prediction information is realized, and the estimated value of the wind speed is dynamically corrected. In this process, the GNSS / INS integrated navigation data of the aircraft is used as the observation reference, avoiding the introduction of additional wind measuring sensors.

[0039] Further, the flight path is combined with wind speed sensing information to model a three-dimensional wind field, and least square method is used to fit parameters to estimate wind gradient information of the wind field.

[0040] Based on optimal control theory, an optimal energy-gaining soaring flight path in a wind field is designed. A three-degree-of-freedom kinematic equation of the aircraft is taken as a state equation, flight boundary constraint conditions and flight performance constraint conditions are added, and the minimum average power consumed in a single cycle is taken as an optimization objective. Chebyshev pseudospectral method is used for discretization processing, and an interior point method is used for online nonlinear programming solution to obtain discrete flight points on the optimal energy-gaining flight path in a single cycle, which is used as a flight path reference for the aircraft to gain energy by using wind field information.

[0041] As shown in Figure 1 , the embodiment of the present application provides an autonomous energy-saving soaring flight path planning method for small and low-cost aircraft, which comprises the following steps.

[0042] In step S1, sensors are configured for the aircraft, and observation parameters are obtained by the sensors. Referring to Figure 2 and Figure 3 , the step S1 specifically comprises the following steps.

[0043] In step S1.1, the sensor configuration comprises setting a global satellite positioning system 1, an inertial navigation system 2 and an airspeed tube 3 on the aircraft. As shown in Figure 2 , the airspeed tube 3 can be arranged at the front of the aircraft nose, and the global satellite positioning system (GNSS) 1 and the inertial navigation system (INS) 2 can be arranged in the middle of the aircraft fuselage.

[0044] In step S1.2, the observation parameters are obtained by the set sensors, including measuring the ground speed component of the aircraft in the inertial coordinate system by the global satellite positioning system 1, measuring the three-axis angular velocity component of the aircraft in the body coordinate system by the inertial navigation system 2, and measuring the airspeed scalar value of the aircraft in the airflow coordinate system by the airspeed tube 3. The specific measurement methods of the observation parameters are as follows.

[0045] The ground speed component of the aircraft in the inertial coordinate system is measured by the global satellite positioning system 1 , wherein the origin of the inertial coordinate system is selected as an arbitrary selected fixed point on the ground, the direction is the north direction, the direction is the east direction, the direction is the vertical downward direction, and the three directions satisfy the right-hand rule, represents the speed of the aircraft along the direction, represents the speed of the aircraft along the direction, represents the speed of the aircraft along Speed ​​in direction.

[0046] The three-axis angular velocity components of the aircraft in the body coordinate system are measured by the inertial navigation system 2 , attitude angle , and the three-axis acceleration components , where the body coordinate system Taking the center of mass of the aircraft as the origin, The axis is located in the symmetry plane of the aircraft and its positive direction is parallel to the fuselage axis. The axis is perpendicular to the symmetry plane and its positive direction is to the right. The axis is located in the symmetry plane of the aircraft and its positive direction is perpendicular to The axis points downward. For the aircraft to orbit The roll angular velocity of the axis, For the aircraft to orbit The pitch angular velocity of the axis, For the aircraft to orbit The yaw rate of the axis, is the yaw angle of the aircraft, is the pitch angle of the aircraft, is the roll angle of the aircraft, The aircraft along the body coordinate system acceleration.

[0047] The airspeed scalar value of the aircraft in the airflow coordinate system is measured by the pitot tube 3 .

[0048] Therefore, the observation parameters are measured by the sensors configured on the aircraft: .

[0049] Since small, low-cost aircraft do not have the conditions for installing an incoming flow angle sensor, the measurement of the angle of attack and the sideslip angle is not considered in this embodiment.

[0050] Step S2: Establishing a state vector of the aircraft to realize the perception of the wind vector, thereby performing state estimation on the flight state of the aircraft.

[0051] In order to accurately reflect the flight state of the aircraft, the airspeed component of the aircraft in the body coordinate system is selected , three-axis angular velocity components , attitude angle As a state quantity. And, in order to accurately estimate the wind speed state, the three-dimensional components of the wind speed in the inertial coordinate system are selected is the state quantity.

[0052] Based on the selected state quantity, establish the state vector of the aircraft: .

[0053] In the following step, the aircraft state equation is estimated to obtain the state vector of the aircraft defined herein.

[0054] Step S3, the aerodynamic force model is established, and the aerodynamic force model is introduced into the dynamic equation and the state transition equation, the aerodynamic force is accurately modeled, and then the accurate estimation of the flight state and the high-precision perception of the wind vector information are realized.

[0055] According to the six-degree-of-freedom dynamics equation of the aircraft and the assumption that the wind field changes slowly, the dynamic equation of the aircraft is established:

[0056] wherein, is a control vector and , is the aileron deflection angle, is the elevator deflection angle, is the rudder deflection angle, is the throttle position, which is obtained from the rudder deflection angle command and the throttle size given by the pilot or the flight control system, denotes the derivative of the state vector of the aircraft with respect to time.

[0057] Step S3 specifically includes the following steps.

[0058] Step S3.1, obtain the inertia constant of the aircraft in the body coordinate system As follows:

[0059]

[0060]

[0061]

[0062]

[0063]

[0064]

[0065]

[0066]

[0067] wherein, are the inertia moments of the aircraft in the body coordinate system, respectively, They are the inertia products of the aircraft in the body coordinate system. These parameters can be obtained by measuring the aircraft model in CATIA during simulation, and can be measured using the torsion pendulum method during the experiment.

[0068] Step S3.2: Assume that the aircraft's engine thrust is only related to the throttle position, and obtain the engine thrust:

[0069] in, is the engine thrust, is the throttle position, It represents the thrust increment corresponding to a unit throttle change after being simplified using the linear thrust model.

[0070] Step S3.3, establishing an aerodynamic model, and integrating the aerodynamic model to establish dynamic equations and state transfer equations, specifically includes the following steps.

[0071] Step S3.3.1, using the estimated airspeed components of the aircraft in the body coordinate system, calculate the airspeed : .

[0072] Step S3.3.2, obtain the coefficients in the aerodynamic model and establish the aerodynamic model.

[0073] First, construct the angle of attack and angle of sideslip:

[0074]

[0075] in, It represents the angle of attack, which is defined as the difference between the projection of the aircraft velocity vector in the aircraft symmetry plane and the body coordinate system. The angle between the axes, under normal flight conditions, the projection line is Above the axis just; represents the sideslip angle, which is defined as the angle between the vehicle's velocity vector and the vehicle's plane of symmetry, with the velocity vector being positive to the right of the plane of symmetry.

[0076] Next, construct the coefficients in the aerodynamic model, including the drag coefficient , lateral force coefficient , lift coefficient , rolling moment coefficient , pitching moment coefficient and yaw moment coefficient :

[0077]

[0078]

[0079]

[0080]

[0081] .

[0082] The aerodynamic force model is thus established. In the following, in the state estimation process, the estimated parameters are substituted into the aerodynamic force model to calculate the aerodynamic force and aerodynamic moment in the discrete time state.

[0083] Step S3.3.3, based on the aerodynamic force model, the aerodynamic force and aerodynamic moment received by the aircraft are obtained:

[0084]

[0085] wherein, is the dynamic pressure, the calculation formula is , is the atmospheric density, is the drag force received by the aircraft, is the side force received by the aircraft, is the lift force received by the aircraft, is the roll moment received by the aircraft, is the pitch moment received by the aircraft, is the yaw moment received by the aircraft, is the wing area of the aircraft, is the span of the aircraft, is the average aerodynamic chord length of the aircraft. Wherein, the aerodynamic force received by the aircraft is ( ); the aerodynamic moment received by the aircraft is ( ).

[0086] Step S3.3.4, the aircraft dynamic equation fused with the aerodynamic force model is established:

[0087]

[0088]

[0089]

[0090]

[0091]

[0092]

[0093]

[0094]

[0095]

[0096]

[0097]

[0098] in, For engine mounting angle, is the mass of the aircraft, is the acceleration due to gravity, They represent the derivatives of the airspeed components of the aircraft with respect to time in the body coordinate system, They represent the derivatives of the angular velocity components of the aircraft with respect to time in the body coordinate system, They represent the derivatives of the aircraft’s attitude angle with respect to time, They represent the time derivatives of the three-dimensional components of wind speed in the inertial coordinate system. The state estimate of the aircraft obtained in step S4.3 in the previous time step is used as an input parameter.

[0099] Through the above aircraft dynamic equations that integrate the aerodynamic model, the aircraft state vector is obtained by time integration. :

[0100] in, For time, express The state vector at time t, represents the state vector at the initial moment, is the integration variable, express The time derivative of the state vector at a given moment, Indicates that the derivative of the state vector with respect to time is arrive The state vector of the aircraft includes the items of the state vector of the aircraft preset in step S2.

[0101] Step S3.3.5, the dynamic equations of the aircraft Use the forward Euler method for discretization and establish the state transfer equation:

[0102] wherein subscript denotes the discrete time step, denotes the state vector at the discrete time step, denotes the discrete time interval, denotes the discrete time step, and the input control vector is provided by the pilot or the autopilot.

[0103] The state vector after discretization is obtained through the above state transition equation, i.e., the prediction result of the state transition equation.

[0104] The aerodynamic force and the aerodynamic moment are obtained by solving the aerodynamic model and the estimated state quantity, and the dynamic equation of the aircraft is established based on the aerodynamic force and the aerodynamic moment, and then the state transition equation is obtained.

[0105] In step S4, the Kalman filter is used for multi-source data fusion, an extended Kalman filter of the nonlinear system is established, and real-time high-precision perception of the wind vector is performed to realize high-precision perception of the state quantity (including the flight state of the aircraft and the wind speed information).

[0106] Relying only on the state transition equation to predict the change of the state quantity will be seriously affected by the model error. For example, since the wind speed is assumed to change slowly, the change rate of the wind speed is modeled as 0, which will cause the estimated value of the wind speed to always be the initial value without change, which is obviously unreasonable. Therefore, the prediction result of the state transition equation needs to be corrected. As shown in Figure 5 , the Kalman filter is used for multi-source data fusion in the present embodiment. The Kalman gain is calculated between the low-confidence prediction state vector with model error and the low-confidence state observation parameter with noise error measured by the sensor, and the multi-source data fusion is performed by using the Kalman gain, so as to obtain a state estimate value with relatively high confidence. The Kalman filter is suitable for linear systems. For nonlinear systems, the extended Kalman filter, the unscented Kalman filter, etc. are evolved. Since the calculation amount of the extended Kalman filter is small, the present embodiment uses the extended Kalman filter to perform multi-source data fusion, which will be described below.

[0107] The extended Kalman filter performs first-order approximation on the original nonlinear system model through first-order Taylor series expansion, further calculates the prediction value of the state quantity through the state prediction step, and calculates the Kalman gain in the observation update step by combining the observation parameter information to balance the prediction state vector and the observation parameter measured by the sensor, so as to obtain a state estimate value with high confidence. Referring to Figure 3 and Figure 5 , step S4 specifically includes the following steps.

[0108] Step S4.1: First, the observation equation of the aircraft is established based on the obtained state vector and state transfer equation of the aircraft.

[0109] Based on the established linearized state transfer equation , combined with the observation parameters of the sensors configured on the aircraft measured in step S1, namely , establish the observation equation to solve the predicted observation vector.

[0110] To establish the relationship between the wind speed in the state vector and the ground speed in the observation vector, a velocity triangle (e.g. Figure 4 shown):

[0111] in, The sum of the three components of the ground speed of the aircraft in the inertial coordinate system is: , Represents the resultant vector of the three components of wind speed in the inertial coordinate system, that is, , The resultant vector of the three components of the aircraft's airspeed in the body coordinate system is: , Respectively axis, axis, axis, axis, axis, The unit vector of the axis.

[0112] The transformation matrix representing the transformation from the body coordinate system to the inertial coordinate system is:

[0113] Based on the state vector of the aircraft obtained in step S3.3.4 , combining the state vector information with the velocity vector triangle relationship (such as Figure 4 As shown), establish the observation equation of the aircraft:

[0114] in, Represents the state vector of the aircraft obtained by The observation vector calculated from the observation equation is represents the observation equation.

[0115] The specific form of the observation equation is as follows:

[0116]

[0117]

[0118]

[0119]

[0120]

[0121]

[0122]

[0123]

[0124]

[0125]

[0126]

[0127]

[0128] where the control vector and the state vector of the aircraft obtained in step S3.3.4 are taken as input parameters.

[0129] After discretization of the observation equation, the following equation is obtained:

[0130] where, denotes the th discrete time step, denotes the observation vector at the th discrete time step.

[0131] The observation vector of the aircraft is calculated as follows:

[0132] In step S4.2, the process noise is set as , which is subject to a Gaussian distribution with a mean of 0 and a covariance matrix of , i.e. ; and the observation noise is set as , which is subject to a Gaussian distribution with a mean of 0 and a covariance matrix of , i.e. Based on the Gaussian distribution, the nonlinear system is established as:

[0133] where the initial time step is set, and the initial state is set as ​​, the covariance matrix of the initial state quantity is .

[0134] Step S4.3, the state estimation of the nonlinear system is carried out by using the extended Kalman filter, the real-time wind vector high-precision sensing is carried out, and the real-time state estimation value of the aircraft is obtained. The flow chart of this step is as shown in Figure 5 , and the specific process is as follows.

[0135] First, the state prediction is carried out, and the state estimation value at the last time is , and the state prediction value at the current time is obtained by the state transition equation: .

[0136] In the first iteration, , the state estimation value uses the initial state set by the above step S4.2.

[0137] Based on the state estimation value at the last time , the Jacobian matrix of the state transition equation with respect to the state quantity is established: ; Based on the Jacobian matrix of the state transition equation with respect to the state quantity and the covariance matrix of the state quantity obtained at the last time, the error covariance matrix is established as:

[0138] , wherein represents the covariance matrix of the process noise at the time , which is set in advance according to experience before the wind speed estimation, and remains a constant value in the operation process, represents the covariance matrix of the state quantity at the last time . In the first iteration, , the covariance matrix of the state quantity at the time uses the initial covariance matrix of the state quantity set by the above step S4.2.

[0139] Based on the state prediction value of the current round and the input control vector at the time of the current round, the Jacobian matrix of the observation equation with respect to the state quantity is established , that is: ; Then, based on the obtained error covariance matrix and the Jacobian matrix of the observation equation with respect to the state quantity, the observation update is carried out, and the Kalman gain value at the time is calculated as: ; wherein, denotes the variance of the observation noise at the time instant, which is set in advance empirically before the wind speed estimation is carried out and remains as a constant value in the calculation process.

[0140] Based on the state prediction value and the Kalman gain value of the current round, observation update is carried out to obtain the state estimation value of the current round at the time instant: ; wherein, is the observation parameter measured by the sensor at the time instant. The obtained state estimation value is the real-time state estimation value. and based on the error covariance matrix of the current round, the Jacobian matrix of the observation equation with respect to the state quantity and the Kalman gain value, observation update is carried out to obtain the covariance matrix of the state quantity of the current round

[0141] at the time instant: .

[0142] wherein, denotes the unit matrix.

[0143] The updated covariance matrix of the state quantity and the state estimation value are used for the iterative estimation of the next round to complete the real-time high-precision perception task of the wind vector.

[0144] Thus, high-precision perception of the state quantity (including the flight state of the aircraft and the wind speed information) is realized.

[0145] Step S5, global wind field modeling is carried out, the wind field environment is estimated to obtain the autonomously perceived wind field information, and the aircraft is enabled to plan a flight path to obtain an optimal enabled soaring flight path planning scheme, which is used to guide the actual flight of the aircraft. The process of this step S5 is as shown in Figure 1 and Figure 6 , and specifically includes the following steps.

[0146] Step S5.1, global wind field modeling is carried out, and based on the real-time state estimation value of the aircraft, wind gradient estimation of the unknown wind field profile is obtained as the estimated wind field environment, as shown in Figure 6 .

[0147] Firstly, the gradient wind field is approximately simplified as a linear wind field, which is described by the following formula:

[0148] wherein, is the ground horizontal wind speed, is the height, ​wind gradient (describing the non-uniformity of wind speed in the height direction), height horizontal wind speed at the height.

[0149] The aircraft is caused to perform a climb or glide flight path to sample wind speed at different height layers, and is provided with a set of observation data , wherein is the height of the first measurement point of the aircraft, is the corresponding horizontal wind speed at the height. Using the wind field perception method of steps S1-S4, the wind speed at

[0150] a set of measurement points is autonomously perceived, the corresponding horizontal wind speed at the height is the horizontal wind speed in the state estimation value estimated by the aircraft at the height, is the synthesis of and , and is the wind speed in the inertial coordinate system direction estimated at the height , is the wind speed in the inertial coordinate system direction estimated at the height .

[0151] The residual sum of squares function of wind field estimation is established as:

[0152] wherein, is the ground wind speed, represents the residual sum of squares function.

[0153] The optimal parameters and are solved by minimizing .

[0154] The partial derivatives of the ground wind speed and the wind gradient are taken respectively and the derivatives are set to 0 to obtain the following equation set:

[0155] After arrangement, the linearized equation set is obtained:

[0156] The above equation set is expressed in matrix form as:

[0157] wherein, the design matrix , parameter vector , the observation vector .

[0158] The least squares solution is: , after expansion, we can get Wind gradient The wind gradient estimation for the wind profile of the unknown wind field is obtained.

[0159] Step S5.2: Obtain a reference trajectory using the interior point method, perform flight trajectory planning in the estimated wind environment, and obtain the optimal flight trajectory planning scheme. Figure 1 and Figure 6 , the specific process of this step is as follows.

[0160] First, based on the aerodynamic model, the three-degree-of-freedom center-of-mass dynamic equation of the aircraft is established:

[0161] in, is the airspeed of the aircraft to be optimized, is the time derivative of the aircraft's airspeed, is the trajectory inclination angle of the aircraft to be optimized, is the time derivative of the aircraft’s track inclination, is the trajectory angle of the aircraft to be optimized, is the time derivative of the aircraft’s track angle, The position of the aircraft to be optimized in the inertial coordinate system Directional component, is the position of the aircraft in the inertial coordinate system The time derivative of the directional component, The position of the aircraft to be optimized in the inertial coordinate system Directional component, is the position of the aircraft in the inertial coordinate system The time derivative of the directional component, The position of the aircraft to be optimized in the inertial coordinate system Directional component, is the position of the aircraft in the inertial coordinate system The time derivative of the directional component, is the roll angle requirement to be optimized, is the thrust control requirement to be optimized, is the resistance to be optimized, is the lift control requirement to be optimized, For the perceived wind speed in the inertial coordinate system, see Figure 6 Assume that the wind speed is along the direction, is the derivative of the wind speed in the inertial coordinate system with respect to time.

[0162] where the lift satisfies , the drag satisfies , is the dynamic pressure, satisfying , is the atmospheric density, is the wing area of the aircraft, which is the same as in step S3.3.3. is the drag coefficient, satisfying , is the friction drag coefficient, which can be estimated empirically or calculated by CFD method, is the induced drag factor, which can be calculated by using vortex lattice method and other fast aerodynamic analysis methods, is the lift coefficient control requirement to be optimized and solved.

[0163] The above is obtained from the following formula: .

[0164] Add flight performance constraints for the energy-saving gliding flight of the aircraft, including establishing constraints for the lift coefficient, airspeed, flight path angle, flight path angle, roll angle and engine thrust of the aircraft in the energy-saving gliding state:

[0165] where, is the stall speed of the aircraft, represents the maximum lift coefficient of the aircraft, represents the maximum coefficient of the aircraft, represents the maximum thrust of the aircraft.

[0166] Add periodic flight boundary constraint conditions in the hovering mode, including establishing constraints for the position, flight path angle, flight path angle and estimated airspeed of the aircraft in the inertial coordinate system in one flight cycle time in the hovering mode:

[0167] where, represents the time of one flight cycle.

[0168] Add periodic flight boundary constraint conditions in the forward mode, including establishing constraints for the position, flight path angle, flight path angle and estimated airspeed of the aircraft in the inertial coordinate system in one flight cycle time in the forward mode:

[0169] wherein, is the displacement of the aircraft in the y direction, indicating the difference of the position of the aircraft in the y coordinate at the initial time and the end time of a flight cycle, indicating the advance of the aircraft in the y direction.

[0170] The cost function of the trajectory optimization problem is taken as the minimum average power consumed in a cycle:

[0171] wherein, represents the minimum average power consumed in a cycle of the optimization target, represents the value of the th point distributed in the interval under the Chebyshev point distribution method, represents the total number of points selected under the Chebyshev point distribution method, represents the duration of the th time interval, represents the airspeed estimate value at the beginning of the th time interval, represents the thrust value at the beginning of the th time interval.

[0172] Further, the fmincon function configured in matlab can be used to solve the nonlinear programming problem of the three-degree-of-freedom mass center dynamics equation of the aircraft, and the optimal energy-gaining soaring flight trajectory information planned under the autonomously perceived wind field condition is obtained, including the cycle flight waypoint position and the corresponding speed, track angle and other information at the waypoint, and the average power consumed in a cycle and the total cycle time. The estimated wind field environment obtained in step S5.1 is taken as the autonomously perceived wind field condition.

[0173] Thus, the small low-cost aircraft autonomously perceives the wind field information and performs energy-saving soaring planning, and the optimal energy-gaining soaring flight trajectory planning scheme is obtained. The scheme can be used to guide the actual flight of the aircraft, so that the reference trajectory can guide the actual flight of the aircraft to realize energy-saving flight by utilizing wind energy, and greatly improve the endurance performance of the aircraft.

[0174] All the optional technical solutions described above can be combined to form optional embodiments of the present application, and will not be described one by one here.

[0175] It should be understood that the size of the serial number of each step in the above embodiments does not mean the order of execution, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0176] ​The above merely describes preferred specific embodiments of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. A method for autonomous energy-saving flight trajectory planning for small low-cost aircraft, characterized by: The following steps are involved: Step S1, equipping the aircraft with sensors and obtaining observation parameters through sensor measurements; Step S2, establishing a state vector of the aircraft, including the airspeed component, three-axis angular velocity components and attitude angle of the aircraft in the body coordinate system, and the three-dimensional component of the wind speed in the inertial coordinate system; Step S3, establishing an aerodynamic model, and introducing the aerodynamic model into the dynamic equation and the state transition equation to accurately model the aerodynamic force; Step S4: Using extended Kalman filtering to fuse multi-source data, establishing an extended Kalman filter for a nonlinear system, performing real-time high-precision wind vector sensing, and achieving high-precision sensing of state quantities; In step S5, global wind field modeling is performed to estimate the wind field environment to obtain autonomously sensed wind field information, and enabled flight trajectory planning is performed to obtain the optimal enabled soaring flight trajectory planning scheme to guide the actual flight of the aircraft.

2. The autonomous energy-saving soaring trajectory planning method for small low-cost aircraft according to claim 1 is characterized in that: Step S1 specifically includes: Step S1.1, configuring sensors for the aircraft includes setting up a global satellite positioning system, an inertial navigation system, and a pitot tube on the aircraft; Step S1.2, measure the ground speed component of the aircraft in the inertial coordinate system through the global satellite positioning system, measure the three-axis angular velocity components of the aircraft in the body coordinate system through the inertial navigation system, and measure the airspeed scalar value of the aircraft in the airflow coordinate system through the pitot tube as the observation parameters measured by the sensors.

3. The autonomous energy-saving soaring trajectory planning method for small low-cost aircraft according to claim 1 is characterized in that: Step S3 specifically includes: Step S3.1, obtaining the inertia constant of the aircraft in the body coordinate system based on the moment of inertia and the product of inertia in the aircraft body coordinate system obtained by simulation or experiment; Step S3.2, obtaining engine thrust based on the throttle position; Step S3.3: Establish an aerodynamic model and integrate the aerodynamic model to establish dynamic equations and state transfer equations.

4. The autonomous energy-saving soaring trajectory planning method for small low-cost aircraft according to claim 1 is characterized in that: Step S3.3 specifically includes: Step S3.3.1, using the estimated airspeed components of the aircraft in the body coordinate system, calculate the airspeed; Step S3.3.2, obtaining coefficients in an aerodynamic model and establishing an aerodynamic model, wherein the coefficients in the aerodynamic model include a drag coefficient, a side force coefficient, a lift coefficient, a rolling moment coefficient, a pitching moment coefficient, and a yaw moment coefficient; Step S3.3.3, based on the aerodynamic model, obtain the aerodynamic forces and aerodynamic moments acting on the aircraft. The aerodynamic forces acting on the aircraft include the drag, side force, and lift acting on the aircraft. The aerodynamic moments acting on the aircraft include the rolling moment, pitching moment, and yaw moment acting on the aircraft. Step S3.3.4, establishing the aircraft dynamic equations integrated with the aerodynamic model to calculate the aircraft state vector; Step S3.3.5, discretize the aircraft dynamic equations using the forward Euler method, establish a state transfer equation, and use it to calculate the discretized state vector.

5. The autonomous energy-saving soaring trajectory planning method for small low-cost aircraft according to claim 4 is characterized in that: Step S4 specifically includes: Step S4.1: Based on the obtained state vector and state transition equation of the aircraft, establish the observation equation of the aircraft to calculate the observation vector and the discretized observation vector; Step S4.2, establishing a nonlinear system of the discretized state vector and observation vector based on Gaussian distribution, and presetting the initial state of the aircraft and the covariance matrix of the initial state quantity; In step S4.3, an extended Kalman filter is used to perform state estimation of the nonlinear system, perform real-time high-precision wind vector perception, and obtain a real-time state estimation value of the aircraft.

6. The autonomous energy-saving soaring trajectory planning method for small low-cost aircraft according to claim 5 is characterized in that: Step S4.3 specifically includes: Through the state transfer equation, the state estimation value of the previous round is input to perform state prediction to obtain the state prediction value of the current round. For the first round of iteration, the preset initial state of the aircraft is used as the state estimation value of the previous round; Based on the state estimation value of the previous round, the Jacobian matrix of the state transfer equation with respect to the state quantity is obtained; Based on the Jacobian matrix of the state transfer equation with respect to the state quantity and the covariance matrix of the state quantity obtained in the previous round, the error covariance matrix is ​​obtained. For the first round of iteration, the covariance matrix of the preset initial state quantity is used as the covariance matrix of the state quantity in the previous round; Based on the state prediction value of the current round and the control vector input at the current round, the Jacobian matrix of the observation equation with respect to the state quantity is obtained; Based on the obtained error covariance matrix and the Jacobian matrix of the observation equation with respect to the state quantity, the observation update is performed to calculate the Kalman gain value of the current round; Based on the state prediction value and Kalman gain value of the current round, the observation update is performed to obtain the state estimation value of the current round as the real-time state estimation value of the aircraft; Based on the error covariance matrix of the current round, the Jacobian matrix of the observation equation with respect to the state quantity and the Kalman gain value, the observation update is performed to obtain the covariance matrix of the state quantity of the current round; The updated covariance matrix and state estimation value of the current round of state quantities are substituted into the next round of iterative estimation to achieve real-time high-precision perception of the wind vector.

7. The autonomous energy-saving soaring trajectory planning method for small low-cost aircraft according to claim 5 is characterized in that: Step S5 specifically includes: Step S5.1, performing global wind field modeling, obtaining a wind gradient estimate for the unknown wind field profile based on the real-time state estimate of the aircraft, as the estimated wind field environment; Step S5.2: Perform energy-generating flight trajectory planning in the estimated wind environment to obtain the optimal energy-generating soaring flight trajectory planning scheme.

8. The autonomous energy-saving soaring trajectory planning method for small low-cost aircraft according to claim 7 is characterized in that: Step S5.2 specifically includes: Establish the three-degree-of-freedom center-of-mass dynamic equations of the aircraft; Add flight performance constraints for the aircraft's energy-saving soaring flight, including establishing constraints for the aircraft's lift coefficient, airspeed, track inclination, track deviation, roll angle, and engine thrust; Add periodic flight boundary constraints in hover mode, including establishing constraints for the aircraft's position, track inclination, track deviation, and estimated airspeed in the inertial coordinate system during a flight cycle. Add periodic flight boundary constraints in forward mode, including the position of the aircraft in the inertial coordinate system during a flight cycle, the track inclination angle, the track deviation angle, and the estimated airspeed to establish constraints; The cost function of the trajectory optimization problem is set to the minimum average power consumed in a single cycle; The three-degree-of-freedom center-of-mass dynamic equations of the aircraft are solved by nonlinear programming problems to obtain the optimal energy-generating soaring flight trajectory information planned under the estimated wind field environment, including the position of the periodic flight waypoints and the corresponding speed and track angle information at the waypoints, as well as the average power consumed in a single cycle and the total cycle time.

Images