A method for optimizing cigarette logistics distribution routes

Abstract

The present invention proposes a method for optimizing cigarette logistics distribution routes, comprising obtaining customer information, marking an electronic map, performing a first division according to the administrative district level, performing cluster area division according to a clustering algorithm, and performing small combination division according to the clustering algorithm. The cluster area and small combination division methods obtained by the clustering algorithm will be optimized according to specific weight scores, so as to obtain a division method that is compounded with actual distribution needs. Finally, according to a neural network algorithm, group cooperation is used as the minimum calculation unit to perform corresponding distribution route planning, and according to the planned route, customer location, interception loading and route segmentation, until all orders are loaded, the distribution route and distribution plan are obtained, which overcomes the problem of large calculation amount of the conventional distribution path calculation method. At the same time, a more reasonable weight scoring method is used to optimize the distribution area division method, so as to obtain a route optimization method that is more compounded with actual distribution needs, and has good application prospects.

Description

Technical Field

[0001] The present invention relates to the technical field of logistics distribution path planning, and in particular to a cigarette logistics distribution route optimization method. Background Art

[0002] With the continuous development of the national economy, the improvement of national consumption levels, the relaxation of tobacco management and operation policies, and the application of network management technology, the number of tobacco cigarette retail sales in my country has increased rapidly, and the cigarette distribution method has begun to change from centralized and large-scale distribution to small-batch and decentralized distribution, which has brought great challenges to cigarette distribution.

[0003] The ideal cigarette delivery process should be to deliver to as many users as possible with the shortest distance. When there are fewer users in need and the delivery area is small, the optimal delivery route can be obtained through a simple cost algorithm. However, when the number of delivery users is large and there are many path choices, the use of a simple path algorithm will bring huge computing tasks, and conventional processors cannot meet the computing requirements. At the same time, even if the delivery is carried out according to the shortest and most efficient path, there will be some other problems, such as cross-regional delivery, large differences in delivery volume between adjacent areas, and inconvenience in the division and arrangement of vehicles.

[0004] In view of the existing unreasonable aspects, there is an urgent need for a cigarette logistics distribution route optimization method that can overcome the above problems. Summary of the invention

[0005] In view of this, the present invention proposes a cigarette logistics distribution route optimization method which can reduce the amount of calculation and make the distribution method more reasonable.

[0006] The technical solution of the present invention is implemented as follows: The present invention provides a method for optimizing cigarette logistics distribution routes, which specifically includes:

[0007] S1. Obtain all customer information within the delivery city and mark them on the electronic map according to the customer information;

[0008] S2. Divide the distribution urban area into several district-level distribution areas by first dividing it into administrative districts;

[0009] S3, divide each district-level distribution area into a second-level division, and divide each district-level distribution area into a corresponding number of cluster areas according to the number of working days. The cluster area division method adopts K-means clustering method;

[0010] S4, dividing the collection area into the third-level distribution area, and dividing the customers in the collection area into several small groups, and the division method of the small groups is divided by K-means clustering method;

[0011] S5, taking group cooperation as the smallest calculation unit, the logistics center as the starting point and the end point, and each group cooperation as the passing point, the Hopfield neural network algorithm is used to calculate the optimal delivery route for each distribution area;

[0012] S6, based on the optimal route obtained in S5, according to the customer sequence on the route and the customer information of each customer, the single vehicle is filled with goods until the rated load of the single vehicle is reached, and the starting small combination of filling is used as the starting point and the small combination that reaches the rated load of the single vehicle is used as the end point as the delivery track of a vehicle, and the next vehicle is filled with the end point of the last fully loaded vehicle as the starting point, and the filling steps of the last vehicle are repeated until the orders of all small combinations are filled into the delivery vehicle;

[0013] S7. All vehicles deliver goods according to the corresponding starting points and end points in S6.

[0014] In the above technical solutions, the K-means clustering algorithm and the Hopfield neural network algorithm are both existing technologies.

[0015] Based on the above technical solution, preferably, the customer information includes: location address, delivery serial number, license number, license name and delivery quantity.

[0016] On the basis of the above technical scheme, preferably, in S1, the method of marking the customer information on the electronic map is specifically to obtain the location address in the customer information, locate on the electronic map according to the location address, and after the positioning is completed, mark the corresponding delivery serial number, license number, license name and delivery quantity in the customer information to the corresponding position.

[0017] On the basis of the above technical scheme, preferably, in S2, firstly, according to the division of administrative areas, the city is divided into several district-level areas, and the district-level areas are used as independent distribution areas, and then each administrative district-level distribution area is further divided. For example, if there are five working days a week, a district-level distribution area is divided into 5 cluster areas according to the K-means clustering method, and one cluster area is distributed every day. Then, each cluster area is divided into several small combinations according to the K-means clustering method. The number of small combinations can be determined according to needs, for example, determined along the main distribution road, or determined according to the number and specific location of intersections and lanes on the distribution route.

[0018] On the basis of the above technical solution, preferably, the S3 also includes performing a first rationality score and a first regulation item score on the cluster areas after clustering, and selecting the cluster area division with the highest sum of the first rationality score and the first regulation item score.

[0019] On the basis of the above technical scheme, preferably, in step S3, the cluster area division result of the K-means clustering method may have multiple coexisting optimization schemes, so it is necessary to further optimize the clustering results. The optimization method adopts the method of influencing factor weight scoring to evaluate, and each division scheme is scored for the first rationality and the first regulatory item. The first rationality score includes whether the classification of the cluster area is reasonable, whether the distribution vehicle can meet the road distribution requirements of the cluster area, whether the number of customer orders is evenly distributed, and whether the number of customers is balanced. Score each cluster area separately and then calculate the total score. According to the total score, make a comprehensive judgment and screen out the best cluster area division scheme.

[0020] On the basis of the above technical solution, preferably, the first rationality scoring method includes:

[0021] Cluster area classification score: Collect the urban area division information of all customers in the corresponding cluster area, and divide it according to the urban area with the highest proportion. For example, if there are 100 customers in a cluster area, 50 of them are in the urban area, 40 are in the suburbs, and 10 are in the mountainous area, then the corresponding cluster area will be classified as an urban cluster area. The cluster areas are classified into urban cluster areas, suburban cluster areas, and mountainous cluster areas. The maximum radius of each cluster area is calculated, and the maximum radius of the cluster area is compared with the maximum allowable radius of the corresponding urban area division. The average value of the location address coordinates of all customers in the cluster area is used as the center point, and the distance of all customers from the center point is calculated, and the maximum distance is used as the maximum radius. Large radius: when the maximum radius of the collection area is not greater than the maximum allowable radius of the corresponding urban area, 5 points will be awarded. Whenever the maximum radius of the collection area exceeds 5% of the maximum allowable radius of the corresponding urban area, 1 point will be deducted. The minimum score is 1 point. The maximum allowable radius of the urban collection area is 10km, the maximum allowable radius of the suburban collection area is 20km, and the maximum allowable radius of the mountainous collection area is 50km. For example, the maximum radius of a certain city collection area is 11.3km, which exceeds 1.3km, accounting for 13% of the maximum allowable radius, so 3 points will be deducted. The part less than 5% will be calculated as 5%. The maximum radius of the collection area can be calculated based on the map boundary coordinate information of the collection area;

[0022] Delivery vehicle feasibility score: Delivery vehicles include electric vehicles, small trucks and medium trucks. They are matched with the traffic data on the delivery routes in the distribution collection area. Whenever a restriction on one of the vehicles appears in the traffic data on the corresponding delivery route, 1 point will be deducted, with a total score of 5 points. For example, if there is a section of the delivery route in the existing distribution collection area where medium trucks are prohibited from passing, 1 point will be deducted. If there is also a section of the road where small trucks are prohibited from passing, another point will be deducted, and the final score is 3 points;

[0023] Customer order quantity balance score: Get the number of orders in each collection area in the same district-level distribution area, take the minimum number of orders as the benchmark, calculate the difference between the number of orders in other collection areas and the minimum number of orders, and get a full score of 5 points when the difference does not exceed 20% of the minimum number of orders. When the difference exceeds 20% of the minimum number of orders, deduct points. For every 5% excess, deduct 1 point. The minimum score is 1 point. For example, there are 5 collection areas, and the number of orders in each collection area is 1000, 1100, 1200, 1300, and 1360 respectively. The minimum order is 1000. The orders in other distribution areas exceed 10%, 20%, 30%, and 40% respectively. The 4 collection areas that exceed the limit need to deduct 0 points, 0 points, 2 points, and 4 points respectively. If it is less than 5%, it will be calculated as 5%;

[0024] Customer number balance score: obtain the arithmetic mean of the number of customers in all collection areas within the same district-level distribution area, collect the urban area division information of all customers in the corresponding collection areas, and classify the collection areas into urban collection areas, suburban collection areas and mountainous collection areas according to the urban area division with the highest proportion. Calculate the difference between the number of customers in the three types of collection areas and the average number of customers. When the number of customers in the urban collection area does not exceed 130% of the average number of customers, no point will be deducted, if it exceeds, 1 point will be deducted; when the number of customers in the suburban collection area is not less than 90% of the average number of customers, no point will be deducted, if it is less than, 1 point will be deducted; when the number of customers in the mountainous collection area is not less than 60% of the average number of customers, no point will be deducted, if it is less than, 1 point will be deducted. The total score is 5 points, and the minimum score is 1 point.

[0025] Further preferably, the delivery vehicle feasibility score also includes calculating whether all customers in the corresponding collection area are more than 50 km away from the logistics center. If so, 1 point will be deducted. Since the endurance of electric vehicles is relatively weak, this deduction item is set in consideration of the endurance of electric vehicles and the limitation of the minimum delivery distance.

[0026] Based on the above technical solution, preferably, the method for scoring the first regulatory item includes:

[0027] First, the monopoly supervision score: compare the law enforcement area to which the customers in each collection area belong with the administrative area corresponding to the customers to determine whether they are the same, collect the monopoly license information of all retail customers in the collection area, and determine whether the license is valid. When the law enforcement area described by all customers in a collection area is the same as the administrative area corresponding to it, and the licenses of all customers are valid, 1 point is awarded, otherwise 0 points are awarded. Specifically, the monopoly license information of all customers in the urban area is filed with the corresponding tobacco bureau, so the time information and business information at that time can be directly compared with the monopoly license information to determine whether it is valid;

[0028] First financial management score: Collect the historical payment methods of all customers in the collection area. The payment methods are divided into cash payment and other payments. Collect the urban area division information of all customers in the corresponding collection area. According to the urban area division with the highest proportion, classify the collection area into urban collection area, suburban collection area and mountainous collection area. Determine whether 100% of the historical payment methods of all customers in the urban collection area are other payments, determine whether other payment methods account for no less than 98% of the historical payment methods of all customers in the suburban collection area, and determine whether other payment methods account for no less than 95% of the historical payment methods of all customers in the mountainous collection area. If the above judgment results are all yes, get 1 point, otherwise get 0 points.

[0029] On the basis of the above technical solution, preferably, S4 also includes performing a second rationality score and a second regulation item score on the small combinations after clustering, and dividing the small combinations with the highest sum of the second rationality score and the second regulation item score.

[0030] On the basis of the above technical solutions, preferably, after clustering, there may also be a variety of preferred small combination division methods. In order to further and more reasonably optimize the small combinations, the second rationality score and the second regulation item score are used to evaluate each small combination, and then the optimal division method is selected according to the score.

[0031] On the basis of the above technical solution, preferably, the second rationality scoring method includes:

[0032] Delivery time scoring: collect the historical delivery time of each customer in each small group and calculate the average delivery time, and compare whether the average delivery time does not exceed 3 minutes. The full score is 5 points. If the average delivery time does not exceed 3 minutes, 5 points will be awarded. If the average delivery time exceeds 3 minutes, 1 point will be deducted for each additional minute. The minimum score is 1 point.

[0033] Delivery distance score: Calculate the total delivery distance of all customers in each small group, collect the urban area division information of all customers in the corresponding small group, and classify the small groups into urban small groups, suburban small groups and mountainous small groups according to the urban area division with the highest proportion. Use the customer coordinate data in the small group and the electronic map as a tool to calculate the delivery distance of all customers in the small group. The total score is 5 points. When the total delivery distance of the urban small group is not more than 2km, it will get full marks, and 1 point will be deducted for every 10% exceeding it, with a minimum score of 1 point. When the total delivery distance of the suburban small group is not more than 5km, it will get full marks, and 1 point will be deducted for every 10% exceeding it, with a minimum score of 1 point. When the total delivery distance of the mountainous small group is not more than 10km, it will get full marks, and 1 point will be deducted for every 10% exceeding it, with a minimum score of 1 point.

[0034] Score for number of delivered households: Count the total number of customers in each small group, collect the urban area division information of all customers in the corresponding small group, and classify the small groups into urban small groups, suburban small groups and mountainous small groups according to the urban area division with the highest proportion. The total score is 5 points. When the number of customers in the urban small group does not exceed 15 households, full marks will be awarded, and 1 point will be deducted for every additional 3 households, with a minimum of 1 point. When the number of customers in the suburban small group does not exceed 10 households, full marks will be awarded, and 1 point will be deducted for every additional 2 households, with a minimum of 1 point. When the number of customers in the mountainous small group does not exceed 5 households, full marks will be awarded, and 1 point will be deducted for every additional household, with a minimum of 1 point.

[0035] On the basis of the above technical solution, preferably, the method for scoring the second regulatory item includes:

[0036] Second Monopoly Supervision Scoring: Compare the law enforcement area to which the customers in each small group belong with the administrative area corresponding to the customers to determine whether they are the same, determine whether the monopoly licenses of all customers in the small group are valid, and determine whether the legal person of all customers in the small group is the same person. If all the above judgments are yes, 1 point is awarded, otherwise 0 point is awarded;

[0037] Second financial management score: Collect the urban area division information of all customers in the corresponding small group, and classify the small groups into urban small groups, suburban small groups and mountainous small groups according to the urban area division with the highest proportion. Collect the historical payment methods of each customer in the small group. The payment methods are divided into cash payment and other payments. Determine whether all customers in the urban small group are other payments, determine whether the number of customers who pay in cash in the suburban small group does not exceed 10% of the total number of customers in the small group, and determine whether the number of customers who pay in cash in the mountainous small group does not exceed 1. When the above judgment results are all yes, get 1 point, otherwise get 0 points.

[0038] On the basis of the above technical solution, preferably, the calculation method of the small combination includes:

[0039] Delivery time or delivery distance restrictions: By collecting the delivery history data of customers in the past 6 months, when the delivery time between two adjacent households is less than 180 seconds, the two adjacent households can be regarded as a small group. When the delivery time between any of the two households and other customers is less than 180 seconds, the other customers will also be included in the small group; when the driving distance or coordinate distance between two adjacent households is less than 200 meters, the two adjacent households can be regarded as a small group. When the driving distance or coordinate distance between any of the two households and other customers is less than 200 meters, the other customers will also be included in the small group.

[0040] On the basis of the above technical solutions, preferably, the small combination also includes a prescribed delivery plan: the fixed delivery plan includes several customers in the same community or several customers sharing a road entrance and exit, that is, all customers in the same community are classified into a small combination, and several customers sharing a road entrance and exit are classified into a small combination.

[0041] Based on the above technical solution, preferably, the combined delivery time of the small group cannot exceed 1 hour, and the upper limit of the number of users in the small group is 30.

[0042] The combined delivery time for a small group is the sum of the service time for the group and all customers within it. The service time includes the driving time for a single household and the door-to-door service time for a single household.

[0043] The cigarette logistics distribution route optimization method of the present invention has the following beneficial effects compared with the prior art:

[0044] (1) The cigarette logistics distribution route optimization method of the present invention adopts two clustering, with group cooperation as the smallest unit, which greatly reduces the computational complexity and difficulty of the clustering algorithm. At the same time, group cooperation is the smallest unit of distribution, which can effectively reduce the distribution difficulty and distribution workload. In order to further make the distribution division result obtained by the clustering algorithm consistent with the actual distribution, the present invention also adopts the cluster area optimization evaluation scheme and the small combination optimization evaluation scheme to further optimize the clustering results, so that the optimized distribution results can better meet the needs of actual distribution, reduce the difficulty of actual distribution, and reduce the difficulties encountered in the actual distribution process;

[0045] (2) The present invention uses group cooperation as the minimum delivery unit. A small group is composed of several customers. The customers in the small group can be satisfied with one delivery. The delivery personnel can complete the delivery by walking. In this way, the minimum calculation unit changes from one customer to multiple customers, and the calculation base is reduced by multiples. At the same time, the small group method can be used for delivery, which can greatly shorten the delivery time and reduce the delivery workload.

[0046] (3) In order to further reduce the problems that may exist in the distribution process, the first rationality score and the first regulatory item score are first given to the divided collection areas. The first rationality score is to give a corresponding weight score to each collection area division method based on the problems that may exist in the actual distribution process. The first regulatory item score is a statistical score based on some supervision and financial risks that exist in actual distribution and customer management. The risk item adopts a one-vote veto system, so as to fundamentally eliminate or minimize risk problems. Combined with the first rationality score and the first regulatory item score, the division method with the highest score is selected as the optimal result;

[0047] (4) A method similar to the first rationality scoring and the first regulatory item scoring of the cluster area is used to further score and optimize the small combinations, and a small combination division method that better meets the actual distribution needs is screened out. The present invention overcomes the problem of large amount of calculation in conventional cigarette logistics distribution routes, and at the same time optimizes the selection of distribution routes, making the optimal distribution route more in line with the actual distribution requirements, reducing the distribution workload of distribution personnel, and reducing the difficulty of the distribution process and subsequent maintenance management. DETAILED DESCRIPTION

[0048] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] Example 1

[0050] The cigarette logistics distribution routes in Shiyan City are optimized, and the specific optimization methods are as follows.

[0051] According to statistics, there are 13,900 retail customers in Shiyan City. The information of these retail customers is collected and marked on the electronic map. Then, according to the administrative divisions, including Shiyan City, Yunyang District, Danjiang District, Yunxi District, Fangxian District, Zhushan District, and Zhuxi District, further clustering and sub-division are carried out for the seven administrative districts. The tobacco working days in Shiyan City are Monday to Friday, so there are 5 working days. The corresponding administrative districts are divided into 5 clustering and sub-districts. The K-means clustering algorithm is used for clustering calculation. After obtaining multiple results of clustering division, the first rationality scoring method and the first regulatory item scoring are used to evaluate the results of clustering division. The specific first rationality scoring method includes clustering and sub-district classification scoring, delivery vehicle feasibility scoring, customer order quantity balance scoring, and customer number balance scoring.

[0052] Specific area classification scores include:

[0053] The customer information in each collection area is counted separately according to the urban area division information of the customer's location. For example, there are 1,000 customers in the first collection area, of which 500 customers are located in urban areas, 400 customers are located in suburban areas, and 100 customers are located in mountainous areas. The area is defined according to the urban area division of the largest number of customers. Therefore, the first collection area is defined as the urban collection area based on the 500 urban customers with the largest proportion. Similarly, the other four collection areas are classified accordingly. Then the maximum radius of each collection area is calculated, and the average value of the location address coordinates of all customers in the collection area is used as the center point. The distance of all customers from the center point is calculated, and the maximum distance is used as the maximum radius. The maximum radius of the urban collection area shall not exceed 10km, and the maximum radius of the suburban area shall not exceed 2 0km, the maximum radius in mountainous area does not exceed 50km, and comparative calculation is carried out. When the requirements are met, the full score of 5 points is obtained. When the requirements are not met, 1 point is deducted for each 5% exceeding the limit. For example, the first set of areas is the urban set area, and its maximum radius is 8km, the second set of areas is the urban set area, and its maximum radius is 12km, the third set of areas is the suburbs, and its maximum radius is 18km, the fourth set of areas is the suburbs, and its maximum radius is 23km, and the fifth set of areas is the mountainous area, and its maximum radius is 53km. Therefore, according to the scoring criteria, the first set of areas gets 5 points, the second set of areas exceeds 20%, 4 points are deducted, and the score is 1, the third set of areas gets 5 points, the fourth set of areas exceeds 15%, 3 points are deducted, and the score is 2, the fifth set of areas exceeds 6%, 2 points are deducted, and the score is 3, and so on. The different set area division results are scored.

[0054] Since urban areas have a denser population, more shops, and more congested traffic, shorter distances may take more time to travel. Therefore, the geographical location of urban areas needs to be reduced as much as possible to balance the actual delivery time. Suburban and mountainous areas have faster traffic, are less prone to congestion, and are more dispersed. A larger area is needed to accommodate more customers, so that the proportion of goods in the vehicle is balanced, and there will be no situation where there are not enough goods to deliver in urban areas and not enough goods to be loaded in suburban areas.

[0055] Specific delivery vehicle feasibility scores include:

[0056] The vehicle traffic rules of the roads where customers are located in different collection areas are counted respectively to determine whether there are any prohibitions on delivery vehicles in the vehicle traffic rules. Delivery vehicles include electric vehicles, small trucks and medium-sized trucks. The vehicle traffic rules can be directly obtained from the background data of Baidu Maps. It is a prior art. When there is one more prohibition on delivery vehicles in the traffic rules, 1 point is deducted for the corresponding one, and the full score is 5 points. At the same time, it is also necessary to determine whether the distance from the distribution center to the nearest customer in the corresponding collection area exceeds 50km. If it exceeds, 1 point is deducted. For example, the vehicle traffic rules for the corresponding roads in the first collection area do not prohibit vehicle types, and there is at least one customer inside that is less than 50km away from the nearest distribution center, so full marks are obtained. In the second collection area, there are regulations for stationary medium-sized vehicles in the vehicle traffic rules for the corresponding roads, so 1 point is deducted. At the same time, no customer inside is less than or equal to 50km away from the nearest distribution center, so 1 point is deducted, and 3 points are obtained. And so on, and the sum of the scores of the five collection areas in different division methods is calculated.

[0057] The feasibility assessment of vehicles can effectively avoid the problem of vehicles turning back to unload and re-arrange vehicles for delivery due to road regulations not allowing the corresponding delivery vehicles to pass during the actual delivery process, ensuring that vehicle delivery is effective.

[0058] The customer order quantity balance score includes:

[0059] Count the number of orders in each set area in each set area division result, find out the set area with the least number of orders, and compare the number of customers in other set areas with the minimum order quantity to determine whether it exceeds 20% of the minimum order number. If it does not exceed, get the full score of 5 points. If it exceeds, 1 point will be deducted for every 5% exceeding, and the minimum score is 1 point. For example, the number of orders in the first set area is 1000, the number of orders in the second set area is 1050, the number of orders in the third set area is 1100, the number of orders in the fourth set area is 1140, and the number of orders in the fifth set area is 1200. Then the first set area has the smallest number of orders and gets the full score of 5 points. The second set area exceeds 5%, 1 point will be deducted, and the score is 4 points. The third set area exceeds 10%, 2 points will be deducted, and the score is 3 points. The fourth set area exceeds 14%, and the part less than 5% is calculated as 5%, 3 points will be deducted, and the score is 2. The fifth set area exceeds 100%, but it is calculated according to the minimum score of 1 point. And so on, the score of order quantity balance in each division method is calculated.

[0060] The order balance score can also ensure that the daily freight demand is relatively balanced, avoiding the situation where there is too much cargo on a certain day and it is difficult to transport it, or too little cargo on a certain day and the vehicles are idle.

[0061] The customer balance score includes:

[0062] Get the arithmetic mean of the number of customers in the five cluster areas. For example, if the total number of customers in the five cluster areas is 5,000, the average value is 1,000. Perform statistics based on the customer information in each cluster area and the urban area division information of the customer's location. For example, there are 1,000 customers in the first cluster area, of which 500 customers are located in the urban area, 400 customers are located in the suburbs, and 100 customers are located in the mountainous area. Then define it according to the urban area division of the largest proportion of customers. Therefore, according to the 500 urban customers with the largest proportion, the first cluster area will be divided into One collection area is defined as the urban collection area, and the other four collection areas are classified similarly. Then, based on the classification of the collection areas and the comparison between the number of customers in the corresponding collection areas and the average value, the points will be calculated: when the number of customers in the urban collection area does not exceed 130% of the average number of customers, no points will be deducted; if it exceeds, 1 point will be deducted; when the number of customers in the suburban collection area is not less than 90% of the average number of customers, no points will be deducted; if it is less than, 1 point will be deducted; when the number of customers in the mountainous collection area is not less than 60% of the average number of customers, no points will be deducted; if it is less than, 1 point will be deducted. The total score is 5 points, and the minimum score is 1 point. For example, the first set of areas is the urban area with 1,400 customers, the second set of areas is the suburban area with 800 customers, the third set of areas is the suburban area with 950 customers, the fourth set of areas is the suburban area with 960 customers, and the fifth set of areas is the mountain area with 890 customers. Then, 1 point will be deducted for exceeding 1,300 customers in the first set, no point will be deducted for not exceeding 900 customers in the second set, 1 point will be deducted for exceeding 900 customers in the third set, 1 point will be deducted for exceeding 900 customers in the fourth set, and 1 point will be deducted for exceeding 600 customers in the fifth set, so the total score is 1 point. And so on, a score will be calculated for each division method.

[0063] A more balanced number of customers can make the daily workload relatively even, which is conducive to the division of labor and work management.

[0064] The scoring method for the first regulation item includes the first monopoly supervision score and the first financial management score, specifically:

[0065] The first monopoly supervision score includes:

[0066] First, determine whether the law enforcement area to which the customers in each collection area belong is the same as the administrative area corresponding to the customers;

[0067] Second, determine whether the exclusive licenses of all customers in the collection area are valid.

[0068] If both of the above judgment results are yes, 1 point is awarded. If one of them is not satisfied, 0 point is awarded.

[0069] For example, if the law enforcement area of ​​the customers in the first, second, third, fourth and fifth areas is the same as the corresponding geographical administrative area, the first item is judged to be satisfied, and the exclusive licenses of all customers in these five areas are judged to be valid. If they are all valid, 1 point is awarded. If the law enforcement area of ​​any customer in any area is different from the corresponding administrative area, or the exclusive license of any customer is invalid, 0 points are awarded.

[0070] First Financial Management Scores include:

[0071] Collect the historical payment methods of all customers in the collection area. The payment methods are divided into cash payment and other payment methods. Other payment methods include Alipay transfer, WeChat transfer, bank transfer and other forms of online transaction methods. Statistics are made based on the customer information in each collection area, and based on the urban area division information of the customer's location. For example, there are 1,000 customers in the first collection area, of which 500 customers are located in the urban area, 400 customers are located in the suburban area, and 100 customers are located in the mountainous area. The area is defined according to the urban area division of the largest number of customers. Therefore, according to the largest number of 500 urban customers, the first collection area is defined as the urban collection area. Similarly, the other four collection areas are classified accordingly. The specific requirements are that the historical payment methods of all customers in the urban area must be 100% other payment, at least 98% of customers in the suburban collection area are other payment, and at least 95% of customers in the mountainous collection area are other payment. When all the above requirements are met, 1 point is obtained. If at least one of the requirements is not met, 0 points are obtained. Specifically, for example, the first area is an urban area, where all customers use other payment methods; the second area is a suburban area, where 99% of customers use other payment methods; the third area is a suburban area, where 98% of customers use other payment methods; the fourth area is a suburban area, where 96% of customers use other payment methods; and the fifth area is a mountain area, where 97% of customers use other payment methods. Therefore, the payment method of the fourth area cannot meet the requirements and gets 0 points. If no less than 98% of the customers in the fourth area use other payment methods, 1 point can be obtained.

[0072] The above first regulatory item score can make the background processing of cigarette logistics easier and facilitate data management and financial management.

[0073] The cluster areas of each division method are scored according to the above scoring method, and the total score is calculated. The one with the highest score is the optimal cluster area division method.

[0074] After the cluster area division is completed, each cluster area is further clustered according to the K-means clustering method. The cluster center can be determined by manually setting a cluster center value or setting a fixed distance according to the geographical location of the main road in the cluster area. In this embodiment, it is determined in a fixed number manner, for example, 2000 cluster centers are set. After clustering calculation, the cluster division results of several small combinations are obtained, and then the divided small combinations are scored by the second rationality score and the second regulation item score. Specifically:

[0075] The second rationality scoring method includes: delivery time scoring, delivery distance scoring, and delivery number scoring.

[0076] Delivery time scores include:

[0077] Collect the historical delivery time of each customer in each small group and calculate the average delivery time, compare whether the average delivery time does not exceed 3 minutes, if not, get the full score of 5 points, if it exceeds, deduct 1 point for every minute exceeded, and count as 1 minute if it is less than 1 minute. For example, small group 1 has 3 customers, the historical delivery time of the first customer is 3 minutes, the historical delivery time of the second customer is 4 minutes, and the historical delivery time of the third customer is 4.5 minutes, then the average delivery time of the three customers is 3 minutes and 50 seconds, if it exceeds 3 minutes and 50 seconds, deduct 1 point, and get 4 points, and so on, the sum of the scores of all small groups in each division;

[0078] A more balanced delivery time is also conducive to the management of delivery tasks, making the rhythm of delivery more coordinated. It can also improve the rationality of small combinations and avoid the situation where the actual delivery time within a small combination is much longer than the vehicle driving time, thereby reducing actual work efficiency.

[0079] The delivery number score includes:

[0080] Count the total number of customers in each small group, and divide them into different areas according to the customers' location. For example, if there are 5 customers in a small group, 3 of them are located in the urban area, 1 is located in the suburbs, and 1 is located in the mountainous area, then the group is defined according to the urban area to which the largest number of customers belongs. Therefore, the group is defined as an urban area group based on the 3 urban area customers that account for the largest proportion. Similarly, all other small groups are classified accordingly, and the number of delivery households is scored based on the classification of the small groups. The total score is 5 points. When the customers in the urban area group When the number of customers does not exceed 15 households, full marks will be awarded, 1 point will be deducted for every additional 3 households, and the minimum score is 1 point. When the number of customers in the suburban small group does not exceed 10 households, full marks will be awarded, 1 point will be deducted for every additional 2 households, and the minimum score is 1 point. When the number of customers in the mountainous small group does not exceed 5 households, full marks will be awarded, 1 point will be deducted for every additional household, and the minimum score is 1 point. For example, when there are 14 households in a small group in an urban area, 5 full marks will be awarded. For another example, when there are 12 households in a small group in the suburbs, 1 point will be deducted for every additional household, and the score is 4. And so on. All small groups in each divided area will be scored and the total score will be calculated.

[0081] Too many delivery households will also reduce the efficiency of a single delivery within a small group. If the number of customers in a small group far exceeds the delivery load that the delivery staff can deliver at one time, then in actual operations, the delivery staff will still need to perform multiple round trips, reducing the actual delivery efficiency.

[0082] The scoring methods for the second regulation item of the small group include: the second monopoly supervision score and the second financial management score.

[0083] Second, the monopoly supervision score includes:

[0084] First, compare the law enforcement area to which the customers in each small group belong with the administrative area corresponding to the customers to determine whether they are the same;

[0085] Second, determine whether the exclusive licenses of all customers in the small group are valid;

[0086] Third, determine whether the legal person of all customers in the small group is the same person.

[0087] If all of the above judgments are true, 1 point will be awarded. If any one or more of the above judgments are not satisfied, 0 points will be awarded.

[0088] Specifically, when there are 3 customers in a small group, the law enforcement area to which the three customers belong is the same as the administrative area in which the three customers are located, and the exclusive licenses of the three customers are all valid, and the actual legal persons of the three customers are the same person, 1 point will be awarded. If at least one customer's exclusive license is invalid, or the legal persons of the three customers are different, or at least one customer's law enforcement area is different from the administrative area, 0 points will be awarded.

[0089] The second financial management score includes:

[0090] Collect the information of the urban area division of all customers in the corresponding small group. According to the urban area division with the highest proportion, classify the small group into urban small group, suburban small group and mountainous small group. Collect the historical payment methods of each customer in the small group. The payment methods are divided into cash payment and other payment. For example, there are 5 customers in a small group, of which three customers are in the urban area, one customer is in the suburbs, and the other customer is in the mountainous area. Then the school group is judged as an urban small group. If there are urban area divisions with the same number of customers, it is preferentially judged as a small group of a lower level. For example, if there are 2 urban customers, 2 suburban customers, and 2 mountainous customers at the same time, then the small group is judged as a mountainous small group. Determine whether all customers in the urban small group are other payments, whether the number of customers who pay in cash in the suburban small group does not exceed 10% of the total number of customers in the small group, and whether the number of customers who pay in cash in the mountainous small group does not exceed 1. When the above judgment results are all yes, 1 point is obtained, otherwise 0 points are obtained. For example, in a mountainous small group, there are 10 customers, and if 1 of them pays in cash, 1 point is obtained. If 2 customers pay in cash, 0 points are obtained. And so on, and calculate the financial management scores of all subcombinations in each subcombination division.

[0091] Calculate the sum of the second rationality score and the second regulatory item score for each division in the corresponding small combination division, and select the small combination division with the highest score as the preferred small combination division method.

[0092] After the corresponding small combination division method is optimized, the group cooperation is used as the minimum calculation unit, the logistics center closest to the corresponding collection area is used as the starting point and midpoint, and each group cooperation is used as the way point. The Hopfield neural network algorithm is an existing technology. The optimal path is used as the distribution route. The delivery vehicles are filled according to the order of customers on the route and the order volume of each customer. For example, the upper limit of the distribution of the delivery vehicle is 1,000. The distribution path includes in order: 300 orders for small combination 1, 200 orders for small combination 2, 50 orders for small combination 3, 100 orders for small combination 4, 500 orders for small combination 5, 300 orders for small combination 6, and 2 orders for small combination 7. 00, small combination 8 has 300 orders, and so on, then in order, small combination 1, small combination 2, small combination 3 and small combination 4 are first arranged to the first vehicle. At this time, if small combination 5 is added to the first vehicle, the delivery upper limit of the delivery vehicle will be exceeded. After filling, small combination 5 is used as the delivery starting point of the second delivery vehicle to fill, and small combination 5, small combination 6 and small combination 7 are filled into the second delivery vehicle respectively. At this time, the second delivery vehicle has just delivered 1000 items. At this time, small combination 8 is used as the delivery starting point of the third delivery vehicle to fill, and so on, until the orders of all small combinations in the collection area are filled, and then the delivery vehicles deliver according to the above-mentioned combination filling method. Delivery can be carried out simultaneously or in batches, and planning and arrangement are carried out according to the number of delivery vehicles.

[0093] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for optimizing cigarette logistics distribution routes, characterized in that: The steps include: S1. Obtain all customer information within the delivery city and mark them on the electronic map according to the customer information; S2. Divide the distribution urban area into several district-level distribution areas by first dividing it into administrative districts; S3, divide each district-level distribution area into a second-level division, and divide each district-level distribution area into a corresponding number of cluster areas according to the number of working days. The cluster area division method adopts K-means clustering method; S4, dividing the collection area into the third-level distribution area, and dividing the customers in the collection area into several small groups, and the division method of the small groups is divided by K-means clustering method; S5, taking group cooperation as the smallest calculation unit, the logistics center as the starting point and the end point, and each group cooperation as the passing point, the Hopfield neural network algorithm is used to calculate the optimal delivery route for each distribution area; S6, based on the optimal route obtained in S5, according to the customer sequence on the route and the customer information of each customer, the single vehicle is filled with goods until the rated load of the single vehicle is reached, and the starting small combination of filling is used as the starting point and the small combination that reaches the rated load of the single vehicle is used as the end point as the delivery track of a vehicle, and the next vehicle is filled with the end point of the last fully loaded vehicle as the starting point, and the filling steps of the last vehicle are repeated until the orders of all small combinations are filled into the delivery vehicle; S7, all vehicles deliver goods according to the corresponding starting points and destinations in S6; The S3 further includes performing a first rationality score and a first regulation item score on the clustered areas after clustering, and selecting the cluster area with the highest sum of the first rationality score and the first regulation item score; The first rationality scoring method includes: Area classification score: Collect the urban area division information of all customers in the corresponding area, and classify the area into urban area area, suburban area area and mountain area area according to the urban area with the highest proportion. Calculate the maximum radius of each area, and compare the maximum radius of the area with the maximum allowable radius of the corresponding urban area. Take the average value of the location address coordinates of all customers in the area as the center point, calculate the distance of all customers from the center point, and take the maximum distance as the maximum radius. When the maximum radius of the area is not greater than the maximum allowable radius of the corresponding urban area, 5 points are awarded. Whenever the maximum radius of the area exceeds 5% of the maximum allowable radius of the corresponding urban area, 1 point is deducted. The minimum score is 1 point. The maximum allowable radius of the urban area is 10km, the maximum allowable radius of the suburban area is 20km, and the maximum allowable radius of the mountain area is 50km. Delivery vehicle feasibility score: Delivery vehicles include electric vehicles, small trucks and medium trucks. They are matched with delivery vehicles based on the traffic data on the delivery routes within the distribution area. Whenever a restriction on one of the vehicles appears in the traffic data on the corresponding delivery route, 1 point is deducted, with a total score of 5 points. Customer order quantity balance score: Get the order quantity of each collection area in the same district-level distribution area, take the minimum order quantity as the benchmark, calculate the difference between the order quantity of other collection areas and the minimum order quantity, and get a full score of 5 points when the difference does not exceed 20% of the minimum order quantity. When the difference exceeds 20% of the minimum order quantity, points will be deducted, and 1 point will be deducted for every 5% over the minimum order quantity. The minimum score is 1 point. Customer number balance score: obtain the arithmetic mean of the number of customers in all collection areas within the same district-level distribution area, collect the urban area division information of all customers in the corresponding collection area, and classify the collection areas into urban collection areas, suburban collection areas and mountainous collection areas according to the urban area division with the highest proportion. Calculate the difference between the number of customers in the three classifications and the average number of customers. When the number of customers in the urban collection area does not exceed 130% of the average number of customers, no point will be deducted, and 1 point will be deducted if it exceeds. When the number of customers in the suburban collection area is not less than 90% of the average number of customers, no point will be deducted, and 1 point will be deducted if it is less than. When the number of customers in the mountainous collection area is not less than 60% of the average number of customers, no point will be deducted, and 1 point will be deducted if it is less than. The total score is 5 points, and the minimum score is 1 point. The scoring method for the first regulatory item includes: First Monopoly Supervision Score: Compare the law enforcement area to which the customers in each collection area belong with the administrative area corresponding to the customers to determine whether they are the same, collect the monopoly license information of all retail customers in the collection area, and determine whether the license is valid. When the law enforcement area described by all customers in a collection area is the same as the administrative area corresponding to it, and the licenses of all customers are valid, 1 point is awarded, otherwise 0 point is awarded; First financial management score: collect the historical payment methods of all customers in the collection area, and the payment methods are divided into cash payment and other payment. Collect the urban area division information of all customers in the corresponding collection area, and classify the collection area into urban collection area, suburban collection area and mountain collection area according to the urban area division with the highest proportion. Determine whether 100% of the historical payment methods of all customers in the urban collection area are other payment methods, determine whether other payment methods account for no less than 98% of the historical payment methods of all customers in the suburban collection area, and determine whether other payment methods account for no less than 95% of the historical payment methods of all customers in the mountain collection area. If all the above judgment results are yes, 1 point will be awarded, otherwise 0 point will be awarded; The S4 further includes performing a second rationality score and a second regulation item score on the small combinations that have been clustered, and dividing the small combinations with the highest sum of the second rationality score and the second regulation item score; The second rationality scoring method includes: Delivery time scoring: collect the historical delivery time of each customer in each small group and calculate the average delivery time, and compare whether the average delivery time does not exceed 3 minutes. The full score is 5 points. If the average delivery time does not exceed 3 minutes, 5 points will be awarded. If the average delivery time exceeds 3 minutes, 1 point will be deducted for each additional minute. The minimum score is 1 point. Delivery distance score: Calculate the total delivery distance of all customers in each small group, collect the urban area division information of all customers in the corresponding small group, and classify the small groups into urban small groups, suburban small groups and mountainous small groups according to the urban area division with the highest proportion. Use the customer coordinate data in the small group and the electronic map as a tool to calculate the delivery distance of all customers in the small group. The total score is 5 points. When the total delivery distance of the urban small group is not more than 2km, it will get full marks, and 1 point will be deducted for every 10% exceeding it, with a minimum score of 1 point. When the total delivery distance of the suburban small group is not more than 5km, it will get full marks, and 1 point will be deducted for every 10% exceeding it, with a minimum score of 1 point. When the total delivery distance of the mountainous small group is not more than 10km, it will get full marks, and 1 point will be deducted for every 10% exceeding it, with a minimum score of 1 point. Delivery household number score: Count the total number of customers in each small group, collect the urban area division information of all customers in the corresponding small group, and classify the small groups into urban small groups, suburban small groups and mountainous small groups according to the urban area division with the highest proportion. The total score is 5 points. When the number of customers in the urban small group does not exceed 15 households, it will get full marks, and 1 point will be deducted for every additional 3 households, with a minimum score of 1 point. When the number of customers in the suburban small group does not exceed 10 households, it will get full marks, and 1 point will be deducted for every additional 2 households, with a minimum score of 1 point. When the number of customers in the mountainous small group does not exceed 5 households, it will get full marks, and 1 point will be deducted for every additional 1 household, with a minimum score of 1 point. The scoring method of the second regulation item includes: Second Monopoly Supervision Scoring: Compare the law enforcement area to which the customers in each small group belong with the administrative area corresponding to the customers to determine whether they are the same, determine whether the monopoly licenses of all customers in the small group are valid, and determine whether the legal person of all customers in the small group is the same person. If the above judgment results are all yes, 1 point is awarded, otherwise 0 point is awarded; Second financial management score: Collect the urban area division information of all customers in the corresponding small group, and classify the small group into urban small group, suburban small group and mountainous small group according to the urban area division with the highest proportion. Collect the historical payment methods of each customer in the small group. The payment methods are divided into cash payment and other payments. Determine whether all customers in the urban small group are other payments, determine whether the number of customers who pay in cash in the suburban small group does not exceed 10% of the total number of customers in the small group, and determine whether the number of customers who pay in cash in the mountainous small group does not exceed 1. When the above judgment results are all yes, get 1 point, otherwise get 0.

2. The cigarette logistics distribution route optimization method according to claim 1, characterized in that: The customer information includes: location address, delivery serial number, license number, license name and delivery quantity.

3. The cigarette logistics distribution route optimization method according to claim 1, characterized in that: The delivery vehicle feasibility score also includes calculating whether all customers in the corresponding collection area are more than 50 km away from the logistics center. If so, 1 point will be deducted.