A truck and drone collaborative logistics distribution method and system suitable for mountainous areas

Abstract

The present invention discloses a truck and drone collaborative logistics distribution method and system applicable to mountainous areas. The method takes into account the environmental factors of mountainous areas, quantifies the risk cost through risk assessment, and optimizes the drone delivery path in combination with a path planning algorithm. On this basis, the advantages of trucks at special nodes in mountainous areas are fully utilized to establish a joint distribution model of trucks and drones, ensuring comprehensive coverage of the distribution area, and effectively improving the efficiency and safety of mountainous logistics distribution. The present invention is applicable to logistics distribution tasks under complex terrain conditions in mountainous areas.

Description

Technical Field

[0001] The present invention relates to the field of logistics and distribution, and in particular to a method and system for completing logistics and distribution in coordination between trucks and unmanned aerial vehicles in mountainous areas. Background Art

[0002] In the current field of logistics and distribution, especially in complex mountainous environments, traditional distribution methods face great challenges. The terrain in mountainous areas is complex and changeable, and there are often a large number of natural obstacles. These natural conditions greatly limit the efficiency and accessibility of traditional ground transportation tools. With the rapid development of drone technology, the application of drones in the field of logistics and distribution has gradually become possible. Drones have the advantages of fast flight speed, high flexibility, and no terrain restrictions. They are particularly suitable for logistics and distribution in complex terrain and severe weather conditions. However, drones have limited endurance and load capacity, which limits their application in long-distance and large-scale logistics and distribution.

[0003] In order to overcome the above challenges, some studies have begun to explore the collaborative logistics distribution model of trucks and drones. This model combines the large load capacity and long-distance transportation capacity of trucks with the flexibility and rapid response capabilities of drones, aiming to improve the efficiency and reliability of logistics distribution in mountainous areas. However, most of the existing research focuses on the collaborative model between trucks and drones, and these collaborative models still have some limitations in practical applications. For example, the existing models often do not consider the constraints of drones when performing tasks in combination with scene characteristics, resulting in low scene adaptability of the model and higher actual distribution costs. Therefore, how to optimize the collaborative logistics distribution model of trucks and drones, especially in complex mountainous environments, to reduce site selection costs and improve distribution efficiency has become an urgent problem to be solved.

[0004] Therefore, the present invention aims to propose a new truck-UAV collaborative logistics distribution model to address the shortcomings of the existing UAV and carrier collaborative operation mode. By optimizing the site selection strategy of the UAV take-off and landing points and combining the obstacle avoidance risk assessment model, the path planning method is improved, thereby significantly reducing the overall cost of collaborative search and rescue while ensuring distribution efficiency and safety. Summary of the invention

[0005] The present invention provides a logistics distribution method and system for the collaboration of trucks and drones suitable for mountainous areas. The method takes into account the obstacle avoidance risk and energy consumption control of drones, performs path planning through an improved A* algorithm, and on this basis considers the distribution routes of trucks and drones and establishes a collaborative planning model for the two, thereby minimizing costs and maximizing efficiency in the distribution process.

[0006] To achieve the above object, the embodiments of the present invention adopt the following technical solutions:

[0007] A logistics distribution method using a truck and a drone in coordination suitable for mountainous areas comprises the following steps:

[0008] S1, considers obstacle avoidance to establish a risk assessment model and quantifies the risk cost E of the drone hitting an obstacle ao Energy consumption of drone under the influence of wind E ec ;

[0009] S2, divide the flight environment into several grids, calculate the risk cost in each grid according to the risk assessment model, and establish a global path planning model for the UAV. Specifically, based on the risk cost E ao Constructing the Cost-Benefit Matrix TC mn , based on the cost-benefit matrix constraints, the energy consumption of the UAV E ec Constraints and the maximum turning angle constraint of the UAV plan the service path with the minimum total cost;

[0010] S3, based on the global path planning model of the UAV, marking the special customer points D that exceed the maximum flight distance and maximum load limit in a single flight of the UAV over , L over , special service points are delivered by truck and / or serve as take-off and landing points for drones;

[0011] S4, respectively plan the paths of drones and trucks, take minimizing the delivery distance as the optimization goal, combine the delivery paths of trucks and drones for overall delivery planning, and establish a truck-drone collaborative path planning model.

[0012] A truck and drone coordinated logistics distribution system suitable for mountainous areas, comprising:

[0013] The risk assessment module inputs the flight path and obstacles and outputs the risk cost E of the drone hitting the obstacle. ao Energy consumption of drone under the influence of wind E ec ;

[0014] The UAV path planning module based on obstacle avoidance risk divides the flight environment into several grids, and the risk cost E in each grid is calculated based on the risk assessment module. ao , based on E ao Constructing the Cost-Benefit Matrix TC mn , based on the cost-benefit matrix constraints, the energy consumption of the UAV E ec Constraints and the maximum turning angle constraint of the UAV are used to plan the service path with the minimum total cost;

[0015] Truck-UAV joint delivery module, marking special customer points D on the service route where the maximum flight distance and maximum load limit of a single UAV flight are exceeded over , L overThe truck and drone path planning are performed in turn to obtain the truck-drone collaborative path.

[0016] Beneficial Effects

[0017] 1. Risk cost quantification and path planning optimization: This paper establishes a risk cost quantification model to effectively quantify the risk cost of obstacle avoidance of drones in complex mountainous environments. Path planning through the improved A* algorithm not only improves the safety of drone flight, but also ensures cost-effectiveness, significantly improving the delivery efficiency and safety of drones in complex terrains. Taking into account the special conditions in mountainous environments, the collaborative delivery system of drones and trucks is not only suitable for general environments, but also can adapt to complex and changeable mountainous environments, providing stable and reliable logistics delivery services.

[0018] 2. Innovative application of synchronous intermodal model: By setting special delivery points and establishing a truck-drone collaborative path planning model, the present invention allows for more flexible load and distance limit settings, effectively solving the delivery challenges brought about by complex mountain terrain and unpredictable environmental changes. This synchronous intermodal model enables trucks and drones to minimize costs and maximize efficiency while ensuring service quality.

[0019] 3. Dual optimization of cost and efficiency: The present invention combines genetic algorithms to optimize the delivery routes of trucks and drones, achieving cost minimization and efficiency maximization in the delivery process. Compared with the traditional logistics delivery model, the present invention can significantly reduce logistics costs and improve delivery efficiency while ensuring delivery safety, especially when dealing with logistics delivery challenges in mountainous environments, showing more obvious advantages.

[0020] 4. Improved flexibility and adaptability: Through precise risk assessment and path planning, the present invention enables the UAV to flexibly respond to various obstacles and risks in mountainous environments. At the same time, the design of the truck as a mobile station further enhances the adaptability and flexibility of the system, which can effectively respond to different customer needs and environmental changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0022] Figure 1 A schematic diagram of the impact of obstacle avoidance risk cost on path planning according to an embodiment of the present invention;

[0023] Figure 2A schematic diagram of a UAV turning angle constraint according to an embodiment of the present invention;

[0024] Figure 3 A schematic diagram of drone-truck joint delivery according to an embodiment of the present invention;

[0025] Figure 4 A schematic diagram of the location of package demand in Yuexi County according to an embodiment of the present invention;

[0026] Figure 5 A schematic diagram of path planning based on obstacle avoidance risk according to an embodiment of the present invention;

[0027] Figure 6 A schematic diagram of a mountainous area UAV-truck joint delivery planning result according to an embodiment of the present invention;

[0028] Figure 7 , Figure 8 The present invention is a method flow chart of an embodiment of the present invention. DETAILED DESCRIPTION

[0029] In order to make those skilled in the art better understand the technical solution of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. The embodiments of the present invention will be described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be interpreted as limiting the present invention. It can be understood by those skilled in the art that, unless specifically stated, the singular forms "one", "one", "said" and "the" used herein may also include plural forms. It should be further understood that the wording "including" used in the specification of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or their groups. It should be understood that when the present solution states that an element is "connected" or "coupled" to another element, it can be directly connected or coupled to other elements, or there may also be intermediate elements. In addition, the "connection" or "coupling" used here may include wireless connection or coupling. The term "and / or" used herein includes any unit and all combinations of one or more associated listed items. It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical terms and scientific terms) used herein have the same meaning as generally understood by those of ordinary skill in the art to which the present invention belongs. It should also be understood that terms such as those defined in general dictionaries should be understood to have meanings consistent with the meanings in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless defined as herein.

[0030] Embodiment 1:

[0031] This embodiment provides a method for trucks and drones to collaboratively deliver packages in mountainous environments. The method is used in helicopter and drone collaborative logistics delivery scenarios in mountainous environments. In this scenario, the delivery process is roughly as follows: consider the obstacle avoidance factors that may affect the operation of the drone, establish an obstacle avoidance risk assessment model, plan the drone path through a cost-benefit matrix, and use an improved A* algorithm to solve the problem, and obtain the optimized drone path length, the drone's maximum load and maximum flight distance limits, and the energy consumption changes caused by considering wind influence and obstacle avoidance. Specifically, Figures 1 to 8 As shown, including:

[0032] S1. Consider obstacle avoidance to establish a risk assessment model to quantify the risk cost associated with mountain obstacles and the energy consumption of the UAV under the influence of wind;

[0033] S2. Establish a UAV path planning model based on obstacle avoidance risk cost;

[0034] S3, marking special customer points that exceed the UAV flight distance and maximum load limit;

[0035] S4. Establish a truck-UAV collaborative path planning model. The UAV is responsible for package delivery. In addition to being responsible for delivery to special points, the truck also serves as a mobile station for the UAV. Combine the delivery paths of the truck and the UAV to carry out overall delivery planning.

[0036] In this embodiment, the flight path of the drone performing logistics services needs to reduce the obstacle avoidance risk cost while ensuring the completion of the mission. Therefore, the cost of drone obstacle avoidance is quantified as risk cost using the established risk cost quantification model to achieve the effect of mountain path optimization. The main environmental factors threatening drone operations can be summarized into two categories: (1) drone collision with obstacles; and (2) drone energy consumption threat under the influence of wind.

[0037] Specifically, in S1, a cost-benefit value calculation method for the flight path can be established by dividing the UAV flight environment into discrete two-dimensional grids. Based on the cost-benefit graph, a UAV path planning model with energy limitation constraints and a search algorithm with heuristic factors are proposed to guide the UAV to serve more customers in complex mountainous environments and avoid high-risk areas. Considering obstacle avoidance, the environment is divided into grids of equal size by reducing the risk cost to optimize the path, and the risk cost in each grid is calculated according to the comprehensive risk model. Different colors can be used to distinguish the distribution of risk costs. Red represents areas with high risk costs, and blue represents areas with low risk costs. As shown in the example Figure 1As shown in the figure, the dotted line represents the shortest path from the starting point to the end point, and the solid line is the path that takes risk cost mitigation into account. The area with lower risk cost is chosen to pass through, and the risk cost of the final path is much lower than the shortest path. The risk cost of the quantified path point relative to the obstacle is:

[0038] Assume that RO is the flight path of the UAV composed of edge e, γ is the number of waypoints on edge e, that is, the waypoints on edge e can be expressed as (1,...,m-1,m,m+1,...,γ), and the risk cost of the waypoints relative to obstacles is quantified as shown in formula (1):

[0039]

[0040] Where R t is the radius of the t-th obstacle, the path point and the center of the t-th obstacle OB t (x c ,y c ,z c ) is D t (m,c), as shown in formula (2).

[0041]

[0042] (x m ,y m ,z m ) is a path point, (x c ,y c ,z c ) is the center of the obstacle, χ∈[0,1] is the control parameter, increasing χ will reduce the cost, and vice versa.

[0043] Specifically, in S2, after the obstacle avoidance cost is determined by S1, the UAV path planning is based on the obstacle avoidance risk cost. According to the risk cost model, a global path planning model is established to accurately evaluate the optimal flight route of the UAV in a variable mountainous environment through the cost-benefit matrix.

[0044] Quantify the risk cost between any two points in the drone path so that the drone can avoid obstacles in the mountainous area and complete the delivery task safely. The graph is represented as an N×N grid, and any point ro m and ro n The cost-benefit matrix TC mn It can be expressed as shown in formula (3).

[0045]

[0046] RO is the flight path of the UAV composed of edges e, where ro m ,m∈{1,2,...,N 2} is the current position of the drone, ro n ,n∈{1,2,...,N 2} is the next position of the drone, d mn It's a little ro m and ro n The Euclidean distance between risk is the risk cost coefficient that affects the path planning strategy. In view of the potential unknown risks in the UAV path planning, a conservative strategy is considered to appropriately reduce η risk Set it higher to amplify the impact of risk costs and ensure safe drone operations.

[0047] The model objective is to plan a service path with the minimum total cost, which includes not only the obstacle avoidance risk cost but also the energy consumption cost of the drone. This objective is achieved through formula (4).

[0048]

[0049] Where RO is the flight path of the UAV composed of edge e, TC(RO) is the total cost of path RO, and TC(e mn ) is the edge e mn According to the UAV energy consumption model, the energy available for flight is limited. Therefore, the logistics UAV must complete the service and reach the destination as soon as possible before consuming the planned available power. The constraint definition is shown in formula (5), where E available Indicates the total available power of the drone.

[0050] E ec ≤E available (5)

[0051]

[0052] Formula (6) represents the maximum turning angle constraint of the UAV, as Figure 1 As shown, A:(x m-1 ,y m-1 ), B:(x m ,y m ), C:(x m+1 ,y m+1 ) represent the coordinates of three consecutive path points, cosβ represents the cosine value of the angle between vector AB and vector BC. The larger the β value, the smaller the cosine value. The red angle β max is the maximum acceptable turning angle, and this constraint is the turning angle limit for the drone.

[0053] Through these precise calculations and constraint definitions, a safe and efficient drone flight path can be planned that takes into account the complex terrain and potential risk factors in the mountainous areas, ensuring that the drone can successfully complete the delivery mission under harsh conditions.

[0054] The UAV-truck joint delivery studied in this embodiment adopts the following method: Figure 3 The simple model shown in the figure, in this model, both drones and trucks have the ability to deliver goods from the distribution center. The truck can deliver the goods to the destination and bring the drone back to the distribution center after the drone delivers. Unlike the traditional waiting method, the truck does not need to stop when the drone delivers, but can perform delivery operations at the same time. The single drone layout shown in the figure allows multiple customer points to be served at the same time. In addition, nodes that exceed the maximum flight distance of the drone are marked in red, indicating that these nodes need to be delivered by trucks because drones cannot complete long round trips. Overweight cargo nodes can only be delivered by trucks because drones cannot carry heavy objects, but drones can fly to overweight nodes and stay.

[0055] The modeling process can be divided into four parts: first, determine the special points, then plan the paths of trucks and drones separately, and finally plan the overall coordination of drones and trucks. Taking minimizing the delivery distance as the optimization goal, a synchronous intermodal model of trucks and drones in mountainous environments is established, and a genetic algorithm is used to solve the problem to plan the paths of drones and trucks to cover all delivery points. Among them, the main variables and parameters involved in the drone and truck joint delivery model can be listed as follows:

[0056] Table 1 Parameters and variables of the UAV and truck joint delivery model

[0057]

[0058] Specifically, in S3, special points that exceed the distance and weight limits are marked respectively. Customer points that exceed the maximum flight distance and maximum load limits in a single flight of the drone are marked as special points. Based on the previous considerations of drone obstacle avoidance, the drone route is optimized for obstacle avoidance. In a drone package delivery cycle, the drone's route points consist of three nodes. In the kth delivery, the drone takes off from the launch point i and carries the package to the drone's customer service point j. In this single flight, the drone needs to perform obstacle avoidance flight. A planning algorithm is established to guide the drone to find the path with the lowest risk cost under the constraints of flight performance indicators such as energy consumption and step length. Based on the obstacle avoidance planning, the points that exceed the maximum flight distance limit of the drone are marked as D over , mark the customer points that exceed the maximum load limit of the drone as L over , these special customer service points can only be delivered by truck and can also serve as take-off and landing points for drones.

[0059]

[0060] Formula (7) indicates that if the starting node is a special point, it must be delivered by truck. Formula (8) indicates that if the arrival node is a special point, it must be delivered by truck. Formula (9) indicates that the node marked as D over The starting point will not be served by the drone; Formula (10) indicates that the starting point marked as D over The arrival node will not be served by the drone; Formula (11) means that the drone and the truck can leave the distribution center independently or leave together with the truck; Formula (12) means that the drone and the truck can enter the distribution center independently or enter together with the truck; Formula (13) means that each delivery serves at least one customer node.

[0061] Next, based on the marking of special points, the path planning of the truck and the drone is required. Specifically, in S4, the paths of the truck and the drone must be planned separately. First, plan the truck path. In addition to serving special customer points, the truck also needs to provide power and cargo for the drone and act as a mobile station. After the drone is launched from the truck, the truck can depart from the launch point and head to the next stop. The truck that arrives at the next stop first has to wait for the other truck, and the end point recorded by the drone's single path planning is used as the end point of the truck delivery.

[0062]

[0063]

[0064] The goal of formula (14) is to maximize the number of customer nodes served by a single truck; while formula (15) means that the truck enters all unserved customer nodes at most once; similarly, formula (16) means that the truck can leave the node multiple times; formula (17) ensures that each time a customer node is assigned to deliver to a truck, it is removed from the previous set of customer nodes.

[0065] Then, plan the drone path. Due to the limited endurance of the drone and the drone flying in the mountainous area, in order to ensure safety and avoid obstacles, the flight path of the drone directly affects the different energy consumption requirements, so it is necessary to plan the drone path reasonably. Under the condition that the drone meets the maximum load limit of the drone and avoids obstacles, the drone reaches the maximum flight distance radius, and the drone is assigned as many customer demand points as possible. After each assignment, the destination reached in a single time is recorded.

[0066]

[0067] Formula (18) aims to optimize the number of client nodes that each drone can cover and maximize service efficiency; while formula (19) ensures that at any given moment, the cargo carried by the drone does not exceed its maximum carrying capacity; considering that the flight distance limit of the drone is actually also the limit of its energy use, we use formula (20) to ensure that the energy consumption of the drone in a single mission does not exceed its maximum energy consumption limit, which is equivalent to ensuring that the flight path length of the drone does not exceed its maximum flight distance; formula (21) ensures that each client node to be served is only visited by the drone once; at the same time, formula (22) ensures that the drone only departs from each node once; formula (23) indicates that if a client node to be served by the drone is assigned, it will be deleted from the previous client node set.

[0068] Finally, the overall planning is carried out. Considering the power consumption of the UAV, the phased power consumption model of the rotary wing UAV is cited. The flight path of the UAV is divided into three parts: vertical take-off segment, horizontal straight segment and vertical landing segment. Each UAV has a fixed battery capacity. The endurance of each UAV (in time units) depends on its speed and payload. Therefore, the power consumption of the UAV in different flight stages is different. Among them, the power equation of the UAV during horizontal cruising can be expressed by formula (24):

[0069]

[0070] Among them, c 1 , c 2 , c 4 , c 5 are model coefficients, and their values ​​are 2.8037 (m / kg), 0.3177 (m / kg), 0.0296 (kg / m) and 0.0279 (Ns / m). air is the horizontal speed of the drone, g is the gravity constant (9.8m / s 2 ), α is the angle of attack, which is generally assumed to be 10°, the weight of the package carried is w, and the weight of the drone is W.

[0071] Assume that the power of the drone in horizontal flight under no-load condition is P c ′、σ 1 is the ratio of the power of the UAV carrying a load when flying horizontally to the power of the UAV flying horizontally without a load, as shown in formula (25). Then the UAV horizontal flight time t under load conditions, corresponding to the UAV range S′ under standard conditions (UAV without load, operating environment temperature is 25°C) is shown in (26);

[0072]

[0073] S'=v air tσ 1 (26)

[0074] Considering obstacle avoidance planning, the range of the drone in horizontal flight under standard conditions (drone empty and operating environment temperature of 25°C) is as follows: As shown in formula (27), d ij ′ is the delivery distance consumed after path planning from node i to node j considering the risk of obstacle avoidance, and the flight distance The formula is shown in formula (27).

[0075]

[0076] The end point of a single delivery route record is used as the starting point of the next delivery route, and the truck and drone path planning steps are repeated until all customer needs are met. The route distance of each delivery is optimized as the objective function.

[0077]

[0078]

[0079] The core purpose of objective function (28) is to reduce the total delivery distance between drones and trucks, thereby reducing logistics costs; the road impedance coefficient ε is the road flow influencing factor, formula (29) means that for each node to be delivered, the drone or truck only leaves once; formula (30) means that for each node to be delivered, the drone or truck only arrives once; equation (31) is a conservation constraint for drone traffic at non-stop points; formula (32) limits the departure frequency of trucks at each node; formula (33) limits the entry frequency of trucks at each node; formula (34) ensures the conservation of truck traffic between nodes; formula (35) means that at the kth allocation, if the drone moves from node i to node j, is 1, otherwise it is 0; Formula (36) indicates that in the kth allocation, if the truck moves from node i to node j, is 1, otherwise it is 0; Formulas (11), (12) and (28)-(33) formulate the starting and ending rules of drones and trucks at the distribution center to ensure that they depart from the distribution center and eventually return without repeated visits.

[0080] In summary, for mountainous environments, this study established a risk cost quantification model to quantify the risk cost of drone obstacle avoidance. This model not only improves the safety of drone path planning, but also ensures cost-effectiveness. At the same time, by setting special delivery points, this study developed a synchronous intermodal model that allows more flexible load and distance limit settings, providing new solutions for the complex terrain and unpredictable environmental changes in mountainous areas.

[0081] Embodiment 2:

[0082] Based on the distribution method disclosed in Example 1, this embodiment conducts example verification from the algorithm level, including using an improved A* algorithm to solve the obstacle avoidance path problem of the drone and using a genetic algorithm to solve the synchronous intermodal model of trucks and drones, so as to plan the paths of drones and trucks to cover all delivery points.

[0083] 1. In the practical application of this embodiment, in the process of selecting the take-off and landing points, the improved A* algorithm can be used to solve the obstacle avoidance path problem of the drone: Specifically, the drone path must take into account the obstacle avoidance benefit. The cost of each grid is different and unevenly distributed. Therefore, it is necessary to formulate a path search rule adapted to it to improve the effectiveness of the search. This article refers to the created path search rule and sets the exploration rule based on the path point of the cost-benefit value. A heuristic factor is set according to the Boltzmann distribution to ensure the complete exploration of the environment. The drone is currently at the path point ro m ,m∈{1,2,...,N 2}, the next possible position of the drone is ro n ,n∈{1,2,...,N 2}, according to TC mn The value is used to calculate the next path point, as shown in formula (37),

[0084]

[0085] In the model, chromosomes represent the joint delivery plan of drones and trucks. Each chromosome consists of a string of integer codes, which represent the delivery sequence of distribution centers, customer points, and trucks and drones. "0" represents the distribution center, and the remaining numbers represent customer points. The order of the numbers represents the delivery order of trucks and drones. For customer points that are beyond the flight distance of the drone or overweight, special marks (such as negative values) are used to distinguish them. For example, "-3" is used to indicate that customer point "3" cannot be served by drones and can only be served by trucks.

[0086] Where T is the temperature parameter that controls the degree of environmental detection, RO is the number of all N 2 At the beginning of exploration, since the drone knows little about the environment, a smaller T value is set to ensure that the drone can quickly explore the environment in the early stage. As the exploration time increases, the drone has enough information about the environment, and the value of T will increase to ensure that the algorithm converges within a specific time. The algorithm includes the following steps:

[0087] 1. Initialization:

[0088] (1) Set the starting point P_{start} and target point P_{target} of the drone.

[0089] (2) Calculate the risk cost ro(m,n) for each grid according to formula (1) and initialize the risk cost map.

[0090] (3) Initialize the open list OpenList, add the starting point to the list, and set the estimated total cost COST f The value is 0.

[0091] 2. Calculate the cost

[0092] (1) For the current point P_{current}, use formula (5) to calculate the COST of all its feasible adjacent points g (known cost), use formula (3) to calculate the estimated cost COST from the current point to the target point h (heuristic estimate), where COST g It includes the cost from the starting point to the current point plus the risk cost R_{current} and energy consumption cost E_{current} passing through the current point.

[0093] (2)COST h is the estimated risk cost and energy cost from the current point to the target point.

[0094] 3. Update open / closed list

[0095] (1) Add adjacent points to the open list and calculate the COST for each point f =COST g +COST h .

[0096] (2) If the new COST g If the value is lower than the previous one, the COST of the point is updated. f .

[0097] 4. Select the next point

[0098] Select the one with the lowest COST from the open list f The point with the value is used as the next path point.

[0099] 5. Check your goals

[0100] The next target path point is calculated according to formula (37). If the target point is added to the open list, it means that the path is found. The path is constructed by backtracking from the target point to the starting point.

[0101] 6. Obstacle avoidance and energy consumption calculation:

[0102] (1) For each path point, calculate the distance to the obstacle and adjust its cost according to the given risk cost function ro(m,n).

[0103] (2) According to formula (5), the energy consumption model is used to calculate the energy consumption cost E_{current} of the path point.

[0104] 7. Corner constraints

[0105] Formula (10) can ensure that each turning angle in the path does not exceed the maximum turning angle β max .

[0106] 8. Path Optimization

[0107] In each iteration, the path is evaluated and optimized according to equation (4) to reduce the total cost until the path with the lowest cost is found.

[0108] 9. Termination Conditions

[0109] The algorithm ends when the open list is empty or the best path has been found.

[0110] In this way, not only the risk cost and energy cost of each path point are considered, but also the path can be dynamically adjusted according to the real-time status of the drone and environmental conditions. This algorithm is highly adaptive and can find a safe and energy-efficient path for the drone in a changing environment.

[0111] 2. In the practical application of this embodiment, a genetic algorithm can be used to solve the drone-truck collaborative delivery model. In view of the problems of low efficiency, easy to fall into local optimal solutions, rough path planning and high cost in traditional genetic algorithms in path planning, this study adopts an improved genetic algorithm. In this algorithm, the selection, crossover and mutation operations are optimized to improve the performance of the algorithm. Each particle represents a site selection sequence, 1 means that the group of take-off and landing points is selected, and 0 means that the group of take-off and landing points is not selected. In order to solve the problem that the traditional binary particle swarm algorithm is prone to fall into local optimal solutions during the iteration process, this embodiment adopts the following method for improvement: first, a method for dynamically adjusting the inertia weight is introduced; secondly, the judgment of the similarity between two particles is added in the process of particle updating, and the position and speed of the particles judged to be similar are updated according to the new rules. These include:

[0112] 1. Chromosome encoding

[0113] In the model, chromosomes represent the joint delivery plan of drones and trucks. Each chromosome consists of a string of integer codes, which represent the delivery sequence of distribution centers, customer points, and trucks and drones. "0" represents the distribution center, and the remaining numbers represent customer points. The order of the numbers represents the delivery order of trucks and drones. For customer points that are beyond the flight distance of the drone or overweight, special marks (such as negative values) are used to distinguish them. For example, "-3" is used to indicate that customer point "3" cannot be served by drones and can only be served by trucks.

[0114] 2. Population fitness function

[0115] The fitness function measures the efficiency of the delivery solution. In this model, the fitness function is defined as the inverse of the delivery distance. For a given chromosome, decoding is performed to determine the customer points served by drones and trucks, and the fitness value is determined based on the total travel distance of drones and trucks.

[0116] 3. Select

[0117] In order to generate a new generation of population, a roulette wheel selection algorithm is used. The probability of each chromosome being selected is proportional to its fitness. The fitness of each individual in the population is calculated, and then the probability of each individual being selected to enter the next generation is determined based on the fitness.

[0118] 4. Cross

[0119] Traditional single-point crossover may produce invalid solutions. Therefore, I use a special crossover strategy to ensure the validity of the solution. This strategy ensures that the position of the distribution center node remains fixed, while allowing the delivery sequences of trucks and drones to be exchanged at a randomly selected point. For example, for two chromosomes A and B, the crossover point is set at the 6th gene, and the gene segments of the two chromosomes from the 6th gene to the end are swapped while keeping the first 5 genes unchanged. This ensures that the newly generated individuals meet the constraints of the distribution center. After the crossover operation, check again to ensure that the nodes served by the drone do not exceed their maximum flight distance and load capacity.

[0120] 5. Mutation

[0121] A controlled mutation strategy is used to maintain the validity in the chromosome. If the selected gene represents a customer point that can be served by a drone, it will be mutated into another customer point that can be served by a drone. The same mutation strategy applies if it is a customer point served by a truck. This ensures that the mutated chromosome is still a valid delivery solution. Specifically, if the gene mutation causes the drone's service node to exceed its maximum flight distance or load, another drone service point will be randomly selected for exchange, or it will be converted to a truck service point, or vice versa.

[0122] By repeatedly iterating this process, we eventually find a set of optimized joint delivery solutions that meet the needs of all customers while minimizing the delivery distance.

[0123] Specific examples of this embodiment in the numerical experiment: This article takes the logistics in Yuexi County, Anqing as an example to analyze and explore the "mountain drone +" application model. Yuexi County is a county under the jurisdiction of Fuyang City, Anhui Province, China, and has 14 towns under its jurisdiction. In order to simplify the problem, the centers of these 14 towns are used as package demand points. The coordinate points of these 14 towns are shown in Table 2. The acquired elevation data of Yuexi County, Anqing, is used to determine the location of the demand point in units of villages. The warehouse is set in the center of Yuexi County with coordinates of (116.36655539021, 30.855227714002). Figure 2 As shown, based on the terrain characteristics, the UAV-truck package delivery route planning in the mountainous area of ​​Yuexi County is carried out.

[0124] Table 2 Coordinates of package demand points

[0125]

[0126]

[0127] from Figure 2 It can be seen that there are undulating mountains between towns. If obstacle avoidance is not considered, the drone path planning for package delivery is likely to be inaccurate. For example, before obstacle avoidance, the round-trip distance between certain customer points can meet the energy consumption limit of the drone. However, after obstacle avoidance planning, the restricted drone range changes, resulting in the drone being unable to return to the truck when delivering the package to the customer point, and these points will be delivered directly by the truck.

[0128] As the towns are located in mountainous areas, as shown in the figure above, there are undulating peaks between the towns, which causes the drone service path to pass through risky areas. Using the method of calculating the cost-benefit value mentioned above, the drone path is planned with energy constraints. According to the algorithm, the path is planned, taking the path from Xiangchang Town to Zhongguan Town as an example, the cost-benefit diagram is as follows Figure 3 As shown:

[0129] Figure 3 The colors above show the distribution of risk costs, red represents areas with high risk costs, and blue represents areas with low risk costs. Under the premise of ensuring service completion, risk aversion is used as a driving force to find the shortest path. Since there are 14 customer points in the distribution area of ​​this chapter, the number of drone planning paths between two nodes is huge. Therefore, this paper adopts the form of grouping, with 4 adjacent nodes as a group. The length of some drone planning paths is shown in Table 3, and the rest of the data is shown in Appendix 1.

[0130] Table 3. Length of the UAV obstacle avoidance planning path

[0131]

[0132] In this embodiment, the SF-ARK series ARK 25 of SF Express UAV is selected as the vehicle for mountain logistics distribution, and the truck is selected as an ordinary truck of the logistics company. The setting parameters are shown in Table 4;

[0133] Table 4 Parameter settings

[0134]

[0135] To simplify the calculation, taking into account the energy of the drone to support it to reach the customer point and return to the truck, the maximum cruising distance with full load is used as the upper limit, and the customer point with a single distance exceeding 23km is set as a special point that exceeds the distance limit, and the customer point with a delivery package weight exceeding 22kg is set as a special point that exceeds the load limit. Considering that the use of drones to deliver large and heavy items is rare, customer point No. 14 is set as a super key point.

[0136] Drones can deliver to multiple customer demand nodes at one time, while trucks can also deliver without waiting in place. In addition, nodes with demand exceeding the limit can be used as parking points for drones, while nodes with distance exceeding the limit can only be delivered by trucks. Drones and trucks leave the warehouse separately, and finally the trucks carry the drones back to the warehouse at the same time. Drones and trucks jointly perform delivery and cooperate to complete the tasks of each customer demand node.

[0137] In order to verify the effectiveness of the proposed truck-drone collaborative delivery model in mountainous environments, this example uses an improved genetic algorithm to conduct simulation tests in 14 town centers in Yuexi County, Anqing. The solution results are as follows: Figure 4 As shown in the figure, since customer nodes in mountainous cities are scattered and relatively far away, and drones are limited by their maximum flight distance and need to consider obstacle avoidance, and a single flight cannot travel a long distance back and forth, there are more customer demand nodes for truck delivery than for drone delivery.

[0138] like Figure 4 As shown in the planning results, customer points 2 and 3 in the figure are points where the path from the previous customer point exceeds the maximum distance limit of the drone after obstacle avoidance planning, and the weight of the package assigned to the customer point exceeds the load capacity of 22kg, so it is determined to be a special point and can only be delivered by trucks. On this basis, the remaining customer points are delivered by trucks and drones in coordination.

[0139] In order to test the effectiveness of the joint distribution model, the distribution model of the traditional TSP problem (i.e., without considering the special point settings for 14 customer points) is compared with the distribution model established in this chapter. The comparison results are shown in Table 3.5:

[0140] Table 5 Comparison of distribution plan optimization

[0141]

[0142] Due to the unoptimized path selection, the total delivery distance under the classic TSP model is 250 kilometers and the total delivery time is 11 hours. In addition, since the algorithm needs to find a solution in a larger search space, the solution process requires more running time. The truck-drone collaborative delivery model proposed in this chapter requires pre-determination of special delivery points and path planning, which takes into account the unique delivery capabilities of trucks and drones. The optimized total delivery distance is shortened to 200 kilometers, the total delivery time is reduced to 8.5 hours, and the algorithm running time is significantly shortened to 15 seconds, greatly improving the delivery efficiency. This shows that by pre-determining special delivery points and optimizing the delivery routes of these points, the overall performance of the truck-drone collaborative delivery system can be significantly improved. This approach has shown positive results in cost savings, service quality improvement, and operational efficiency.

[0143] This embodiment is based on the provided logistics distribution system and method of truck and drone collaboration in mountainous environment. This embodiment aims to solve the problems of low efficiency and high cost in logistics distribution in mountainous areas. In the traditional logistics distribution model, especially in complex terrain such as mountainous areas, logistics distribution faces huge challenges, including difficulty in avoiding natural obstacles, high energy consumption and risk costs. The present invention introduces an improved A* algorithm for path planning and combines genetic algorithms to optimize the delivery routes of trucks and drones, aiming to minimize costs and maximize efficiency in the distribution process.

[0144] In this embodiment, we adopt an innovative logistics distribution system and method of truck and drone collaboration in mountainous environment, aiming to solve the main challenges in mountainous logistics distribution, including the difficulty in avoiding natural obstacles, high energy consumption and risk cost. By combining the improved A* algorithm for precise path planning and genetic algorithm to optimize the delivery routes of trucks and drones, the present invention achieves cost minimization and efficiency maximization in the distribution process. Specifically, the method first considers obstacle avoidance to establish a risk assessment model, quantifies the risk cost associated with mountain obstacles and the energy consumption of drones under the influence of wind. Then, a drone path planning model is established based on the obstacle avoidance risk cost, and special customer points that exceed the drone flight distance and maximum load limit are marked. The marking and processing of these special points ensure that the drone can efficiently complete the delivery task within its performance range, while the truck is used as a supplement to ensure that all customer needs are met. By comparing the distribution model of the traditional TSP problem, the present invention has significant improvements in the total distribution distance, total distribution time and algorithm running time. In the context of 50 search and rescue points, the site selection cost was reduced by 56.08%, the total delivery distance was shortened from 260 kilometers to 200 kilometers, the total delivery time was reduced from 11 hours to 8.5 hours, and the algorithm running time was shortened from 30 seconds to 15 seconds. These data fully demonstrate the significant advantages of the present invention in terms of cost savings, service quality improvement and operational efficiency.

[0145] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. The above is only a specific implementation method of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or replacements that can be easily thought of by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A truck and drone collaborative logistics distribution method suitable for mountainous areas, characterized in that: The following steps are involved: S1, considers obstacle avoidance to establish a risk assessment model and quantifies the risk cost E of the drone hitting an obstacle ao Energy consumption of drone under the influence of wind E ec , where the risk cost Where, the path edge e constitutes the flight path of the UAV, and the waypoints on e are represented as 1,...,m-1,m,m+1,...,γ, where γ is the number of waypoints on edge e, and R t is the radius of the t-th obstacle, the path point and the center of the t-th obstacle OB t (x c ,y c ,z c ) is D t (m,c), χ is a control parameter with a value range of [0,1], w is the number of flight path points, and z is the number of obstacle center points; Energy cost: Where e is the route edge, and the waypoints on e are represented by 1,...,m-1,m,m+ 1,...,γ, γ is the number of path points on edge e, and E is the energy cost; S2, divide the flight environment into several grids, calculate the risk cost in each grid according to the risk assessment model, and establish a global path planning model for the UAV. Specifically, based on the risk cost E ao Constructing the Cost-Benefit Matrix TC mn , based on the cost-benefit matrix constraints, the energy consumption of the UAV E ec The service path with the minimum total cost is planned by the constraint and the maximum turning angle constraint of the UAV. Specifically, the risk cost between any two points in the UAV path is quantified so that the UAV can avoid obstacles in the mountainous area. m and ro n The cost-benefit matrix TC mn It is expressed as: Among them, m ,m∈{1,2,...,N 2 } is the current position of the drone, ro n ,n∈{1,2,...,N 2 } is the next position of the drone, d mn It's a little ro m and ro n The Euclidean distance between risk is the risk cost coefficient that affects the path planning strategy; The cost-benefit matrix constraints are: Where RO is the flight path of the UAV composed of edge e, TC(RO) is the total cost of path RO, and TC(e mn ) is the edge e mn Costs; UAV energy consumption cost constraint E ec ≤E available , where E available Indicates the total available power of the drone; UAV maximum turning angle constraint Where A: (x m-1 ,y m-1 ), B:(x m ,y m ), C:(x m+1 ,y m+1 ) represent the coordinates of three consecutive path points, cosβ represents the cosine value of the angle between vector AB and vector BC, β max is the maximum acceptable turning angle; S3, based on the global path planning model of the UAV, marking the special customer points D that exceed the maximum flight distance and maximum load limit in a single flight of the UAV over , L over , special service points are delivered by truck and / or serve as take-off and landing points for drones; S4, respectively plan the paths of drones and trucks, take minimizing the delivery distance as the optimization goal, combine the delivery paths of trucks and drones for overall delivery planning, and establish a truck-drone collaborative path planning model.

2. The method according to claim 1, characterized in that In S3, if the starting node or the arrival node of the delivery is a special point, it must be delivered by truck, marked D over The origin or arrival nodes will not be served by drones, and trucks and drones will be able to enter and exit the distribution center independently or together.

3. The method according to claim 1, characterized in that In S4, in a delivery cycle, the drone passes through route points i, j. In the kth delivery, the drone takes off from the launch point i and carries the package to the drone's customer service point j. The truck path planning goal is: U represents the set of unserved customer demand nodes. In the kth allocation, if the truck moves from node i to node j, then is 1, otherwise is 0; based on the truck routing planning objective, the number of customer nodes in a single truck service is maximized; The UAV path planning objectives are: In the kth allocation, if the drone moves from node i to node j, is 1, otherwise is 0; based on the UAV path planning goal, the number of customer nodes covered by each UAV is maximized; The end point of a single delivery route is used as the starting point of the next delivery route, and the truck and drone path planning steps are repeated until all customer needs are met; The overall optimization objective function of the collaboration between drones and trucks is expressed as: N represents the set of all nodes, represents the obstacle avoidance flight range of the drone under standard conditions from point i to service point j under the load condition of the drone, d ij represents the Euclidean distance from node i to node j, and ε represents the road impedance coefficient.

4. The method according to claim 3, characterized in that Considering the power consumption of drones, the endurance of each drone in a time unit depends on its speed and payload. The range of a drone under load conditions starting from point i and arriving at service point j corresponds to the range of a drone in horizontal flight under standard conditions. It is expressed as: Where, d ij ′ is the delivery distance consumed after path planning from node i to node j considering the risk of obstacle avoidance, and σ1 is the ratio of the power of the UAV when it is flying horizontally with a load to the power of the UAV when it is flying horizontally without a load.

5. A truck and drone collaborative logistics distribution system suitable for mountainous areas, characterized in that: include: The risk assessment module inputs the flight path and obstacles and outputs the risk cost E of the drone hitting the obstacle. ao Energy consumption of drone under the influence of wind E ec , where the risk cost Where, the path edge e constitutes the flight path of the UAV, and the waypoints on e are represented as 1,...,m-1,m,m+1,...,γ, where γ is the number of waypoints on edge e, and R t is the radius of the t-th obstacle, the path point and the center of the t-th obstacle OB t (x c ,y c ,z c ) is D t (m,c), χ is a control parameter with a value range of [0,1], w is the number of flight path points, and z is the number of obstacle center points; Energy cost: Where e is the path edge, the waypoints on e are represented by 1,...,m-1,m,m+1,...,γ, γ is the number of path points on edge e, and E is the energy cost; The UAV path planning module based on obstacle avoidance risk divides the flight environment into several grids, and the risk cost E in each grid is calculated based on the risk assessment module. ao , based on E ao Constructing the Cost-Benefit Matrix TC mn , based on the cost-benefit matrix constraints, the energy consumption of the UAV E ec The service path with the minimum total cost is output by the constraint planning and the maximum turning angle constraint of the UAV. The cost-benefit matrix TC mn It is expressed as: Among them, m ,m∈{1,2,...,N 2 } is the current position of the drone, ro n ,n∈{1,2,...,N 2 } is the next position of the drone, d mn It's a little ro m and ro n The Euclidean distance between risk is the risk cost coefficient that affects the path planning strategy; The cost-benefit matrix constraints are: Where RO is the flight path of the UAV composed of edge e, TC(RO) is the total cost of path RO, and TC(e mn ) is the edge e mn Costs; UAV energy consumption cost constraint E ec ≤E available , where E available Indicates the total available power of the drone; UAV maximum turning angle constraint Where A: (x m-1 ,y m-1 ), B:(x m ,y m ), C:(x m+1 ,y m+1 ) represent the coordinates of three consecutive path points, cosβ represents the cosine value of the angle between vector AB and vector BC, β max is the maximum acceptable turning angle; Truck-UAV joint delivery module, marking special customer points D on the service route where the maximum flight distance and maximum load limit of a single UAV flight are exceeded over , L over The truck and drone path planning are performed in turn to obtain the truck-drone collaborative path.

6. The system according to claim 5, characterized in that In the truck-drone joint delivery module, special customer points are delivered by trucks, and special customer points can also serve as take-off and landing points for drones. Truck path planning is as follows: the truck provides power and cargo for drones and acts as a mobile station. After the drone is launched from the truck, the truck can depart from the launch point and head to the next stop. The truck that arrives at the next stop first has to wait for another truck, and the end point recorded by the drone's single path planning is used as the end point of the truck's delivery. The drone path planning is: under the condition that the drone meets the maximum load limit of the drone and avoids obstacles, the drone reaches the longest flight distance radius and allocates as many customer demand points as possible to the drone; The truck-drone collaborative path planning is as follows: considering the drone’s power consumption and obstacle avoidance planning, taking the end point of a single delivery route record as the starting point of the next delivery route, and cyclically repeating the truck and drone path planning until all customer needs are met, thus optimizing the route distance for each delivery.

Images