CN-122018523-A - Cloud control-based unmanned aerial vehicle low-altitude flight intelligent management method and system

CN122018523ACN 122018523 ACN122018523 ACN 122018523ACN-122018523-A

Abstract

The invention provides a cloud control-based unmanned aerial vehicle low-altitude flight intelligent management method and system, which are characterized in that static obstacle information is acquired through a three-dimensional environment map, a global guiding corridor is generated by combining a task starting point, a task ending point and a flight energy consumption model, GPS, IMU and vision sensor data are fused in real time, an extended Kalman filter is utilized to estimate the pose and covariance matrix of an aerial vehicle, in the flight process, local obstacles are detected in the corridor, a flight track is predicted, a space-time obstacle avoidance area is constructed, an environment complexity index is calculated, when the collision risk exceeds a threshold value, global path re-planning is triggered, static obstacle repulsive force potential field safety margin is adjusted according to the pose covariance matrix, multiple target weights are optimized based on the complexity index, an optimal local speed vector and an optimal cost value are solved, the controller gain is adjusted by combining the optimal cost value, and a control instruction is generated, so that the aerial vehicle stably follows an optimal path in a complex environment.

Inventors

HAO SHENJUN
XING RUIXIANG
Qiang Shaobo
LIU PANPAN
LUO CHAOYANG

Assignees

郑州北斗低空经济发展有限公司

Dates

Publication Date: 20260512
Application Date: 20260122

Claims (10)

1. The intelligent management method for the low-altitude flight of the unmanned aerial vehicle based on cloud control is characterized by comprising the following steps of: The method comprises the steps of receiving a three-dimensional environment map and a flight task instruction which are issued by a cloud control platform and contain static obstacle information, and carrying out global path search based on a task starting point, a task ending point and a flight energy consumption model to generate a three-dimensional global guide corridor; when the conflict risk value of the predicted track of the obstacle and the guiding corridor exceeds a first preset threshold, triggering global path rescheduling by taking the current position of the unmanned aerial vehicle as a new starting point, and updating the path result after rescheduling to a cloud control platform; based on the local environment complexity index, adjusting the weight coefficients of the global guiding corridor gravitation, the static obstacle repulsive force and the space-time obstacle avoidance item in the multi-objective optimization problem, and solving to obtain an optimal local expected speed vector and an optimal cost value of the obtained solution; And adjusting gain parameters of the nonlinear controller according to the optimized cost value to generate a control instruction based on the deviation of the optimal local expected speed vector and the current pose state of the unmanned aerial vehicle.
2. The method of claim 1, wherein the performing a global path search based on the mission start point, the mission end point, and the flight energy expenditure model to generate a three-dimensional global guidance corridor comprises: In a cloud control platform or an unmanned aerial vehicle onboard computer, discretizing the three-dimensional environment map into grid units, and adopting a search algorithm to obtain path integration cost The minimum target is searched, where P is the path, ds is the path infinitesimal, And The cloud control platform is provided with preset weights; The energy consumption per unit distance According to the current load m of the unmanned aerial vehicle, presetting cruising speed And climbing angle Calculation in the form of Wherein The air resistance is air resistance, g is gravity acceleration; and expanding the searched optimal path point sequence outwards to form a tubular region serving as the three-dimensional global guiding corridor and issuing the tubular region to a local planning module.
3. The method of claim 1, wherein estimating the current pose state and pose estimation covariance matrix of the unmanned aerial vehicle using the extended kalman filter comprises: in the predicting step, based on the six-degree-of-freedom kinematic model of the unmanned aerial vehicle and the pose state at the last moment, predicting the prior pose state and the prior covariance at the current moment; In the updating step, the three-dimensional position provided by the global positioning system, the triaxial acceleration and the angular velocity provided by the inertial measurement unit and the depth information of the feature points provided by the vision sensor are used as measured values to calculate the Kalman gain, and the prior pose state and the prior covariance are updated to obtain the posterior pose state estimated value and the pose estimated covariance matrix at the current moment.
4. The method of claim 1, wherein the predicting the future trajectory of the obstacle to construct a spatiotemporal obstacle avoidance region comprises: Predicting a position sequence of each detected obstacle in a future period of time by adopting a motion model; and expanding each predicted position point into a safety envelop body, wherein the safety envelop body set of all predicted time steps forms a space-time obstacle avoidance area of the obstacle.
5. The method according to claim 1, wherein when the collision risk value of the predicted trajectory of the obstacle and the guiding corridor exceeds a first preset threshold value, triggering global path re-planning with the current position of the unmanned aerial vehicle as a new starting point comprises: Calculating the minimum distance between the predicted track of the obstacle and the central line of the three-dimensional global guiding corridor; Calculating a collision risk value based on the minimum distance, the collision risk value increasing with decreasing minimum distance; And triggering global path re-planning when the conflict risk value exceeds a first preset threshold value issued by the cloud control platform.
6. The method of claim 1, wherein adjusting the safety margin of the static obstacle repulsive potential field based on determinant values of the pose estimation covariance matrix comprises: estimating covariance matrix according to the pose Is a determinant value of (2) The safety margin of the static obstacle repulsive potential field is adjusted according to the following formula Wherein K and as a base safety margin Is a positive tuning parameter.
7. The method of claim 1, wherein adjusting weight coefficients of global guidance corridor gravitation, static obstacle repulsion, and space-time obstacle avoidance in a multi-objective optimization problem based on the local environmental complexity index comprises: Local environmental complexity index Sum of all obstacle speeds in the detection range of the sensor: Where N is the number of obstacles, A velocity vector for the ith obstacle; According to the index And adjusting the weight coefficient so as to increase the weight coefficient of the space-time obstacle avoidance item and reduce the weight coefficient of the global guide corridor gravitation item when the index is increased.
8. The method of claim 1, wherein adjusting the gain parameter of the nonlinear controller according to the optimization cost value to generate the control command comprises: adopting a proportional-integral-derivative PID controller as the nonlinear controller; according to the optimized cost value Reverse adjusting the proportional gain of the PID controller by the following functional relationship : Wherein A reference proportional gain set for the cloud control platform, Is a positive adjustment coefficient.
9. Unmanned vehicles low latitude flight intelligent management system based on high in clouds control, its characterized in that includes following module: the first generation module is used for receiving a three-dimensional environment map containing static obstacle information and a flight task instruction issued by the cloud control platform, performing global path search based on a task starting point, a task ending point and a flight energy consumption model, and generating a three-dimensional global guiding corridor; the method comprises the steps of integrating the global positioning system, an inertial measurement unit and vision sensor data in real time, estimating the current pose state and a pose estimation covariance matrix of the unmanned aerial vehicle by adopting an extended Kalman filter, and uploading the current pose state to a cloud control platform in real time for monitoring; The system comprises a three-dimensional global guiding corridor, a calculation module, a cloud control platform, a three-dimensional global guiding corridor, a three-dimensional global navigation corridor, a three-dimensional navigation corridor and a three-dimensional navigation corridor, wherein the three-dimensional global guiding corridor is used for guiding a vehicle to a three-dimensional navigation corridor; The adjustment module is used for adjusting the safety margin of the static obstacle repulsive force potential field according to the determinant value of the pose estimation covariance matrix, adjusting the weight coefficients of the global guiding corridor gravitation, the static obstacle repulsive force and the space-time obstacle avoidance item in the multi-objective optimization problem based on the local environment complexity index, and solving to obtain an optimal local expected speed vector and the optimal cost value of the obtained solution; the second generation module is used for adjusting gain parameters of the nonlinear controller according to the optimization cost value to generate a control instruction based on the deviation of the optimal local expected speed vector and the current pose state of the unmanned aerial vehicle.
10. The system of claim 9, wherein the generating a three-dimensional global guidance corridor based on the global path search for the mission start point, the mission end point, and the flight energy expenditure model comprises: In a cloud control platform or an unmanned aerial vehicle onboard computer, discretizing the three-dimensional environment map into grid units, and adopting a search algorithm to obtain path integration cost Searching for minimum target, wherein For a path, ds is the path infinitesimal, And The cloud control platform is provided with preset weights; The energy consumption per unit distance According to the current load m of the unmanned aerial vehicle, presetting cruising speed And climbing angle Calculation in the form of Wherein The air resistance is air resistance, g is gravity acceleration; and expanding the searched optimal path point sequence outwards to form a tubular region serving as the three-dimensional global guiding corridor and issuing the tubular region to a local planning module.

Description

Cloud control-based unmanned aerial vehicle low-altitude flight intelligent management method and system Technical Field The application belongs to the field of control, and particularly relates to an intelligent management method and system for low-altitude flight of an unmanned aerial vehicle based on cloud control. Background Unmanned aerial vehicles are increasingly widely used in the fields of mapping, logistics, inspection and the like, and generally rely on cloud platforms for unified supervision and task scheduling. At present, most of unmanned aerial vehicle flight control methods based on cloud architecture adopt layered frames, namely cloud global path planning and airborne local real-time obstacle avoidance are combined. The system acquires static map information of the cloud environment in advance, and adopts search rules such as A, RRT and the like to draw a collision-free path from a starting point to an end point. However, most of the methods are planned by taking a static environment as an object, and a rigid optimal path rather than a flexible guiding area is often generated by a cloud. In the on-board obstacle avoidance, common methods such as an artificial potential field method, a dynamic window method and the like are used for coping with unknown obstacles detected by a sensor. However, the artificial potential field method is easy to fall into a local minimum value, and most methods remain in reactive avoidance for the treatment of the obstacle, and lack of reliable prediction of the movement trend of the obstacle, so that the obstacle avoidance behavior is short-lived, and safety is difficult to ensure in a high-density scene. The obstacle avoidance strategy is too conservative, sacrificing flight efficiency, and when pose estimation accuracy is reduced, safety margin may not be enough to avoid risks. Regardless of whether the unmanned aerial vehicle is in an open and simple environment or faces an obstacle avoidance situation requiring complex maneuvering, the interaction between the cloud and the airborne terminal is often limited to a simple instruction issuing, and the airborne control response characteristic is constant. The control strategy can not be adjusted according to the current cloud task requirements and the real-time difficulty or cost of the local planning task, so that the performance of the aircraft under different environments is limited, and the balance among safety, stability and maneuverability is difficult to achieve. Disclosure of Invention Aiming at the problems, the invention provides an intelligent management method and system for low-altitude flight of an unmanned aerial vehicle based on cloud control, comprising the following steps: The method comprises the steps of receiving a three-dimensional environment map and a flight task instruction which are issued by a cloud control platform and contain static obstacle information, and carrying out global path search based on a task starting point, a task ending point and a flight energy consumption model to generate a three-dimensional global guide corridor; when the conflict risk value of the predicted track of the obstacle and the guiding corridor exceeds a first preset threshold, triggering global path rescheduling by taking the current position of the unmanned aerial vehicle as a new starting point, and updating the path result after rescheduling to a cloud control platform; based on the local environment complexity index, adjusting the weight coefficients of the global guiding corridor gravitation, the static obstacle repulsive force and the space-time obstacle avoidance item in the multi-objective optimization problem, and solving to obtain an optimal local expected speed vector and an optimal cost value of the obtained solution; And adjusting gain parameters of the nonlinear controller according to the optimized cost value to generate a control instruction based on the deviation of the optimal local expected speed vector and the current pose state of the unmanned aerial vehicle. Optionally, the generating a three-dimensional global guiding corridor based on the task starting point, the task ending point and the flight energy consumption model to perform global path searching includes: In a cloud control platform or an unmanned aerial vehicle onboard computer, discretizing the three-dimensional environment map into grid units, and adopting a search algorithm to obtain path integration cost The minimum target is searched, where P is the path, ds is the path infinitesimal,AndThe cloud control platform is provided with preset weights; The energy consumption per unit distance According to the current load m of the unmanned aerial vehicle, presetting cruising speedAnd climbing angleCalculation in the form ofWhereinThe air resistance is air resistance, g is gravity acceleration; and expanding the searched optimal path point sequence outwards to form a tubular region serving as the three-dimensional global guidi