A multi-rotor unmanned aerial vehicle rotor tension detection and balance control method

Abstract

The application discloses a kind of multi-rotor unmanned aerial vehicle rotor tension detection and balance control method, belong to unmanned aerial vehicle flight control technical field. Including: by the force sensor installed on each rotor force transmission path directly measure the real-time tension of each rotor, and synchronously obtain the inertial measurement information of unmanned aerial vehicle;According to flight instruction and current state calculation total tension demand;Based on control efficiency matrix and constraint optimization dynamically allocate each rotor target tension;Using feedforward-feedback compound control to realize the direct closed-loop tracking of tension with high-frequency driving motor;The output torque instruction of independently running attitude control loop is used as the input of total tension distribution, and the double-loop collaborative architecture is formed.The application directly measures rotor tension, constraint optimization distribution and high-frequency force balance closed loop, fundamentally solves the phenomenon of drawing a dragon based on rotation speed indirect control existing in traditional multi-rotor unmanned aerial vehicle, significantly improves flight stability, anti-disturbance ability and task execution quality.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) flight control technology, and in particular to a method for detecting and balancing the rotor thrust of a multi-rotor UAV based on real-time measurement using distributed force sensors. Background Technology

[0002] Multirotor drones, with their advantages of simple structure, vertical takeoff and landing, and stable hovering, have been widely used in many fields such as aerial surveying, power line inspection, and agricultural plant protection. The core of flight control for these drones lies in the precise adjustment of the thrust of each rotor—by changing the speed of each motor, different rotor thrusts are generated, thereby achieving control over the aircraft's attitude and spatial position.

[0003] Traditional multi-rotor UAV flight control systems generally employ an open-loop "rotor speed-thrust" mapping relationship, indirectly controlling rotor thrust by controlling motor speed through an electronic speed controller. This method suffers from the following drawbacks: aerodynamic disturbance uncertainties, such as the slipstream effect between rotors and ground effects, leading to a non-linear and time-varying mapping relationship between speed and thrust; system parameter perturbations, such as propeller wear, motor aging, and battery voltage drops, causing actual thrust drift at the same speed; insufficient control allocation accuracy, as pseudo-inverse-based control allocation methods cannot handle motor saturation constraints, easily resulting in control commands becoming infeasible during extreme maneuvers; and an excessively long control response chain, where the controller often uses a large gain to quickly respond to attitude deviations, leading to overshoot and subsequent reverse adjustment, resulting in continuous oscillations. This phenomenon of repeatedly correcting course on the flight trajectory is figuratively called "dragon drawing" in engineering, manifesting as a serpentine flight path with significant attitude fluctuations.

[0004] Instability in the underlying control directly limits the mission execution quality of drones: In aerial photography and mapping, micro-vibrations and low-frequency swaying of the drone body are transmitted to the gimbal system. Although gimbal stabilization technology can improve image quality, it is only a passive compensation and cannot eliminate the attitude fluctuations of the drone itself, resulting in "jelly effect" and reduced image clarity. In agricultural spraying drone operations, the nozzle trajectory fluctuates with the body's sway, affecting the operation effect. In lidar scanning, point cloud data noise increases, reducing modeling accuracy. At the same time, continuous over-adjustment correction increases motor fatigue wear, consumes extra power, and reduces endurance.

[0005] To address the aforementioned issues, the industry has undertaken various improvement attempts, primarily including some solutions employing a total thrust sensor or indirectly estimating thrust based on accelerometers. However, the former cannot distinguish the contributions of each rotor, while the latter's accuracy is limited due to the influence of airframe dynamics coupling. Therefore, there is an urgent need for a method that can directly measure the thrust of each rotor and achieve high-frequency closed-loop control. Summary of the Invention

[0006] The technical problem to be solved by this invention is to overcome the shortcomings of the prior art and provide a method for detecting and balancing the rotor force of a multi-rotor UAV. By installing force sensors on the force transmission path of each rotor, the direct measurement and high-frequency closed-loop tracking of the rotor force can be realized, forming a dual-loop collaborative architecture of "force balance inner loop + attitude outer loop". At the same time, the precise control allocation considering motor saturation can be achieved through constraint optimization.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A method for detecting rotor tension and controlling balance in a multi-rotor unmanned aerial vehicle (UAV), comprising: S1. Distributed real-time acquisition of tension and inertia information.

[0008] S11. Force sensor arrangement and installation: A force sensor is installed on the force transmission path of each rotor of the UAV. The force sensor is a strain gauge type force sensor, including an elastic body base and a strain gauge bridge attached to the surface of the elastic body. One end of the elastic body base is fixed to the arm and the other end is fixed to the motor base, so that the tension generated by the rotor is completely transmitted to the elastic body to cause deformation. The strain gauge bridge adopts a full-bridge connection method and outputs a differential voltage signal proportional to the tension.

[0009] S12. Synchronous acquisition trigger: At the beginning of each control cycle, the main controller generates a synchronous trigger signal and simultaneously latches the original electrical signals of all force sensors; the control cycle is the tension control cycle T. f Satisfying T f =T a / k, where T a The attitude control cycle is k, where k is an integer greater than or equal to 2, preferably k=4 or k=5.

[0010] S13. Signal processing and conversion: The original electrical signal is sequentially subjected to bandpass filtering, instrument amplification, and analog-to-digital conversion to obtain the real-time thrust measurement value T of each rotor. i meas , where i=1,2,......,N, and N is the total number of rotors.

[0011] S14. Inertial data synchronous acquisition: At the moment the synchronization trigger signal is generated, data from the inertial measurement unit is simultaneously read to obtain the real-time attitude angle Θ=[ϕ,θ,ψ] of the UAV. T Angular velocity ω = [ω x ,ω y ,ω z ] T and linear acceleration a=[a x ,a y ,a z ] TThe data from the inertial measurement unit and the data from the force sensor are aligned using timestamps to ensure that the data synchronization error is less than 1ms.

[0012] S2. Calculation of total tensile force requirement.

[0013] S21. Flight command reception: Based on the current flight mode of the UAV, receive flight commands from the remote controller or autonomous navigation module; the flight commands include the target altitude h. cmd Target attitude angle Θ cmd =[ϕ cmd ,θ cmd ,ψ cmd ] T and target horizontal velocity v xy cmd =[v x cmd ,v y cmd ] T .

[0014] S22. Position-velocity control: Based on the flight command and the current state, the target horizontal acceleration a is calculated by the position controller. xy cmd The vertical acceleration a of the target is calculated using the altitude controller. z cmd The position controller and height controller adopt a PID or cascaded PID structure.

[0015] S23. Total pulling force calculation, based on the total mass m of the UAV, gravitational acceleration g, and the vertical acceleration a of the target. z cmd Calculate the total thrust F required to maintain the current flight state. req =m⋅(g+a z cmd ).

[0016] S3. Dynamic allocation of target tension based on constraint optimization.

[0017] Define the target thrust T for each rotor i target (i=1,2,…,N) represents the target thrust value of the i-th rotor in the current control cycle; T target =[T1 target T2 target ,...,T N target ] T Let be the target tension vector.

[0018] S31. Construction of the control efficiency matrix: Based on the geometric layout of the UAV and the rotation direction of each rotor, construct the control efficiency matrix B; the i-th column vector of matrix B is: B i =[1,−y i ,x i ,c i ⋅k Q ] T , where (x i ,y i Let be the position coordinates of the i-th rotor in the body coordinate system, and c be the position coordinates of the i-th rotor. i k is the rotor steering coefficient. Q The ratio of tensile force to counter-torque is obtained through offline calibration.

[0019] S32. Construction of the generalized force demand vector: The total tensile force demand F calculated in step S23 is used to construct the vector. req The target control torque M calculated in step S5 req =[M x req M y req M z req ] T The combination forms the generalized flight dynamics demand vector Γ. req =[F req M x req M y req M z req ] T This vector is related to the target thrust T of each rotor. i target Satisfying relation: Γ req =B⋅T target .

[0020] S33. Constrained optimization problem establishment, based on the rotational inertia matrix J∈R of the UAV. 3×3 Establish the following constrained optimization problem: Constraints: ; where T i nom =F req / N represents the nominal uniformly distributed tensile force, λ≥0 is the power consumption weighting coefficient, and T i max This represents the maximum available thrust of the i-th rotor in its current state. The maximum available tensile force T i max Estimate online using the following methods: T i max =k T ⋅ωi max (V bat ) 2 , where k T ω is the tensile coefficient. i max (V bat (V) represents the current battery voltage. bat The maximum speed of the motor is obtained by looking up the motor characteristic curve in a table. S34. Solve the optimization problem using the active set method or interior point method to solve the quadratic programming problem established in step S33; if a feasible solution exists, the target tension value T is obtained. i target If no feasible solution is found, then the priority fallback strategy is executed.

[0021] The priority fallback strategy includes, First priority: Meet the attitude torque requirement M req Allowable total tensile force F req Adjust within ±20%; Second priority: If there is still no feasible solution, maintain the target pull of the previous cycle and trigger an alarm.

[0022] S4. High-frequency tensile closed-loop tracking control.

[0023] S41. Calculation of thrust tracking error: Calculate the thrust tracking error of each rotor: e i =T i target -T i meas .

[0024] S42. Feedback control quantity calculation, based on the tension tracking error e i The feedback component U of the motor control command is calculated through the force balance loop controller. i fb The force balance loop controller employs a PID algorithm with anti-saturation integral: K p ,K i ,K d The proportional, integral, and differential gains are tuned using pole placement or trial and error.

[0025] S43. Feedforward compensation calculation, calculating the rate of change of the target tension: High-frequency noise is suppressed by a low-pass filter, and a feedforward compensation component is generated. K ff For feedforward gain, based on the first-order inertial model of the motor-propeller system. Tuning, take K ff ≈τ m / Km .

[0026] S44. Generate composite control command, add the feedback component and the feedforward compensation component, and perform output limiting: U i cmd =sat(U i fb +U i ff U min U max ), where sat(⋅) is the saturation function, U min U max These are the upper and lower limits of the ESC input signal.

[0027] S45. High-frequency drive execution, to control cycle T f (frequency f) f =1 / T f Preferred f f ≥2kHz) will transmit the composite control command U i cmd The signal is sent to the corresponding electronic speed controller, which drives the motor to adjust the real-time thrust measurement value T of the rotor. i meas Rapidly track target tension T i target .

[0028] S5. Independent attitude control loop operation.

[0029] S51. Attitude Data Reading: In a separate control thread, with attitude control period T a Read the attitude data from the inertial measurement unit; the attitude control period T a For the tension control period T f Integer multiples of T, i.e. a =k⋅T f .

[0030] S52. Attitude Deviation Calculation: Calculate the deviation between the current attitude angle and the target attitude angle: ΔΘ=Θ cmd −Θ, where Θ cmd Θ is the target attitude angle from the flight command, and Θ is the real-time attitude angle obtained in step S14.

[0031] S53. Target angular acceleration calculation: Based on the attitude deviation, the target angular acceleration required to correct the attitude is calculated by the attitude loop controller: α. cmd =K Θ ⋅ΔΘ+K ω ⋅(ω cmd −ω), where K Θ ,K ωFor the control gain of the attitude loop and angular velocity loop, a cascaded PID or PD control structure is adopted; ω cmd ω is the target angular velocity, which is usually taken as zero (when the attitude is stable) or given by the upper guidance command; ω is the real-time angular velocity obtained in step S14.

[0032] S54. Control torque calculation: Based on the UAV's moment of inertia matrix J, the target angular acceleration is converted into the required airframe control torque: M att =J⋅α cmd .

[0033] S55. Torque command output: output the control torque M of the machine body. att The target control torque M in the current control cycle req The torque command is provided to step S3 for dynamic allocation of the target tension; the torque command update frequency is f. a It forms a cascade structure with the tension ring: the attitude ring is the outer ring and the tension ring is the inner ring.

[0034] The beneficial effects of the technical solution in this application are as follows: 1. Direct measurement and closed-loop control: By installing force sensors on each rotor, the thrust can be directly measured, forming a "measurement-comparison-drive" closed-loop control, which eliminates the uncertainty and time-varying nature of the speed-thrust mapping, and significantly improves the thrust control accuracy.

[0035] 2. Achieve optimal allocation for multiple objectives: By solving constrained quadratic programming problems online, motor saturation constraints are explicitly handled. While accurately meeting flight requirements, the load balance and energy efficiency of each rotor are taken into account, avoiding long-term overload of some motors and improving the overall system performance.

[0036] 3. Eliminate the "dragon drawing" phenomenon: Invent a dual-loop collaborative architecture of "force balance inner loop + attitude outer loop", with high-frequency force balance loop directly controlling the thrust, greatly shortening the response delay of the control loop, so that the thrust of each rotor always maintains dynamic balance, and the flight trajectory is smooth and oscillating.

[0037] 4. Improve mission execution quality: Aircraft attitude fluctuations and micro-vibrations are fundamentally suppressed, significantly improving the quality of aerial images, spraying operation precision, and mapping data accuracy, and fully leveraging the performance of external mission equipment. Attached Figure Description

[0038] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1 This is a schematic diagram of the overall control point architecture of a quadcopter unmanned aerial vehicle system provided by the present invention.

[0040] Figure 2 This is a schematic diagram of a quadcopter unmanned aerial vehicle system provided by the present invention.

[0041] Figure 3 This is a flowchart of a constraint optimization allocation method for a multi-rotor UAV rotor tension detection and balance control method.

[0042] Figure 4 This is a comparison curve of the anti-disturbance response of the control method of the present invention and the traditional PID control method.

[0043] Explanation of the labels in the diagram: 1. Rotor; 2. Rotor; 3. Rotor; 4. Rotor; 01. Force sensor; 02. Force sensor; 03. Force sensor; 04. Force sensor. Detailed Implementation

[0044] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0045] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0046] like Figure 1 As shown, a method for rotor tension detection and balance control of a multi-rotor unmanned aerial vehicle includes: S1. Distributed real-time acquisition of tension and inertia information.

[0047] S11. Force sensor arrangement and installation: Install one force sensor on the force transmission path of each rotor of the UAV. Furthermore, such as Figure 2 The diagram shown is a specific embodiment of a quadcopter unmanned aerial vehicle (UAV) system, illustrating that force sensor 01 is arranged in the force transmission path of rotor 1, force sensor 02 is arranged in the force transmission path of rotor 2, force sensor 03 is arranged in the force transmission path of rotor 3, and force sensor 04 is arranged in the force transmission path of rotor 4. Figure 1The diagram shown is a schematic diagram of the overall control point architecture of force sensors 01, 02, 03, and 04 corresponding to rotors 1, 2, 3, and 4 of a quadcopter unmanned aerial vehicle system provided in this application.

[0048] S12. Synchronous acquisition trigger: At the beginning of each control cycle, the main controller generates a synchronous trigger signal and simultaneously latches the original electrical signals of all force sensors; the control cycle is the tension control cycle T. f Satisfying T f =T a / k, where T a The attitude control cycle is k, where k is an integer greater than or equal to 2, preferably k=4 or k=5.

[0049] S13. Signal processing and conversion: The original electrical signal is sequentially subjected to bandpass filtering (passband 0.1Hz~50Hz, suppressing high-frequency vibration noise and low-frequency temperature drift), instrumentation amplification (gain 100~500 times), and analog-to-digital conversion (sampling accuracy ≥16 bits) to obtain the real-time thrust measurement value T of each rotor. i meas , where i=1,2,......,N, and N is the total number of rotors.

[0050] S14. Inertial data synchronous acquisition: At the moment the synchronization trigger signal is generated, data from the inertial measurement unit (IMU) is simultaneously read to acquire the real-time attitude angle Θ=[ϕ,θ,ψ] of the UAV. T Angular velocity ω = [ω x ,ω y ,ω z ] T and linear acceleration a=[a x ,a y ,a z ] T The data from the inertial measurement unit and the data from the force sensor are aligned using timestamps to ensure that the data synchronization error is less than 1ms.

[0051] S2. Calculation of total tensile force requirement.

[0052] S21. Flight command reception: Based on the current flight mode of the UAV, receive flight commands from the remote controller or autonomous navigation module; the flight commands include the target altitude h. cmd Target attitude angle Θ cmd =[ϕ cmd ,θ cmd ,ψ cmd ] T and target horizontal velocity v xy cmd =[v x cmd,v y cmd ] T .

[0053] S22. Position-velocity control: Based on the flight command and the current state, the target horizontal acceleration a is calculated by the position controller. xy cmd The vertical acceleration a of the target is calculated using the altitude controller. z cmd The position controller and height controller adopt a PID or cascaded PID structure.

[0054] S23. Total pulling force calculation, based on the total mass m of the UAV, gravitational acceleration g, and the vertical acceleration a of the target. z cmd Calculate the total thrust F required to maintain the current flight state. req =m⋅(g+a z cmd ).

[0055] S3. Dynamic allocation of target tension based on constraint optimization, combined with Figure 3 As shown.

[0056] Define the target thrust T for each rotor i target (i=1,2,…,N) represents the target thrust value of the i-th rotor in the current control cycle; T target =[T1 target T2 target ,...,T N target ] T Let be the target tension vector.

[0057] S31. Construction of the control efficiency matrix: Based on the geometric layout of the UAV and the rotation direction of each rotor, construct the control efficiency matrix B; the i-th column vector of matrix B is: B i =[1,−y i ,x i ,c i ⋅k Q ] T , where (x i ,y i Let be the position coordinates of the i-th rotor in the body coordinate system, and c be the position coordinates of the i-th rotor. i k is the rotor steering coefficient (+1 for clockwise, -1 for counterclockwise). Q The ratio of tensile force to counter-torque is obtained through offline calibration.

[0058] S32. Construction of the generalized force demand vector: The total tensile force demand F calculated in step S23 is used to construct the vector. reqThe target control torque M calculated in step S5 req =[M x req M y req M z req ] T The combination forms the generalized flight dynamics demand vector Γ. req =[F req M x req M y req M z req ] T This vector is related to the target thrust T of each rotor. i target Satisfying relation: Γ req =B⋅T target .

[0059] S33. Constrained optimization problem establishment, based on the rotational inertia matrix J∈R of the UAV. 3×3 Establish the following constrained optimization problem: Constraints: ; where T i nom =F req / N represents the nominal uniformly distributed tensile force, λ≥0 is the power consumption weighting coefficient, and T i max This represents the maximum available thrust of the i-th rotor in its current state. The maximum available tensile force T i max Estimate online using the following methods: T i max =k T ⋅ω i max (V bat ) 2 , where k T ω is the tensile coefficient. i max (V bat (V) represents the current battery voltage. bat The maximum speed of the motor is obtained by looking up the motor characteristic curve in a table. S34. Solve the optimization problem using the active set method or interior point method to solve the quadratic programming problem established in step S33; if a feasible solution exists, the target tension value T is obtained. i target If no feasible solution is found, then the priority fallback strategy is executed.

[0060] The priority fallback strategy includes, First priority: Meet the attitude torque requirement M req Allowable total tensile force F req Adjust within ±20%; Second priority: If there is still no feasible solution, maintain the target pull of the previous cycle and trigger an alarm.

[0061] S4. High-frequency tensile closed-loop tracking control.

[0062] S41. Calculation of thrust tracking error: Calculate the thrust tracking error of each rotor: e i =T i target -T i meas .

[0063] S42. Feedback control quantity calculation, based on the tension tracking error e i The feedback component U of the motor control command is calculated through the force balance loop controller. i fb The force balance loop controller employs a PID algorithm with anti-saturation integral: K p ,K i ,K d The proportional, integral, and differential gains are tuned using pole placement or trial and error.

[0064] S43. Feedforward compensation calculation, calculating the rate of change of the target tension: High-frequency noise is suppressed by using a low-pass filter (cutoff frequency 10~20Hz) to generate feedforward compensation components: K ff For feedforward gain, based on the first-order inertial model of the motor-propeller system. Tuning, take K ff ≈τ m / K m .

[0065] S44. Generate composite control command, add the feedback component and the feedforward compensation component, and perform output limiting: U i cmd =sat(U i fb +U i ff U min U max ), where sat(⋅) is the saturation function, U min U max These are the upper and lower limits of the ESC input signal.

[0066] S45. High-frequency drive execution, to control cycle T f (frequency f)f =1 / T f Preferred f f ≥2kHz) will transmit the composite control command U i cmd The signal is sent to the corresponding electronic speed controller, which drives the motor to adjust the real-time thrust measurement value T of the rotor. i meas Rapidly track target tension T i target .

[0067] S5. Independent attitude control loop operation.

[0068] S51. Attitude Data Reading: In a separate control thread, with attitude control period T a (frequency f) a =1 / T a Preferred f a =400Hz or f a =500Hz) to read attitude data from the inertial measurement unit; the attitude control period T a For the tension control period T f Integer multiples of T, i.e. a =k⋅T f .

[0069] S52. Attitude Deviation Calculation: Calculate the deviation between the current attitude angle and the target attitude angle: ΔΘ=Θ cmd −Θ, where Θ cmd Θ is the target attitude angle from the flight command, and Θ is the real-time attitude angle obtained in step S14.

[0070] Furthermore, when ΔΘ > 0, it means that the current angle is less than the target angle and positive adjustment is needed; when ΔΘ < 0, it means that the current angle is greater than the target angle and negative adjustment is needed; when ΔΘ = 0, it means that the target attitude has been achieved.

[0071] S53. Target angular acceleration calculation: Based on the attitude deviation, the target angular acceleration required to correct the attitude is calculated by the attitude loop controller: α. cmd =K Θ ⋅ΔΘ+K ω ⋅(ω cmd −ω), where K Θ ,K ω For the control gain of the attitude loop and angular velocity loop, a cascaded PID or PD control structure is adopted; ω cmd ω is the target angular velocity, which is usually taken as zero (when the attitude is stable) or given by the upper guidance command; ω is the real-time angular velocity obtained in step S14.

[0072] S54. Control torque calculation: Based on the UAV's moment of inertia matrix J, the target angular acceleration is converted into the required airframe control torque: M att =J⋅α cmd .

[0073] S55. Torque command output: output the control torque M of the machine body. att The target control torque M in the current control cycle req The torque command is provided to step S3 for dynamic allocation of the target tension; the torque command update frequency is f. a It forms a cascade structure with the tension ring: the attitude ring is the outer ring (slow speed), and the tension ring is the inner ring (fast speed).

[0074] One specific embodiment provided in this application is as follows: Figure 2 As shown, taking an X-shaped quadcopter UAV as an example, the steady-state performance of the quadcopter UAV in hovering state obtained by the method of this application is shown.

[0075] 1. Geometric parameter settings: Number of rotors: N=4; Rotor positions: (x1,y1)=(d,d),(x2,y2)=(−d,d),(x3,y3)=(−d,−d),(x4,y4)=(d,−d), where d=0.15m; Rotation direction: c1=−1,c2=+1,c3=−1,c4=+1 (adjacent rotors rotate in opposite directions); Thrust-to-torque coefficient: kQ=0.015m (obtained through offline calibration); UAV mass: m=1.5 kgm=1.5kg; Moment of inertia: Jxx=0.015kg·m², Jyy=0.015kg·m², Jzz=0.025kg·m²; Force sensor: Range 0~20N, accuracy 0.02%FS, sampling synchronization error <0.1ms; IMU: Gyroscope range ±2000° / s, accelerometer range ±16g, output frequency 1kHz.

[0076] 2. Control parameter settings: f f =1 / T f =2000Hz (period T) f =0.5ms); Attitude control frequency: f a =1 / T a =400Hz (period T) a =2.5ms, k=5); Force ring PID parameter settings: K p =2.0, K i =0.5, K d =0.01; Feedforward gain: Based on the first-order inertial model of the motor-propeller system (time constant τm = 80 ms, gain Km = 1.2), K is taken as... ff=τmKm≈0.067 s; Attitude loop controller: adopts a cascaded PID structure, outer loop gain K Θ =8.0, inner loop gain K ω =0.5; Constraint optimization weight: λ=0.1.

[0077] 3. The control efficiency matrix is ​​constructed as follows: .

[0078] Steady-state performance during hovering: Hovering settings: target height 2m, target attitude angle zero, no external disturbances.

[0079] Total tensile force requirement: F req =mg=1.5×9.8=14.7 N; Nominal uniform tensile force: T i nom =F req / 4=3.675 N; Constraint optimization solution: Since there is no attitude torque requirement, M req =[0,0,0] T The equality constraint is: Γ req =B⋅T target =[14.7,0,0,0] T The target tension force, T, is obtained by solving the problem. target =[3.675,3.675,3.675,3.675] T .

[0080] One specific embodiment provided in this application is as follows: Figure 4 The figure shows a comparison of the anti-disturbance response curves of the control method of this invention and the traditional PID control method, demonstrating the attitude response comparison of the two control methods when encountering gust disturbances. The roll angle response (dashed line) of the traditional PID control shows a large deviation (2.5°) after the disturbance, accompanied by multiple overshoot oscillations, requiring a relatively long time (2.4s) to recover stability. The response curve (solid line) of the method of this invention shows a maximum deviation of only 1.4°, and quickly and smoothly returns to steady state (0.7s) without oscillation, intuitively demonstrating the superior anti-disturbance capability of this invention.

[0081] It is worth noting that, such as Figures 1-4 As shown in the figure, this embodiment only illustrates one implementation of the multi-rotor UAV rotor tension detection and balance control method, but it does not mean that the multi-rotor UAV rotor tension detection and balance control method provided in this application has only this one implementation; the specific settings can be designed according to the needs of actual application.

[0082] In the description of this application, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "inner," "outer," "surface," etc., indicating orientation or positional relationships based on the orientation or positional relationships shown in the accompanying drawings, are only for the convenience of describing this application and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application. Those skilled in the art can understand the specific meaning of the above terms according to the specific circumstances. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. The term "multiple" means two or more, unless otherwise explicitly specified.

[0083] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "connected," "linked," "fixed," and "set" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art will understand the specific meaning of the above terms in this application based on the specific circumstances.

[0084] The embodiments described above are merely preferred embodiments of this application and are not intended to limit the scope of this application. Any modifications and improvements made to the technical solutions of this application by those skilled in the art without departing from the spirit of this application shall fall within the protection scope defined by the claims of this application.

Claims

1. A method for detecting rotor thrust and controlling balance in a multi-rotor unmanned aerial vehicle (UAV), characterized in that, include: S1. Distributed real-time acquisition of tension and inertia information; S2. Calculation of total tensile force requirement; S3. Dynamic allocation of target tension based on constraint optimization; S4. High-frequency tensile closed-loop tracking control; S5. Independent attitude control loop operation.

2. The method for detecting and controlling the rotor thrust of a multi-rotor UAV according to claim 1, characterized in that, Step S1 includes, S11. Force sensor arrangement and installation: Install one force sensor on the force transmission path of each rotor of the UAV. S12. Synchronous acquisition trigger: At the beginning of each control cycle, the main controller generates a synchronous trigger signal and simultaneously latches the original electrical signals of all force sensors. S13. The original electrical signal is filtered, amplified, and converted from analog to digital to obtain the real-time thrust measurement value T of each rotor. i meas , where i=1,2,......,N, and N is the total number of rotors; S14. Inertial data synchronous acquisition: At the moment the synchronization trigger signal is generated, the data of the inertial measurement unit is read simultaneously to acquire the real-time attitude angle, angular velocity and linear acceleration information of the UAV.

3. The method for detecting and controlling rotor thrust and balance of a multi-rotor UAV according to claim 2, characterized in that, Step S2 includes, S21. Flight command reception: Based on the current flight mode of the UAV, receive flight commands from the remote controller or autonomous navigation module; the flight commands include at least the target altitude, target attitude angle, and target speed; S22. Position-velocity control: Based on the flight command and the current state, the target horizontal acceleration a is calculated by the position controller. xy cmd The vertical acceleration a of the target is calculated using the altitude controller. z cmd ; S23. Total pulling force calculation, based on the total mass m of the UAV, gravitational acceleration g, and the vertical acceleration a of the target. z cmd Calculate the total thrust F required to maintain the current flight state. req =m⋅(g+a z cmd ).

4. The method for detecting and controlling the rotor thrust of a multi-rotor UAV according to claim 3, characterized in that, Step S3 includes defining the target thrust T for each rotor. i target (i=1,2,…,N) represents the target thrust value of the i-th rotor in the current control cycle; T target =[T1 target T2 target ,...,T N target ] T The target tension vector; S31. Construction of the control efficiency matrix: Based on the geometric layout of the UAV and the rotation direction of each rotor, construct the control efficiency matrix B; the i-th column vector of matrix B is: B i =[1,−y i ,x i ,c i ⋅k Q ] T , where (x i ,y i Let be the position coordinates of the i-th rotor in the body coordinate system, and c be the position coordinates of the i-th rotor. i k is the rotor steering coefficient. Q The ratio of tensile force to counter-torque is obtained through offline calibration. S32. Construction of the generalized force demand vector: The total tensile force demand F calculated in step S23 is used to construct the vector. req The target control torque M calculated in step S5 req =[M x req M y req M z req ] T The combination forms the generalized flight dynamics demand vector Γ. req =[F req M x req M y req M z req ] T This vector is related to the target thrust T of each rotor. i target Satisfying relation: Γ req =B⋅T target ; S33. Constrained optimization problem establishment, based on the rotational inertia matrix J∈R of the UAV. 3×3 Establish the following constrained optimization problem: Constraints: , ; where T i nom =F req / N represents the nominal uniformly distributed tensile force, λ≥0 is the power consumption weighting coefficient, and T i max This represents the maximum available thrust of the i-th rotor in its current state. S34. Solve the optimization problem using the active set method or interior point method to solve the quadratic programming problem established in step S33; if a feasible solution exists, the target tension value T is obtained. i target ; If no feasible solution is found, the priority fallback strategy is executed.

5. The method for detecting and controlling the rotor thrust of a multi-rotor UAV according to claim 4, characterized in that, Step S4 includes, S41. Calculation of thrust tracking error: Calculate the thrust tracking error of each rotor: e i =T i target -T i meas ; S42. Feedback control quantity calculation, based on the tension tracking error e i The feedback component U of the motor control command is calculated through the force balance loop controller. i fb The force balance loop controller employs a PID algorithm with anti-saturation integral: K p ,K i ,K d The proportional, integral, and differential gains are tuned using pole placement or trial and error. S43. Calculation of feedforward compensation amount; calculation of the rate of change of target tension. High-frequency noise is suppressed by a low-pass filter, and a feedforward compensation component is generated. K ff For the feedforward gain, based on the first-order inertial model of the motor-propeller system... Tuning, take K ff ≈τ m / K m ; S44. Generate composite control command, add the feedback component and the feedforward compensation component, and perform output limiting: U i cmd =sat(U i fb +U i ff U min U max ), where sat(⋅) is the saturation function, U min U max These are the upper and lower limits of the ESC input signal; S45. High-frequency drive execution, to control cycle T f The composite control command U i cmd The signal is sent to the corresponding electronic speed controller, which drives the motor to adjust the real-time thrust measurement value T of the rotor. i meas Rapidly track target tension T i target .

6. The method for detecting and controlling the rotor thrust of a multi-rotor UAV according to claim 5, characterized in that, Step S5 includes, S51. Attitude Data Reading: In a separate control thread, with attitude control period T a Read the attitude data from the inertial measurement unit; the attitude control period T a For the tension control period T f Integer multiples of; S52. Attitude Deviation Calculation: Calculate the deviation between the current attitude angle and the target attitude angle: ΔΘ=Θ cmd −Θ, where Θ cmd Θ is the target attitude angle from the flight command, and Θ is the real-time attitude angle obtained in step S14; S53. Target angular acceleration calculation: Based on the attitude deviation, the target angular acceleration required to correct the attitude is calculated by the attitude loop controller: α. cmd =K Θ ⋅ΔΘ+K ω ⋅(ω cmd −ω), where K Θ ,K ω ω represents the control gain of the attitude loop and the angular velocity loop; ω represents the real-time angular velocity obtained in step S14. S54. Control torque calculation: Based on the UAV's moment of inertia matrix J, the target angular acceleration is converted into the required airframe control torque: M att =J⋅α cmd ; S55. Torque command output: output the control torque M of the machine body. att The target control torque M in the current control cycle req This information is provided to step S3 for dynamic distribution of the target tension.

7. The method for detecting and controlling the rotor thrust of a multi-rotor UAV according to claim 2, characterized in that, The force sensor is a strain gauge type force sensor, including an elastic body base and a strain gauge bridge attached to the surface of the elastic body; one end of the elastic body base is fixed to the arm and the other end is fixed to the motor base, so that the tension generated by the rotor is completely transmitted to the elastic body to cause deformation; the strain gauge bridge adopts a full-bridge connection method and outputs a differential voltage signal proportional to the tension.

8. The method for detecting and controlling rotor thrust and balance of a multi-rotor UAV according to claim 4, characterized in that, The i-th column vector of matrix B is: B i =[1,−y i ,x i ,c i ⋅k Q ] T , where (x i ,y i Let be the position coordinates of the i-th rotor in the body coordinate system, and c be the position coordinates of the i-th rotor. i k is the rotor steering coefficient. Q The ratio of tensile force to counter-torque is obtained through offline calibration.

9. The method for detecting and controlling the rotor thrust of a multi-rotor UAV according to claim 4, characterized in that, The maximum available tensile force T i max Estimate online using the following methods: T i max =k T ⋅ω i max (V bat ) 2 , where k T ω is the tensile coefficient. i max (V bat (V) represents the current battery voltage. bat The maximum speed of the motor is obtained by looking up the motor characteristic curve in a table.

10. A method for detecting and controlling the rotor thrust of a multi-rotor unmanned aerial vehicle according to claim 4, characterized in that, The priority fallback strategy includes: first priority: satisfying the attitude torque requirement M req Allowable total tensile force F req Adjust within ±20%; Second priority: If there is still no feasible solution, maintain the target pull force of the previous cycle and trigger an alarm.

Images