An Optimization Algorithm for UAV Safety Scheduling

Abstract

The present invention discloses an optimization algorithm for the safe scheduling of unmanned aerial vehicles. The present invention relates to the technical field of warehouse scheduling. Based on the position, shape, and size information of the shelves and mechanical equipment in the warehouse, a warehouse environment is modeled to establish a preliminary environment model. The lidar carried by the unmanned aerial vehicle is used to scan the environment to obtain the environmental data in the warehouse, and the preliminary environment model is combined to update the environment model in real time. This optimization algorithm for the safe scheduling of unmanned aerial vehicles significantly improves the accuracy and efficiency of the flight path by integrating path planning technology. Combining the actual conditions of the terrain and obstacle distribution in the warehouse area, an optimal flight path is customized for the unmanned aerial vehicle, reducing redundancy and waste during the flight. At the same time, it can also dynamically adjust the path according to real-time environmental changes to effectively respond to emergencies, thereby further improving the flight efficiency and safety.

Description

Technical Field

[0001] The present invention relates to the technical field of warehousing scheduling, and specifically to an optimized algorithm for the safe scheduling of unmanned aerial vehicles (UAVs). Background Art

[0002] With the booming development of e-commerce and the continuous improvement of consumers' requirements for logistics speed, the logistics industry is facing unprecedented challenges. Digital transformation is an important trend that all industries are actively promoting. In the logistics industry, digital transformation means achieving the intelligence, automation, and visualization of warehousing, transportation, distribution, etc. by introducing advanced information technologies and intelligent devices. As an important part of digital transformation, the application of UAV technology in the warehousing field can significantly improve management efficiency and accuracy and reduce labor costs.

[0003] The warehousing environment is relatively complex. When using UAVs to identify and schedule goods in the warehousing environment, due to obstacles such as shelves, mechanical equipment, and personnel in the warehouse, it will interfere with or pose a threat to the flight of UAVs. Moreover, during the flight of UAVs, encountering obstacles with irregular shapes will further increase the flight risk of UAVs. Therefore, how to achieve the positioning and safe scheduling of UAVs with the ultimate goal of warehousing safety, relying on high-precision positioning and navigation and radar obstacle avoidance technologies to prevent the occurrence of collision accidents is the problem we need to solve. For this reason, an optimized algorithm for the safe scheduling of UAVs is proposed herein. Summary of the Invention

[0004] To achieve the above objectives, the present invention is realized through the following technical solutions: An optimized algorithm for the safe scheduling of UAVs, including the following steps:

[0005] Step 1: Based on the position, shape, and size information of shelves and mechanical equipment in the warehouse, model the warehousing environment to establish a preliminary environmental model;

[0006] Step 2: Use the lidar carried by the UAV to scan the environment, obtain the environmental data in the warehouse, and combine it with the preliminary environmental model to update the environmental model in real time. Among them, the environmental data in the warehouse includes moving shelves or personnel;

[0007] Step 3: According to the real-time updated environmental model and obstacle information, in cooperation with the positioning and navigation technologies of GPS and IMU, obtain the position, speed, and attitude information of the UAV in real time, and use a path planning algorithm to formulate a flight path for the UAV;

[0008] Step 4: Based on the lidar-based radar obstacle avoidance technology, continuously monitor the obstacles around the UAV. Through the radar beam scanning and reflection of the lidar, obtain the distance, speed, and direction information of the obstacles. Combine the preset safety distance threshold and radar data to adjust the flight speed and direction of the UAV, avoiding collisions with obstacles;

[0009] Step 5: According to the results of the path planning, navigate and control the UAV, and continuously monitor the flight status of the UAV and environmental changes in real time. Dynamically adjust the flight plan and obstacle avoidance strategy of the UAV to ensure warehouse safety.

[0010] Preferably, in Step 1, the process of establishing the preliminary environmental model includes:

[0011] Collect detailed information on all fixed and moving obstacles in the warehouse, including the locations, shapes, and sizes of shelves, mechanical equipment, and personnel passages. Conduct on-site measurements and refer to the warehouse layout plan to obtain the precise location information of the shelves and mechanical equipment, and record the shapes, sizes, and relative positions between the shelves and mechanical equipment;

[0012] Divide the warehouse environment into different areas, namely storage areas, operation areas, and personnel passages, and set the personnel passages as no-fly zones;

[0013] Preprocess the collected data to ensure the accuracy and integrity of the data. For each obstacle, create three-dimensional models of warehouse elements such as shelves, mechanical equipment, and personnel passages in the modeling tool according to the preprocessed data, ensuring that the shapes, sizes, and positions of the models are consistent with the actual situation;

[0014] Divide the warehouse space into grids, and mark the shelves, mechanical equipment, and other obstacles in each grid space. Then set the height limits, obstacle types, and safety distance parameters for each grid space;

[0015] Integrate all the constructed three-dimensional models and grid cells into a unified model to obtain the preliminary environmental model.

[0016] Preferably, in Step 2, the process of real-time updating the environmental model includes:

[0017] Use UAV route planning software on the computer side. According to the warehouse layout plan and actual environment, plan the flight route of the UAV. The UAV takes off according to the planned route and uses lidar to scan the warehouse environment during flight. The lidar measures the distance and obtains three-dimensional information of the warehouse environment by emitting laser beams and receiving the reflected signals;

[0018] The scanned data of the warehouse environment is transmitted to the computer in real time for storage and processing, and the Shuttle and gAirHawk software are used to solve the original data;

[0019] In the Shuttle software, data sorting and POS solution steps are carried out to obtain a POS result file. The POS result file contains the position and attitude information of the UAV. In the gAirHawk software, the POS result file, Lidar original data file and photo file are imported, and coordinate system setting, flight path division and point cloud calculation steps are carried out to obtain three-dimensional laser point cloud data, and then the point cloud volume, slice area and height range of the point cloud are analyzed;

[0020] The three-dimensional laser point cloud data obtained by the solution is compared and fused with the preliminary environment model, and the preliminary environment model is updated in real time according to the position and shape information of the mobile shelves or personnel in the point cloud data, and the updated environment model is obtained. The updated environment model will more accurately reflect the actual situation in the warehouse and provide strong support for subsequent environmental monitoring and management.

[0021] Preferably, the point cloud volume is calculated by the slicing method, and its calculation expression is:

[0022] ;

[0023] where V is the point cloud volume, is the slice area of the i-th slice, N is the number of slices, and h is the slice spacing;

[0024] The calculation expression of the slice area is:

[0025] ;

[0026] where, is the slice area of the i-th slice, m is the number of boundary points, is the boundary point of slice i;

[0027] The calculation expression of the height range of the point cloud is:

[0028] ;

[0029] where H is the height range of the point cloud, is the maximum height of the points in the point cloud, is the minimum height of the points in the point cloud.

[0030] Preferably, in step three, the process of formulating the UAV flight path includes:

[0031] The position information of the UAV is obtained by using GPS technology, and the angular velocity and acceleration of the UAV are measured using an IMU (Inertial Measurement Unit). Combining the initial attitude information, the real-time attitude and speed information of the UAV are deduced. The data of GPS and IMU are fused to improve the accuracy and stability of UAV positioning and navigation. Among them, GPS provides the longitude, latitude and altitude position data of the UAV;

[0032] According to the real-time updated environmental model and obstacle information, as well as the current position, speed and attitude information of the UAV, the Dijkstra algorithm is selected for path planning;

[0033] In the path planning algorithm, the current position coordinates of the UAV, the target position coordinates, and the real-time updated environmental model and obstacle information are input, and the turning angle and reference distance of the current position are analyzed to calculate an optimal path that avoids obstacles and meets the flight safety requirements. Among them, the flight path includes the starting point, the ending point, intermediate nodes, flight altitude and speed parameters;

[0034] The flight path is converted into control commands for the UAV, and the flight mission is executed through the control system of the UAV. During the flight process, the UAV adjusts its flight attitude and speed in real time through the closed-loop control algorithm of PID control to ensure stable flight along the planned path. By integrating path planning technology, the accuracy and efficiency of the flight path are significantly improved. Combining the actual conditions of the terrain and obstacle distribution in the warehousing area, an optimal flight path is customized for the UAV, reducing redundancy and waste during the flight process. At the same time, it can also dynamically adjust the path according to real-time environmental changes to effectively respond to emergencies, thereby further improving flight efficiency and safety.

[0035] Preferably, the cost function expression of the optimal path is:

[0036] ;

[0037] ;

[0038] ;

[0039] Among them, is the cost function of the optimal path, indicating minimizing the cost function, is the actual cost from the starting point to the current position coordinates, is the heuristic estimated cost from the current position coordinates to the target position coordinates, is the current position coordinates, is the target position coordinates, is the turning angle of the current position coordinates, is the cost weight coefficient of power consumption, is the cost weight coefficient of the steering angle, is the reference distance for scaling in the logarithmic function, increases as the distance between nodes increases, increases as the distance from the current node to the target point increases, and are positive constants used to adjust the weights of power consumption and steering angle in the cost function.

[0040] Preferably, in the fourth step, the adjustment process of the flight speed and direction of the UAV includes:

[0041] During the flight of the UAV, the lidar emits laser beams to the target area through the laser diodes inside it. Using phased array technology, by adjusting the phase difference of the laser beams, rapid scanning of the beams is achieved to cover the target area, enabling the lidar to obtain three-dimensional information of the target area;

[0042] The receiver of the lidar receives the reflected laser signals, thereby obtaining obstacle information and converting it into an electrical signal for subsequent signal amplification, filtering, and demodulation steps to identify the position, shape, and size information of the obstacles;

[0043] After identifying the obstacles, re-plan the path according to the current environmental information and the motion state of the UAV, find an optimal path from the current position to the target position while avoiding all known obstacles, and monitor the surrounding obstacle information in real time. Combine the preset safety distance threshold to calculate the safety factor of the flight path to determine whether to re-plan the flight path to ensure flight safety;

[0044] Preset a warning threshold, compare the safety factor with the warning threshold, and according to the comparison result, determine whether to issue a warning to remind the staff to intervene, and then adjust the path planning of the UAV;

[0045] According to the result of the path planning, navigate and control the UAV to achieve safe obstacle avoidance, including adjusting the flight speed and direction of the UAV to ensure its stable flight along the planned path while avoiding all obstacles. By real-time monitoring and dynamically adjusting the flight path, the safety and reliability of the UAV in complex environments are significantly improved. Combining the radar obstacle avoidance technology of the lidar and the high-precision positioning and navigation technology, it can obtain the position, speed, and attitude information of the UAV in real time, as well as the detailed information of the surrounding obstacles, and re-plan the path when necessary to ensure that the UAV can respond quickly when encountering unknown obstacles or environmental changes, thereby reducing the probability of accidents and improving the reliability of task execution.

[0046] Preferably, the specific process of the staff's intervention includes:

[0047] Monitor the safety factor of the flight path in real time, compare it with the preset warning threshold, and analyze whether the safety factor is lower than the preset warning threshold;

[0048] If the safety factor is lower than the warning threshold, it indicates that the safety of the current path is insufficient. Mark the current path and issue a warning;

[0049] Based on the triggered warning mechanism, issue a prompt on the console to remind the staff to intervene and take actions, and start the path replanning program. Input the latest environmental information and obstacle data, and use the path planning algorithm to calculate a new safe path until the safety factor exceeds the warning threshold, that is, determine the result of path planning.

[0050] Preferably, the calculation expression of the safety factor is:

[0051] ;

[0052] where S is the safety factor, and the value range of S is between 0 and 1, is the preset safety distance threshold, which is used to ensure that there is enough distance between the drone and the obstacle to avoid collision, is the distance from the drone to the nearest obstacle, The smaller the value of, the closer the drone is to the obstacle, and the lower the safety factor. v is the current speed of the drone, which affects the calculation of the safety factor. Since the faster the speed, the shorter the reaction time and the lower the safety factor, is the maximum speed of the drone, which is used to standardize the current speed.

[0053] Preferably, in step five, the process of navigating and controlling the drone includes:

[0054] Convert the path planning result into navigation instructions, send them to the flight control system of the drone, and adjust the flight attitude and speed of the drone through the actuators (motors, propellers, and servos);

[0055] Real-time monitor the flight state information of the drone's position, speed, and attitude to make the drone fly along the predetermined path;

[0056] Use lidar to perceive the environment around the drone in real time, including static and dynamic obstacles, and dynamically adjust the flight plan of the drone according to the real-time monitored environmental changes.

[0057] The present invention provides a drone safety scheduling optimization algorithm. It has the following beneficial effects:

[0058] 1. The UAV safety scheduling optimization algorithm significantly improves the accuracy and efficiency of the flight path by integrating path planning technology. Considering the actual conditions of the terrain and obstacle distribution in the storage area, it customizes the optimal flight path for the UAV, reducing redundancy and waste during flight. At the same time, it can dynamically adjust the path according to real-time environmental changes, effectively coping with emergencies, thereby further improving flight efficiency and safety.

[0059] 2. The UAV safety scheduling optimization algorithm significantly improves the safety and reliability of the UAV in complex environments through real-time monitoring and dynamic adjustment of the flight path. Combining lidar radar obstacle avoidance technology and high-precision positioning and navigation technology, it can obtain real-time information on the UAV's position, speed, and attitude, as well as detailed information on surrounding obstacles, and re-plan the path when necessary, ensuring that the UAV can respond quickly when encountering unknown obstacles or environmental changes, thereby reducing the probability of accidents and improving the reliability of task execution. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 is a method flowchart of a UAV safety scheduling optimization algorithm of the present invention;

[0061] Figure 2 is a flowchart of real-time updating of the environmental model of the present invention;

[0062] Figure 3 is a flowchart of formulating the UAV flight path of the present invention;

[0063] Figure 4 is a flowchart of adjusting the UAV flight speed and direction of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0064] The present invention will be further described in detail below with reference to the drawings and specific embodiments. The embodiments of the present invention are given for purposes of illustration and description, and are not intended to be exhaustive or to limit the invention to the disclosed form. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments are selected and described in order to better illustrate the principles of the present invention and its practical applications, and to enable those of ordinary skill in the art to understand the present invention and design various embodiments with various modifications suitable for specific purposes.

[0065] The first embodiment, as Figures 1 to 3 shown, the present invention provides a technical solution: a UAV safety scheduling optimization algorithm, including the following steps:

[0066] Step 1: Based on the position, shape, and size information of the shelves and mechanical equipment in the warehouse, model the warehousing environment to establish a preliminary environment model. Collect detailed information on all fixed and moving obstacles in the warehouse, including the position, shape, and size of the shelves, mechanical equipment, and personnel passages. Conduct on-site measurements and consult the warehouse layout plan to obtain the accurate position information of the shelves and mechanical equipment, and record the shape, size, and relative position relationship between the shelves and mechanical equipment. Divide the warehouse environment into different areas, namely the storage area, operation area, and personnel passage, and set the personnel passage as a no-fly zone. Preprocess the collected data to ensure its accuracy and integrity. For each obstacle, create 3D models of the warehouse elements such as shelves, mechanical equipment, and personnel passages in the modeling tool, ensuring that the shape, size, and position of the models match the actual situation. Divide the warehouse space into grids and mark the shelves, mechanical equipment, and other obstacles in each grid space, and then set the height limit, obstacle type, and safety distance parameters for each grid space. Integrate all the constructed 3D models and grid cells into a unified model to obtain a preliminary environment model;

[0067] Step 2: Use the lidar carried by the drone to scan the environment to obtain the environmental data in the warehouse, and combine it with the preliminary environment model to update the environment model in real time. Among them, the environmental data in the warehouse includes moving shelves or personnel. Use the drone route planning software on the computer side to plan the flight route of the drone according to the warehouse layout plan and the actual environment. The drone takes off according to the planned route and uses the lidar to scan the warehouse environment during flight. The lidar measures the distance and obtains the 3D information of the warehouse environment by emitting laser beams and receiving the reflected signals. Transmit the scanned data of the warehouse environment to the computer side in real time for storage and processing, and use Shuttle and gAirHawk software to solve the original data. In the Shuttle software, perform data sorting and POS solution steps to obtain the POS result file, which contains the position and attitude information of the drone. In the gAirHawk software, import the POS result file, Lidar original data file, and photo file, and perform coordinate system setting, flight route division, and point cloud calculation steps to obtain the 3D laser point cloud data. Then analyze the point cloud volume, slice area, and height range of the 3D laser point cloud data. Compare and fuse the solved 3D laser point cloud data with the preliminary environment model, and update the preliminary environment model in real time according to the position and shape information of the moving shelves or personnel in the point cloud data to obtain the updated environment model. The updated environment model will more accurately reflect the actual situation in the warehouse and provide strong support for subsequent environmental monitoring and management;

[0068] Furthermore, the point cloud volume is calculated using the slicing method, and its calculation expression is:

[0069] ;

[0070] where V is the point cloud volume, is the slicing area of the i-th slice, N is the number of slices, and h is the slice spacing;

[0071] The calculation expression for the slicing area is:

[0072] ;

[0073] where, is the slicing area of the i-th slice, m is the number of boundary points, are the boundary points of slice i;

[0074] The calculation expression for the height range of the point cloud is:

[0075] ;

[0076] where H is the height range of the point cloud, is the maximum height of the points in the point cloud, is the minimum height of the points in the point cloud;

[0077] Step 3: According to the real-time updated environmental model and obstacle information, in cooperation with the positioning and navigation technologies of GPS and IMU, the position, speed and attitude information of the UAV are obtained in real time. The path planning algorithm is used to formulate a flight path for the UAV. The position information of the UAV is obtained by using GPS technology, and the angular velocity and acceleration of the UAV are measured by using IMU (Inertial Measurement Unit). Combining with the initial attitude information, the real-time attitude and speed information of the UAV are deduced. The data of GPS and IMU are fused to improve the accuracy and stability of UAV positioning and navigation. Among them, GPS provides the longitude, latitude and altitude position data of the UAV. According to the real-time updated environmental model and obstacle information, as well as the current position, speed and attitude information of the UAV, the Dijkstra algorithm is selected for path planning. In the path planning algorithm, the current position coordinates of the UAV, the target position coordinates, and the real-time updated environmental model and obstacle information are input, and the turning angle and reference distance of the current position are analyzed to calculate an optimal path that avoids obstacles and meets the flight safety requirements. Among them, the flight path includes the starting point, the ending point, intermediate nodes, flight altitude and speed parameters. The flight path is converted into control instructions for the UAV, and the flight mission is executed through the control system of the UAV. During the flight, the UAV adjusts the flight attitude and speed in real time through the closed-loop control algorithm of PID control to ensure stable flight along the planned path. By integrating path planning technology, the accuracy and efficiency of the flight path are significantly improved. Combining with the actual conditions of the terrain and obstacle distribution in the storage area, an optimal flight path is customized for the UAV, reducing redundancy and waste during the flight. At the same time, the path can be dynamically adjusted according to real-time environmental changes to effectively respond to emergencies, thereby further improving flight efficiency and safety;

[0078] Furthermore, the cost function expression of the optimal path is:

[0079] ;

[0080] ;

[0081] ;

[0082] Among them, is the cost function of the optimal path, indicating minimizing the cost function, is the actual cost from the starting point to the current position coordinates, is the heuristic estimated cost from the current position coordinates to the target position coordinates, is the current position coordinates, is the target position coordinates, is the turning angle of the current position coordinates, is the cost weight coefficient of power consumption, is the cost weight coefficient of the steering angle, is the reference distance for scaling in the logarithmic function, increases as the distance between nodes increases, increases as the distance from the current node to the target point increases, and are positive constants used to adjust the weights of power consumption and steering angle in the cost function;

[0083] Step Four: Based on the lidar-based radar obstacle avoidance technology, the situation of obstacles around the UAV is monitored in real time. Through the radar beam scanning and reflection of the lidar, the distance, speed, and direction information of the obstacles are obtained. Combining the preset safety distance threshold and radar data, the flight speed and direction of the UAV are adjusted to avoid collisions with obstacles;

[0084] Step Five: According to the result of path planning, the UAV is navigated and controlled, and the flight state and environmental changes of the UAV are monitored in real time. The flight plan and obstacle avoidance strategy of the UAV are dynamically adjusted to ensure warehouse safety.

[0085] Second Embodiment, based on the First Embodiment, please refer to Figure 4 as shown. In Step Four, the adjustment process of the flight speed and direction of the UAV includes:

[0086] During the flight of the UAV, the lidar emits laser beams to the target area through its internal laser diodes. Using phased array technology, by adjusting the phase difference of the laser beams, rapid scanning of the beams is achieved to cover the target area, enabling the lidar to obtain three-dimensional information of the target area. The receiver of the lidar receives the reflected laser signals, thereby obtaining obstacle information and converting it into electrical signals for subsequent signal amplification, filtering, and demodulation steps to identify the position, shape, and size information of the obstacles. After identifying the obstacles, based on the current environmental information and the motion state of the UAV, path planning is re-performed to find an optimal path from the current position to the target position while avoiding all known obstacles, and the obstacle information around is monitored in real time. Combining the preset safety distance threshold, the safety factor of the flight path is calculated to determine whether to re-plan the flight path to ensure flight safety. A warning threshold is preset in advance, and the safety factor is compared with the warning threshold. According to the comparison result, it is determined whether to issue a warning to remind the staff to intervene, and then the path planning of the UAV is adjusted. According to the result of the path planning, the UAV is navigated and controlled to achieve safe obstacle avoidance, including adjusting the flight speed and direction of the UAV to ensure its stable flight along the planned path while avoiding all obstacles;

[0087] Further, the specific process of the staff's intervention includes:

[0088] The safety factor of the flight path is monitored in real time and compared with a preset warning threshold. It is analyzed whether the safety factor is lower than the preset warning threshold. If the safety factor is lower than the warning threshold, it indicates that the safety of the current path is insufficient. The current path is marked and a warning is issued. Based on the triggered warning mechanism, a prompt is sent on the console to remind the staff to intervene and take actions, and a path replanning program is started. The latest environmental information and obstacle data are input, and a new safe path is calculated using the path planning algorithm until the safety factor exceeds the warning threshold, that is, the result of the path planning is determined. By monitoring and dynamically adjusting the flight path in real time, the safety and reliability of the unmanned aerial vehicle (UAV) in complex environments are significantly improved. Combining the radar obstacle avoidance technology of lidar and the high-precision positioning and navigation technology, it can obtain the position, speed, and attitude information of the UAV in real time, as well as detailed information about surrounding obstacles, and replan the path when necessary, ensuring that the UAV can respond quickly when encountering unknown obstacles or environmental changes, thereby reducing the probability of accidents and improving the reliability of mission execution;

[0089] The calculation expression of the safety factor is:

[0090] ;

[0091] where S is the safety factor, and the value range of S is between 0 and 1. is the preset safety distance threshold, which is used to ensure that there is enough distance between the UAV and the obstacle to avoid collision. is the distance from the UAV to the nearest obstacle. The smaller the value of, the closer the UAV is to the obstacle, and the lower the safety factor. v is the current speed of the UAV, which affects the calculation of the safety factor. Since the faster the speed, the shorter the reaction time and the lower the safety factor. is the maximum speed of the UAV, which is used to standardize the current speed to ensure that the safety factor calculation is not affected by the speed unit. When is much greater than , that is, the UAV is far from the obstacle, is close to 0, and the safety factor S is close to 1, indicating very safe. When is close to , the safety factor S decreases rapidly, indicating a decrease in safety. For the part where the UAV speed v is close to the maximum speed , the value of the logarithmic function increases, and the safety factor S decreases, indicating a decrease in safety during high-speed flight;

[0092] In step five, the process of navigating and controlling the UAV includes:

[0093] Convert the path planning result into navigation instructions, send them to the flight control system of the drone, and adjust the flight attitude and speed of the drone through actuators (motors, propellers, and servos). Real-time monitor the flight state information of the drone's position, speed, and attitude to enable the drone to fly along the predetermined path. Use lidar to perceive the environment around the drone in real time, including static and dynamic obstacles, and dynamically adjust the flight plan of the drone according to the real-time monitored environmental changes.

[0094] Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art and related fields without creative efforts shall fall within the protection scope of the present invention. The structures, devices, and operation methods not specifically described and explained in the present invention shall be implemented according to the conventional means in the art without special instructions and limitations.

Claims

1. A drone safety scheduling optimization algorithm, characterized in that: The following steps are involved: Step 1: Model the storage environment based on the location, shape and size information of the shelves and mechanical equipment in the warehouse, and establish a preliminary environment model; Step 2: Use the laser radar carried by the drone to scan the environment, obtain the environmental data in the warehouse, and update the environmental model in real time based on the preliminary environmental model. The environmental data in the warehouse includes moving shelves or personnel; Step 3: Based on the real-time updated environment model and obstacle information, in conjunction with GPS and IMU positioning and navigation technology, obtain the position, speed and attitude information of the drone in real time, and use the path planning algorithm to formulate a flight path for the drone; Step 4: Obtain the distance, speed and direction information of the obstacle through the laser radar beam scanning and reflection, and adjust the flight speed and direction of the drone based on the preset safety distance threshold and radar data; Step 5: Navigate and control the UAV based on the path planning results, monitor the flight status and environmental changes of the UAV in real time, and dynamically adjust the flight plan and obstacle avoidance strategy of the UAV; In the step 1, the process of establishing the preliminary environment model includes: Collect detailed information on all fixed and mobile obstacles in the warehouse, including the location, shape and size of shelves, mechanical equipment and personnel passages; conduct on-site measurements and consult the warehouse layout diagram to obtain the location information of shelves and mechanical equipment, and record the shape, size and relative position relationship between shelves and mechanical equipment; The warehouse environment is divided into different areas, namely storage area, operation area and personnel passage, and the personnel passage is set as a no-fly zone; Preprocess the collected data, and for each obstacle, create a three-dimensional model of warehouse elements such as shelves, mechanical equipment, and personnel passages in the modeling tool based on the preprocessed data; Grid the warehouse space and mark the shelves, machinery and other obstacles in each grid space, and then set the height limit, obstacle type and safety distance parameters for each grid space; Integrate all constructed three-dimensional models and grid units into a unified model to obtain a preliminary environment model; In the step 2, the process of updating the environment model in real time includes: Use the drone route planning software on the computer to plan the drone's flight route according to the warehouse layout and actual environment. The drone takes off according to the planned route and uses the laser radar to scan the warehouse environment during the flight. The laser radar measures the distance and obtains the three-dimensional information of the warehouse environment by emitting laser beams and receiving reflected signals. The scanned data of the warehouse environment is transmitted to the computer in real time for storage and processing, and the raw data is solved using Shuttle and gAirHawk software; In the Shuttle software, data sorting and POS solution steps are performed to obtain the POS result file, which contains the position and attitude information of the drone. In the gAirHawk software, the POS result file, Lidar original data file and photo file are imported to perform coordinate system setting, route division and point cloud computing steps to obtain 3D laser point cloud data, and then analyze the point cloud volume, slice area and height range of the 3D laser point cloud data. Compare and fuse the solved 3D laser point cloud data with the preliminary environment model, and update the preliminary environment model in real time according to the position and shape information of the mobile shelves or personnel in the point cloud data to obtain an updated environment model; The point cloud volume is calculated using the slicing method, and its calculation expression is: Where V is the point cloud volume, A i is the slice area of ​​the i-th slice, N is the number of slices, and h is the slice spacing; The calculation expression of the slice area is: Among them, A i is the slice area of ​​the i-th slice, m is the number of boundary points, (x j ,y j ) is the boundary point of slice i; The calculation expression of the height range of the point cloud is: H=z max -z min ; Where H is the height range of the point cloud, z max is the maximum height of a point in the point cloud, z min is the minimum height of a point in the point cloud; In step 3, the process of formulating the flight path of the drone includes: Use GPS technology to obtain the location information of the drone, and use IMU to measure the angular velocity and acceleration of the drone. Combined with the initial attitude information, the real-time attitude and speed information of the drone are calculated. Among them, GPS provides the longitude, latitude and altitude position data of the drone; Based on the real-time updated environment model and obstacle information, as well as the current position, speed and attitude information of the drone, the Dijkstra algorithm is selected for path planning; In the path planning algorithm, the current position coordinates of the drone, the target position coordinates, and the real-time updated environmental model and obstacle information are input, and the steering angle and reference distance of the current position are analyzed to calculate the optimal path that avoids obstacles and meets flight safety requirements. The flight path includes the starting point, end point, intermediate nodes, flight altitude and speed parameters; The flight path is converted into the control instructions of the UAV, and the flight mission is executed through the control system of the UAV. During the flight, the UAV adjusts the flight attitude and speed in real time through the closed-loop control algorithm of PID control; The cost function expression of the optimal path is: f(n)=g(n)+h(n); Among them, f(n) is the cost function of the optimal path, g(n) is the actual cost from the starting point to the current position coordinates, h(n) is the heuristic estimated cost from the current position coordinates to the target position coordinates, and x n ,y n , z n ) is the current position coordinate, (x g ,y g , z g ) is the target position coordinate, θ n is the steering angle of the current position coordinate, λ is the cost weight coefficient of power consumption, α is the cost weight coefficient of the steering angle, and d0 is the reference distance; In step 4, the process of adjusting the flight speed and direction of the drone includes: During the flight of the drone, the laser radar emits a laser beam to the target area through its internal laser diode. By using phased array technology, the phase difference of the laser beam is adjusted to achieve rapid scanning of the beam, covering the target area, so that the laser radar can obtain three-dimensional information of the target area. The laser radar receiver receives the reflected laser signal, obtains obstacle information, and converts it into an electrical signal for subsequent signal amplification, filtering and demodulation steps to identify the location, shape and size of the obstacle; After identifying obstacles, the path is replanned based on the current environmental information and the motion status of the drone to find the optimal path from the current location to the target location, while avoiding all known obstacles, and monitoring the surrounding obstacle information in real time. Combined with the preset safety distance threshold, the safety factor of the flight path is calculated to determine whether to replan the flight path; Pre-set the warning threshold, compare the safety factor with the warning threshold, and determine whether to issue a warning based on the comparison result to remind staff to intervene and adjust the drone's path planning; Based on the results of path planning, the UAV is navigated and controlled to achieve safe obstacle avoidance, including adjusting the flight speed and direction of the UAV.

2. The UAV safety scheduling optimization algorithm according to claim 1 is characterized by: The specific process of staff intervention includes: Monitor the safety factor of the flight path in real time and compare it with the preset warning threshold to analyze whether the safety factor is lower than the preset warning threshold; If the safety factor is lower than the warning threshold, it indicates that the safety of the current path is insufficient, and the current path is marked and a warning is issued; Based on the triggered early warning mechanism, a prompt is issued on the console to remind the staff to intervene and take action, and start the path replanning program, input the latest environmental information and obstacle data, and use the path planning algorithm to calculate a new safe path until the safety factor exceeds the early warning threshold, thus determining the result of the path planning.

3. The unmanned aerial vehicle safety scheduling optimization algorithm according to claim 2 is characterized in that: The calculation expression of the safety factor is: Where S is the safety factor, d safe is the preset safety distance threshold, d obs is the distance from the drone to the nearest obstacle, d obs The smaller the value, v is the current speed of the drone, v max is the maximum speed of the drone.

4. The unmanned aerial vehicle safety scheduling optimization algorithm according to claim 3 is characterized in that: In step 5, the process of navigating and controlling the drone includes: The path planning results are converted into navigation instructions, sent to the UAV's flight control system, and the UAV's flight attitude and speed are adjusted through the actuator; Monitor the position, speed and attitude of the drone in real time to ensure that the drone flies along the planned path; Use lidar to perceive the environment around the drone in real time, including static and dynamic obstacles, and dynamically adjust the drone's flight plan based on real-time monitored environmental changes.

Images