Express delivery method based on low-altitude logistics

Abstract

The invention discloses an express delivery method based on low-altitude logistics, and belongs to the technical field of express delivery, and the method comprises the steps: obtaining the position information of express delivery target points, distributing a different dimension for each express delivery target point, carrying out the coding of the express delivery target points through employing a continuous coding method, determining a plurality of different path codes, and carrying out the coding of a plurality of different paths. Then, according to the length of a low-altitude logistics path between any two express delivery target points, an express delivery constraint condition and a target optimization function, optimizing the plurality of different path codes, determining an optimal path code, and finally, on the basis of the optimal path code, obtaining a low-altitude logistics path between any two express delivery target points. And the unmanned aerial vehicle is dispatched to distribute the express on each express distribution target point, so that the express distribution efficiency can be effectively improved, the manpower cost consumption in the express distribution process is reduced, and the influence of traffic conditions is eliminated.

Description

Technical Field

[0001] The present invention belongs to the technical field of express delivery, and in particular relates to an express delivery method based on low-altitude logistics. Background Art

[0002] Express delivery is an efficient and convenient logistics service. Through professional courier companies, packages are delivered safely and quickly from senders to recipients. Senders simply fill out a waybill and hand the package to the courier, who then takes care of the subsequent transportation, sorting, and delivery. The express delivery network offers extensive coverage, enabling fast delivery within the same city, across different regions, and even across borders. Courier companies often provide real-time tracking services, allowing senders to track the status of their packages at all times. In recent years, with the booming growth of e-commerce, the express delivery industry has continued to innovate, offering diverse services such as same-day and next-day delivery to meet growing logistics needs. Traditional ground delivery methods face challenges such as low efficiency, high costs, and significant traffic impacts. Summary of the Invention

[0003] The present invention provides an express delivery method based on low-altitude logistics, which is used to solve the problems faced by traditional ground delivery methods such as low efficiency, high cost, and being greatly affected by traffic conditions.

[0004] A method for express delivery based on low-altitude logistics, comprising:

[0005] Obtaining the location information of the express delivery target points, and determining the length of the low-altitude logistics path between any two express delivery target points based on the location information of the express delivery target points;

[0006] Assign a different dimension to each express delivery destination point, and use a continuous coding method to encode the express delivery destination points to determine multiple different path codes;

[0007] Constructing express delivery constraints and a target optimization function, and optimizing the multiple different path codes based on the low-altitude logistics path length between any two express delivery destinations, the express delivery constraints, and the target optimization function to determine the optimal path code;

[0008] Based on the optimal path coding, drones are dispatched to deliver the express at each express delivery target point, completing the express delivery based on low-altitude logistics.

[0009] In a possible implementation, the method further includes:

[0010] The K-means clustering algorithm is used to cluster the location information of express delivery target points and determine multiple cluster sets;

[0011] For any cluster set, a drone is assigned to perform express delivery, and the optimal path code corresponding to the drone is determined so that the drone can deliver the express to each express delivery target point in the cluster set.

[0012] In one possible implementation, obtaining location information of express delivery destination points, and determining the length of a low-altitude logistics path between any two express delivery destination points based on the location information of the express delivery destination points, includes:

[0013] Obtaining the location information of the express delivery destination point input by the staff through human-computer interaction; wherein the location information is set to Beidou satellite positioning information;

[0014] For any two express delivery destinations, if there is no no-fly zone between the two express delivery destinations, then the straight-line distance between the two express delivery destinations is used as the low-altitude logistics path length between the two express delivery destinations based on their location information;

[0015] For any two express delivery destination points, if there is a no-fly zone between the two express delivery destination points, the two express delivery destination points are first connected by a line, and the line is bent in the no-fly zone to circle around the edge of the no-fly zone. The length of the line after the bend is determined according to the location information of the express delivery destination points, and the length of the low-altitude logistics path between the two express delivery destination points is obtained.

[0016] In one possible implementation, a different dimension is assigned to each express delivery destination point, and a continuous coding method is used to encode the express delivery destination points to determine multiple different path codes, including:

[0017] Based on the number of express delivery destinations, a multi-dimensional vector is constructed, and the parameters of each dimension in the vector are initialized between (0, 1) to obtain a path code;

[0018] Repeat the initialization multiple times to obtain multiple different path encodings; among them, each express delivery destination occupies a fixed dimension.

[0019] In one possible implementation, constructing express delivery constraints and a target optimization function includes:

[0020] The constraints are: when the power consumption of transportation to the next express delivery destination and the remaining power after reaching the next express delivery destination are greater than or equal to the power required to return to the next express delivery destination, the drone returns to replace or recharge the battery;

[0021] The target optimization function is constructed as a fusion function of time consumption and cost consumption.

[0022] In one possible implementation, the multiple different path codes are optimized to determine the optimal path code based on the low-altitude logistics path length between any two express delivery destinations, express delivery constraints, and the target optimization function, including:

[0023] For any path code, the target optimization function value of the path code is obtained based on the low-altitude logistics path length between any two express delivery destinations and the express delivery constraints.

[0024] According to the target optimization function values corresponding to all path codes, the path code with the smallest target optimization function value is determined as the optimal path code;

[0025] Based on the optimal path code, a coding collaborative search mechanism is used to perform a neighborhood search on the path code to obtain a path code after the neighborhood search;

[0026] Adopting an adaptive hybrid search mechanism to perform adaptive balanced search on the path coding after the neighborhood search, and obtaining the path coding after the adaptive balanced search;

[0027] A coding diffusion search mechanism is used to perform a global diffusion search on the path coding after the adaptive balance search to obtain the path coding after the global diffusion search;

[0028] The coding collaborative search mechanism, the adaptive hybrid search mechanism, and the coding diffusion search mechanism are repeatedly executed until the number of optimizations reaches the maximum number. Then, the optimal path coding is re-determined and output based on the path coding after the global diffusion search in the last optimization process.

[0029] In one possible implementation, for any path code, based on the length of the low-altitude logistics path between any two express delivery destinations and subject to express delivery constraints, the target optimization function value of the path code is obtained, including:

[0030] For any path code, arrange the elements of each dimension in the path code in descending order. If there are elements of the same size, arrange them in the order of the dimensions to obtain the delivery path corresponding to the path code.

[0031] For any two adjacent express delivery destination points in the delivery path corresponding to the path code, determine the transportation power consumption between any two adjacent express delivery destination points based on the low-altitude logistics path length between the any two express delivery destination points;

[0032] Based on the transportation power consumption between any two adjacent express delivery destinations, determine whether the express delivery constraint conditions are met. If so, based on the target optimization function, obtain the target optimization function value corresponding to the delivery path corresponding to the path code. Otherwise, temporarily place the return target node before the express delivery destination that does not meet the express delivery constraint conditions, and then based on the target optimization function, obtain the target optimization function value corresponding to the delivery path corresponding to the path code;

[0033] The target optimization function value corresponding to the delivery path corresponding to the path coding is used as the target optimization function value of the path coding.

[0034] In a possible implementation, based on the optimal path code, a code collaborative search mechanism is used to perform a neighborhood search on the path code to obtain the path code after the neighborhood search, including:

[0035] Determining an elite pool based on the target optimization function value of the path code; wherein the number of elite path codes in the elite pool is at least greater than 3;

[0036] According to the elite path code in the elite pool, an elite region search is performed on the path code using a balanced search factor to obtain an elite region search code;

[0037] According to the optimal path code, a quantum search method is used to perform a quantum search on the path code to obtain a quantum search code;

[0038] The elite area search code and the quantum search code are used to perform a coordinated code search on the path code to obtain the path code after the neighborhood search.

[0039] In a possible implementation, an adaptive hybrid search mechanism is used to perform an adaptive balanced search on the path coding after the neighborhood search to obtain the path coding after the adaptive balanced search, including:

[0040] According to the historical optimal value and optimal path coding corresponding to the path coding, an adaptive memory factor is used to obtain the optimization speed of the path coding after the neighborhood search in the current optimization process;

[0041] Obtaining a first neighborhood sensing position and a second neighborhood sensing position in the solution space according to the optimization speed in the previous optimization process, and obtaining a neighborhood search amount in a better direction according to the first neighborhood sensing position and the second neighborhood sensing position;

[0042] According to the optimization speed of the path coding in the current optimization process and the neighborhood search amount in the better direction, an adaptive balanced search is performed on the path coding after the neighborhood search to obtain the path coding after the adaptive balanced search.

[0043] In a possible implementation, a coding diffusion search mechanism is used to perform a global diffusion search on the path coding after the adaptive balance search to obtain the path coding after the global diffusion search, including:

[0044] Obtain optimized stagnation evaluation factor and adaptive diffusion factor;

[0045] When it is determined that the optimization stagnation evaluation factor is less than a preset optimization stagnation evaluation factor threshold, a diffusion quantity factor is obtained according to the target optimization function value corresponding to the path code after the adaptive balance search; otherwise, the path code after the adaptive balance search is directly used as the path code after the global diffusion search;

[0046] generating a plurality of diffusion codes for the path code after the adaptive balance search using the adaptive diffusion factor according to the diffusion quantity factor;

[0047] According to the diffusion code corresponding to the path code, the path code after the global diffusion search is obtained.

[0048] The present invention provides an express delivery method based on low-altitude logistics. By acquiring the position information of the express delivery target points, a different dimension is assigned to each express delivery target point, and the express delivery target points are encoded using a continuous coding method to determine a plurality of different path codes. Then, according to the low-altitude logistics path length between any two express delivery target points, the express delivery constraint conditions and the target optimization function, the plurality of different path codes are optimized to determine the optimal path code. Finally, based on the optimal path code, a drone is dispatched to deliver the express at each express delivery target point. This method can effectively improve the express delivery efficiency, reduce the labor cost consumption in the express delivery process and eliminate the influence of traffic conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0050] Figure 1 A flowchart of an express delivery method based on low-altitude logistics provided in an embodiment of the present invention.

[0051] Figure 2 A flowchart of determining optimal path coding provided by an embodiment of the present invention.

[0052] Figure 3 A flowchart of obtaining a target optimization function value for path coding provided by an embodiment of the present invention.

[0053] The above drawings illustrate specific embodiments of the present invention, which will be described in more detail below. These drawings and the accompanying description are not intended to limit the scope of the present invention in any way, but rather to illustrate the concept of the present invention to those skilled in the art by reference to specific embodiments. DETAILED DESCRIPTION

[0054] Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. In the following description, when referring to the drawings, like numbers in different figures represent like or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present invention, as detailed in the appended claims.

[0055] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0056] like Figure 1 As shown, an embodiment of the present invention provides an express delivery method based on low-altitude logistics, comprising:

[0057] S101. Acquire location information of express delivery destination points, and determine the length of a low-altitude logistics path between any two express delivery destination points based on the location information of the express delivery destination points;

[0058] The location information of express delivery destinations is generally positioning information, which can be obtained through third-party maps. Once the location information of the express delivery destinations is determined, the length of the low-altitude logistics path between any two express delivery destinations can be determined, providing supporting data for subsequent delivery routes.

[0059] S102, assigning a different dimension to each express delivery destination point, and encoding the express delivery destination points using a continuous encoding method to determine a plurality of different path codes;

[0060] In existing technologies, the ant colony algorithm (ACO) is commonly used for path planning. The ant colony algorithm (ACO) is an optimization algorithm based on biomimetic principles and is widely used in fields such as path planning and combinatorial optimization. However, this algorithm also has some significant drawbacks, which are analyzed in detail below: 1. Slow convergence: In the initial stages of the ACO, due to the uniform distribution of pheromones, ants tend to choose paths randomly. This results in the algorithm taking a long time to gradually accumulate pheromones and form an effective path selection. This slow convergence makes the algorithm less efficient when dealing with large-scale or complex problems. 2. Susceptibility to local optimality: While the positive feedback mechanism of the ACO helps quickly find a better path, it can also cause the algorithm to converge prematurely to a local optimal solution. Once excessive pheromones accumulate on a particular path, the opportunity to explore other paths is significantly reduced, thus limiting the ability to search for a global optimal solution. 3. Parameter sensitivity: The performance of the ACO is sensitive to parameter settings. Parameters such as the pheromone volatility rate and the number of ants need to be adjusted according to the specific problem. Improper parameter settings can lead to poor performance or even inability to find an effective solution. 4. High computational complexity: Since the ant colony algorithm involves parallel searches and pheromone updates by a large number of ants, its computational complexity is usually high. Especially when there are many express delivery destinations, the algorithm's running time and resource consumption will increase significantly. 5. Insufficient adaptability to complex environments: In complex and changing scenarios (such as dynamic environments or situations with multiple constraints), the basic ant colony algorithm may not be able to respond effectively. For example, in dynamic path planning, the algorithm needs to quickly adjust the pheromone distribution to adapt to environmental changes, and the basic ant colony algorithm has limited adjustment capabilities.

[0061] It can be seen that although the ant colony algorithm can realize path planning, there are many problems that may lead to inaccurate final path planning results, and other optimization algorithms are difficult to apply to path planning. Therefore, the embodiment of the present invention combines the ideas of path planning and optimization algorithm for encoding so that the optimization algorithm can realize path planning.

[0062] S103: Constructing express delivery constraints and a target optimization function, and optimizing the plurality of different path codes based on the low-altitude logistics path length between any two express delivery destinations, the express delivery constraints, and the target optimization function to determine an optimal path code;

[0063] The process of optimizing these multiple different path codes essentially involves finding a path code that minimizes the target optimization function. Express delivery constraints are used to enforce this optimization process, ensuring the reliability of the optimization results. Furthermore, the length of the low-altitude logistics path between any two express delivery destinations can be used to determine the target optimization function, thereby achieving optimization.

[0064] S104: Based on the optimal path code, dispatch drones to deliver the express at each express delivery target point, completing the express delivery based on low-altitude logistics.

[0065] The optimal path coding is essentially a delivery sequence. Express delivery can be achieved based on the optimized optimal path coding.

[0066] In a possible implementation, the express delivery method based on low-altitude logistics further includes:

[0067] The K-means clustering algorithm is used to cluster the location information of express delivery target points and determine multiple cluster sets;

[0068] For any cluster set, a drone is assigned to perform express delivery, and the optimal path code corresponding to the drone is determined so that the drone can deliver the express to each express delivery target point in the cluster set.

[0069] The location information of the express delivery target points is clustered by the above-mentioned K-means clustering algorithm, and then a drone is assigned to each cluster for express delivery. This can realize the coordinated delivery of multiple drones and achieve automated distribution.

[0070] In one possible implementation, obtaining location information of express delivery destination points, and determining the length of a low-altitude logistics path between any two express delivery destination points based on the location information of the express delivery destination points, includes:

[0071] Obtaining the location information of the express delivery destination point input by the staff through human-computer interaction; wherein the location information is set to Beidou satellite positioning information;

[0072] For example, staff can dispatch third-party maps to obtain the location information of express delivery destination points, and then determine the length of the low-altitude logistics path between the two express delivery destination points.

[0073] For any two express delivery destinations, if there is no no-fly zone between the two express delivery destinations, then the straight-line distance between the two express delivery destinations is used as the low-altitude logistics path length between the two express delivery destinations based on their location information;

[0074] For any two express delivery destination points, if there is a no-fly zone between the two express delivery destination points, the two express delivery destination points are first connected by a line, and the line is bent in the no-fly zone to circle around the edge of the no-fly zone. The length of the line after the bend is determined according to the location information of the express delivery destination points, and the length of the low-altitude logistics path between the two express delivery destination points is obtained.

[0075] In one possible implementation, a different dimension is assigned to each express delivery destination point, and a continuous coding method is used to encode the express delivery destination points to determine multiple different path codes, including:

[0076] Based on the number of express delivery destinations, a multi-dimensional vector is constructed, and the parameters of each dimension in the vector are initialized between (0, 1) to obtain a path code;

[0077] Repeat the initialization multiple times to obtain multiple different path encodings; among them, each express delivery destination occupies a fixed dimension.

[0078] Optionally, a uniform initialization mechanism for the solution space can be used to obtain multiple path codes to improve the uniformity of the distribution of path codes in the solution space, which is more conducive to finding the optimal solution and improving path planning capabilities. The uniform initialization mechanism for the solution space may include:

[0079] Based on the number of express delivery destinations, a multi-dimensional vector is constructed, and the parameters of each dimension in the vector are initialized between (0, 1) to obtain a path code;

[0080] Based on the path code after initialization, other path codes are obtained as follows:

[0081]

[0082] Among them, X g,d represents the d-th dimension element of the g-th path code, and when g=1, X g,d represents the d-th dimension element of the path encoding after initialization, d = 1, 2, ..., D, D represents the total dimension of the elements in the path encoding, X g+1,d It represents the d-th dimension element of the g+1-th path encoding. The total dimension of the elements is the total number of express delivery destinations that the drone needs to deliver. π represents the number of pi.

[0083] In one possible implementation, constructing express delivery constraints and a target optimization function includes:

[0084] The constraints are: when the power consumption of transportation to the next express delivery destination and the remaining power after reaching the next express delivery destination are greater than or equal to the power required to return to the next express delivery destination, the drone returns to replace or recharge the battery;

[0085] Optionally, the return target node can be set as a courier station or a drone charging station, and battery replacement technology can be used to achieve rapid energy replenishment, or energy replenishment can be achieved through charging technology.

[0086] The target optimization function is constructed as a fusion function of time consumption and cost consumption.

[0087] For example, the time consumption part can be set to the time required to fly the entire delivery route at a fixed speed plus the landing to take-off time of each express delivery target point, where the landing to take-off time can be set to a fixed time, or the time can be determined according to the number of express deliveries (for example, one express takes 1 minute and two express deliveries take 2 minutes, then when the landing to take-off time required to deliver one express at the express delivery target point is 2 minutes, then the landing to take-off time required to deliver two express deliveries should be 3 minutes).

[0088] The cost consumption portion can refer to power consumption. The power consumption can be calculated based on the fixed path length, and the power consumed by landing and taking off can be added to obtain the cost consumption portion. However, in reality, the heavier the cargo, the greater the power consumption may be. Therefore, the power consumption can also be calculated based on the weight of the cargo carried. For example, the power consumed per unit distance of a 1kg express delivery is h1, and the power consumed per unit distance of a 2kg express delivery is h2. Assuming that the distance between the two express delivery destinations is q, the power consumed by flying a 1kg express delivery is q×h1, and the power consumed by flying a 2kg express delivery is q×h2. Adding the power required for landing and taking off, the power consumption for transportation can be obtained.

[0089] Then, the time consumption part and the cost consumption part are weighted and fused to obtain the fusion function of time consumption and cost consumption. For example, a weight parameter w between (0, 1) can be set for time consumption, and the weight parameter for cost consumption should be 1-w.

[0090] like Figure 2 As shown, according to the low-altitude logistics path length between any two express delivery destinations, the express delivery constraints and the target optimization function, the multiple different path codes are optimized to determine the optimal path code, including:

[0091] S201. For any path code, obtain the target optimization function value of the path code based on the low-altitude logistics path length between any two express delivery destination points and the express delivery constraint conditions as constraints;

[0092] S202: Determine, based on the target optimization function values corresponding to all path codes, the path code with the smallest target optimization function value as the optimal path code;

[0093] S203: Based on the optimal path code, a coding collaborative search mechanism is used to perform a neighborhood search on the path code to obtain a path code after the neighborhood search;

[0094] S204, using an adaptive hybrid search mechanism to perform an adaptive balanced search on the path code after the neighborhood search to obtain the path code after the adaptive balanced search;

[0095] S205, using a code diffusion search mechanism to perform a global diffusion search on the path code after the adaptive balance search to obtain the path code after the global diffusion search;

[0096] S206. Repeat the coding collaborative search mechanism, the adaptive hybrid search mechanism, and the coding diffusion search mechanism until the number of optimizations reaches the maximum number of optimizations. Then, based on the path coding after the global diffusion search in the last optimization process, the optimal path coding is re-determined and output.

[0097] While many existing optimization algorithms offer better results for optimizing numerical parameters than traditional path planning algorithms such as genetic algorithms and ant colony algorithms, they can also easily fall into local optima and experience poor optimization results, making them unsuitable for path planning. Therefore, the encoding and optimization methods provided by the embodiments of the present invention not only transform path planning into continuous numerical optimization but also leverage the powerful optimization capabilities of optimization algorithms to improve path planning performance.

[0098] At the same time, the embodiment of the present invention proposes a new optimization algorithm to solve the problems of the existing technology that it is easy to fall into local optimality and the optimization effect is poor, and further improve the path planning capability.

[0099] like Figure 3 As shown in the figure, for any path code, according to the low-altitude logistics path length between any two express delivery target points, and with the express delivery constraint as a restriction, the target optimization function value of the path code is obtained, including:

[0100] S301. For any path code, arrange the elements of each dimension in the path code in descending order. If there are elements of the same size, arrange them in order of the dimensions to obtain the delivery path corresponding to the path code.

[0101] For example, there are five express delivery destinations a1, a2, a3, a4, and a5. In a certain path code, the elements corresponding to a, b, c, d, and e are: 0.1, 0.2, 0.5, 0.2, and 0.3, respectively. Then the delivery path corresponding to the path code should be: a3, a5, a2, a4, and a1.

[0102] S302: For any two adjacent express delivery destination points in the delivery path corresponding to the path code, determine the transportation power consumption between the any two adjacent express delivery destination points based on the low-altitude logistics path length between the any two express delivery destination points;

[0103] The power consumption per unit distance of transporting packages of varying weights can be obtained in advance. Once the delivery route corresponding to the route code is determined, the transport power consumption can be calculated based on the weight of the drone and the length of the low-altitude logistics path between two adjacent delivery destinations. To make route planning more accurate, the power consumed during landing and takeoff can also be included in the transport power consumption, thus preventing the drone from running out of power.

[0104] For example, the amount of electricity consumed per unit distance of transporting 50g, the amount of electricity consumed per unit distance of transporting 100g, the amount of electricity consumed per unit distance of transporting 150g, and so on can be obtained in advance. After determining the weight of the transported goods, a table can be looked up to determine the amount of electricity consumed per unit distance of the drone flying between two adjacent express delivery destinations. Combined with the length of the low-altitude logistics path between the two adjacent express delivery destinations, the amount of electricity consumed between any two adjacent express delivery destinations can be determined. Then, by adding the preset amount of electricity consumed for landing and takeoff, the amount of electricity consumed for transportation can be determined. It is worth noting that in order to make the calculation of transportation power consumption more accurate, after obtaining the amount of electricity consumed per unit distance of the transportation weight, interpolation processing can be performed to improve the query accuracy.

[0105] Under ideal conditions, the amount of electricity consumed per unit weight carried and per unit distance flown should be the same and cumulative. Therefore, we can first measure the amount of electricity consumed by the drone per unit weight carried and per unit distance flown, and then convert the transportation electricity consumption between any two adjacent express delivery destinations based on the carried weight and the length of the low-altitude logistics path.

[0106] S303. Determine whether the express delivery constraint conditions are met based on the transportation power consumption between any two adjacent express delivery destinations. If so, obtain the target optimization function value corresponding to the delivery path corresponding to the path code based on the target optimization function. Otherwise, temporarily place the return destination node before the express delivery destination that does not meet the express delivery constraint conditions, and then obtain the target optimization function value corresponding to the delivery path corresponding to the path code based on the target optimization function.

[0107] For example, when there are five express delivery destinations a1, a2, a3, a4, and a5, and the return destination node is express delivery station b, assuming that the drone fails to meet the constraint after flying from a2 to a3, that is, the drone has sufficient power to fly from a2 to b, but after flying to a3, the power is insufficient to fly to b. In this case, b should be inserted before a3, that is, a1, a2, b, a3, a4, a5. It is worth noting that the return destination node is only temporarily placed in the path code and is only used to obtain the target optimization function value corresponding to the delivery path corresponding to the path code to avoid changes in the length of the path code.

[0108] S304. The target optimization function value corresponding to the delivery path corresponding to the path code is used as the target optimization function value of the path code.

[0109] In a possible implementation, based on the optimal path code, a code collaborative search mechanism is used to perform a neighborhood search on the path code to obtain the path code after the neighborhood search, including:

[0110] S2031. Determine an elite pool based on the target optimization function value of the path code; wherein the number of elite path codes in the elite pool is at least greater than 3;

[0111] For example, in addition to the optimal path encoding, the three path encodings with the largest target optimization function values can also be selected to form an elite pool. However, it is worth noting that the number here is only an example of an embodiment of the present invention, and can also be set to other numbers such as 5, 10, 20, etc., which can be set according to actual needs.

[0112] S2032: Based on the elite path code in the elite pool, an elite region search is performed on the path code using a balanced search factor, and the elite region search code is obtained as follows:

[0113]

[0114] in, Indicates the elite region search code corresponding to the mth path code in the tth optimization process, represents the i-th elite path code selected for the m-th path code during the t-th optimization process, m = 1, 2, ..., M, M represents the total number of path codes, i = 1, 2, 3; λ i represents the influence weight of the i-th elite path code corresponding to the m-th path code in the t-th optimization process, f mi Indicates the influence degree of the i-th elite path code corresponding to the m-th path code in the t-th optimization process, f mimin represents the minimum impact degree of all elite path codes corresponding to the mth path code in the tth optimization process, fmimax Indicates the maximum influence degree of all elite path codes corresponding to the mth path code in the tth optimization process;

[0115] The degree of influence is obtained by adding the target optimization function value to a very small constant (such as 0.0001) and taking the inverse.

[0116] S2033. Based on the optimal path code, a quantum search method is used to perform a quantum search on the path code, and the quantum search code is obtained as follows:

[0117]

[0118] in, represents the quantum search code corresponding to the m-th path code in the t-th optimization process, represents the optimal path encoding, α represents the shrinkage coefficient, Indicates the encoding of the mth path in the tth optimization process, represents the historical optimal value corresponding to the mth path encoding in the tth optimization process, r1 represents the first random number between (0,1), cos represents the cosine function, π represents pi, and T represents the preset maximum number of optimization times;

[0119] S2034. Use the elite area search code and the quantum search code to perform a coordinated search on the path code, and obtain the path code after the neighborhood search as follows:

[0120]

[0121]

[0122] in, represents the path encoding after the mth neighborhood search, β m1 represents the collaborative search weight corresponding to the elite regional search code, β m2 represents the collaborative search weight corresponding to the quantum search code, f m1 Indicates the influence degree corresponding to the elite area search code, f m2 Indicates the degree of influence corresponding to the quantum search coding.

[0123] The embodiment of the present invention adopts a coding collaborative search mechanism to perform neighborhood search on the path code, which not only enables the path code to search for a more optimal area in the solution space, but also can learn the optimal path code and the location information of the historical state. At the same time, it uses quantum behavior for search, which can enable the algorithm to have the ability to jump out of the local optimum. The factor balances the global search and local search capabilities while ensuring the optimization speed.

[0124] In a possible implementation, an adaptive hybrid search mechanism is used to perform an adaptive balanced search on the path coding after the neighborhood search to obtain the path coding after the adaptive balanced search, including:

[0125] S2041. Based on the historical optimal value and the optimal path code corresponding to the path code, and using the adaptive memory factor to obtain the path code after the neighborhood search, the optimization speed in the current optimization process is:

[0126]

[0127] in, represents the path code after the nth neighborhood search in the tth optimization process, n=1,2,…,M, M represents the total number of path codes, It represents the optimization speed corresponding to the path encoding after the nth neighborhood search in the tth optimization process, that is, the optimization speed in the previous optimization process; represents the optimization speed corresponding to the path encoding after the nth neighborhood search in the t+1th optimization process, that is, the optimization speed in the current optimization process; c1 represents the first learning factor, c2 represents the second learning factor, r2 represents the second random number between (0,1), and r3 represents the third random number between (0,1). represents the optimal path encoding, represents random path encoding; ω represents inertia weight, ω max Indicates the preset maximum value of the inertia weight, ω min Indicates the preset minimum value of inertia weight;

[0128] S2042. Obtain a first neighborhood sensing position and a second neighborhood sensing position in the solution space based on the optimization speed in the previous optimization process, and obtain a neighborhood search amount in a better direction based on the first neighborhood sensing position and the second neighborhood sensing position:

[0129]

[0130] in, It represents the neighborhood search amount corresponding to the path encoding after the nth neighborhood search in the tth optimization process, represents the search step length during the t+1th optimization process, and represents the search step length during the t-th optimization process, and At the initial moment, it is randomly generated from (0,1) or preset to a fixed value, κ1 represents the first attenuation factor, sign represents the sign function, represents the first neighbor-aware position, represents the second neighbor perception position, f n1Indicates the influence degree corresponding to the first neighborhood perception position, f n2 Indicates the influence degree corresponding to the second neighborhood perception position, η t+1 represents the neighborhood search distance control parameter in the t+1th optimization process, and η t+1 =κ2η t , η t represents the neighborhood search distance control parameter in the t-th optimization process, κ2 represents the second attenuation factor;

[0131] S2043. Based on the optimization speed of the path code in the current optimization process and the amount of neighborhood search in the optimal direction, an adaptive balanced search is performed on the path code after the neighborhood search, and the path code after the adaptive balanced search is obtained as follows:

[0132]

[0133] in, represents the path code after the nth adaptive balance search, and r4 represents the fourth random number between (0.2, 0.8).

[0134] The embodiment of the present invention adopts an adaptive hybrid search mechanism to perform an adaptive balanced search on the path coding after the neighborhood search, which can adaptively search in a better direction of the neighborhood, and perform adaptive social learning and individual learning at the same time, thereby ensuring the solution space traversal and neighborhood exploration of the algorithm, improving the search accuracy and search speed, and providing a certain global optimization capability.

[0135] In a possible implementation, a coding diffusion search mechanism is used to perform a global diffusion search on the path coding after the adaptive balance search to obtain the path coding after the global diffusion search, including:

[0136] S2051. Obtain the optimized stagnation evaluation factor and the adaptive diffusion factor as follows:

[0137]

[0138] Among them, φ k represents the optimization stagnation evaluation factor, represents the path encoding after the kth adaptive balance search during the tth training process, represents the path encoding after the jth adaptive balance search in the tth training process, k=1,2,…,M, M represents the total number of path encodings, express and The Euclidean distance between them; σ represents the adaptive diffusion factor, e represents the natural constant, r5 represents the fifth random number between (0,1), and sin represents the sine function;

[0139] S2052: When it is determined that the optimization stagnation evaluation factor is less than a preset optimization stagnation evaluation factor threshold, a diffusion quantity factor is obtained based on the target optimization function value corresponding to the path code after the adaptive balance search; otherwise, the path code after the adaptive balance search is directly used as the path code after the global diffusion search;

[0140] The diffusion number factor is:

[0141]

[0142] Among them, S k represents the diffusion factor corresponding to the path encoding after the kth adaptive balance search, floor represents the rounding function, f k represents the influence degree of the path encoding after the kth adaptive balance search in the tth training process, f worst Indicates the impact degree corresponding to the worst path encoding, f best Indicates the influence degree corresponding to the optimal path encoding, S max represents the upper limit of the diffusion number factor, S min Indicates the lower limit of the diffusion number factor;

[0143] S2053: Based on the diffusion quantity factor, the adaptive diffusion factor is used to generate multiple diffusion codes for the path code after the adaptive balance search:

[0144]

[0145] in, represents the d-th dimension element of the path encoding after the k-th adaptive balance search in the t-th training process, d = 1, 2, ..., D, D represents the total dimension of the elements in the path encoding, represents the d-th dimension element of the diffusion coding, γ1 represents the first diffusion control factor, γ2 represents the second diffusion control factor (for example, it can be set to a constant between (0, 0.2) or (0, 0.5), which can be set according to actual needs), Cauchy (0, σ 2 ) represents the Cauchy variation, Gauss(0,σ 2 ) represents Gaussian variation;

[0146] S2054: Obtain the path code after the global diffusion search according to the diffusion code corresponding to the path code.

[0147] For example, the code with the smallest target optimization function value may be selected from the path code and the corresponding diffusion code as the path code after the global diffusion search.

[0148] This embodiment of the present invention uses a coded diffusion search mechanism to perform a global diffusion search on the path codes after the adaptive balance search. Based on the path codes, this diffusion search integrates Cauchy and Gaussian mutations. The adaptive diffusion factor decreases with each optimization iteration, ensuring algorithm convergence. Furthermore, the adaptive diffusion factor fluctuates during the search process, making it easier to escape local optimal solutions and find the global optimal solution.

[0149] Optionally, after each update of the path code, the path code can be processed for out-of-bounds conditions to ensure that the path code is always between (0, 1). For example, the out-of-bounds elements can be randomly generated between (0, 1), or the upper bound elements can be set to 0.99 and the lower bound elements can be set to 0.01.

[0150] Through the mutual cooperation of the above mechanisms, we can effectively escape from the local optimum and the convergence ability of the algorithm, so that the algorithm can find the global optimal solution in the solution space, thereby improving the path planning ability and reducing the consumption of time and cost.

[0151] The present invention provides an express delivery method based on low-altitude logistics. By acquiring the position information of the express delivery target points, a different dimension is assigned to each express delivery target point, and the express delivery target points are encoded using a continuous coding method to determine a plurality of different path codes. Then, according to the low-altitude logistics path length between any two express delivery target points, the express delivery constraint conditions and the target optimization function, the plurality of different path codes are optimized to determine the optimal path code. Finally, based on the optimal path code, a drone is dispatched to deliver the express at each express delivery target point. This method can effectively improve the express delivery efficiency, reduce the labor cost consumption in the express delivery process and eliminate the influence of traffic conditions.

[0152] Those skilled in the art will readily appreciate other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. The present invention is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of the present invention and include common knowledge or customary techniques in the art not disclosed herein. It should be understood that the present invention is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and variations can be made without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. A method for express delivery based on low-altitude logistics, characterized in that: include: Obtaining the location information of the express delivery target points, and determining the length of the low-altitude logistics path between any two express delivery target points based on the location information of the express delivery target points; Assign a different dimension to each express delivery destination point, and use a continuous coding method to encode the express delivery destination points to determine multiple different path codes; Constructing express delivery constraints and a target optimization function, and optimizing the multiple different path codes based on the low-altitude logistics path length between any two express delivery destinations, the express delivery constraints, and the target optimization function to determine the optimal path code; Based on the optimal path coding, drones are dispatched to deliver the express at each express delivery target point, completing the express delivery based on low-altitude logistics.

2. The express delivery method based on low-altitude logistics according to claim 1, characterized in that: Also includes: The K-means clustering algorithm is used to cluster the location information of express delivery target points and determine multiple cluster sets; For any cluster set, a drone is assigned to perform express delivery, and the optimal path code corresponding to the drone is determined so that the drone can deliver the express to each express delivery target point in the cluster set.

3. The express delivery method based on low-altitude logistics according to claim 1, characterized in that: Obtaining the location information of the express delivery target points, and determining the length of the low-altitude logistics path between any two express delivery target points based on the location information of the express delivery target points, including: Obtaining the location information of the express delivery destination point input by the staff through human-computer interaction; wherein the location information is set to Beidou satellite positioning information; For any two express delivery destinations, if there is no no-fly zone between the two express delivery destinations, then the straight-line distance between the two express delivery destinations is used as the low-altitude logistics path length between the two express delivery destinations based on their location information; For any two express delivery destination points, if there is a no-fly zone between the two express delivery destination points, the two express delivery destination points are first connected by a line, and the line is bent in the no-fly zone to circle around the edge of the no-fly zone. The length of the line after the bend is determined according to the location information of the express delivery destination points, and the length of the low-altitude logistics path between the two express delivery destination points is obtained.

4. The express delivery method based on low-altitude logistics according to claim 1, characterized in that: Assign a different dimension to each express delivery destination point, and use a continuous coding method to encode the express delivery destination points to determine multiple different path codes, including: Based on the number of express delivery destinations, a multi-dimensional vector is constructed, and the parameters of each dimension in the vector are initialized between (0, 1) to obtain a path code; Repeat the initialization multiple times to obtain multiple different path encodings; among them, each express delivery destination occupies a fixed dimension.

5. The express delivery method based on low-altitude logistics according to claim 4 is characterized in that: Construct express delivery constraints and target optimization functions, including: The constraints are: when the power consumption of transportation to the next express delivery destination and the remaining power after reaching the next express delivery destination are greater than or equal to the power required to return to the next express delivery destination, the drone returns to replace or recharge the battery; The target optimization function is constructed as a fusion function of time consumption and cost consumption.

6. The express delivery method based on low-altitude logistics according to claim 5, characterized in that: Optimizing the plurality of different path codes to determine the optimal path code based on the low-altitude logistics path length between any two express delivery destinations, express delivery constraints, and the target optimization function, including: For any path code, the target optimization function value of the path code is obtained based on the low-altitude logistics path length between any two express delivery destinations and the express delivery constraints. According to the target optimization function values corresponding to all path codes, the path code with the smallest target optimization function value is determined as the optimal path code; Based on the optimal path code, a coding collaborative search mechanism is used to perform a neighborhood search on the path code to obtain a path code after the neighborhood search; Adopting an adaptive hybrid search mechanism to perform adaptive balanced search on the path coding after the neighborhood search, and obtaining the path coding after the adaptive balanced search; A coding diffusion search mechanism is used to perform a global diffusion search on the path coding after the adaptive balance search to obtain the path coding after the global diffusion search; The coding collaborative search mechanism, the adaptive hybrid search mechanism, and the coding diffusion search mechanism are repeatedly executed until the number of optimizations reaches the maximum number. Then, the optimal path coding is re-determined and output based on the path coding after the global diffusion search in the last optimization process.

7. The express delivery method based on low-altitude logistics according to claim 6, characterized in that: For any path code, based on the low-altitude logistics path length between any two express delivery destinations and subject to express delivery constraints, the target optimization function value of the path code is obtained, including: For any path code, arrange the elements of each dimension in the path code in descending order. If there are elements of the same size, arrange them in the order of the dimensions to obtain the delivery path corresponding to the path code. For any two adjacent express delivery destination points in the delivery path corresponding to the path code, determine the transportation power consumption between any two adjacent express delivery destination points based on the low-altitude logistics path length between the any two express delivery destination points; Based on the transportation power consumption between any two adjacent express delivery destinations, determine whether the express delivery constraint conditions are met. If so, based on the target optimization function, obtain the target optimization function value corresponding to the delivery path corresponding to the path code. Otherwise, temporarily place the return target node before the express delivery destination that does not meet the express delivery constraint conditions, and then based on the target optimization function, obtain the target optimization function value corresponding to the delivery path corresponding to the path code; The target optimization function value corresponding to the delivery path corresponding to the path coding is used as the target optimization function value of the path coding.

8. The express delivery method based on low-altitude logistics according to claim 6, characterized in that: Based on the optimal path code, a coding collaborative search mechanism is used to perform a neighborhood search on the path code to obtain the path code after the neighborhood search, including: Determining an elite pool based on the target optimization function value of the path code; wherein the number of elite path codes in the elite pool is at least greater than 3; According to the elite path code in the elite pool, an elite region search is performed on the path code using a balanced search factor to obtain an elite region search code; According to the optimal path code, a quantum search method is used to perform a quantum search on the path code to obtain a quantum search code; The elite area search code and the quantum search code are used to perform a coordinated code search on the path code to obtain the path code after the neighborhood search.

9. The express delivery method based on low-altitude logistics according to claim 8, characterized in that: An adaptive hybrid search mechanism is used to perform an adaptive balanced search on the path coding after the neighborhood search, and the path coding after the adaptive balanced search is obtained, including: According to the historical optimal value and optimal path coding corresponding to the path coding, the optimization speed of the path coding after the neighborhood search in the current optimization process is obtained by using an adaptive memory factor; Obtaining a first neighborhood sensing position and a second neighborhood sensing position in the solution space according to the optimization speed in the previous optimization process, and obtaining a neighborhood search amount in a better direction according to the first neighborhood sensing position and the second neighborhood sensing position; According to the optimization speed of the path coding in the current optimization process and the neighborhood search amount in the better direction, an adaptive balanced search is performed on the path coding after the neighborhood search to obtain the path coding after the adaptive balanced search.

10. The express delivery method based on low-altitude logistics according to claim 9, characterized in that: The coding diffusion search mechanism is used to perform a global diffusion search on the path coding after the adaptive balance search, and the path coding after the global diffusion search is obtained, including: Obtain optimized stagnation evaluation factor and adaptive diffusion factor; When it is determined that the optimization stagnation evaluation factor is less than a preset optimization stagnation evaluation factor threshold, a diffusion quantity factor is obtained according to the target optimization function value corresponding to the path code after the adaptive balance search; otherwise, the path code after the adaptive balance search is directly used as the path code after the global diffusion search; generating a plurality of diffusion codes for the path code after the adaptive balance search using the adaptive diffusion factor according to the diffusion quantity factor; According to the diffusion code corresponding to the path code, the path code after the global diffusion search is obtained.

Images