Material distribution service scheduling management system based on cooperation of unmanned aerial vehicle and unmanned vehicle

Abstract

The invention relates to a material distribution service scheduling management system based on cooperation of an unmanned aerial vehicle and an unmanned vehicle. The system comprises a dual-navigation visual system module used for acquiring environment image data of an unmanned aerial vehicle and an unmanned vehicle; the algorithm operation module is used for processing and analyzing the selected logistics data, and carrying out path planning based on an SLAM algorithm and an optimization DWA algorithm based on reinforcement learning according to the environment image data in combination with the actual traffic topology to obtain transportation route information; the positioning module is used for positioning based on the UWB technology to obtain real-time position information of the unmanned aerial vehicle and the unmanned vehicle; the distribution management module is used for matching the distribution materials; and analysis is carried out based on the transportation route information and the real-time position information to obtain distribution logistics information. According to the system, a distribution path is dynamically optimized according to real-time traffic and environment changes by integrating dual-navigation vision, an SLAM and reinforcement learning optimization algorithm, a UWB technology and the like, and the cooperative efficiency of an unmanned aerial vehicle and an unmanned vehicle in material distribution is effectively improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicles and unmanned vehicles, and in particular to a material distribution service dispatching and management system based on the collaboration of unmanned aerial vehicles and unmanned vehicles. Background Art

[0002] With the rapid development of the modern logistics industry, especially in the context of the growing demand for e-commerce and express delivery, traditional delivery methods have gradually faced a series of problems such as low efficiency, high cost, and environmental pollution. In order to improve logistics efficiency and reduce costs, intelligent and automated delivery methods have emerged, among which the coordinated delivery of drones and unmanned vehicles based on unmanned driving technology has become a solution that has attracted much attention.

[0003] As an emerging delivery tool, drones and unmanned vehicles have shown great application potential in a variety of scenarios. Due to their high flexibility and flight capabilities, drones are suitable for short-distance, high-frequency, and small-scale delivery tasks; while unmanned vehicles have strong load capacity, stability, and endurance, and are more suitable for the delivery of bulk materials on the ground. Therefore, the coordinated transportation based on drones and unmanned vehicles can fully utilize the advantages of both and improve the efficiency and reliability of the overall logistics system.

[0004] However, traditional logistics transportation systems based on drones and unmanned vehicles usually rely on fixed routes and simple scheduling rules, and are unable to dynamically optimize delivery routes based on real-time traffic conditions and environmental changes. Summary of the invention

[0005] Based on this, it is necessary to address the above-mentioned technical problems. The purpose of the present invention is to provide a material distribution service scheduling and management system based on the collaboration of drones and unmanned vehicles, so as to dynamically plan route transportation according to real-time traffic conditions and environmental changes, realize the collaborative work of drones and unmanned vehicles, and improve distribution efficiency and reliability.

[0006] In the first aspect, the present application provides a material distribution service dispatching and management system based on the collaboration of drones and unmanned vehicles, the system comprising:

[0007] Dual navigation vision system module, used to obtain environmental image data of drones and unmanned vehicles;

[0008] The algorithm operation module is used to process and analyze the selected logistics data, combine the actual traffic topology, and perform path planning based on the SLAM algorithm and the optimized DWA algorithm based on the combination of reinforcement learning and multi-objective optimization according to the environmental image data to obtain the transportation route information of the drone and unmanned vehicle collaboration;

[0009] The positioning module is used to locate the UAV and the unmanned vehicle based on UWB technology to obtain the real-time location information of the UAV and the unmanned vehicle;

[0010] The distribution management module is used to match each distribution material with each drone and unmanned vehicle; and to obtain distribution logistics information based on analysis of transportation route information and real-time location information.

[0011] In one embodiment, the algorithm execution module includes:

[0012] The map building submodule is used to build an environmental map through the SLAM algorithm based on logistics data and traffic topology;

[0013] The path calculation submodule is used to calculate the optimal path based on the delivery vehicle status information and the environment map through the optimized DWA algorithm based on reinforcement learning and multi-objective optimization. The delivery vehicle status information is the status information of the drone and the unmanned vehicle;

[0014] The route planning submodule is used to obtain the transportation route information of the UAV and the unmanned vehicle based on the kinematic constraints and optimal paths of the UAV and the unmanned vehicle.

[0015] In one embodiment, the map construction submodule includes:

[0016] Data collection unit, used to obtain logistics data and traffic topology;

[0017] A map construction unit is used to extract features from environmental image data, logistics data and traffic topology to obtain feature data; and to construct an environmental map based on the feature data;

[0018] The map optimization unit is used to fuse and optimize the acquired environmental image data with the environmental map during the transportation process of drones and unmanned vehicles to obtain an optimized environmental map.

[0019] In one embodiment, the path calculation submodule includes:

[0020] An evaluation function setting unit, used to set parameters of the evaluation function, the parameters including at least one of the weight of the heading score, the weight of the distance score, and the weight of the speed score, the evaluation function is used to quantify the advantages and disadvantages of different paths;

[0021] A reward and punishment factor setting unit is used to set a reward and punishment function, which is used to optimize the transportation paths of drones and unmanned vehicles;

[0022] The path optimization unit is used to adjust the weights of the parameters of the evaluation function based on reinforcement learning according to the optimization information obtained through the reward and punishment function to obtain the best path.

[0023] In one embodiment, the evaluation function includes:

[0024] G(s,a)=w hH(s,a)+w d D(s,a)+w v V(s,a)

[0025] Among them, G(s,a) is the comprehensive evaluation score, s represents the current state, a represents the action, that is, the selected path or driving direction, and w h is the heading score ratio, w d is the distance score ratio, w v is the speed score ratio, H(s,a) is the heading score, D(s,a) is the distance score, and V(s,a) is the speed score.

[0026] In one embodiment, the UWB positioning module includes:

[0027] A signal transmitting unit, used for transmitting a UWB positioning signal;

[0028] A signal receiving unit, used for receiving the reflected UWB positioning signal;

[0029] The positioning calculation unit is used to calculate the real-time position information of the unmanned vehicle and the unmanned aerial vehicle based on the UWB positioning signal and the reflected UWB positioning signal.

[0030] In one embodiment, the dual navigation vision system module includes:

[0031] A first visual navigation system, configured to acquire first environment image data based on a first sensor, wherein the first environment image data includes image information based on color and texture;

[0032] A second visual navigation system is used to obtain second environment image data based on a second sensor, where the second environment image data includes depth information and distance information;

[0033] The visual information fusion unit is used to fuse the first environment image data and the second environment image data to obtain environment image data.

[0034] In one embodiment, the delivery management module further includes:

[0035] A distribution task allocation unit is used to allocate distribution tasks to drones and unmanned vehicles based on material requirements and distribution sites;

[0036] A real-time monitoring unit is used to analyze the transportation route information and the real-time location information to obtain the distribution logistics information, which includes the distribution progress and distribution status;

[0037] The visual display unit is used to visualize the delivery information in the form of charts. The delivery information includes delivery tasks, delivery progress and delivery status.

[0038] In one embodiment, it also includes:

[0039] The material management module is used to manage the distribution of materials. The material management module includes a material information recording unit and a material sorting and loading unit;

[0040] Among them, the material information recording unit is used to record and manage the distribution material information, including addition and deletion operations and classification records;

[0041] The material sorting and loading unit is used to automatically sort and distribute materials according to the distribution tasks and material characteristics.

[0042] In a second aspect, the present invention further provides a material distribution service scheduling management method based on the collaboration of unmanned aerial vehicles and unmanned vehicles, the method comprising:

[0043] Obtain environmental image data from drones and unmanned vehicles;

[0044] Based on the selected logistics data, combined with the actual traffic topology, path planning is performed based on the SLAM algorithm and the optimized DWA algorithm based on the combination of reinforcement learning and multi-objective optimization according to the environmental image data, and the transportation route information of the coordinated UAV and unmanned vehicle is obtained;

[0045] Based on UWB technology, drones and unmanned vehicles are positioned to obtain real-time location information of drones and unmanned vehicles;

[0046] Match each delivery material with each drone and unmanned vehicle; and analyze based on transportation route information and real-time location information to obtain delivery logistics information.

[0047] The above-mentioned material distribution service dispatching management system and method based on the collaboration of drones and unmanned vehicles obtains the environmental image data of drones and unmanned vehicles through the dual navigation vision system, and combines the actual traffic topology, and performs path planning through the SLAM algorithm and the DWA algorithm optimized by reinforcement learning, thereby realizing the efficient collaborative distribution of drones and unmanned vehicles. The system uses UWB technology for real-time positioning to ensure the accurate position tracking of drones and unmanned vehicles. At the same time, the distribution management module intelligently matches materials with drones and unmanned vehicles, and optimizes distribution scheduling based on real-time location information and transportation route information.

[0048] Compared with traditional logistics and transportation systems based on drones and unmanned vehicles, this system can obtain environmental data in real time and perform intelligent analysis to automatically adjust the transportation path by integrating dual navigation vision systems and advanced path planning algorithms. By combining the SLAM algorithm with the DWA algorithm optimized by reinforcement learning, the system can dynamically plan the optimal route according to changes in real-time traffic, terrain and obstacles, realize the collaborative work of drones and unmanned vehicles, flexibly respond to complex and changing distribution needs, and significantly improve distribution efficiency and reliability. In addition, the real-time positioning capability of UWB technology ensures the accurate tracking of the positions of drones and unmanned vehicles, allowing the system to adjust the scheduling strategy in real time, improving the system's adaptability and emergency handling capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related technologies, the drawings required for use in the embodiments or the related technical descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0050] Figure 1 A schematic diagram of the structure of a material distribution service dispatching and management system based on the collaboration of unmanned aerial vehicles and unmanned vehicles provided by an exemplary embodiment of the present invention;

[0051] Figure 2 A flow chart of a method for obtaining environmental image data provided by an exemplary embodiment of the present invention;

[0052] Figure 3 A flow chart of a method for obtaining an optimal path provided for an exemplary embodiment of the present invention;

[0053] Figure 4 A flow chart of a method for constructing and optimizing an environment map based on a SLAM algorithm provided by an exemplary embodiment of the present invention;

[0054] Figure 5 A flow chart of a material distribution service scheduling management method based on the collaboration of unmanned aerial vehicles and unmanned vehicles is provided as an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0056] First, the application scenario of the embodiment of the present application is described. In the embodiment of the present application, a material distribution service dispatching and management system based on the collaboration of drones and unmanned vehicles is provided, which is applicable to but not limited to such scenarios. For example, in urban logistics distribution, there are tall buildings and traffic congestion, and traditional distribution vehicles often encounter long delays on the road. Drones can use their air advantage to quickly cross obstacles and transport emergency or small materials from logistics centers to areas close to the destination, and then the unmanned vehicles that flexibly shuttle on the ground complete the final "last mile" delivery. For example, in some large commercial areas, drones can transport goods from warehouses in the suburbs to temporary stops on the roof of shopping malls. Unmanned vehicles then deliver goods accurately to the door of each store based on the road planning and store location information inside the shopping mall, greatly improving the distribution efficiency, reducing customer waiting time, and improving the overall service level of urban logistics distribution.

[0057] In addition, the system is also suitable for medical supplies distribution and emergency rescue scenarios. For example, in large hospitals or emergency centers, drones can quickly deliver urgently needed medicines, medical equipment or samples from supply stations or warehouses outside the hospital to the hospital's designated helipad, while unmanned vehicles are responsible for accurately delivering these supplies to various departments or wards to ensure that life-saving supplies arrive in the shortest time. Especially during peak traffic hours or post-disaster rescue scenarios, traditional delivery methods may face serious delays, and the collaborative operation of drones and unmanned vehicles can break through road restrictions and improve the efficiency and timeliness of medical supplies delivery. In remote mountainous or island areas, where transportation infrastructure is relatively weak, drones can cross geographical obstacles such as mountains and rivers to transport urgently needed medical supplies, daily necessities, etc. to open areas near mountain villages or island settlements, and then unmanned vehicles will relay and deliver them to specific households.

[0058] Illustratively, the material distribution service scheduling and management system based on the collaboration of drones and unmanned vehicles provided in the embodiments of the present application can also be applied to other application scenarios. This is only an example and is not limited to the specific application scenarios.

[0059] In an exemplary embodiment, Figure 1 As shown, a material distribution service scheduling management system 100 based on the collaboration of drones and unmanned vehicles is provided. In the embodiment of the present application, the system is applied to the controller of drones and unmanned vehicles as an example. It can be understood that the controller can also be a control terminal or control server connected to the website device, which is not limited here. The system includes:

[0060] The dual navigation vision system module 101 is used to obtain environmental image data of the UAV and the unmanned vehicle.

[0061] Specifically, the dual navigation vision system module 101 obtains environmental image data in real time through the dual navigation vision system to help drones and unmanned vehicles accurately perceive their surroundings. Exemplarily, multi-sensor fusion technology can be used in the dual navigation vision system to combine binocular vision cameras with laser radars to obtain environmental image data. Among them, the binocular vision camera is based on the principle of parallax and can accurately calculate the distance, shape, size and other information of the target object, and has unique advantages in identifying road signs, small obstacles and judging terrain undulations. The laser radar constructs a three-dimensional point cloud map of the surrounding environment by emitting a laser beam and receiving reflected light, which can efficiently detect large obstacles at a distance, such as buildings, mountains, etc., and model the surrounding space with high precision, providing drones and unmanned vehicles with more comprehensive and accurate environmental perception data. The environmental image data obtained by this module has the characteristics of high frame rate and high definition to ensure that environmental changes can be captured in time, and after preliminary image preprocessing, such as noise reduction, contrast enhancement and other operations, the availability and accuracy of the data are improved.

[0062] The algorithm operation module 102 is used to process and analyze the selected logistics data, combine the actual traffic topology, and perform path planning based on the SLAM algorithm and the optimized DWA algorithm based on the combination of reinforcement learning and multi-objective optimization according to the environmental image data to obtain the transportation route information of the drone and the unmanned vehicle.

[0063] Schematically, when the algorithm is running in module 102, the received logistics data, such as the type, weight, volume, urgency and other information of the materials, will be comprehensively considered. For example, for special materials such as fragile goods or fresh food, a smooth and fast route will be preferred in path planning to reduce losses during transportation. Secondly, by combining the actual traffic topology and environmental image data, based on the SLAM (Simultaneous Localization and Mapping, SLAM) algorithm, a dynamically updated environmental map can be constructed, so that drones and unmanned vehicles can identify and avoid obstacles, buildings, and road inequality factors, especially in complex urban environments or unknown areas. The DWA (Dynamic Window Approach, DWA) algorithm is an algorithm for local path planning of robots or unmanned equipment. By combining reinforcement learning technology, the model can be continuously trained to optimize path selection based on historical data and real-time feedback, that is, after each path planning, the system will evaluate the effect of the path (such as transportation timeliness, cost, risk, etc.), and adjust the algorithm through a feedback mechanism to gradually improve the planning accuracy of the system.

[0064] When the optimal path planning is completed, the algorithm operation module 102 will generate the final delivery route information. This information usually includes the path and key nodes, estimated delivery time, resource consumption information and risk warnings. Path and key nodes: including the starting point, stop points along the way, material handover points, and destination of drones and unmanned vehicles. Estimated delivery time: the estimated arrival time calculated after comprehensive consideration of traffic, equipment status and material characteristics. Resource consumption information: including information such as power or fuel consumption, speed, distance, etc. of unmanned equipment, to help monitor resources during the delivery process. Risk warning: If there are potential risks in path planning (such as traffic congestion, abnormal weather, etc.), the system will automatically issue a warning and provide alternative paths or adjustment plans.

[0065] The positioning module 103 is used to locate the UAV and the unmanned vehicle based on the UWB technology to obtain the real-time position information of the UAV and the unmanned vehicle.

[0066] UWB (Ultra Wide Band, UWB) is a wireless communication technology that performs high-precision positioning by transmitting signals within an extremely wide frequency band. For example, based on UWB technology, its high-precision ranging principle can be used to set up multiple UWB base stations in a certain range of drones, unmanned vehicles and the surrounding area to measure the distance information between the drones and unmanned vehicles and the base stations, and then calculate the three-dimensional coordinate position information of the drones and unmanned vehicles based on triangulation or multilateral positioning algorithms. UWB technology has the characteristics of strong anti-interference ability and high positioning accuracy. Even in a complex urban environment, facing multipath effects, signal shielding and other problems, it can still stably and accurately provide real-time location information of drones and unmanned vehicles. In addition, the module has a real-time dynamic update function, which can refresh the location data at a very high frequency (such as several times per second), ensuring that the system can grasp the latest location status of drones and unmanned vehicles in a timely manner, and provide a reliable location basis for the precise scheduling of the distribution management module 104.

[0067] The distribution management module 104 is used to match each distribution material with each drone and unmanned vehicle; and to analyze the transportation route information and real-time location information to obtain distribution logistics information.

[0068] Specifically, when matching each delivery material with each drone and unmanned vehicle, intelligent matching will be performed based on the attributes and delivery requirements of the materials, as well as the performance parameters of the drone and unmanned vehicle, such as load capacity, cruising range, and transportation speed. This step usually includes material characteristics analysis, unmanned equipment capability analysis, and task priority and timeliness. Material characteristics analysis: Different delivery materials have different characteristics such as size, weight, and volume. For example, some materials may be suitable for rapid air transportation by drones, especially small and light items, while other heavier and larger items require ground delivery by unmanned vehicles. Unmanned equipment capability analysis: Each drone and unmanned vehicle has its own specific load capacity, flight or driving speed, travel range, and other parameters. Task priority and timeliness: For different delivery tasks, there may be different timeliness requirements. The system will consider the priority of the task to ensure that urgent or time-sensitive materials can be given priority to the appropriate equipment for transportation.

[0069] When the materials and unmanned equipment are matched, the distribution management module 104 can calculate the expected delivery time, material handover location, resource consumption in the entire distribution process and other distribution logistics information based on the transportation route information and real-time location information, taking into account the traffic condition forecast on the transportation route (such as judging the degree of road congestion based on historical data and real-time road condition information), the remaining power or fuel of the drone and unmanned vehicle, the delivery time window and other factors, and visualize it to facilitate dispatchers, managers or customers to monitor, optimize and adjust the distribution plan in real time to ensure that the distribution process is efficient, safe and meets the expected time requirements.

[0070] The above-mentioned material distribution service dispatching and management system based on the collaboration of drones and unmanned vehicles optimizes the collaborative work of drones and unmanned vehicles in material distribution by integrating multiple efficient modules. The system obtains environmental image data of drones and unmanned vehicles through the dual navigation vision system module 101, and then perceives the surrounding environment, identifies dynamic obstacles or changes, and provides visual information for path planning; secondly, the algorithm operation module 102 combines the actual traffic topology and environmental images, and uses the SLAM algorithm and the DWA algorithm optimized by reinforcement learning to perform path planning, thereby generating collaborative transportation route information, which can minimize transportation time, avoid congestion, and improve overall efficiency. And through the positioning module 103, the UWB technology is used to locate the drones and unmanned vehicles in real time to ensure that the system can accurately obtain their current positions, reducing the impact of positioning errors on distribution tasks. Finally, the distribution management module 104 intelligently matches the distribution materials with drones and unmanned vehicles, and generates accurate distribution logistics information based on the transportation route and real-time location, helping dispatchers to understand the distribution progress and status in a timely manner, further optimize the distribution process, and reduce unnecessary delays and conflicts.

[0071] Through the design of this system, the delivery routes of drones and unmanned vehicles can be dynamically adjusted under complex traffic and environmental conditions, improving the efficiency and flexibility of material distribution. Compared with traditional logistics systems, this system can not only optimize route planning, but also adjust distribution strategies in real time, effectively improve collaborative efficiency, and ensure the safety and timeliness of logistics transportation.

[0072] In an exemplary embodiment, the dual navigation vision system module 101 includes:

[0073] A first visual navigation system, configured to acquire first environment image data based on a first sensor, wherein the first environment image data includes image information based on color and texture;

[0074] A second visual navigation system is used to obtain second environment image data based on a second sensor, where the second environment image data includes depth information and distance information;

[0075] The visual information fusion unit is used to fuse the first environment image data and the second environment image data to obtain environment image data.

[0076] Indicatively, Figure 2 As shown, the first visual navigation system can obtain the first environment image data based on the binocular vision camera, and the second visual navigation system can obtain the second environment image data based on the laser camera. The binocular vision camera is composed of two cameras with a certain distance between them. The surrounding environment is photographed by these two cameras at the same time to obtain two sets of original images with different perspectives, and image information based on color and texture is extracted from them. Among them, color information can help identify different objects, such as green vegetation, gray roads, etc.; texture information helps to distinguish the features of the surface of the object, such as rough ground, smooth building walls, etc. According to the principle of parallax and triangulation, the distance between the two cameras and the parallax size are known (for the same object in the environment, due to the difference in the positions of the two cameras, the position difference of the object in the images taken by the two cameras) can be calculated. By performing this calculation on multiple objects or feature points in the image, their relative distance information can be obtained, thereby obtaining the depth structure information of the environment to a certain extent, and the depth structure information is the first environment image data.

[0077] Since the laser radar will emit laser pulse beams into the surrounding space at a certain angle range and frequency, and will be reflected back after encountering objects in the environment, the distance between the laser radar and the reflection point on the surface of the object can be calculated by accurately measuring the flight time (Time of Flight, TOF) experienced from the emission to the reception of the laser pulse. At the same time, the laser radar itself has an accurate angle measurement mechanism that can record the horizontal and vertical angle information of each beam of emitted laser. Based on the distance and corresponding angle information obtained above, for each reflected laser pulse, the following spatial coordinate conversion formula can be used to calculate the three-dimensional coordinates (x, y, z) of the reflection point on the surface of the object in the laser radar's own coordinate system.

[0078]

[0079] Among them, the distance between a certain reflection point and the laser radar is r, the horizontal angle is θ, and the vertical angle is x, y and z are its coordinates in rectangular coordinates.

[0080] As the laser radar continuously emits and receives laser pulses, a large number of three-dimensional coordinate data of reflection points on the surface of the object will be obtained in a short time. Many such coordinate points are collected together to construct a three-dimensional point cloud map of the surrounding environment, which is the second environmental image data. This three-dimensional point cloud map can very intuitively and accurately present the spatial distribution, shape and relative position relationship of objects in the environment, especially for detecting large obstacles at a long distance. Finally, the first environmental image data and the second environmental image data are fused by the visual information fusion unit, including data normalization, feature fusion based on Kalman filtering or image registration, etc., to obtain environmental image data that contains detailed appearance visual information such as color and texture, and has accurate spatial position information such as depth and distance. Through this environmental image data, a comprehensive and accurate environmental perception basis can be provided for drones and unmanned vehicles, so that in the process of performing tasks, whether it is path planning, judging whether there are obstacles ahead, and choosing appropriate obstacle avoidance strategies, etc., more reasonable and safe decisions can be made based on better environmental cognition.

[0081] In an exemplary embodiment, the algorithm execution module 102 includes:

[0082] The map building submodule is used to build an environmental map through the SLAM algorithm based on logistics data and traffic topology;

[0083] The path calculation submodule is used to calculate the optimal path based on the delivery vehicle status information and the environment map through the optimized DWA algorithm based on reinforcement learning and multi-objective optimization. The delivery vehicle status information is the status information of the drone and the unmanned vehicle;

[0084] The route planning submodule is used to obtain the transportation route information of the UAV and the unmanned vehicle based on the kinematic constraints and optimal paths of the UAV and the unmanned vehicle.

[0085] Specifically, logistics data includes information such as warehouse location, cargo distribution points, and cargo volume, which reflect the key nodes in logistics activities. Traffic topology includes traffic infrastructure information such as the connection relationship of the road network and the conditions of intersections. The environmental map is comprehensively constructed based on these data through the Simultaneous Localization and Mapping (SLAM) algorithm. By integrating multi-source data into the calculation, the map is not only a simple geographical environment information, but also integrates important elements related to logistics and transportation. The delivery vehicle status information is the current status of drones and unmanned vehicles, such as position, speed, acceleration, remaining power (for electric vehicles), etc., which can reflect the current operating capabilities and limitations of the vehicle. It is combined with the environmental map through the optimized DWA (Dynamic Window Approach) algorithm based on reinforcement learning and multi-objective optimization to plan the best path. For example, the optimized DWA (Dynamic Window Approach) algorithm based on reinforcement learning and multi-objective optimization generates a series of possible trajectories in the speed-steering angle space by combining the DWA algorithm with multi-objective optimization. The generation of these trajectories comprehensively considers multiple objectives such as path length, safety, and stability. Then, based on the reinforcement learning mechanism, the parameters and strategies in the multi-objective optimization are dynamically adjusted according to the current state of the intelligent agent, the actions taken, and the rewards obtained, so as to better adapt to the environment and select a trajectory with the best comprehensive evaluation, i.e., the optimal path, from the generated trajectories. By constructing a system dynamics model through the kinematic constraints of drones and unmanned vehicles and combining it with the optimal path, the transportation route information of drones and unmanned vehicles can be obtained.

[0086] In an exemplary embodiment, the path calculation submodule includes:

[0087] An evaluation function setting unit, used to set parameters of the evaluation function, the parameters including at least one of the weight of the heading score, the weight of the distance score, and the weight of the speed score, the evaluation function is used to quantify the advantages and disadvantages of different paths;

[0088] A reward and punishment factor setting unit is used to set a reward and punishment function, which is used to optimize the transportation paths of drones and unmanned vehicles;

[0089] The path optimization unit is used to adjust the weights of the parameters of the evaluation function based on reinforcement learning according to the optimization information obtained through the reward and punishment function to obtain the best path.

[0090] Indicatively, Figure 3 As shown, a flow chart of a method for obtaining the best path is shown. Among them, the state space definition includes state variables such as the position, speed, attitude, and target distance of the drone and the unmanned vehicle, and the action space includes operations such as acceleration, deceleration, and turning. The current state information of the drone and the unmanned vehicle includes the current speed and acceleration, etc. to calculate the possible speed, that is, the current speed window. Based on this information, sensor information, maximum speed, and steering limit, etc., the possible actions, that is, the action window, are calculated, and these actions are evaluated using an evaluation function. The evaluation function may consider multiple factors, such as path length, safety (the degree of avoiding obstacles), efficiency, etc., to determine the pros and cons of each action and select the optimal action in the current state. According to the selected optimal action, check whether the current path meets the transportation demand, and give corresponding rewards and penalties through the reward function. For example, if the optimal action can make the drone and the unmanned vehicle closer to the target and avoid obstacles, a reward will be given; on the contrary, if it causes the vehicle to fall into danger or move away from the target, a penalty will be given. In this way, optimization information is obtained, and the evaluation function is adjusted through continuous trial and error and feedback through reinforcement learning to obtain the best path when the transportation demand is met and complete the path planning. Compared with the traditional DWA algorithm, this method has stronger adaptability and global optimization capabilities. Based on reinforcement learning, it can adjust strategies through continuous trial and error and feedback, better cope with uncertainties in complex environments, optimize local ranges and global paths, and thus improve the efficiency and safety of task execution.

[0091] In an exemplary embodiment, the evaluation function includes:

[0092] G(s,a)=w h H(s,a)+w d D(s,a)+w v V(s,a)

[0093] Among them, G(s,a) is the comprehensive evaluation score, s represents the current state, a represents the action, that is, the selected path or driving direction, and w h is the heading score ratio, w d is the distance score ratio, w v is the speed score ratio, H(s,a) is the heading score, D(s,a) is the distance score, and V(s,a) is the speed score.

[0094] The above comprehensive evaluation score can reflect the pros and cons of taking action a in the current state s. Schematically, the current state s includes the position, speed, posture and other related state variables of the drone or unmanned vehicle. Based on multi-objective optimization, different weights w h 、w d and w vIt determines the relative importance of heading, distance and speed in the comprehensive evaluation, thereby improving the comprehensive performance of the optimized path, effectively coping with complex and changing environments, and improving the robustness and resource utilization of the system.

[0095] In an exemplary embodiment, the map construction submodule includes:

[0096] Data collection unit, used to obtain logistics data and traffic topology;

[0097] A map construction unit is used to extract features from environmental image data, logistics data and traffic topology to obtain feature data; and to construct an environmental map based on the feature data;

[0098] The map optimization unit is used to fuse and optimize the acquired environmental image data with the environmental map during the transportation process of drones and unmanned vehicles to obtain an optimized environmental map.

[0099] Indicatively, Figure 4 As shown, a flow chart of a method for building and optimizing an environmental map based on a SLAM algorithm is shown. In general visual SLAM, key points (such as feature points such as corner points) and corresponding descriptors are extracted from images to determine the corresponding relationship between different images. By extracting features from environmental image data, logistics data, and traffic topology, and fusing them through weighted fusion or based on deep learning, feature data can be obtained, which is helpful for subsequent map construction and positioning. Based on the feature data, an environmental map can be constructed by a graph optimization algorithm or a Kalman filter algorithm. Exemplarily, a point cloud map or a topological map of the environment can be constructed based on the relative position and direction of the feature points in the feature data. Since drones and unmanned vehicles will continuously acquire new environmental image data during transportation, the image relationship in their data can be added to the existing environmental map through a fusion function, thereby optimizing the environmental map. Through the environmental map, various geographic information and obstacle distribution in the transportation route can be reflected in detail, which helps drones and unmanned vehicles to better understand and adapt to the environment to obtain more intelligent transportation decisions.

[0100] In an exemplary embodiment, the UWB positioning module 103 includes:

[0101] A signal transmitting unit, used for transmitting a UWB positioning signal;

[0102] A signal receiving unit, used for receiving the reflected UWB positioning signal;

[0103] The positioning calculation unit is used to calculate the real-time position information of the unmanned vehicle and the unmanned aerial vehicle based on the UWB positioning signal and the reflected UWB positioning signal.

[0104] Specifically, since the UWB signal will be reflected back when encountering obstacles (such as walls, ceilings, and the ground) after being transmitted, the UWB positioning module 103 mainly determines the distance by measuring the flight time of the UWB signal from transmission to reception. Among them, the UWB signal is a broadband wireless signal, which usually covers the frequency band of 3.1GHz to 10.6GHz. It can achieve high-precision positioning by transmitting a large amount of data simultaneously in a wide frequency range. When positioning the calculation unit, the UWB positioning signal and the reflected UWB positioning signal can be used to calculate the distance between the device and multiple base stations using technologies such as time difference positioning (Time of Arrival, TOA) and angle of arrival (Angle of Arrival, AOA), and the accurate position of the device can be obtained through the triangulation positioning principle or other positioning methods.

[0105] In an exemplary embodiment, the delivery management module 104 further includes:

[0106] A distribution task allocation unit is used to allocate distribution tasks to drones and unmanned vehicles based on material requirements and distribution sites;

[0107] A real-time monitoring unit is used to analyze the transportation route information and the real-time location information to obtain the distribution logistics information, which includes the distribution progress and distribution status;

[0108] The visual display unit is used to visualize the delivery information in the form of charts. The delivery information includes delivery tasks, delivery progress and delivery status.

[0109] Indicatively, the above-mentioned distribution task allocation unit operates mainly based on the material demand and the relevant information of the distribution site. First, by obtaining the specific demand for materials, including parameters such as the type, quantity, weight, volume, and urgency of the materials, as well as the location of each distribution site, the busyness of the site, the receiving capacity of the site, etc., the distribution tasks are allocated to drones and unmanned vehicles. For example, for urgent and light materials, priority is given to drones with fast speed and strong maneuverability; while for materials with heavier weight and larger volume but slightly lower timeliness requirements, they are allocated to unmanned vehicles with large load capacity.

[0110] Secondly, by comparing the obtained transportation route information with the real-time location information obtained based on UWB technology, we analyze whether there are deviations from the route, delays, etc., and then calculate the delivery progress (for example, how many delivery tasks have been completed, how far the currently delivered goods are from the destination) and status (for example, whether the delivery is carried out as planned, whether there are obstacles, problems or delays). The delivery progress can include information such as whether the task is carried out as expected, the remaining time, and the amount of materials delivered. The delivery status can include information such as whether the current delivery is smooth, whether there are problems such as failures or traffic jams, and whether the route needs to be adjusted. Finally, it can be displayed through a visualization unit. For example, the delivery task can be displayed through the task list view, dashboard view, and map view. The delivery progress of each delivery task can be displayed through a progress bar, a path on the map, or a timetable. The delivery status of the delivery task can be marked with a color (green for normal, red for problems) or an icon.

[0111] In an exemplary embodiment, the system further includes a material management module 105 for managing the distribution of materials, and the material management module includes a material information recording unit and a material sorting and loading unit;

[0112] Among them, the material information recording unit is used to record and manage the distribution material information, including addition and deletion operations and classification records;

[0113] The material sorting and loading unit is used to automatically sort and distribute materials according to the distribution tasks and material characteristics.

[0114] Specifically, the material information recording unit can add, delete and classify material information. For example, when new materials are received or expired or damaged materials are no longer needed, records can be added or deleted accordingly, and they can be classified according to type, weight, size, purpose of use or other characteristics. The material sorting and loading unit may include two parts: sorting and loading. Sorting: Automatically sort out the materials that need to be delivered according to the specific requirements of the distribution task, such as delivery to a specific location, delivery according to customer needs, or delivery in sequence. Loading: Load according to the characteristics of each material to ensure that the materials can be delivered safely and efficiently. For example, heavy materials need to be placed at the bottom, and fragile materials need to be specially marked or placed in a specific location to avoid damage during transportation. Through the material management module 105, it is ensured that the management and processing of materials in the distribution process are more efficient and intelligent, and the materials are sorted and loaded reasonably to complete the distribution task quickly and accurately.

[0115] Based on the same inventive concept, Figure 5 As shown, the embodiment of the present application also provides a material distribution service scheduling management method based on the collaboration of drones and unmanned vehicles, the method comprising the following steps:

[0116] S501: Acquire environmental image data of the UAV and the unmanned vehicle;

[0117] S502: Based on the selected logistics data, combined with the actual traffic topology, path planning is performed based on the SLAM algorithm and the optimized DWA algorithm based on the combination of reinforcement learning and multi-objective optimization according to the environmental image data to obtain the transportation route information of the UAV and the unmanned vehicle collaboration;

[0118] S503: positioning the UAV and the unmanned vehicle based on the UWB technology to obtain real-time location information of the UAV and the unmanned vehicle;

[0119] S504: Match each delivery material with each drone and unmanned vehicle; and analyze based on the transportation route information and real-time location information to obtain delivery logistics information.

[0120] This method uses the environmental image data obtained by the dual navigation vision system, and uses the SLAM algorithm and the optimized DWA algorithm based on reinforcement learning and multi-objective optimization to plan the path. It fully considers the actual traffic topology. Compared with the traditional path planning method based only on fixed maps or simple navigation algorithms, it can more accurately adapt to the complex and changeable distribution environment, improve the scientificity and flexibility of transportation route planning, and effectively reduce the risk of delivery delays caused by unreasonable paths. Based on the positioning of UWB technology, the real-time position of drones and unmanned vehicles can be determined with high precision. Its positioning accuracy far exceeds that of traditional GPS positioning. In the environment where signals are easily interfered with, such as in the city with tall buildings, the location information of drones and unmanned vehicles can still be accurately obtained, which provides a solid foundation for accurate distribution and real-time scheduling. In the distribution management link, by intelligently matching the distribution materials with drones and unmanned vehicles, and combining the transportation route information and real-time location information to analyze the distribution logistics information, the entire distribution process can be finely controlled. Compared with traditional manual scheduling or simple automatic scheduling, the utilization efficiency of distribution resources is greatly improved, and the occurrence of problems such as material mismatch and missing distribution is reduced.

[0121] This material distribution service scheduling and management method based on the collaboration of drones and unmanned vehicles not only optimizes resource allocation, reduces the need for manual intervention, and improves the efficiency and accuracy of material distribution, but also can cope with complex traffic and environmental changes, ensuring efficient and reliable logistics operations under dynamic conditions.

[0122] Furthermore, the method further comprises obtaining the transportation route information by the following steps:

[0123] Based on logistics data and traffic topology, the environment map is constructed through SLAM algorithm;

[0124] According to the delivery vehicle status information and the environment map, the optimal path is calculated through the optimized DWA algorithm based on reinforcement learning and multi-objective optimization. The delivery vehicle status information is the status information of the drone and the unmanned vehicle.

[0125] Based on the kinematic constraints and optimal paths of UAVs and unmanned vehicles, the transportation route information of UAVs and unmanned vehicles is obtained.

[0126] Furthermore, the method further comprises constructing and optimizing the environment map by the following steps:

[0127] Obtain logistics data and traffic topology;

[0128] Extract features from environmental image data, logistics data, and traffic topology to obtain feature data; and construct an environmental map based on the feature data;

[0129] During the transportation process of drones and unmanned vehicles, the acquired environmental image data is fused and optimized with the environmental map to obtain an optimized environmental map.

[0130] Furthermore, the method further comprises obtaining the best path by the following steps:

[0131] Setting parameters of the evaluation function, the parameters including at least one of the weight of the heading score, the weight of the distance score, and the weight of the speed score, the evaluation function is used to quantify the advantages and disadvantages of different paths;

[0132] Set up reward and penalty functions, which are used to optimize the transportation paths of drones and unmanned vehicles;

[0133] According to the optimization information obtained through the reward and punishment function, the weights of the parameters of the evaluation function are adjusted based on reinforcement learning to obtain the optimal path.

[0134] Furthermore, the evaluation function includes:

[0135] G(s,a)=w h H(s,a)+w d D(s,a)+w v V(s,a)

[0136] Among them, G(s,a) is the comprehensive evaluation score, s represents the current state, a represents the action, that is, the selected path or driving direction, and w h is the heading score ratio, w d is the distance score ratio, w v is the speed score ratio, H(s,a) is the heading score, D(s,a) is the distance score, and V(s,a) is the speed score.

[0137] Furthermore, the method further comprises obtaining the real-time location information of the unmanned machine and the unmanned aerial vehicle by the following steps:

[0138] Transmit UWB positioning signal;

[0139] Receive the reflected UWB positioning signal;

[0140] Based on the UWB positioning signal and the reflected UWB positioning signal, the real-time position information of the UAV and the drone is calculated.

[0141] Furthermore, the method further comprises obtaining environmental image data by the following steps:

[0142] Acquire first environment image data based on the first sensor, where the first environment image data includes image information based on color and texture;

[0143] Acquire second environment image data based on the second sensor, where the second environment image data includes depth information and distance information;

[0144] The first environment image data and the second environment image data are fused to obtain environment image data.

[0145] Furthermore, the method further includes obtaining and visualizing the distribution logistics information through the following steps:

[0146] Assign delivery tasks to drones and unmanned vehicles based on material needs and delivery sites;

[0147] Based on the analysis of transportation route information and real-time location information, distribution logistics information is obtained, and the distribution logistics information includes distribution progress and distribution status;

[0148] The delivery information is visualized in the form of charts, including delivery tasks, delivery progress and delivery status.

[0149] Furthermore, the method also includes managing the distribution of materials through the following steps:

[0150] Record and manage the distribution material information, including addition, deletion and classification operations;

[0151] Automatically sort and distribute materials according to distribution tasks and material characteristics.

[0152] In an exemplary embodiment, the present invention further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the material distribution service scheduling and management system based on the collaboration of drones and unmanned vehicles in the present application are implemented. A multi-core processor is preferred to improve the parallel processing capability of the system. Memory: Provides sufficient temporary storage space to support the running of programs and the processing of data. The memory capacity should be large enough to accommodate a large amount of supply information and computing tasks.

[0153] In an exemplary embodiment, the present invention also provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the material distribution service scheduling and management system based on the collaboration of drones and unmanned vehicles in this application are implemented. The computer-readable storage medium may include: a read-only memory (ROM), a random access memory (RAM), a solid-state drive (SSD), or an optical disk, etc. Among them, the random access memory may include a resistance random access memory (ReRAM) and a dynamic random access memory (DRAM).

[0154] The above-mentioned embodiments only express several implementation methods of the embodiments of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the patent of the embodiments of the present application. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the embodiments of the present application, and these all belong to the protection scope of the embodiments of the present application.

Claims

1. A material distribution service dispatching and management system based on the collaboration of drones and unmanned vehicles, characterized by: The system comprises: A dual navigation vision system module, used to obtain environmental image data of the UAV and the unmanned vehicle; The algorithm operation module is used to process and analyze the selected logistics data, combine the actual traffic topology, perform path planning based on the SLAM algorithm and the optimized DWA algorithm based on the combination of reinforcement learning and multi-objective optimization according to the environmental image data, and obtain the transportation route information of the UAV and the unmanned vehicle collaboration; A positioning module, used to locate the UAV and the unmanned vehicle based on UWB technology to obtain real-time location information of the UAV and the unmanned vehicle; The distribution management module is used to match each distribution material with each of the drones and the unmanned vehicles; and to analyze the transportation route information and the real-time location information to obtain distribution logistics information.

2. The material distribution service dispatching and management system based on the collaboration of unmanned aerial vehicles and unmanned vehicles according to claim 1 is characterized in that: The algorithm operation module includes: A map construction submodule, used to construct an environment map through a SLAM algorithm based on the logistics data and the traffic topology; A path calculation submodule is used to calculate the best path according to the delivery vehicle status information and the environmental map by using an optimized DWA algorithm based on a combination of reinforcement learning and multi-objective optimization, wherein the delivery vehicle status information is the status information of the drone and the unmanned vehicle; The route planning submodule is used to obtain the transportation route information of the drone and the unmanned vehicle based on the kinematic constraints of the drone and the unmanned vehicle and the optimal path.

3. The material distribution service dispatching and management system based on the collaboration of unmanned aerial vehicles and unmanned vehicles according to claim 2 is characterized in that: The map construction submodule includes: A data acquisition unit, used to obtain the logistics data and the traffic topology; A map construction unit, configured to extract features from the environmental image data, the logistics data and the traffic topology to obtain feature data; and to construct the environmental map based on the feature data; The map optimization unit is used to fuse and optimize the environmental image data acquired with the environmental map during the transportation process of the UAV and the unmanned vehicle to obtain an optimized environmental map.

4. The material distribution service dispatching and management system based on the collaboration of unmanned aerial vehicles and unmanned vehicles according to claim 2 is characterized in that: The path calculation submodule includes: An evaluation function setting unit, used to set parameters of the evaluation function, wherein the parameters include at least one of a weight of a heading score, a weight of a distance score, and a weight of a speed score, and the evaluation function is used to quantify the advantages and disadvantages of different paths; A reward and punishment factor setting unit, used to set a reward and punishment function, wherein the reward and punishment function is used to optimize the transportation paths of the drone and the unmanned vehicle; A path optimization unit is used to adjust the weights of the parameters of the evaluation function based on reinforcement learning according to the optimization information obtained through the reward and punishment function to obtain the optimal path.

5. The material distribution service dispatching and management system based on the collaboration of unmanned aerial vehicles and unmanned vehicles according to claim 4 is characterized in that: The evaluation function includes: G(s,a)=w h H(s,a)+w d D(s,a)+w v V(s,a) Among them, G(s,a) is the comprehensive evaluation score, s represents the current state, a represents the action, that is, the selected path or driving direction, and w h is the heading score ratio, w d is the distance score ratio, w v is the speed score ratio, H(s,a) is the heading score, D(s,a) is the distance score, and V(s,a) is the speed score.

6. The material distribution service dispatching and management system based on the collaboration of unmanned aerial vehicles and unmanned vehicles according to claim 1 is characterized in that: The UWB positioning module includes: A signal transmitting unit, used for transmitting a UWB positioning signal; A signal receiving unit, used for receiving the reflected UWB positioning signal; A positioning calculation unit is used to calculate the real-time position information of the unmanned machine and the unmanned aerial vehicle based on the UWB positioning signal and the reflected UWB positioning signal.

7. The material distribution service dispatching and management system based on the collaboration of unmanned aerial vehicles and unmanned vehicles according to claim 1 is characterized in that: The dual navigation vision system module includes: A first visual navigation system, configured to acquire first environment image data based on a first sensor, wherein the first environment image data includes image information based on color and texture; A second visual navigation system is used to obtain second environment image data based on a second sensor, wherein the second environment image data includes depth information and distance information; A visual information fusion unit is used to fuse the first environment image data and the second environment image data to obtain the environment image data.

8. The material distribution service dispatching and management system based on the collaboration of unmanned aerial vehicles and unmanned vehicles according to claim 1 is characterized in that: The delivery management module also includes: A distribution task allocation unit, used to allocate distribution tasks to the drone and the unmanned vehicle according to material requirements and distribution sites; A real-time monitoring unit, configured to analyze the transportation route information and the real-time location information to obtain the distribution logistics information, wherein the distribution logistics information includes a distribution progress and a distribution status; The visual display unit is used to visually display the delivery information in the form of a chart, wherein the delivery information includes the delivery task, the delivery progress and the delivery status.

9. The material distribution service dispatching and management system based on the collaboration of unmanned aerial vehicles and unmanned vehicles according to claim 8 is characterized in that: Also includes: A material management module, used to manage the distribution of materials, the material management module includes a material information recording unit and a material sorting and loading unit; The material information recording unit is used to record and manage the distribution material information, including addition and deletion operations and classification records; The material sorting and loading unit is used to automatically sort and distribute materials according to the distribution tasks and material characteristics.

10. A material distribution service scheduling management method based on the collaboration of unmanned aerial vehicles and unmanned vehicles, characterized in that: The method comprises: Acquiring environmental image data of the UAV and the unmanned vehicle; Based on the selected logistics data, combined with the actual traffic topology, path planning is performed based on the SLAM algorithm and the optimized DWA algorithm based on the combination of reinforcement learning and multi-objective optimization according to the environmental image data to obtain the transportation route information of the UAV and the unmanned vehicle collaboration; Positioning the UAV and the unmanned vehicle based on UWB technology to obtain real-time location information of the UAV and the unmanned vehicle; Each delivery material is matched with each of the drones and the unmanned vehicles; and based on the transportation route information and the real-time location information, analysis is performed to obtain delivery logistics information.

Images