Dynamic departure sorting method and device for the terminal area of a logistics drone take-off and landing field
Through dynamic departure sorting method, fuzzy clustering and genetic algorithms are used to optimize the departure order of drones, which solves the departure sorting problem of logistics drone terminal area, improves operation efficiency and safety, and realizes personalized cargo orders Serve.
Patent Information
- Application Number
- CN202211399976.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-11-09
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2042-11-09
AI Technical Summary
The current problem of departure sorting in logistics drone terminal areas needs to be solved urgently, resulting in possible congestion, safety hazards and economic losses. It is difficult for the existing technology to effectively manage the order of departure of drones and improve operational efficiency.
A dynamic departure sorting method for logistics drone take-off and landing terminal area is proposed. By obtaining the drone flight plan and cargo order information, fuzzy clustering analysis is carried out to determine the delivery task category, based on the value, distance and the late one of the categories Set priority for delivery time, build a departure sorting model, and use genetic algorithm to optimize the sorting scheme to target the minimum delay cost, ensuring safe time intervals and maximum allowable delay time constraints.
It realizes effective management of the order of departure of drones, improves the efficiency of airspace operation in terminal areas, meets the personalized needs of different cargo orders, ensures the safety and efficiency of distribution services, and provides an economical and practical resource. 整合方案。
Smart Images

Figure CN115730787B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of information control technology, and in particular relates to a method and device for dynamic departure sorting in a terminal area of a logistics drone take-off and landing field. Background Art
[0002] Civilian drones have been in a rapid development stage in recent years. The establishment of a low-altitude unmanned logistics distribution system has become an important part of building a smart community. The development of the market and the national economy is providing more and more demand for logistics drone delivery services. In order to meet the needs of the distribution market and prevent congestion problems in the normalization of logistics drone delivery services in the future, causing safety hazards and economic losses, logistics drone terminal area traffic management is crucial. The main methods of terminal area traffic management currently include ground waiting, entry and departure sorting, and capacity flow allocation. As one of the important management methods, drone departure sorting is still a big gap in its current research. The problem of departure sorting in urban logistics drone terminals needs to be solved urgently. Summary of the invention
[0003] In response to the above problems, the present invention proposes a dynamic departure sorting method and device for the terminal area of a logistics drone take-off and landing field, which can effectively manage the departure order of drones and improve the operating efficiency of the drone terminal area airspace.
[0004] In order to achieve the above technical objectives and the above technical effects, the present invention is implemented through the following technical solutions:
[0005] In a first aspect, the present invention provides a method for dynamically sorting departures in a terminal area of a logistics drone take-off and landing field, comprising:
[0006] Obtain drone flight plan information and cargo order information at the terminal area of the logistics drone take-off and landing field within a certain period of time;
[0007] Determine an initial departure sorting plan based on the flight plan information and cargo order information, the initial sorting plan including the initial departure order of the logistics drones and the initial departure time of each drone;
[0008] Performing fuzzy clustering analysis on all cargo orders according to the cargo order information, taking one cluster as a delivery task category, and obtaining multiple delivery task categories;
[0009] Based on the value of the goods in each delivery task category, the distance from the delivery point to the terminal area of the logistics drone landing and take-off field, and the latest delivery time, the priority of each delivery task category is determined, and different delay penalty coefficients are set for each priority;
[0010] Based on the priority of each delivery task category, a departure sorting model is constructed; the departure sorting model aims to minimize the total delay cost, and the constraints include the drone departure safety time interval constraint and the maximum allowable delay time constraint;
[0011] The initial departure sorting scheme is encoded as a chromosome, and the departure sorting model is iteratively solved with a genetic algorithm to obtain an optimal departure sorting scheme, which includes the optimal departure order of logistics drones and the optimal departure time of each drone.
[0012] Optionally, performing fuzzy clustering analysis on all cargo orders according to the cargo order information, taking one cluster as a delivery task category, and obtaining multiple delivery task categories, comprises the following steps:
[0013] The distance from the delivery point to the terminal area of the logistics drone landing and take-off field, the latest delivery time and the value of the goods in the cargo order information are used as indicators;
[0014] The entropy weight method is used to determine the indicator weights of the distance from the receiving point to the terminal area of the logistics drone take-off and landing field, the latest receiving time and the value of the goods;
[0015] Based on the indicator weight of the goods value, the first weighted similarity between order i and order j regarding the goods value is calculated;
[0016] Based on the index weight of the distance from the delivery point to the terminal area of the logistics drone landing field, the second weighted similarity of order i and order j regarding the distance from the delivery point to the terminal area of the logistics drone landing field is calculated;
[0017] Based on the indicator weight of the latest delivery time, the third weighted similarity between order i and order j regarding the latest delivery time is calculated;
[0018] Based on the first weighted similarity, the second weighted similarity and the third weighted similarity, a multi-index comprehensive similarity between order i and order j is calculated;
[0019] Generate a similarity matrix based on the multi-index comprehensive similarity;
[0020] Based on the similarity matrix, the clustering results of all cargo orders are obtained by using the transitive closure method, so that one cluster is used as a delivery task category, and multiple delivery task categories are obtained.
[0021] Optionally, the index weight of the cargo value, the index weight of the distance from the delivery point to the terminal area of the logistics drone landing and take-off field, and the index weight of the latest delivery time are calculated by the following formula:
[0022]
[0023]
[0024] In the formula, when k = 1, ω k is the indicator weight of the goods value, P ik is the index value ratio of order i to the value of goods, E k is the entropy value of the goods; when k = 2, ω k is the indicator weight of the distance from the receiving point to the terminal area of the logistics drone landing and take-off field, P ik It represents the weight of the index value of the distance from the receiving point to the terminal area of the logistics drone landing field for order i, E k is the entropy value of the distance from the receiving point to the terminal area of the logistics drone landing field; when k = 3, ω k is the indicator weight of the latest delivery time, P ik represents the weight of the index value of order i to the latest delivery time, E k is the entropy value of the latest delivery time.
[0025] Optionally, the calculation formula of the first weighted similarity is:
[0026]
[0027]
[0028]
[0029] In the formula, r ij1 is the first weighted similarity, i is order i, j is order j, ω1 is the indicator weight of the goods value, is the triangular fuzzy number of the value index of goods in order i, α l i , α m i , α u i and α l j , α m j , α u j are the lower limit, maximum possibility value and upper limit of the triangular fuzzy number of the goods value index of order i and order j respectively, and n is the total number of orders.
[0030] Optionally, the calculation formulas of the second weighted similarity and the third weighted similarity are the same, both of which are:
[0031]
[0032]
[0033] In the formula, when k = 2, rijk is the second weighted similarity, ω k is the indicator weight of the distance from the receiving point to the terminal area of the logistics drone landing field; when k = 3, r ijk is the third weighted similarity, ω k is the indicator weight of the latest delivery time, is the normalized index value of order i corresponding to the kth index, and n is the total number of orders; when k = 2, is the actual index value of the distance from the receiving point to the take-off and landing field corresponding to order i; when k = 3, is the actual indicator value corresponding to the latest delivery time of order i.
[0034] Optionally, the calculation formula for the multi-index comprehensive similarity between order i and order j is:
[0035]
[0036] The expression of the similarity matrix is:
[0037] R=(S ij ) n×n .
[0038] Optionally, the priority of each delivery task category is determined based on the value of the goods of each delivery task category, the distance from the delivery point to the terminal area of the logistics drone landing and take-off field, and the latest delivery time, and different delay penalty coefficients are set for each priority, including the following steps:
[0039] Select the average cargo value X for each delivery task category i , the average distance Y from the receiving point to the terminal area of the logistics drone landing field i and the average latest delivery time Z i As characteristic parameters;
[0040] Calculate the values of the three characteristic parameters of each delivery task category;
[0041] Based on the values of the three characteristic parameters of each delivery task category, the priority of each delivery task category is calculated;
[0042] Based on the priority of each delivery task category, the corresponding delay penalty coefficient is calculated.
[0043] Optionally, the calculation formula for the values of the three characteristic parameters of each delivery task category is:
[0044]
[0045]
[0046]
[0047] in, represents the average distance from all order receiving points in the delivery task category g to the terminal area of the logistics drone landing and take-off field, represents the average latest delivery time of all orders in the delivery task category g, α m i is the maximum possible value of a triangular fuzzy number, X g , Y g , Z g They represent the values of the three characteristic parameters of the delivery task category g, namely, the value of the goods, the distance from the receiving point to the terminal area of the logistics drone landing and take-off field, and the latest receiving time. They represent the average value of all order goods in task category g, the distance from the delivery point to the terminal area of the logistics drone landing and take-off field, and the latest delivery time;
[0048] The calculation formula for the priority of each delivery task category is:
[0049] prior g =(p g -1)(p g -2)(p g -3) / 6+(2p g -X g -2)(X g -1) / 2+Y g
[0050] p g =X g +Y g +Z g
[0051] The calculation formula for the delay penalty coefficient of each delivery task category is:
[0052]
[0053] In the formula, c ag is the delay penalty coefficient of category g, prior g is the task priority of category g.
[0054] Optionally, the objective function of the departure sorting model is:
[0055]
[0056]
[0057] Where, T i is the actual departure time of the drone of order i, L iis the latest departure time of the drone for order i, LT i is the latest arrival time of the drone for order i, d i is the distance from the order i receiving point to the terminal area of the logistics drone landing field, v is the flight speed of the drone, c ai is the delay penalty coefficient of order i, and n is the number of orders;
[0058] The constraints of the departure sorting model are:
[0059] T i -L i ≤lt i
[0060] T i -T j ≥t ij
[0061] In the formula, lt i is the maximum allowable delay time of the delivery point of order i, T i is the actual departure time of the drone of order i, T j is the actual departure time of the drone of order j, t ij is the safe time interval between the takeoffs of the two drones of order i and order j.
[0062] In a second aspect, the present invention provides a dynamic departure sorting device for a terminal area of a logistics drone take-off and landing field, comprising:
[0063] The acquisition module is used to obtain the UAV flight plan information and cargo order information in the terminal area of the logistics UAV take-off and landing field within a certain period of time;
[0064] An initial departure sorting scheme determination module is used to determine an initial departure sorting scheme according to the flight plan information and the cargo order information, wherein the initial sorting scheme includes the initial departure order of the logistics drones and the initial departure time of each drone;
[0065] A clustering module, used to perform fuzzy clustering analysis on all cargo orders according to the cargo order information, wherein one cluster is regarded as a delivery task category, and a plurality of delivery task categories are obtained;
[0066] The priority and delay penalty coefficient determination module is used to determine the priority of each delivery task category based on the value of the goods in each delivery task category, the distance from the delivery point to the terminal area of the logistics drone landing and take-off field, and the latest delivery time, and set different delay penalty coefficients for each priority;
[0067] A departure sorting model module is constructed to construct a departure sorting model based on the priority of each delivery task category; the departure sorting model aims to minimize the total delay cost, and the constraints include the safety time interval constraint of the UAV departure and the maximum allowable delay time constraint;
[0068] The optimal departure sorting scheme acquisition module is used to encode the initial departure sorting scheme as a chromosome, cooperate with the genetic algorithm to iteratively solve the departure sorting model, and obtain the optimal departure sorting scheme. The optimal sorting scheme includes the optimal departure order of logistics drones and the optimal departure time of each drone.
[0069] Compared with the prior art, the present invention has the following beneficial effects:
[0070] The present invention fully considers the low-altitude environmental restrictions and delivery service requirements of logistics drone terminals, and provides personalized services for different cargo orders by clustering cargo orders and prioritizing delivery tasks. On this basis, the departure order of drones is sorted with the lowest delay cost as the goal, which not only ensures the safety and efficiency of delivery services, but also meets the personalized needs of different cargo orders, providing an economical and practical solution for integrating logistics resources and improving delivery efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0071] In order to make the content of the present invention more clearly understood, the present invention is further described in detail below based on specific embodiments and in conjunction with the accompanying drawings, wherein:
[0072] Figure 1 This is a flow chart of a method for dynamic departure sorting in a terminal area of a logistics drone take-off and landing field according to an embodiment of the present invention;
[0073] Figure 2 A process of clustering goods orders according to an embodiment of the present invention;
[0074] Figure 3 This is a schematic diagram of a vertical take-off and landing field for an unmanned aerial vehicle according to an embodiment of the present invention. DETAILED DESCRIPTION
[0075] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the scope of protection of the present invention.
[0076] The application principle of the present invention is described in detail below in conjunction with the accompanying drawings.
[0077] Example 1
[0078] The embodiment of the present invention provides a method for dynamic departure sorting in the terminal area of a logistics drone take-off and landing field, comprising the following steps:
[0079] (1) Obtaining drone flight plan information and cargo order information at the terminal area of the logistics drone take-off and landing field within a certain period of time;
[0080] (2) determining an initial departure sequencing plan based on the flight plan information and cargo order information, the initial sequencing plan including the initial departure order of the logistics drones and the initial departure time of each drone;
[0081] (3) performing fuzzy clustering analysis on all cargo orders according to the cargo order information, wherein one cluster is regarded as a delivery task category, and a plurality of delivery task categories are obtained;
[0082] (4) Based on the value of the goods in each delivery task category, the distance from the delivery point to the terminal area of the logistics drone landing and take-off field, and the latest delivery time, determine the priority of each delivery task category and set different delay penalty coefficients for each priority;
[0083] (5) Based on the priority of each delivery task category, a departure sorting model is constructed; the departure sorting model aims to minimize the total delay cost, and the constraints include the UAV departure safety time interval constraint and the maximum allowable delay time constraint;
[0084] (6) The initial departure sorting scheme is encoded as a chromosome, and the departure sorting model is iteratively solved with a genetic algorithm to obtain an optimal departure sorting scheme. The optimal sorting scheme includes the optimal departure order of the logistics drones and the optimal departure time of each drone.
[0085] The following is a detailed description of the dynamic departure sorting method for the terminal area of the logistics drone take-off and landing field in the embodiment of the present invention in conjunction with a specific implementation method. Figure 1 As shown, the specific steps include:
[0086] Step A, obtaining the drone flight plan information and cargo order information in the terminal area of the logistics drone take-off and landing field within a certain period of time in the current urban environment;
[0087] The UAV flight plan information includes all departure flights within the period, the earliest and latest planned departure times of each departure flight, and the types and quantities of logistics UAVs involved;
[0088] The cargo order information includes cargo weight, latest delivery time, distance from the delivery point to the terminal area of the logistics drone take-off and landing field, and cargo value;
[0089] Step B, according to the flight plan information and cargo order information, using the FCFS (First Come First Served) algorithm to determine the initial departure order and initial departure time of the terminal area of the logistics drone landing and take-off field within the period; and performing fuzzy clustering analysis on all cargo orders according to the cargo order information, with one cluster as a delivery task category, and obtaining multiple delivery task categories; specifically including the following sub-steps:
[0090] Step B1: Use the entropy weight method to determine the indicator weights of the distance from the receiving point to the terminal area of the logistics drone take-off and landing field, the latest receiving time and the value of the goods.
[0091] The three indicators of cargo value, distance from the receiving point to the terminal area of the logistics drone landing field, and the latest receiving time are selected as key indicators. Indicates the indicator value of order i corresponding to the kth key indicator.
[0092] The language description of the qualitative indicator goods value is converted into triangular fuzzy number form. If there are n orders in total, the expression of the language description of the goods value of order i after being converted into triangular fuzzy number is:
[0093]
[0094] In the formula, is the triangular fuzzy number of the value index of goods in order i, α l i , α m i , α u i are the lower limit, maximum possibility value and upper limit of the triangular fuzzy number of the value index of goods for order i, and n is the total number of orders.
[0095] The evaluation rules of triangular fuzzy numbers are shown in Table 1:
[0096] Table 1 Triangular fuzzy numbers for goods value evaluation
[0097]
[0098] The distance from the delivery point of order i to the terminal area of the logistics drone landing field and the latest delivery time can be transformed by the following formula to standardize the data:
[0099]
[0100] For the three indicators of cargo value, distance from the delivery point to the take-off and landing field, and the latest delivery time, the proportion of the index value of order i to the kth indicator P ik It can be expressed by the following formula:
[0101]
[0102] In the formula, when k = 1,
[0103] The entropy value of the kth indicator can be calculated by the following formula:
[0104]
[0105] If P ik =0, then define:
[0106]
[0107] The weight of the kth indicator can be calculated using the following formula:
[0108]
[0109] Step B2: Based on the weight of the goods value, calculate the first weighted similarity between order i and order j regarding the goods value.
[0110] The first weighted similarity between order i and order j regarding the value of goods can be calculated using the following formula:
[0111]
[0112]
[0113] Step B3: The calculation formula for calculating the second weighted similarity between order i and order j regarding the distance from the delivery point to the terminal area of the logistics drone landing and take-off field and the latest delivery time is the same, and both can be calculated using the following formula:
[0114]
[0115] Step B4: Calculate the multi-index comprehensive similarity between goods order i and goods order j. The calculation formula of the multi-index comprehensive similarity is as follows:
[0116]
[0117] Step B5: After calculating the multi-index comprehensive similarity between the goods orders, generate a similarity matrix R = (S ij ) n×n , and then use the transitive closure method to find the clustering results of all orders, so that one cluster is used as a delivery task category, and multiple delivery task categories are obtained.
[0118] Step C, based on the value of the goods in each delivery task category, the distance from the delivery point to the terminal area of the logistics drone landing and take-off field, and the latest delivery time, determine the priority of each delivery task category, and set different delay penalty coefficients for each priority; specifically, it includes the following sub-steps:
[0119] Step C1: Select the average cargo value X for each delivery task category g , the average distance Y from the receiving point to the terminal area of the logistics drone landing field g and the average latest delivery time Z g As characteristic parameters, the following formula can be used to determine the values of the three characteristic parameters of the delivery task category g:
[0120]
[0121]
[0122]
[0123] in, represents the average distance from the delivery point to the take-off and landing field for all orders in the delivery task category g, represents the average latest delivery time of all orders in the delivery task category g, α m i is the maximum possible value of a triangular fuzzy number, X g , Y g , Z g They represent the values of the three characteristic parameters of the task category g: the value of the goods, the distance from the receiving point to the take-off and landing field, and the latest receiving time. They represent the average value of the goods value, the distance from the delivery point to the take-off and landing field, and the latest delivery time of all orders in task category g;
[0124] Step C2: Calculate the priorities of different delivery task categories; specifically, the following steps are included:
[0125] Calculate the delivery task category g priority prior g The formula is:
[0126] prior g =(p g -1)(p g -2)(p g -3) / 6+(2p g -X g -2)(X g -1) / 2+Y g
[0127] p g =X g+Y g +Z g
[0128] The dynamic priority of the delivery task is reflected by the delay penalty coefficient. The calculation formula of the delay penalty coefficient is:
[0129]
[0130] Step D: Based on the priority of each delivery task category, a departure sorting model is constructed; the departure sorting model aims to minimize the total delay cost, and the constraints include the drone departure safety time interval constraint and the maximum allowable delay time constraint; specifically:
[0131] The objective function of the departure sorting model is:
[0132]
[0133]
[0134] Among them, T i is the actual departure time of the drone of order i, L i is the latest departure time of the drone for order i, LT i is the latest arrival time of the drone for order i, d i is the distance from the order receiving point to the terminal area of the logistics drone take-off and landing field, v is the flight speed of the drone, c ai is the delay penalty coefficient of order i, and n is the number of orders.
[0135] The constraints of the departure sorting model include:
[0136] (1) Maximum allowable delay time constraint.
[0137] The delay time of the logistics drone arriving at the delivery point needs to be within the maximum allowable delay time range. The maximum allowable delay time of the customer point of order i is lt i , then the constraint formula is:
[0138] T i -L i ≤lt i
[0139] In the formula, lt i is the maximum allowable delay time of the delivery point of order i, T i The actual departure time of the drone of order i;
[0140] (2) Minimum takeoff safety time interval constraint.
[0141] When logistics drones take off vertically, the aircraft that take off in sequence need to meet the safety interval requirements. The safety distance interval is converted into a time interval. If the actual sorting position of order i is k, and the actual sorting position of order j is k-1, the take-off safety time interval between order i and order j is t. ij , the drones that take off one after another need to meet the following constraints:
[0142] T i -T j ≥t ij
[0143] Where, T j is the actual departure time of the drone of order j, t ij is the safe time interval between the takeoffs of the two drones of order i and order j.
[0144] Step E: According to the objective function and constraints, a genetic algorithm is used to iteratively solve the dynamic departure sorting model of the logistics drone terminal area, and the initial departure sorting scheme is encoded as one of the chromosomes to participate in the iterative operation until the maximum number of iterations is reached to obtain the optimal departure sorting scheme, that is, the optimal departure order and optimal departure time of the drone.
[0145] Example 2
[0146] In an embodiment of the present invention, a dynamic departure sorting device for a terminal area of a logistics drone take-off and landing field is provided.
[0147] The acquisition module is used to obtain the UAV flight plan information and cargo order information in the terminal area of the logistics UAV take-off and landing field within a certain period of time;
[0148] An initial departure sequencing scheme determining module, used to determine an initial departure sequencing scheme according to the flight plan information and cargo order information;
[0149] A clustering module, used to perform fuzzy clustering analysis on all cargo orders according to the cargo order information, wherein one cluster is regarded as a delivery task category, and a plurality of delivery task categories are obtained;
[0150] The priority and delay penalty coefficient determination module is used to determine the priority of each delivery task category based on the value of the goods in each delivery task category, the distance from the delivery point to the terminal area of the logistics drone landing and take-off field, and the latest delivery time, and set different delay penalty coefficients for each priority;
[0151] A departure sorting model module is constructed to construct a departure sorting model based on the priority of each delivery task category; the departure sorting model aims to minimize the total delay cost, and the constraints include the safety time interval constraint of the UAV departure and the maximum allowable delay time constraint;
[0152] The optimal departure sorting scheme acquisition module is used to encode the initial departure sorting scheme as a chromosome, cooperate with the genetic algorithm to iteratively solve the departure sorting model, and obtain the optimal departure sorting scheme.
[0153] The rest are the same as in Example 1.
[0154] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.
Claims
1. A dynamic departure sorting method for the terminal area of a logistics drone take-off and landing field, characterized in that: include: Obtain drone flight plan information and cargo order information at the terminal area of the logistics drone take-off and landing field within a certain period of time; Determine an initial departure sequencing plan based on the flight plan information and cargo order information, wherein the initial departure sequencing plan includes the initial departure order of the logistics drones and the initial departure time of each drone; Performing fuzzy clustering analysis on all cargo orders according to the cargo order information, taking one cluster as a delivery task category, and obtaining multiple delivery task categories; Based on the cargo value of each delivery task category, the distance from the delivery point to the terminal area of the logistics drone take-off and landing field, and the latest delivery time, the priority of each delivery task category is determined, and different delay penalty coefficients are set for each priority; based on the priority of each delivery task category, a departure sorting model is constructed; the departure sorting model aims to minimize the total delay cost, and the constraints include the drone departure safety time interval constraint and the maximum allowable delay time constraint; the initial departure sorting scheme is encoded as a chromosome, and the departure sorting model is iteratively solved with a genetic algorithm to obtain the optimal departure sorting scheme, which includes the optimal departure order of logistics drones and the optimal departure time of each drone; The priority of each delivery task category is determined based on the value of the goods of each delivery task category, the distance from the delivery point to the terminal area of the logistics drone landing and take-off field, and the latest delivery time, and different delay penalty coefficients are set for each priority, including the following steps: Select the average cargo value X for each delivery task category i , the average distance Y from the receiving point to the terminal area of the logistics drone landing field i and the average latest delivery time Z i As characteristic parameters; Calculate the values of the three characteristic parameters of each delivery task category; Based on the values of the three characteristic parameters of each delivery task category, the priority of each delivery task category is calculated; Based on the priority of each delivery task category, calculate the corresponding delay penalty coefficient; The calculation formula for the values of the three characteristic parameters of each delivery task category is: in, represents the average distance from all order receiving points in the delivery task category g to the terminal area of the logistics drone landing and take-off field, represents the average latest delivery time of all orders in the delivery task category g, α m i is the maximum possible value of a triangular fuzzy number, X g , Y g , Z g They represent the values of the three characteristic parameters of the delivery task category g, namely, the value of the goods, the distance from the receiving point to the terminal area of the logistics drone landing and take-off field, and the latest receiving time. They represent the average value of all order goods in task category g, the distance from the delivery point to the terminal area of the logistics drone landing and take-off field, and the latest delivery time; The calculation formula for the priority of each delivery task category is: prior g =(p g -1)(p g -2)(p g -3) / 6+(2p g -X g -2)(X g -1) / 2+Y g p g =X g +Y g +Z g The calculation formula for the delay penalty coefficient of each delivery task category is: In the formula, c ag is the delay penalty coefficient of category g, prior g is the task priority of category g; The objective function of the departure sorting model is: Where, T i is the actual departure time of the drone of order i, L i is the latest departure time of the drone for order i, LT i is the latest arrival time of the drone for order i, d i is the distance from the order i receiving point to the terminal area of the logistics drone landing field, v is the flight speed of the drone, c ai is the delay penalty coefficient of order i, and n is the number of orders; The constraints of the departure sorting model are: T i -L i ≤lt i T i -T j ≥t ij In the formula, lt i is the maximum allowable delay time of the delivery point of order i, T i is the actual departure time of the drone of order i, T j is the actual departure time of the drone of order j, t ij is the safe time interval between the takeoffs of the two drones of order i and order j.
2. A method for dynamic departure sorting in the terminal area of a logistics drone take-off and landing field according to claim 1, characterized in that: The method of performing fuzzy clustering analysis on all cargo orders according to the cargo order information, wherein one cluster is used as a delivery task category, and obtaining multiple delivery task categories includes the following steps: The distance from the delivery point to the terminal area of the logistics drone landing and take-off field, the latest delivery time and the value of the goods in the cargo order information are used as indicators; The entropy weight method is used to determine the indicator weights of the distance from the receiving point to the terminal area of the logistics drone take-off and landing field, the latest receiving time and the value of the goods; Based on the indicator weight of the goods value, the first weighted similarity between order i and order j regarding the goods value is calculated; Based on the index weight of the distance from the delivery point to the terminal area of the logistics drone landing field, the second weighted similarity of order i and order j regarding the distance from the delivery point to the terminal area of the logistics drone landing field is calculated; Based on the indicator weight of the latest delivery time, the third weighted similarity between order i and order j regarding the latest delivery time is calculated; Based on the first weighted similarity, the second weighted similarity and the third weighted similarity, a multi-index comprehensive similarity between order i and order j is calculated; Generate a similarity matrix based on the multi-index comprehensive similarity; Based on the similarity matrix, the clustering results of all cargo orders are obtained by using the transitive closure method, so that one cluster is used as a delivery task category, and multiple delivery task categories are obtained.
3. A method for dynamic departure sorting in the terminal area of a logistics drone take-off and landing field according to claim 2, characterized in that: The index weight of the cargo value, the index weight of the distance from the receiving point to the terminal area of the logistics drone landing and take-off field, and the index weight of the latest receiving time are calculated by the following formula: In the formula, when k = 1, ω k is the indicator weight of the goods value, P ik is the index value ratio of order i to the value of goods, E k is the entropy value of the goods; when k = 2, ω k is the indicator weight of the distance from the receiving point to the terminal area of the logistics drone landing and take-off field, P ik It represents the weight of the index value of the distance from the receiving point to the terminal area of the logistics drone landing field for order i, E k is the entropy value of the distance from the receiving point to the terminal area of the logistics drone landing field; when k = 3, ω k is the indicator weight of the latest delivery time, P ik represents the weight of the index value of order i to the latest delivery time, E k is the entropy value of the latest delivery time, and n is the total number of orders.
4. A method for dynamic departure sorting in the terminal area of a logistics drone take-off and landing field according to claim 3, characterized in that: The calculation formula of the first weighted similarity is: In the formula, r ij1 is the first weighted similarity, i is order i, j is order j, ω1 is the indicator weight of the goods value, is the triangular fuzzy number of the value index of goods in order i, α l i , α m i , α u i and α l j , α m j , α u j are the lower limit, maximum possibility value and upper limit of the triangular fuzzy number of the goods value index of order i and order j respectively, and n is the total number of orders.
5. A method for dynamic departure sorting in the terminal area of a logistics drone take-off and landing field according to claim 4, characterized in that: The calculation formulas of the second weighted similarity and the third weighted similarity are the same, both of which are: In the formula, when k = 2, r ijk is the second weighted similarity, ω k is the indicator weight of the distance from the receiving point to the terminal area of the logistics drone landing field; when k = 3, r ijk is the third weighted similarity, ω k is the indicator weight of the latest delivery time, is the normalized index value of order i corresponding to the kth index, and n is the total number of orders; when k = 2, is the actual index value of the distance from the receiving point to the take-off and landing field corresponding to order i; when k = 3, is the actual indicator value corresponding to the latest delivery time of order i.
6. A method for dynamic departure sorting in the terminal area of a logistics drone take-off and landing field according to claim 5, characterized in that: The calculation formula of the multi-index comprehensive similarity of order i and order j is: The expression of the similarity matrix is: R=(S ij ) n×n 。 7. A dynamic departure sorting device for the terminal area of a logistics drone take-off and landing field, characterized in that: include: The acquisition module is used to obtain the UAV flight plan information and cargo order information in the terminal area of the logistics UAV take-off and landing field within a certain period of time; An initial departure sequencing scheme determination module is used to determine an initial departure sequencing scheme according to the flight plan information and the cargo order information, wherein the initial departure sequencing scheme includes the initial departure sequence of the logistics drones and the initial departure time of each drone; A clustering module, used to perform fuzzy clustering analysis on all cargo orders according to the cargo order information, wherein one cluster is regarded as a delivery task category, and a plurality of delivery task categories are obtained; The priority and delay penalty coefficient determination module is used to determine the priority of each delivery task category based on the value of the goods in each delivery task category, the distance from the delivery point to the terminal area of the logistics drone landing and take-off field, and the latest delivery time, and set different delay penalty coefficients for each priority; A departure sorting model module is constructed to construct a departure sorting model based on the priority of each delivery task category; the departure sorting model aims to minimize the total delay cost, and the constraints include the safety time interval constraint of the UAV departure and the maximum allowable delay time constraint; The optimal departure sorting scheme acquisition module is used to encode the initial departure sorting scheme as a chromosome, and iteratively solve the departure sorting model with a genetic algorithm to obtain the optimal departure sorting scheme, which includes the optimal departure order of the logistics drones and the optimal departure time of each drone; The priority of each delivery task category is determined based on the value of the goods of each delivery task category, the distance from the delivery point to the terminal area of the logistics drone landing and take-off field, and the latest delivery time, and different delay penalty coefficients are set for each priority, including the following steps: Select the average cargo value X for each delivery task category i , the average distance Y from the receiving point to the terminal area of the logistics drone landing field i and the average latest delivery time Z i As characteristic parameters; Calculate the values of the three characteristic parameters of each delivery task category; Based on the values of the three characteristic parameters of each delivery task category, the priority of each delivery task category is calculated; Based on the priority of each delivery task category, calculate the corresponding delay penalty coefficient; The calculation formula for the values of the three characteristic parameters of each delivery task category is: in, represents the average distance from all order receiving points in the delivery task category g to the terminal area of the logistics drone landing and take-off field, represents the average latest delivery time of all orders in the delivery task category g, α m i is the maximum possible value of a triangular fuzzy number, X g , Y g , Z g They represent the values of the three characteristic parameters of the delivery task category g, namely, the value of the goods, the distance from the receiving point to the terminal area of the logistics drone landing and take-off field, and the latest receiving time. They represent the average value of all order goods in task category g, the distance from the delivery point to the terminal area of the logistics drone landing and take-off field, and the latest delivery time; The calculation formula for the priority of each delivery task category is: prior g =(p g -1)(p g -2)(p g -3) / 6+(2p g -X g -2)(X g -1) / 2+Y g p g =X g +Y g +Z g The calculation formula for the delay penalty coefficient of each delivery task category is: In the formula, c ag is the delay penalty coefficient of category g, prior g is the task priority of category g; The objective function of the departure sorting model is: Where, T i is the actual departure time of the drone of order i, L i is the latest departure time of the drone for order i, LT i is the latest arrival time of the drone for order i, d i is the distance from the order i receiving point to the terminal area of the logistics drone landing field, v is the flight speed of the drone, c ai is the delay penalty coefficient of order i, and n is the number of orders; The constraints of the departure sorting model are: T i -L i ≤lt i T i -T j ≥t ij In the formula, lt i is the maximum allowable delay time of the delivery point of order i, T i is the actual departure time of the drone of order i, T j is the actual departure time of the drone of order j, t ij is the safe time interval between the takeoffs of the two drones of order i and order j.
