A smart logistics delivery method and system based on electronic map

Abstract

The present invention relates to the technical field of logistics delivery, and specifically to an intelligent logistics delivery method and system based on electronic maps, comprising: obtaining a set of customers to be delivered, obtaining a set of locations to be delivered of the customer set to be delivered based on an electronic map of a delivery area, clustering the set of locations to be delivered to obtain a set of points to be delivered of the set of locations to be delivered; obtaining a logistics distribution center of the delivery area, and constructing a delivery connectivity network of the delivery area based on the electronic map of the delivery area, the set of points to be delivered, and the logistics distribution center. The present invention can better integrate delivery resources, reduce unnecessary repeated routes in the delivery process, and improve the overall efficiency of logistics delivery by clustering the locations to be delivered, and can effectively calculate the optimal delivery path by using a weighted delivery connectivity network and a solution algorithm, thereby reducing transportation costs and shortening delivery time.

Description

Technical Field

[0001] The present invention relates to the technical field of logistics delivery, and in particular to an intelligent logistics delivery method and system based on electronic maps. Background Art

[0002] An electronic map is a tool that displays geographic information in digital form. It is usually provided in the form of a computer program or mobile application and can display various information such as geographic locations, roads, buildings, natural landscapes, etc.

[0003] Traditional methods often rely on manual planning and experience, which may lead to less than optimized routes, increased transportation time and costs; traditional methods lack real-time data and information sharing, resulting in poor coordination in the delivery process and difficulty in quickly responding to changing demands or emergencies; traditional methods are difficult to capture and meet customers' personalized needs in a timely manner, which may lead to decreased customer satisfaction, and due to the lack of accurate demand forecasting and route planning, vehicles may run empty or repeat transportation, increasing resource waste; traditional methods are usually fixed and lack the ability to flexibly adjust in the face of market changes or order fluctuations, and manual scheduling and management processes are complex and prone to errors, increasing management costs and the burden on human resources. Summary of the invention

[0004] The technical problem to be solved by the present invention is to overcome the shortcomings of the above-mentioned prior art and provide an intelligent logistics delivery method and system based on electronic maps.

[0005] The technical solution adopted to solve the above technical problems is: a smart logistics delivery method based on electronic maps, including:

[0006] Acquire a set of customers to be delivered, acquire a set of locations to be delivered of the set of customers to be delivered based on an electronic map of the delivery area, and perform location clustering on the set of locations to be delivered to obtain a set of points to be delivered of the set of locations to be delivered;

[0007] Acquire the logistics distribution center of the delivery area, and construct a delivery connection network of the delivery area based on the electronic map of the delivery area, the set of points to be delivered and the logistics distribution center;

[0008] Obtaining the needs of the customers to be delivered and the transportation efficiency of the delivery area, and weighting the delivery connection network based on the needs of the customers to be delivered and the transportation efficiency of the delivery area to obtain a weighted delivery connection network;

[0009] Building a first logistics delivery model based on the weighted delivery connection network, solving the first logistics delivery model based on reinforcement learning to obtain an optimal first logistics delivery path, and delivering the logistics to the to-be-delivered point based on the optimal first logistics delivery path;

[0010] A second logistics delivery model is constructed based on the set of delivery points and the set of locations to be delivered, and the second logistics delivery model is solved based on the finder optimization algorithm to obtain an optimal second logistics delivery path, and the logistics is delivered to the locations to be delivered based on the optimal second logistics delivery path.

[0011] Preferably, performing location clustering on the set of locations to be delivered to obtain a set of points to be delivered of the set of locations to be delivered includes:

[0012] Acquire the location text of the location to be dispatched, and perform a word segmentation operation on the location text to obtain a location word sequence corresponding to the location text;

[0013] Performing stop word filtering on the position word sequence to obtain a standard position word sequence, and performing word frequency statistics on the standard position word sequence to obtain a word frequency bigram of the position word sequence, wherein the word frequency bigram includes the position word and the word frequency of the position word;

[0014] Performing feature extraction on the word frequency bigrams of the position word sequence to obtain a word feature sequence of the position word sequence;

[0015] Repeat the above operation until the set of locations to be dispatched is traversed to obtain a set of word feature sequences of the set of locations to be dispatched;

[0016] The word feature sequence set is clustered to obtain clusters corresponding to the word feature sequence set, and the set of locations to be dispatched is classified into different points to be dispatched based on the clusters corresponding to the word feature sequence set.

[0017] Preferably, the word feature sequence includes a first feature extraction formula, a second feature extraction formula, a third feature extraction formula and a fourth feature extraction formula, and the first feature extraction formula is as follows:

[0018] ;

[0019] in, Words indicating location In the location word library The frequency of occurrence of the position in the text, Words indicating location In the location word library The number of occurrences in the position text, Indicates the location of the word library The total number of words in the text at each position;

[0020] The second feature extraction formula is as follows:

[0021] ;

[0022] in, Indicates the inverse frequency of position text, Indicates that a location word appears in the location word database The number of times, Indicates the size of the location word library;

[0023] The third feature extraction formula is as follows:

[0024] ;

[0025] in, Words indicating location In the location word library The first occurrence of a position in the text, Words indicating location Previously in The total number of other words that appear in the text at each position;

[0026] The fourth feature extraction formula is as follows:

[0027] ;

[0028] in, Words indicating location The average position in the position text, Words indicating location At each occurrence of the position in the position text, Words indicating location Number of occurrences in the location text.

[0029] Preferably, clustering the word feature sequence set to obtain a cluster corresponding to the word feature sequence set includes:

[0030] Calculating the local density of the word feature sequence based on information entropy, and constructing the neighborhood space of each word feature sequence based on K neighborhood and inverse K neighborhood;

[0031] Calculating the local density of the word feature sequence in the neighborhood space, and determining the affiliation of the word feature sequence based on the local density of the word feature sequence to obtain a density center;

[0032] Based on changing the neighborhood space to obtain multi-level density centers, a density graph structure is constructed based on the membership relationship of the multi-level density centers in the neighborhood space, and the graph connected components of the density graph structure are initial clusters;

[0033] The unclustered word feature sequences are divided into initial clusters based on the membership relationship of the neighborhood space, and the above operation is repeated until all clusters corresponding to the word feature sequence set are obtained.

[0034] Preferably, constructing a delivery connection network of the delivery area based on the electronic map of the delivery area, the set of points to be delivered and the logistics distribution center includes:

[0035] Extracting a road network from the electronic map, wherein the road network includes a plurality of nodes and a plurality of connecting edges, wherein the nodes correspond to the points to be dispatched and the logistics distribution centers, and the connecting edges correspond to connection relationships between roads;

[0036] The road network is traversed based on a depth-first search algorithm to check connectivity between the points to be dispatched and the distribution centers to obtain a reachable point set, and a dispatch connectivity network is constructed based on the reachable point set.

[0037] Preferably, building a first logistics delivery model based on the weighted delivery connectivity network includes:

[0038] A first objective function is constructed based on the weighted delivery connectivity network, wherein the first objective function is used to describe the minimization of the total driving distance of the delivery vehicle, and the first objective function is as follows:

[0039] ;

[0040] in, represents the first objective function, Indicates the point where the shipment is to be dispatched To the delivery point The distance Indicates delivery vehicle Whether it has passed the delivery point and delivery point , and when the delivery vehicle Passing by the delivery point and delivery point , and wait for the delivery point and delivery point When there is a road ,otherwise, , represents the set of points to be dispatched in the weighted dispatch network. Indicates the delivery vehicle collection. Represents the set of points to be dispatched in each class;

[0041] A second objective function is constructed based on the weighted delivery connection network, wherein the second objective function is used to describe the maximization of the delivery vehicle loading rate, and the second objective function is as follows:

[0042] ;

[0043] in, represents the second objective function, Indicates the needs of each customer to be delivered. Indicates delivery vehicle Whether it passes through the delivery point , and when the delivery vehicle Passing by delivery point hour ,otherwise, , Indicates delivery vehicle The capacity limit of

[0044] A third objective function is constructed based on the weighted delivery connection network, wherein the third objective function is used to describe the minimization of delivery time cost, and the third objective function is as follows:

[0045] ;

[0046] in, represents the third objective function, Indicates the point where the shipment is to be dispatched To the delivery point The driving time, Indicates the average time required to process each logistics. Indicates the point where the shipment is to be dispatched The number of express deliveries, Indicates the average consumption cost of each delivery vehicle;

[0047] A first overall constraint condition is constructed based on the weighted delivery connectivity network, and a first logistics delivery model is constructed based on the first objective function, the second objective function, the third objective function and the first overall constraint condition. The first overall constraint condition is as follows:

[0048] .

[0049] Preferably, building a second logistics delivery model based on the delivery point set and the to-be-delivered location set includes:

[0050] A fourth objective function is constructed based on the delivery point and the location to be delivered, wherein the fourth objective function is used to describe the minimization of the terminal delivery cost of the delivery vehicle, and the fourth objective function is as follows:

[0051] ;

[0052] in, represents the fourth objective function, Indicates whether the delivery point is connected to the location to be delivered. If so, , otherwise, , Indicates delivery vehicle fixed costs, Indicates the location of the shipment to be delivered The penalty cost, Indicates delivery vehicle Arrived at the location to be dispatched time, Indicates delivery vehicle Leaving the delivery location time, Indicates the distance from the delivery point to the location to be delivered. represents the unit transportation distance cost, Indicates the price of unit logistics, It indicates the cargo damage rate per unit time during transportation. It represents the cargo damage rate per unit time in the unloading process, Indicates the location of the shipment to be delivered The demand, represents the unloading rate, Indicates delivery vehicle Arrived at the location to be dispatched The remaining amount of logistics, represents the service cost per time;

[0053] A second overall constraint condition is constructed based on the delivery point and the location to be delivered, and a second logistics delivery model is constructed based on the fourth objective function and the second overall constraint condition. The second overall constraint condition is as follows:

[0054] .

[0055] Preferably, solving the second logistics delivery model based on a finder optimization algorithm to obtain an optimal second logistics delivery path includes:

[0056] Set the discoverer ratio , the proportion of investigators and warning value , and initialize the population, the population initialization formula is as follows:

[0057] ;

[0058] in, and Indicates the population and Individuals, represents the remainder function;

[0059] Calculate the fitness value of each individual in the population, sort them according to the fitness value, select the discoverer from the individuals whose fitness value is greater than the preset fitness value threshold, and update the position of the discoverer. The update formula of the discoverer's position is as follows:

[0060] ;

[0061] in, and Indicates that the coordinates in the population are The discoverer of Second and The position in the d-dimensional solution space in the iteration, Represents a random number that determines the distance the finder moves. Represents a random number that determines the direction in which the finder moves. Indicates The current finder's optimal position for iterations, and represents the division coefficient, Indicates the safety threshold;

[0062] After the discoverer is selected from the population, the remaining individuals in the population are taken as joiners, and the positions of the joiners are updated. The joiner position update formula is as follows:

[0063] ;

[0064] in, and Indicates that the coordinates in the population are The first member Second and The position in the d-dimensional solution space in the iteration, Indicates The global worst position of the current joiner in iterations, Indicates The optimal position of the current joiner in iterations, Represents a matrix composed of random numbers 1 or -1. represents the identity matrix, Represents the total number of individuals in the population.

[0065] Preferably, solving the second logistics delivery model based on a finder optimization algorithm to obtain an optimal second logistics delivery path also includes:

[0066] A scout is randomly selected from the population, and the position of the scout is updated. The scout position update formula is as follows:

[0067] ;

[0068] in, and Indicates that the coordinates in the population are The Investigator Second and The position in the d-dimensional solution space in the iteration, Indicates The global worst position of the current scout at iteration, Indicates The optimal position of the current scout at iteration, represents a random number that follows a Gaussian distribution with an expectation of 0 and a variance of 1. Take 10E-8, The random coefficients representing the direction and step size, represents the fitness value of the scout, and Indicates the best and worst fitness values ​​of the scout;

[0069] The current optimal solution is disturbed, the population fitness is recalculated, and the optimal solution after disturbance is found. The current optimal solution disturbance formula is as follows:

[0070] ;

[0071] in, Indicates that the coordinates in the population are Individual The position in the d-dimensional solution space in the iteration, represents the current optimal solution, represents the standard Cauchy distribution function;

[0072] Determine whether the termination condition is met. If the termination condition is met, output the optimal individual position and its fitness value. Otherwise, repeat the above operation to obtain the optimal second logistics delivery path.

[0073] The technical solution adopted to solve the above technical problems is: an intelligent logistics delivery system based on electronic maps, which is applicable to the intelligent logistics delivery method based on electronic maps, including:

[0074] A location clustering unit, the location clustering unit is used to obtain a set of customers to be delivered, obtain a set of locations to be delivered of the set of customers to be delivered based on an electronic map of a delivery area, and perform location clustering on the set of locations to be delivered to obtain a set of points to be delivered of the set of locations to be delivered;

[0075] A network construction unit, the network construction unit is used to obtain the logistics distribution center of the delivery area, and build a delivery connection network of the delivery area based on the electronic map of the delivery area, the set of points to be delivered and the logistics distribution center;

[0076] A network weighting unit, the network weighting unit is used to obtain the needs of the customers to be delivered and the transportation efficiency of the delivery area, and weight the delivery connection network based on the needs of the customers to be delivered and the transportation efficiency of the delivery area to obtain a weighted delivery connection network;

[0077] A first dispatch unit, the first dispatch unit is used to build a first logistics dispatch model based on the weighted dispatch connection network, solve the first logistics dispatch model based on reinforcement learning to obtain an optimal first logistics dispatch path, and deliver the logistics to the to-be-delivered point based on the optimal first logistics dispatch path;

[0078] The second dispatch unit is used to construct a second logistics dispatch model based on the dispatch point set and the set of locations to be dispatched, solve the second logistics dispatch model based on the finder optimization algorithm to obtain an optimal second logistics dispatch path, and deliver the logistics to the locations to be dispatched based on the optimal second logistics dispatch path.

[0079] The beneficial effects of the present invention are as follows: (1) By clustering the locations to be delivered, the present invention can better integrate delivery resources, reduce unnecessary repeated routes in the delivery process, and improve the overall efficiency of logistics distribution. By utilizing a weighted delivery connection network and advanced solution algorithms, the optimal delivery path can be effectively calculated, thereby reducing transportation costs and shortening delivery time. (2) By acquiring the needs of customers to be delivered, the present invention can achieve personalized delivery and improve customer satisfaction while ensuring the rational allocation of delivery resources. By considering transportation benefits and performing network weighting, the delivery decision is more in line with the actual transportation situation, which can reduce costs while improving transportation efficiency. (3) By optimizing routes and reducing transportation time, the present invention helps to reduce vehicle carbon emissions and energy consumption, thereby promoting the development of green logistics. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] Figure 1 A schematic flow chart of the steps of an overall method in an embodiment of the present invention;

[0081] Figure 2 A schematic diagram of the system architecture of an overall system in an embodiment of the present invention.

[0082] Figure numerals: 1. location clustering unit; 2. network construction unit; 3. network weighting unit; 4. first dispatching unit; 5. second dispatching unit. DETAILED DESCRIPTION

[0083] Embodiment 1, as Figure 1 As shown, the present invention proposes a smart logistics delivery method based on electronic maps, including:

[0084] S1. Obtain a set of customers to be delivered, obtain a set of locations to be delivered of the set of customers to be delivered based on an electronic map of the delivery area, and perform location clustering on the set of locations to be delivered to obtain a set of points to be delivered of the set of locations to be delivered;

[0085] S2. Obtain the logistics distribution center of the delivery area, and build a delivery connection network for the delivery area based on the electronic map of the delivery area, the set of points to be delivered, and the logistics distribution center;

[0086] S3, obtaining the needs of customers to be delivered and the transportation efficiency of the delivery area, and weighting the delivery connection network based on the needs of customers to be delivered and the transportation efficiency of the delivery area to obtain a weighted delivery connection network;

[0087] S4. constructing a first logistics delivery model based on the weighted delivery connection network, solving the first logistics delivery model based on reinforcement learning to obtain an optimal first logistics delivery path, and delivering the logistics to the to-be-delivered point based on the optimal first logistics delivery path;

[0088] S5. Construct a second logistics delivery model based on the delivery point set and the set of locations to be delivered, solve the second logistics delivery model based on the finder optimization algorithm to obtain the optimal second logistics delivery path, and deliver the logistics to the locations to be delivered based on the optimal second logistics delivery path.

[0089] In the present invention, the customer set to be delivered refers to a list of customers who need delivery, including information on all customers who need delivery; the electronic map of the delivery area refers to geographic information represented by digital technology, which can be used to display various information of the delivery area, such as roads, terrain, buildings, etc.; the location set to be delivered refers to the specific location coordinates of all items to be delivered extracted from the customer set to be delivered, which is used for subsequent logistics scheduling; location clustering refers to a data analysis technology that uses an algorithm to group the locations in the location set to be delivered according to their spatial characteristics, so that similar points are clustered together to facilitate subsequent delivery optimization; the point set to be delivered refers to a smaller, representative location set obtained after clustering, which is used to reduce the complexity of the delivery path; the logistics distribution center refers to the central node responsible for managing the distribution process, usually a cargo A centralized storage and distribution location; the delivery connectivity network is a network diagram constructed based on the electronic map of the delivery area, the points to be delivered and the logistics distribution center, which represents the connection relationship and accessibility between different nodes (such as the distribution center and the points to be delivered); the first logistics delivery model is a model constructed based on the weighted delivery connectivity network, which is used to describe and optimize the delivery process from the distribution center to the points to be delivered; reinforcement learning refers to a machine learning method that learns the optimal strategy through interaction with the environment. Reinforcement learning is used to solve the first logistics delivery model and find the optimal delivery path; the second logistics delivery model is a new model constructed based on the first model based on the set of delivery points and the set of locations to be delivered, which is used to further optimize the logistics delivery path; the finder optimization algorithm refers to a swarm intelligence algorithm that optimizes the problem by simulating the behavior of the finder.

[0090] Embodiment 2, a smart logistics delivery method based on an electronic map proposed by the present invention, compared with embodiment 1, this embodiment further includes: performing location clustering on the set of locations to be delivered to obtain a set of points to be delivered of the set of locations to be delivered, including:

[0091] A1. Obtain the location text of the location to be delivered, and perform word segmentation on the location text to obtain a location word sequence corresponding to the location text;

[0092] A2. Perform stop word filtering on the position word sequence to obtain a standard position word sequence, and perform word frequency statistics on the standard position word sequence to obtain a word frequency bigram of the position word sequence, wherein the word frequency bigram includes the position word and the word frequency of the position word;

[0093] A3, extracting features from the word frequency bigrams of the position word sequence to obtain a word feature sequence of the position word sequence;

[0094] A4. Repeat the above operation until the set of locations to be dispatched is traversed to obtain a set of word feature sequences of the set of locations to be dispatched;

[0095] A5. Clustering the word feature sequence set to obtain clusters corresponding to the word feature sequence set, and classifying the to-be-delivered location set into different to-be-delivered points based on the clusters corresponding to the word feature sequence set.

[0096] In this embodiment, location text refers to text information describing the location of the package to be delivered, usually including street, city, region and other information; word segmentation operation refers to a technology in natural language processing, which divides continuous text into independent words for subsequent processing; location word sequence refers to a sequence composed of location-related words obtained after word segmentation, which can represent the specific information of the location of the package to be delivered; stop word filtering refers to removing common words that are not very meaningful in text analysis to retain words with more information value.

[0097] In an optional embodiment, the word feature sequence includes a first feature extraction formula, a second feature extraction formula, a third feature extraction formula and a fourth feature extraction formula, and the first feature extraction formula is as follows:

[0098] ;

[0099] in, Words indicating location In the location word library The frequency of occurrence of the position in the text, Words indicating location In the location word library The number of occurrences in the position text, Indicates the location of the word library The total number of words in the text at each position;

[0100] The second feature extraction formula is as follows:

[0101] ;

[0102] in, Indicates the inverse frequency of position text, Indicates that a location word appears in the location word database The number of times, Indicates the size of the location word library;

[0103] The third feature extraction formula is as follows:

[0104] ;

[0105] in, Words indicating location In the location word library The first occurrence of a position in the text, Words indicating location Previously in The total number of other words that appear in the text at each position;

[0106] The fourth feature extraction formula is as follows:

[0107] ;

[0108] in, Words indicating location The average position in the position text, Words indicating location At each occurrence of the position in the position text, Words indicating location Number of occurrences in the location text.

[0109] In an optional embodiment, clustering the word feature sequence set to obtain a cluster corresponding to the word feature sequence set includes:

[0110] B1. Calculate the local density of the word feature sequence based on information entropy, and construct the neighborhood space of each word feature sequence based on K neighborhood and inverse K neighborhood;

[0111] B2. Calculate the local density of the word feature sequence in the neighborhood space, and determine the affiliation of the word feature sequence based on the local density of the word feature sequence to obtain the density center;

[0112] B3. Based on the change of the neighborhood space, a multi-level density center is obtained, and a density graph structure is constructed based on the affiliation of the multi-level density center in the neighborhood space. The graph connected component of the density graph structure is the initial clustering cluster;

[0113] B4. The unclustered word feature sequences are divided into initial clusters based on the membership relationship in the neighborhood space, and the above operation is repeated until all clusters corresponding to the word feature sequence set are obtained.

[0114] It should be noted that information entropy refers to a measure of quantitative uncertainty, usually used to evaluate the complexity of information. In text processing, it can be used to measure the importance of words; local density refers to the density of word feature sequences in a specific neighborhood, usually used to measure the importance of a feature in space; K neighborhood refers to the K other feature sequences closest to a given feature sequence, which constitute the local environment of the feature; inverse K neighborhood refers to how many feature sequences can be regarded as neighbors among the K neighbors of a feature sequence, reflecting the density of the feature sequence; neighborhood space refers to the set of feature sequences formed by K neighborhoods or inverse K neighborhoods, and this space is used for local analysis; affiliation refers to the relationship between a feature sequence and its cluster or density. The relationship between degree centers determines the affiliation of the feature sequence in the cluster; density center refers to a representative point calculated based on local density, which represents a high-density area in the feature space and is usually regarded as the center of the cluster; multi-level density center refers to extracting density centers of multiple different levels by changing the neighborhood space, so as to facilitate more detailed clustering analysis; density graph structure is a graph structure constructed based on the density center and its neighborhood relationship, which is usually used to visualize the clustering of feature sequences; graph connected component refers to a subset formed by all interconnected points in the density graph structure, which represents a group of closely related feature sequences and usually represents a preliminary clustering cluster.

[0115] In an optional embodiment, a delivery connection network of the delivery area is constructed based on an electronic map of the delivery area, a set of points to be delivered, and a logistics distribution center, including:

[0116] C1. Extracting a road network from an electronic map, wherein the road network includes a plurality of nodes and a plurality of connecting edges, wherein the nodes correspond to the delivery points and the logistics distribution centers, and the connecting edges correspond to the connection relationships of the roads;

[0117] C2. Traverse the road network based on the depth-first search algorithm, check the connectivity between the delivery points and the distribution centers to obtain a set of reachable points, and build a delivery connectivity network based on the set of reachable points.

[0118] It should be noted that the depth-first search (DFS) algorithm refers to a graph traversal algorithm that starts from the starting node and goes as deep as possible along the connecting edges until there are no unvisited nodes, and then backtracks to visit other nodes. DFS can be used to explore all nodes and edges of the graph; the reachable point set refers to the set of all nodes that can be reached by the traversal algorithm starting from a certain starting node, representing the reachable target points in the network.

[0119] In an optional embodiment, building a first logistics delivery model based on a weighted delivery connectivity network includes:

[0120] D1. Construct the first objective function based on the weighted delivery connection network, where the first objective function is used to describe the minimization of the total driving distance of the delivery vehicle. The first objective function is as follows:

[0121] ;

[0122] in, represents the first objective function, Indicates the point where the shipment is to be dispatched To the delivery point The distance Indicates delivery vehicle Whether it has passed the delivery point and delivery point , and when the delivery vehicle Passing by the delivery point and delivery point , and wait for the delivery point and delivery point When there is a road ,otherwise, , represents the set of points to be dispatched in the weighted dispatch network. Indicates the delivery vehicle collection. Represents the set of points to be dispatched in each class;

[0123] D2. Construct a second objective function based on the weighted delivery connection network, where the second objective function is used to describe the maximization of the delivery vehicle loading rate. The second objective function is as follows:

[0124] ;

[0125] in, represents the second objective function, Indicates the needs of each customer to be delivered. Indicates delivery vehicle Whether it passes through the delivery point , and when the delivery vehicle Passing by delivery point hour ,otherwise, , Indicates delivery vehicle The capacity limit of

[0126] D3. Construct a third objective function based on the weighted delivery connectivity network, where the third objective function is used to describe the minimization of delivery time cost. The third objective function is as follows:

[0127] ;

[0128] in, represents the third objective function, Indicates the point where the shipment is to be dispatched To the delivery point The driving time, Indicates the average time required to process each logistics. Indicates the point where the shipment is to be dispatched The number of express deliveries, Indicates the average consumption cost of each delivery vehicle;

[0129] D4. Construct the first overall constraint condition based on the weighted delivery connectivity network, and construct the first logistics delivery model based on the first objective function, the second objective function, the third objective function and the first overall constraint condition. The first overall constraint condition is as follows:

[0130] .

[0131] In an optional embodiment, building a second logistics delivery model based on the delivery point set and the to-be-delivered location set includes:

[0132] E1. Construct a fourth objective function based on the delivery point and the location to be delivered, wherein the fourth objective function is used to describe the minimization of the terminal delivery cost of the delivery vehicle. The fourth objective function is as follows:

[0133] ;

[0134] in, represents the fourth objective function, Indicates whether the delivery point is connected to the location to be delivered. If so, , otherwise, , Indicates delivery vehicle fixed costs, Indicates the location of the shipment to be delivered The penalty cost, Indicates delivery vehicle Arrived at the location to be dispatched time, Indicates delivery vehicle Leaving the delivery location time, Indicates the distance from the delivery point to the location to be delivered. represents the unit transportation distance cost, Indicates the price of unit logistics, It indicates the cargo damage rate per unit time during transportation. It represents the cargo damage rate per unit time in the unloading process, Indicates the location of the shipment to be delivered The demand, represents the unloading rate, Indicates delivery vehicle Arrived at the location to be dispatched The remaining amount of logistics, represents the service cost per time;

[0135] E2. Construct a second overall constraint condition based on the delivery point and the location to be delivered, and construct a second logistics delivery model based on the fourth objective function and the second overall constraint condition. The second overall constraint condition is as follows:

[0136] .

[0137] In an optional embodiment, solving the second logistics delivery model based on the finder optimization algorithm to obtain the optimal second logistics delivery path includes:

[0138] F1. Set the proportion of discoverers , the proportion of investigators and warning value , and initialize the population. The population initialization formula is as follows:

[0139] ;

[0140] in, and Indicates the population and Individuals, represents the remainder function;

[0141] F2. Calculate the fitness value of each individual in the population, sort them according to the fitness value, select the discoverer from the individuals whose fitness value is greater than the preset fitness value threshold, and update the position of the discoverer. The formula for updating the position of the discoverer is as follows:

[0142] ;

[0143] in, and Indicates that the coordinates in the population are The discoverer of Second and The position in the d-dimensional solution space in the iteration, Represents a random number that determines the distance the finder moves. Represents a random number that determines the direction in which the finder moves. Indicates The current finder's optimal position for iterations, and represents the division coefficient, Indicates the safety threshold;

[0144] F3. After selecting the discoverer from the population, the remaining individuals in the population are taken as joiners, and the positions of the joiners are updated. The formula for updating the joiner positions is as follows:

[0145] ;

[0146] in, and Indicates that the coordinates in the population are The first member Second and The position in the d-dimensional solution space in the iteration, Indicates The global worst position of the current joiner in iterations, Indicates The optimal position of the current joiner in iterations, Represents a matrix composed of random numbers 1 or -1. represents the identity matrix, Represents the total number of individuals in the population.

[0147] In an optional embodiment, solving the second logistics delivery model based on the finder optimization algorithm to obtain the optimal second logistics delivery path also includes:

[0148] F4. Randomly select a scout from the population and update the scout's position. The scout's position update formula is as follows:

[0149] ;

[0150] in, and Indicates that the coordinates in the population are The Investigator Second and The position in the d-dimensional solution space in the iteration, Indicates The global worst position of the current scout at iteration, Indicates The optimal position of the current scout at iteration, represents a random number that follows a Gaussian distribution with an expectation of 0 and a variance of 1. Take 10E-8, The random coefficients representing the direction and step size, represents the fitness value of the scout, and Indicates the best and worst fitness values ​​of the scout;

[0151] F5. Perturb the current optimal solution, recalculate the population fitness, and find the optimal solution after disturbance. The current optimal solution disturbance formula is as follows:

[0152] ;

[0153] in, Indicates that the coordinates in the population are Individual The position in the d-dimensional solution space in the iteration, represents the current optimal solution, represents the standard Cauchy distribution function;

[0154] F6. Determine whether the termination condition is met. If the termination condition is met, output the optimal individual position and its fitness value. Otherwise, repeat the above operation to obtain the optimal second logistics delivery path.

[0155] Embodiment three, as Figure 2 As shown, the present invention proposes an intelligent logistics delivery system based on electronic maps, which is applicable to the intelligent logistics delivery method based on electronic maps, including:

[0156] Position clustering unit 1, the position clustering unit 1 is used to obtain a set of customers to be delivered, obtain a set of locations to be delivered of the set of customers to be delivered based on an electronic map of the delivery area, and perform position clustering on the set of locations to be delivered to obtain a set of points to be delivered of the set of locations to be delivered;

[0157] Network construction unit 2, network construction unit 2 is used to obtain the logistics distribution center of the delivery area, and build a delivery connection network of the delivery area based on the electronic map of the delivery area, the set of points to be delivered and the logistics distribution center;

[0158] The network weighting unit 3 is used to obtain the needs of the customers to be delivered and the transportation efficiency of the delivery area, and weight the delivery connection network based on the needs of the customers to be delivered and the transportation efficiency of the delivery area to obtain a weighted delivery connection network;

[0159] The first dispatch unit 4 is used to construct a first logistics dispatch model based on the weighted dispatch connection network, solve the first logistics dispatch model based on reinforcement learning to obtain an optimal first logistics dispatch path, and deliver the logistics to the point to be dispatched based on the optimal first logistics dispatch path;

[0160] The second dispatch unit 5 is used to construct a second logistics dispatch model based on the dispatch point set and the set of locations to be dispatched, solve the second logistics dispatch model based on the finder optimization algorithm to obtain the optimal second logistics dispatch path, and deliver the logistics to the locations to be dispatched based on the optimal second logistics dispatch path.

[0161] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited thereto, and various changes can be made within the knowledge scope of technicians in the relevant technical field without departing from the purpose of the present invention.

Claims

1. A smart logistics delivery method based on electronic maps, characterized in that: include: Acquire a set of customers to be delivered, acquire a set of locations to be delivered of the set of customers to be delivered based on an electronic map of the delivery area, and perform location clustering on the set of locations to be delivered to obtain a set of points to be delivered of the set of locations to be delivered; Acquire the logistics distribution center of the delivery area, and construct a delivery connection network of the delivery area based on the electronic map of the delivery area, the set of points to be delivered and the logistics distribution center; Obtaining the needs of the customers to be delivered and the transportation efficiency of the delivery area, and weighting the delivery connection network based on the needs of the customers to be delivered and the transportation efficiency of the delivery area to obtain a weighted delivery connection network; Building a first logistics delivery model based on the weighted delivery connection network, solving the first logistics delivery model based on reinforcement learning to obtain an optimal first logistics delivery path, and delivering the logistics to the to-be-delivered point based on the optimal first logistics delivery path; A second logistics delivery model is constructed based on the set of delivery points and the set of locations to be delivered, and the second logistics delivery model is solved based on the finder optimization algorithm to obtain an optimal second logistics delivery path, and the logistics is delivered to the locations to be delivered based on the optimal second logistics delivery path.

2. According to the electronic map-based intelligent logistics delivery method of claim 1, it is characterized in that: Performing location clustering on the set of locations to be dispatched to obtain a set of points to be dispatched of the set of locations to be dispatched, including: Acquire the location text of the location to be dispatched, and perform a word segmentation operation on the location text to obtain a location word sequence corresponding to the location text; Performing stop word filtering on the position word sequence to obtain a standard position word sequence, and performing word frequency statistics on the standard position word sequence to obtain a word frequency bigram of the position word sequence, wherein the word frequency bigram includes the position word and the word frequency of the position word; Performing feature extraction on the word frequency bigrams of the position word sequence to obtain a word feature sequence of the position word sequence; Repeat the above operation until the set of locations to be dispatched is traversed to obtain a set of word feature sequences of the set of locations to be dispatched; The word feature sequence set is clustered to obtain clusters corresponding to the word feature sequence set, and the set of locations to be dispatched is classified into different points to be dispatched based on the clusters corresponding to the word feature sequence set.

3. The method for intelligent logistics delivery based on electronic maps according to claim 2 is characterized in that: The word feature sequence includes a first feature extraction formula, a second feature extraction formula, a third feature extraction formula and a fourth feature extraction formula, wherein the first feature extraction formula is as follows: ； in, Words indicating location In the location word library The frequency of occurrence of the position in the text, Words indicating location In the location word library The number of occurrences in the position text, Indicates the location of the word library The total number of words in the text at each position; The second feature extraction formula is as follows: ； in, Indicates the inverse frequency of position text, Indicates that a location word appears in the location word database The number of times Indicates the size of the location word library; The third feature extraction formula is as follows: ； in, Words indicating location In the location word library The first occurrence of a position in the text, Words indicating location Previously in The total number of other words that appear in the text at each position; The fourth feature extraction formula is as follows: ； in, Words indicating location The average position in the position text, Words indicating location At each occurrence of the position in the position text, Words indicating location The number of occurrences in the location text.

4. The method for intelligent logistics delivery based on electronic maps according to claim 2 is characterized in that: Clustering the word feature sequence set to obtain a cluster corresponding to the word feature sequence set includes: Calculating the local density of the word feature sequence based on information entropy, and constructing the neighborhood space of each word feature sequence based on K neighborhood and inverse K neighborhood; Calculating the local density of the word feature sequence in the neighborhood space, and determining the affiliation of the word feature sequence based on the local density of the word feature sequence to obtain a density center; Based on changing the neighborhood space to obtain multi-level density centers, a density graph structure is constructed based on the membership relationship of the multi-level density centers in the neighborhood space, and the graph connected components of the density graph structure are initial clusters; The unclustered word feature sequences are divided into initial clusters based on the membership relationship of the neighborhood space, and the above operation is repeated until all clusters corresponding to the word feature sequence set are obtained.

5. The method for intelligent logistics delivery based on electronic maps according to claim 1 is characterized in that: Building a delivery connection network for the delivery area based on the electronic map of the delivery area, the set of points to be delivered, and the logistics distribution center includes: Extracting a road network from the electronic map, wherein the road network includes a plurality of nodes and a plurality of connecting edges, wherein the nodes correspond to the points to be dispatched and the logistics distribution centers, and the connecting edges correspond to connection relationships between roads; The road network is traversed based on a depth-first search algorithm to check connectivity between the points to be dispatched and the distribution centers to obtain a reachable point set, and a dispatch connectivity network is constructed based on the reachable point set.

6. The method for intelligent logistics delivery based on electronic maps according to claim 1, characterized in that: Building a first logistics delivery model based on the weighted delivery connectivity network includes: A first objective function is constructed based on the weighted delivery connectivity network, wherein the first objective function is used to describe the minimization of the total driving distance of the delivery vehicle, and the first objective function is as follows: ； in, represents the first objective function, Indicates the point where the shipment is to be dispatched To the delivery point The distance Indicates delivery vehicle Whether it has passed the delivery point and delivery point , and when the delivery vehicle Passing by the delivery point and delivery point , and wait for the delivery point and delivery point When there is a road ,otherwise, , represents the set of points to be dispatched in the weighted dispatch network. Indicates the delivery vehicle collection. Represents the set of points to be dispatched in each class; A second objective function is constructed based on the weighted delivery connection network, wherein the second objective function is used to describe the maximization of the delivery vehicle loading rate, and the second objective function is as follows: ； in, represents the second objective function, Indicates the needs of each customer to be delivered. Indicates delivery vehicle Whether it passes through the delivery point , and when the delivery vehicle Passing by delivery point hour ,otherwise, , Indicates delivery vehicle The capacity limit of A third objective function is constructed based on the weighted delivery connection network, wherein the third objective function is used to describe the minimization of delivery time cost, and the third objective function is as follows: ； in, represents the third objective function, Indicates the point where the shipment is to be dispatched To the delivery point The driving time, Indicates the average time required to process each logistics. Indicates the point where the shipment is to be dispatched The number of express deliveries, Indicates the average consumption cost of each delivery vehicle; A first overall constraint condition is constructed based on the weighted delivery connectivity network, and a first logistics delivery model is constructed based on the first objective function, the second objective function, the third objective function and the first overall constraint condition. The first overall constraint condition is as follows: 。 7. The method for intelligent logistics delivery based on electronic maps according to claim 6 is characterized in that: Constructing a second logistics delivery model based on the delivery point set and the to-be-delivered location set includes: A fourth objective function is constructed based on the delivery point and the location to be delivered, wherein the fourth objective function is used to describe the minimization of the terminal delivery cost of the delivery vehicle, and the fourth objective function is as follows: ； in, represents the fourth objective function, Indicates whether the delivery point is connected to the location to be delivered. If so, , otherwise, , Indicates delivery vehicle fixed costs, Indicates the location of the shipment to be delivered The penalty cost, Indicates delivery vehicle Arrived at the location to be dispatched time, Indicates delivery vehicle Leaving the delivery location time, Indicates the distance from the delivery point to the location to be delivered. represents the unit transportation distance cost, Indicates the price of unit logistics, It indicates the cargo damage rate per unit time during transportation. It represents the cargo damage rate per unit time in the unloading process, Indicates the location of the shipment to be delivered The demand, represents the unloading rate, Indicates delivery vehicle Arrived at the location to be dispatched The remaining amount of logistics, represents the service cost per time; A second overall constraint condition is constructed based on the delivery point and the location to be delivered, and a second logistics delivery model is constructed based on the fourth objective function and the second overall constraint condition. The second overall constraint condition is as follows: 。 8. The method for intelligent logistics delivery based on electronic maps according to claim 7 is characterized in that: Solving the second logistics delivery model based on the finder optimization algorithm to obtain the optimal second logistics delivery path includes: Set the discoverer ratio , the proportion of investigators and warning value , and initialize the population, the population initialization formula is as follows: ； in, and Indicates the population and Individuals, represents the remainder function; Calculate the fitness value of each individual in the population, sort them according to the fitness value, select the discoverer from the individuals whose fitness value is greater than the preset fitness value threshold, and update the position of the discoverer. The update formula of the discoverer's position is as follows: ； in, and Indicates that the coordinates in the population are The discoverer of Second and In the iteration d The location of the solution space, Represents a random number that determines the distance the finder moves. Represents a random number that determines the direction in which the finder moves. Indicates The current finder's optimal position for iterations, and represents the division coefficient, Indicates the safety threshold; After the discoverer is selected from the population, the remaining individuals in the population are taken as joiners, and the positions of the joiners are updated. The joiner position update formula is as follows: ； in, and Indicates that the coordinates in the population are Joined by Second and The position in the d-dimensional solution space in the iteration, Indicates The global worst position of the current joiner in iterations, Indicates The optimal position of the current joiner in iterations, Represents a matrix composed of random numbers 1 or -1. represents the identity matrix, Represents the total number of individuals in the population.

9. The method for intelligent logistics delivery based on electronic maps according to claim 8, characterized in that: Solving the second logistics delivery model based on the finder optimization algorithm to obtain the optimal second logistics delivery path also includes: A scout is randomly selected from the population, and the position of the scout is updated. The scout position update formula is as follows: ； in, and Indicates that the coordinates in the population are The Investigator Second and The position in the d-dimensional solution space in the iteration, Indicates The global worst position of the current scout at iteration, Indicates The optimal position of the current scout at iteration, represents a random number that follows a Gaussian distribution with an expectation of 0 and a variance of 1. Take 10E-8, The random coefficients representing the direction and step size, represents the fitness value of the scout, and Indicates the best and worst fitness values ​​of the scout; The current optimal solution is disturbed, the population fitness is recalculated, and the optimal solution after disturbance is found. The current optimal solution disturbance formula is as follows: ； in, Indicates that the coordinates in the population are Individual The position in the d-dimensional solution space in the iteration, represents the current optimal solution, represents the standard Cauchy distribution function; Determine whether the termination condition is met. If the termination condition is met, output the optimal individual position and its fitness value. Otherwise, repeat the above operation to obtain the optimal second logistics delivery path.

10. An intelligent logistics delivery system based on electronic maps, which is applicable to an intelligent logistics delivery method based on electronic maps as described in any one of claims 1 to 9, characterized in that: include: A location clustering unit (1), the location clustering unit (1) being used to obtain a set of customers to be delivered, obtain a set of locations to be delivered of the set of customers to be delivered based on an electronic map of a delivery area, and perform location clustering on the set of locations to be delivered to obtain a set of points to be delivered of the set of locations to be delivered; A network construction unit (2), the network construction unit (2) being used to obtain a logistics distribution center in the delivery area, and to construct a delivery connection network in the delivery area based on an electronic map of the delivery area, the set of points to be delivered and the logistics distribution center; A network weighting unit (3), the network weighting unit (3) being used to obtain the needs of the customers to be delivered and the transportation efficiency of the delivery area, and to weight the delivery connection network based on the needs of the customers to be delivered and the transportation efficiency of the delivery area, so as to obtain a weighted delivery connection network; A first dispatch unit (4), the first dispatch unit (4) is used to construct a first logistics dispatch model based on the weighted dispatch connection network, solve the first logistics dispatch model based on reinforcement learning to obtain an optimal first logistics dispatch path, and deliver the logistics to the to-be-delivered point based on the optimal first logistics dispatch path; A second dispatch unit (5), the second dispatch unit (5) is used to construct a second logistics dispatch model based on the dispatch point set and the set of locations to be dispatched, solve the second logistics dispatch model based on a finder optimization algorithm to obtain an optimal second logistics dispatch path, and deliver the logistics to the locations to be dispatched based on the optimal second logistics dispatch path.

Images