Speed-adaptive variable-overload half-rolling inversion control method for unmanned aerial vehicle

Abstract

The invention discloses an unmanned aerial vehicle speed self-adaptive variable overload half-rolling inversion control method, which comprises the following steps: dividing half-rolling inversion maneuver into a rolling stage and an inversion pull-up stage, setting a switching logic, and adopting different control strategies in each stage, such as attack angle control in the rolling stage, rolling angle rate control based on a robust servo theory and the like, so that the speed self-adaptive variable overload half-rolling inversion maneuver of an unmanned aerial vehicle is realized. Overload control in the reverse pull-up stage and the like. Meanwhile, various technical means such as real-time updating of an aerodynamic model, introduction of machine learning auxiliary parameter estimation, multi-model fusion, a high-precision sensor, sensor data fusion and filtering, real-time compensation of measurement delay, an adaptive robust control algorithm, a multi-modal control strategy and interference suppression based on model prediction are adopted; the control precision, robustness and adaptability of the unmanned aerial vehicle in the semi-rolling inversion process are improved, and the application range of the unmanned aerial vehicle in a complex environment is expanded.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles, and specifically to a variable overload half-roll and inverted control method for speed adaptation of unmanned aerial vehicles. Background Technique

[0002] With the rapid development of unmanned aerial vehicle technology, unmanned aerial vehicles are increasingly widely used in military, civilian and other fields. In some complex tasks, such as military reconnaissance, disaster relief, etc., unmanned aerial vehicles need to have fast and flexible maneuvering capabilities. As a typical complex maneuver, the half-roll and inverted maneuver can enable the unmanned aerial vehicle to quickly change its flight direction and attitude, and has important application value. However, there are many deficiencies in the existing half-roll and inverted control methods for unmanned aerial vehicles. For example, the aerodynamic model is difficult to accurately describe the complex aerodynamic phenomena in actual flight, and the uncertainty of the airframe parameters affects the control accuracy; the accuracy limitation and measurement delay of the sensors lead to inaccurate and lagging control decisions; the robustness of the control algorithm is poor, and it is difficult to adapt to external disturbances and complex task scenarios; the real-time computing requirements are high, and the hardware performance and reliability restrict the control effect, etc.

[0003] Therefore, there is an urgent practical need to develop a variable overload half-roll and inverted control method for speed adaptation of unmanned aerial vehicles that can overcome the above problems. Summary of the Invention

[0004] The present invention provides a variable overload half-roll and inverted control method for speed adaptation of unmanned aerial vehicles, aiming to solve the problems proposed in the above background technique.

[0005] The present invention is implemented as follows. A variable overload half-roll and inverted control method for speed adaptation of unmanned aerial vehicles includes the following steps:

[0006] Maneuver phase division and switching: accurately divide the half-roll and inverted maneuver into a roll phase and an inverted pull-up phase; set the switching logic, when the absolute value of the roll angle of the unmanned aerial vehicle reaches a preset threshold, automatically switch from the roll phase to the inverted pull-up phase, and the preset threshold is 160°.

[0007] Implementation of control strategy: In the roll phase, angle-of-attack control is adopted longitudinally to keep the angle of attack at 0° to avoid the coupling risk of sideslip angle and angle of attack; the aileron uses roll angular rate control based on robust servo theory; the rudder maintains the sideslip angle at 0° through coordinated turning; the engine executes indicated airspeed closed-loop control to stabilize the flight speed; in the inverted pull-up phase, overload control is adopted longitudinally to protect the angle of attack, a feedforward value is designed to improve the control rapidity, and an angle-of-attack protection term is introduced into the inner-loop angular rate setting; the aileron keeps the wings of the aircraft level within a specific pitch angle range; the control mode of the roll angle is switched; the rudder adopts stability augmentation control to maintain the sideslip angle at 0°; the engine also adopts indicated airspeed closed-loop control;

[0008] Control law construction: Design the angle of attack control law, and calculate the elevator control command based on the angle of attack command, the current angle of attack, and the pitch rate; the roll rate control law calculates the aileron control signal according to the target value; the rudder stability augmentation control law calculates the control signal based on the sideslip angle and the yaw rate; the overload control law calculates the pitch rate command based on the overload command, the current true airspeed, etc., and then calculates the elevator control command.

[0009] Preferably, the aerodynamic model is updated in real time, that is, the air pressure, temperature, and wind speed data are collected in real time by the on-board sensors, and combined with the online identification algorithm to dynamically correct the aerodynamic model to fit the influence of the atmospheric environment and the UAV's own state changes during flight on the aerodynamic force.

[0010] Preferably, the introduction of machine learning for auxiliary parameter estimation is to collect a large amount of data on the attitude, speed, and overload of the UAV under different flight conditions, and use machine learning algorithms such as neural networks or support vector machines to estimate and update parameters such as the body mass and moment of inertia, improve the accuracy of parameter estimation, and reduce the interference of parameter uncertainty on the control performance.

[0011] Preferably, for the multi-model fusion, multiple aerodynamic sub-models are respectively constructed for different flight stages and working conditions of the UAV during takeoff, cruise, and half-roll inverted flight. During flight, according to the real-time state, the sub-models are intelligently switched or weighted and fused to improve the accuracy of the model within the entire flight envelope.

[0012] Preferably, high-precision sensors are used. A high-precision MEMS inertial measurement unit is selected to replace the traditional inertial measurement sensor, and a laser Doppler velocimeter is used to replace the traditional airspeed indicator to reduce the measurement error and provide more accurate original data for the control algorithm.

[0013] Preferably, for the sensor data fusion and filtering, algorithms such as Kalman filtering or particle filtering are used to fuse the multi-sensor data. At the same time, a low-pass filter digital filter is designed to filter the noise and outliers in real time to improve the reliability and accuracy of the data.

[0014] Preferably, for the real-time compensation of the measurement delay, by establishing a measurement delay model, combined with the sensor data transmission and processing delay, the prediction algorithm is used to estimate the true state of the current UAV, and the delayed data is corrected to the current moment to ensure that the control algorithm makes decisions based on the accurate real-time state and avoid control lag.

[0015] Preferably, for the adaptive robust control algorithm, the adaptive sliding mode control algorithm is adopted to estimate the upper bound of the disturbance online and dynamically adjust the sliding mode surface parameters, so that the system can still ensure the accuracy and stability of the half-roll inverted action in the face of external disturbances and system uncertainties.

[0016] Preferably, for the multi-modal control strategy, various control modes are designed for different task scenarios and interference conditions, including the conventional speed adaptive variable overload control mode and the anti-wind interference control mode under strong wind interference, and they are automatically switched according to the real-time state of the UAV to meet the complex task requirements.

[0017] Preferably, for the interference suppression based on model prediction, model predictive control technology is used to predict the state of the UAV and possible interferences in the future for a period of time, plan the control input in advance, optimize the control command, suppress the influence of the interference on the flight of the UAV, and ensure the successful completion of the half-roll inverted maneuver in the interference environment.

[0018] Due to the adoption of the above solutions, the beneficial effects of the present invention are as follows: By improving the aerodynamic model, enhancing the sensor performance, and adopting advanced control algorithms, the control error can be significantly reduced. In a complex wind field environment, the improved control method can reduce the speed control error to within ±2 m / s, the overload control error to within ±0.5 g, and the attitude control error to within ±2°, making the speed, overload, and attitude control of the UAV more accurate during the half-roll inverted process and improving the completion quality of the maneuver.

[0019] Measures such as the adaptive robust control algorithm, multi-modal control strategy, and interference suppression based on model prediction enable the UAV to better cope with external interferences and complex task scenarios. When encountering strong wind interference (wind speed above 15 m / s), the UAV can still complete the half-roll inverted maneuver according to the predetermined trajectory, while traditional methods may lead to maneuver failure or UAV out of control.

[0021] Upgrading the hardware equipment, adopting a distributed computing architecture, and means such as hardware fault diagnosis and fault-tolerant control effectively improve the real-time computing ability and reliability of the UAV control system. In the case of a high load on the flight control computer, the distributed computing architecture can shorten the execution time of the control algorithm by more than 30%, and the hardware fault diagnosis and fault-tolerant control can start the corresponding fault-tolerant measures within 1 s after detecting a fault to ensure the UAV continues to fly safely.

[0022] The improved control method enables the UAV to perform excellently under various complex environments and task requirements, thus expanding the application fields of the UAV. In areas such as urban canyons where the GPS signal is weak and the airflow is complex, the UAV can rely on the improved control technology to complete complex maneuvers such as half-roll inverted and achieve more accurate inspection, mapping, and other tasks, while traditional control methods are often difficult to handle in these scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 [[ID=2l]]It is a schematic diagram of the system flow of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0024] To make the objectives, technical solutions and advantages of the present invention more clear and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0025] As Figure 1 shown: A variable-overload half-roll inversion control method for the speed adaptation of an unmanned aerial vehicle, comprising the following steps:

[0026] Maneuver phase division and switching: The half-roll inversion maneuver is accurately divided into a roll phase and an inversion pull-up phase; a switching logic is set, and when the absolute value of the roll angle of the unmanned aerial vehicle reaches a preset threshold, it automatically switches from the roll phase to the inversion pull-up phase, and the preset threshold is 160°.

[0027] Implementation of control strategy: In the roll phase, angle of attack control is adopted longitudinally to keep the angle of attack at 0° to avoid the coupling risk of sideslip angle and angle of attack; the aileron uses roll angle rate control based on robust servo theory; the rudder maintains the sideslip angle at 0° through coordinated turning; the engine executes indicated airspeed closed-loop control to stabilize the flight speed; in the inversion pull-up phase, overload control is adopted longitudinally to protect the angle of attack, a feedforward value is designed to improve the control rapidity, and an angle of attack protection term is introduced into the inner-loop angle rate setting; the aileron keeps the wings of the aircraft level within a specific pitch angle range; the roll angle control method is switched; the rudder adopts stability augmentation control to maintain the sideslip angle at 0°; the engine also adopts indicated airspeed closed-loop control;

[0028] Construction of control law: Design an angle of attack control law to calculate the elevator control command according to the angle of attack command, the current angle of attack and the pitch angle rate; the roll angle rate control law calculates the aileron control signal according to the target value; the rudder stability augmentation control law calculates the control signal according to the sideslip angle and the yaw angle rate; the overload control law calculates the pitch angle rate command based on the overload command, the current true airspeed, etc., and then calculates the elevator control command.

[0029] The real-time updated aerodynamic model, that is, using on-board sensors to collect air pressure, temperature, and wind speed data in real time, and combining with an online identification algorithm to dynamically correct the aerodynamic model to fit the influence of the atmospheric environment during flight and the state change of the unmanned aerial vehicle itself on the aerodynamic force.

[0030] The introduction of machine learning-assisted parameter estimation is to collect a large amount of data on the attitude, speed, and overload of the unmanned aerial vehicle under different flight conditions, and use machine learning algorithms such as neural networks or support vector machines to estimate and update parameters such as the body mass and moment of inertia, improve the parameter estimation accuracy, and reduce the interference of parameter uncertainty on the control performance.

[0031] The multi-model fusion constructs multiple aerodynamic sub-models for different flight phases and working conditions of the UAV during takeoff, cruise, and half-roll inversion. During flight, according to the real-time state, the sub-models are intelligently switched or weighted and fused to improve the accuracy of the model within the entire flight envelope.

[0032] High-precision sensors are adopted. A high-precision MEMS inertial measurement unit is selected to replace the traditional inertial measurement sensor, and a laser Doppler velocimeter is used to replace the traditional airspeed indicator to reduce measurement errors and provide more accurate original data for the control algorithm.

[0033] The sensor data fusion and filtering uses algorithms such as Kalman filtering or particle filtering to fuse multi-sensor data. At the same time, a low-pass filter and a digital filter are designed to filter noise and outliers in real time, improving the reliability and accuracy of the data.

[0034] The real-time compensation for measurement delay is achieved by establishing a measurement delay model, combining the sensor data transmission and processing delays, using a prediction algorithm to estimate the true state of the current UAV, and correcting the delayed data to the current moment to ensure that the control algorithm makes decisions based on the accurate real-time state and avoid control lag.

[0035] The adaptive robust control algorithm adopts an adaptive sliding mode control algorithm to estimate the upper bound of the disturbance online and dynamically adjust the parameters of the sliding mode surface, enabling the system to ensure the accuracy and stability of the half-roll inversion action in the face of external disturbances and system uncertainties.

[0036] The multi-modal control strategy designs multiple control modes for different task scenarios and interference conditions, such as the conventional speed adaptive variable overload control mode and the anti-wind interference control mode under strong wind interference, and automatically switches according to the real-time state of the UAV to meet the requirements of complex tasks.

[0037] The interference suppression based on model prediction uses model predictive control technology to predict the state of the UAV and possible interferences in the future for a period of time, plan the control input in advance, optimize the control command, and suppress the influence of the interference on the UAV flight to ensure the successful completion of the half-roll inversion maneuver in an interference environment.

[0038] Example 1: Half-roll inversion control in a conventional flight environment

[0039] Preparation work: Select a multi-rotor UAV with a certain load capacity and flight performance, install sensors such as a high-precision MEMS inertial measurement unit and a laser Doppler velocimeter, and a flight control computer with strong computing power. According to the body parameters of the UAV and the requirements of the flight mission, initialize the relevant parameters of the aerodynamic model and the control algorithm.

[0040] Flight process: When the drone is in the cruise flight state, it receives a half-roll inverted maneuver command. First, it enters the roll phase. The control system maintains the angle of attack at 0° according to the angle-of-attack control law, quickly controls the ailerons to achieve roll through the roll angular rate control law based on robust servo theory, coordinates the turn with the rudder to maintain the sideslip angle at 0°, and the engine executes the indicated airspeed closed-loop control to stabilize the flight speed. When the absolute value of the roll angle reaches 160°, it switches to the inverted pull-up phase. Longitudinally, overload control is used to protect the angle of attack. The designed feedforward value enables the overload control to respond quickly, and at the same time, an angle-of-attack protection term is introduced to prevent the angle of attack from exceeding the limit. The ailerons keep the wings level, switch the roll angle control mode within a specific pitch angle range, the rudder stability augmentation control maintains the sideslip angle at 0°, and the engine continues with the indicated airspeed closed-loop control. Throughout the process, the aerodynamic model is updated in real time, machine learning is used to assist in parameter estimation, sensor data is fused and filtered, and the measurement delay is compensated in real time to ensure the accuracy and real-time performance of the control algorithm.

[0041] Result analysis: Through the collection and analysis of flight test data, the drone can accurately and stably complete the half-roll inverted maneuver in a conventional flight environment. The speed control error is within ±1.5 m / s, the overload control error is within ±0.3 g, and the attitude control error is within ±1.5°. This verifies the effectiveness of the control method of the present invention in a conventional environment.

[0042] Example 2: Half-roll inverted control in a strong wind interference environment

[0043] Preparation: Similar to Example 1, but a simulated strong wind environment is set up at the test site, and the wind speed can be adjusted to more than 15 m / s.

[0044] Flight process: After the drone enters the strong wind interference area, the adaptive robust control algorithm is activated to online estimate the upper bound of the interference and dynamically adjust the sliding mode surface parameters. At the same time, the multi-modal control strategy automatically switches to the anti-wind interference control mode, and according to the interference situation predicted by the model predictive control technology, the control input is planned in advance. Other control strategies such as stage division and switching, and the implementation of control laws in each stage are the same as in Example 1, and operations such as continuous update of the aerodynamic model, parameter estimation, and sensor data processing are carried out.

[0045] Result analysis: Despite being in a strong wind interference environment, the drone can still complete the half-roll inverted maneuver according to the predetermined trajectory. The speed control error is within ±2 m / s, the overload control error is within ±0.5 g, and the attitude control error is within ±2°. This shows that the control method of the present invention has good robustness and adaptability and can effectively cope with external interferences such as strong winds.

[0046] The above description of the embodiments is to enable those of ordinary skill in the art to understand and use the present invention. It is obvious that those skilled in the art can easily make various modifications to these embodiments and apply the general principles described herein to other embodiments without creative efforts. Therefore, the present invention is not limited to the above embodiments. Any improvements and modifications made by those skilled in the art based on the principles of the present invention without departing from the scope of the present invention should be within the protection scope of the present invention. The above is only the preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A variable overload half-roll and inverted control method for an unmanned aerial vehicle with speed self-adaptation, characterized in that, It includes the following steps: Maneuvering phase division and switching: Precisely divide the half-roll inverted maneuver into a roll phase and an inverted pull-up phase; Set the switching logic. When the absolute value of the roll angle of the UAV reaches a preset threshold, automatically switch from the roll phase to the inverted pull-up phase, and the preset threshold is 160°. Implementation of control strategy: In the roll phase, angle of attack control is adopted longitudinally to keep the angle of attack at 0° to avoid the coupling risk of sideslip angle and angle of attack; The aileron uses roll angular rate control based on robust servo theory; The rudder maintains the sideslip angle at 0° through coordinated turning; The engine executes indicated airspeed closed-loop control to stabilize the flight speed; In the inverted pull-up phase, overload control is adopted longitudinally to protect the angle of attack, design a feedforward value to improve the control rapidity, and introduce an angle of attack protection term to the inner-loop angular rate setting; The aileron keeps the aircraft wings level within a specific pitch angle range; Switch the roll angle control method; The rudder adopts stability augmentation control to maintain the sideslip angle at 0°; The engine also adopts indicated airspeed closed-loop control; Construction of control law: Design an angle of attack control law to calculate the elevator control command based on the angle of attack command, the current angle of attack, and the pitch angular rate; The roll angular rate control law calculates the aileron control signal according to the target value; The rudder stability augmentation control law calculates the control signal based on the sideslip angle and the yaw angular rate; The overload control law calculates the pitch angular rate command based on the overload command, the current true airspeed, etc., and then calculates the elevator control command.

2. The variable overload half-roll inverted control method for a drone with speed adaptation according to claim 1, characterized in that, The real-time updated aerodynamic model, that is, use on-board sensors to collect air pressure, temperature, and wind speed data in real time, and combine with online identification algorithms to dynamically correct the aerodynamic model to fit the influence of the atmospheric environment and the UAV's own state changes during flight on aerodynamic forces.

3. The variable overload half-roll and inverted control method for an unmanned aerial vehicle with speed adaptation according to claim 1, characterized in that, The introduction of machine learning-assisted parameter estimation is to collect a large amount of data on the attitude, speed, and overload of the UAV under different flight conditions, and use machine learning algorithms such as neural networks or support vector machines to estimate and update parameters such as the aircraft mass and moment of inertia, improve the parameter estimation accuracy, and reduce the interference of parameter uncertainty on control performance.

4. The variable-overload half-roll inverted control method for a drone with speed adaptability according to claim 1, characterized in that, The multi-model fusion is to construct multiple aerodynamic sub-models for different flight phases and working conditions of the UAV during takeoff, cruise, and half-roll inversion. During flight, according to the real-time state, intelligently switch or weighted fuse the sub-models to improve the accuracy of the model within the full flight envelope.

5. The variable overload half-roll inverted control method for the speed self-adaptive unmanned aerial vehicle according to claim 1, characterized in that, Adopt high-precision sensors, select high-precision MEMS inertial measurement units to replace traditional inertial measurement sensors, and use laser Doppler velocimeters to replace traditional airspeed indicators to reduce measurement errors and provide more accurate original data for control algorithms.

6. The variable overload half-roll inverted control method for a drone with speed self-adaptation according to claim 1, characterized in that The sensor data fusion and filtering is to fuse multi-sensor data using algorithms such as Kalman filtering or particle filtering, and at the same time design low-pass filters and digital filters to filter out noise and outliers in real time to improve the reliability and accuracy of the data.

7. The variable overload half-roll inverted control method for the speed self-adaptive unmanned aerial vehicle according to claim 1, wherein, The real-time compensation of measurement delay is to establish a measurement delay model, combine the sensor data transmission and processing delay, use a prediction algorithm to estimate the true state of the current UAV, and correct the delayed data to the current moment to ensure that the control algorithm makes decisions based on the accurate real-time state and avoid control lag.

8. The variable overload half-roll inverted control method for an unmanned aerial vehicle with speed self-adaptation according to claim 1, characterized in that, The described adaptive robust control algorithm adopts the adaptive sliding mode control algorithm to online estimate the upper bound of interference and dynamically adjust the sliding mode surface parameters, enabling the system to ensure the accuracy and stability of the half-roll inversion maneuver in the face of external interference and system uncertainties.

9. The variable overload half-roll inverted control method for an unmanned aerial vehicle with speed adaptability according to claim 1, characterized in that, The described multi-modal control strategy designs multiple control modes, namely the conventional speed adaptive variable overload control mode and the anti-wind interference control mode under strong wind interference, for different task scenarios and interference conditions, and automatically switches according to the real-time state of the UAV to meet the requirements of complex tasks.

10. The variable overload half-roll and inverted control method for an unmanned aerial vehicle with speed self-adaptation according to claim 1, characterized in that, The described interference suppression based on model prediction uses model predictive control technology to predict the UAV state and possible interference in the future for a period of time, plan the control input in advance, optimize the control command, and suppress the impact of interference on the UAV flight to ensure the successful completion of the half-roll inversion maneuver in an interference environment.

Images