Method, system and equipment for automatically learning and scheduling orders

Abstract

The invention discloses an automatic order learning and scheduling method and system, and the method comprises the steps: creating a unique index in a database through a newly-added pickup scheduling and learning detail table and a pickup scheduling and learning table; the method comprises the following steps of: acquiring information of a scheduling network point and a scheduling salesman through order pickup information, storing the acquired information into a database, and accumulating a network point pickup amount every day; the method comprises the following steps of: acquiring each province, city, district and a website with the maximum express taking quantity in the near 15 days at a detailed address through timed task calling, and sending data to a Rocket MQ server; data are pulled from the Rocket MQ server, and learning data are stored in a database; after the order is placed, the newly added order data is pushed to the Kafka server through the Oracle GoldenGate; pulling data from the Kafka server and querying province, city, district and detailed addresses corresponding to the order; learning data stored in the database is inquired in real time through provincial, municipal, district and detailed addresses of the order, and a corresponding pick-up branch, namely a predicted pick-up branch obtained after learning, is obtained. According to the method and the system, the manual operation frequency can be remarkably reduced, and the scheduling accuracy is improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence, and in particular to a method, system and device for realizing automatic learning and scheduling of orders in the express delivery industry. Background Art

[0002] With the rapid development of express delivery business and the access to e-commerce business systems in various channels, the number of orders is increasing. The business needs to be able to more quickly identify which outlet should pick up each order. However, the current method of manually maintaining the pickup outlets for each address is not universal and accurate enough, and the labor and time costs of manual intervention are relatively high. Due to the inaccuracy and untimeliness of existing scheduling, there are many failed collections. In order to improve the collection timeliness and the timeliness of the overall logistics, it is necessary to improve the accuracy and timeliness of automatic scheduling and reduce labor costs and time costs. Enterprises need a practical system solution for automatic learning scheduling of orders.

[0003] In the prior art, the invention patent with publication number CN116109234B discloses an intelligent order allocation method for an express delivery system based on the integration of prediction and decision-making, the method comprising: constructing a prediction model, the prediction model obtains the predicted order quantity in a subsequent time period based on the historical order quantity; constructing a clustering optimization model, the clustering optimization model uses the predicted order quantity and the geographic location information of the area of ​​interest to cluster the area of ​​interest into multiple categories under the set optimization objectives and constraints; constructing an end-to-end order prediction deep learning model based on the prediction model and the clustering optimization model; training the order prediction deep learning model, wherein in the forward propagation, the clustering result is converted into a discrete solution to obtain the final clustering result, and the goal of the back propagation is to back propagate the loss function through the clustering optimization model and the prediction model, thereby improving the efficiency and accuracy of order allocation.

[0004] If the above existing technology is used in the pickup and delivery of express delivery, it cannot effectively and quickly predict orders, and cannot provide push operations for branch information. In addition, the application operation is relatively complicated, and even exceeds the time of manual input, and the application efficiency is lower. Therefore, it is necessary to find a simpler and more effective solution to replace it. Summary of the invention

[0005] The purpose of the present invention is to overcome the deficiencies in the above-mentioned prior art and provide a method and system for automatic learning and scheduling of express delivery orders. The method and system of the present invention are applied in the express delivery industry to improve the versatility and accuracy of automatic order scheduling, thereby improving service efficiency.

[0006] In order to achieve the above invention purpose, the technical solution provided by the present invention is as follows:

[0007] The present invention first provides a method for automatic learning scheduling of orders, the method comprising the following steps:

[0008] The first step is to add a pickup scheduling learning details table and a pickup scheduling learning table to the database of the local server, and create a unique index for the pickup scheduling learning details table according to province, city, district, detailed address, outlet and date;

[0009] The second step is to obtain the dispatching network point information through the order pickup information, store the obtained dispatching network point information in the pickup scheduling learning details table in the database on the local server, and accumulate the daily network point pickup volume;

[0010] The third step is to obtain each province, city, district and detailed address through scheduled task calls, so as to determine the outlets with the largest pickup volume in the past 15 days, send the called data to the RocketMQ server for transfer, and generate the final scheduling learning data from the RocketMQ server, pull the scheduling learning data and update and save it in the pickup scheduling learning table in the database on the local server;

[0011] The fourth step is to push the new order data to the Kafka server after the order is placed, and query the province, city, district, and detailed address corresponding to the new order from the Oracle database of the Kafka server. The Oracle database is matched with the pickup scheduling learning table, and the learning data stored in the Oracle database is queried in real time through the province, city, district and detailed address of the new order to obtain the corresponding pickup point. The obtained pickup point is the predicted pickup point obtained after learning, and the obtained point data is saved in the order table of the Oracle database.

[0012] In the first step, indexes are created in the database of the local server, including the pickup scheduling learning details table and the pickup scheduling learning table, and UNIQUE constraints are used to ensure that the data in one or more columns is unique and no duplicate values ​​appear.

[0013] By creating a pickup scheduling learning details table in the database, and adding a unique index based on province, city, district, detailed address, pickup point, and date to the table;

[0014] Obtain the actual pickup point information through the pickup information data;

[0015] Count the actual pickup volume data of each outlet in the system on a daily basis;

[0016] The acquired dispatching point information is stored in the pickup dispatching learning details table in the database, the daily pickup volume of the points is accumulated, the daily pickup dispatching learning details data is obtained, and it is updated to the pickup dispatching learning table to form learning data.

[0017] In the third step, the outlets with the largest pickup volume in the past 15 days in each province, city, district, and detailed address are counted every day, and the data is sent to the RocketMQ server; data is pulled from the RocketMQ server and saved in the database, and data is pulled from the RocketMQ server to compare the pickup volume in the past 15 days, and updated to the pickup scheduling learning table, thus generating the final pickup scheduling learning record.

[0018] In the fourth step, after the operator places the order, the new order data is pushed to the Kafka server through Oracle GoldenGate, and the province, city, district, and detailed address corresponding to the order are pulled from the Kafka server, matched with the data in the pickup scheduling learning table saved in the Oracle database, and the pickup point corresponding to the province, city, district, and detailed address is obtained, which is the pickup point corresponding to the new order.

[0019] The present invention also provides an order automatic learning scheduling system, which comprises a pickup information collection module, a pickup quantity statistics module, a scheduling learning module and an order automatic scheduling module connected in sequence, wherein:

[0020] The pickup information collection module is used to collect the pickup information of the order in real time, the pickup information includes the pickup outlet and the pickup clerk, and save the collected pickup information into the pickup scheduling learning details table;

[0021] The pickup volume statistics module obtains pickup information from the pickup scheduling learning details table, and counts the network point information with the largest pickup volume corresponding to each province, city, district, and detailed address in the past 15 days, and sends the statistical results to the RocketMQ server;

[0022] The scheduling learning module pulls the data in the RabbitMQ server, updates the pickup quantity data in the pickup scheduling learning table, and updates the data to the pickup scheduling learning table to form learning data;

[0023] The order automatic scheduling module pushes the order data to the Kafka server in real time through Oracle GoldenGate and queries the province, city, district and detailed address corresponding to the order. It queries the learning data stored in the pickup scheduling learning table in the database in real time through the province, city, district and detailed address of the order, thereby obtaining the pickup point corresponding to the order data.

[0024] Furthermore, a database is provided in the above-mentioned Kafka server, and the database obtains a pickup scheduling learning table from a database on a local server through Oracle GoldenGate, and the pickup scheduling learning table is updated with learning data by a scheduling learning module.

[0025] At the same time, the present invention also provides a set of automatic order learning and scheduling equipment, which includes a server side, a computer platform terminal and a handheld terminal. The computer terminal is communicatively connected to the server side, and the handheld terminal is communicatively connected to the server side. The server side stores the automatic order learning and scheduling system described in claim 6, executes the above-mentioned automatic order learning and scheduling method, and realizes the entry of pickup point data through the computer platform terminal or the handheld terminal, and then saves the obtained point data into the order table of the Oracle database.

[0026] Based on the above prior art, the method, system and equipment for automatic learning and scheduling of orders in the present invention have been applied in the express delivery industry and have achieved the following technical effects:

[0027] 1. The method of automatic learning and scheduling of orders of the present invention realizes the automatic learning scheduling function by collecting the pickup information of the salesperson, thereby realizing the automatic scheduling function of orders with high reliability, strong flexibility and high accuracy, quickly supporting business development and improving work efficiency.

[0028] 2. In the system of automatic learning and scheduling of orders of the present invention, a pickup information collection module is set up to collect the pickup information of the order in real time, including the pickup outlets and pickup clerks, and save it in the pickup scheduling learning details table. The pickup quantity statistics module is used to count the outlet information with the largest pickup quantity corresponding to each province, city, district, and detailed address in the past 15 days, and send it to the RocketMQ server. The scheduling learning module pulls the data in the above RabbitMQ server and updates the pickup quantity data in the learning table. The order automatic scheduling module pushes the order data to the Kafka server in real time through Oracle GoldenGate, and queries the province, city, district, and detailed address corresponding to the order. The province, city, district, and detailed address of the order are used to query the learning data stored in the database in real time to obtain the corresponding pickup outlet. This outlet is the predicted pickup outlet obtained after learning. The system replaces manual input operations and realizes automatic order mobilization through automatic learning, which greatly improves reliability and scheduling efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a schematic diagram of the implementation flow of a method for automatic order learning and scheduling of the present invention.

[0030] Figure 2 It is a schematic diagram of the composition architecture of a system for automatic learning and scheduling of orders according to the present invention.

[0031] Figure 3 It is a schematic diagram of the equipment composition of an automatic order learning and scheduling system of the present invention. DETAILED DESCRIPTION

[0032] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all 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.

[0033] As attached Figure 1 As shown, an order automatic learning scheduling system includes the following steps:

[0034] S1, add a pickup scheduling learning details table and a pickup scheduling learning table to the database of the local server, and create a unique index for the pickup scheduling learning details table according to province, city, district, detailed address, outlet and date; in the database of the local server, when creating indexes for the pickup scheduling learning details table and the pickup scheduling learning table, use the UNIQUE constraint to ensure that the data in one or more columns is unique and no duplicate values ​​appear. By setting the UNIQUE constraint in the database, it can be ensured that no duplicate data appears in one or more columns in the table, thereby avoiding the risk of data inconsistency and data loss.

[0035] S2, obtaining the dispatching network point information through the order pickup information, storing the obtained dispatching network point information in the pickup scheduling learning details table saved in the database on the local server, and accumulating the daily network point pickup volume.

[0036] S3, through the scheduled task, calls the data in the pickup scheduling learning details table, obtains the province, city, district and detailed address, and calculates the outlet with the largest pickup volume in the past 15 days, sends the called data to the RocketMQ server for transfer, and then generates the final scheduling learning data from the RocketMQ server, pulls the scheduling learning data and saves it to the pickup scheduling learning table in the local database. In the specific operation, it is necessary to count the outlets with the largest pickup volume in the past 15 days for each province, city, district, and detailed address every day, and send the data to the RocketMQ server; pull data from the RocketMQ server and save the learning data to the database, pull data from the RocketMQ server to compare the pickup volume in the past 15 days, and update it to the pickup scheduling learning table, that is, to generate the final pickup scheduling learning record.

[0037] The steps for generating the above pickup scheduling learning table are described as follows:

[0038] 1) Create a pickup scheduling learning details table through the database, and add a unique index based on province, city, district, detailed address, pickup point, and date to the table;

[0039] 2) Obtain the actual pickup point information through the pickup information data;

[0040] 3) Calculate the actual pickup volume data of each outlet in the system on a daily basis;

[0041] 4) The acquired dispatching network point information is stored in the pickup dispatching learning detailed table in the database, and the daily pickup volume of the network points is accumulated. Among them, the operation of storing the dispatching network point information in the database includes:

[0042] Determine whether the data for the current day exists by day:

[0043] If it does not exist, add a new data entry of province, city, district, detailed address, actual pickup point, and date, and set the pickup quantity to 1.

[0044] If it exists, the pickup quantity corresponding to the province, city, district, detailed address, actual pickup point, and date will be increased by 1.

[0045] 5) Call the data in the pickup scheduling learning details table through the scheduled task to obtain the province, city, district and detailed address, so as to calculate and determine the outlet with the largest pickup volume in the past 15 days, send the called data to the RocketMQ server for transfer, and then generate the final scheduling learning data from the RocketMQ server. The storage step includes:

[0046] Determine the outlet with the largest pickup volume in the past 15 days by province, city, district, and detailed address

[0047] The data is updated into the pickup scheduling learning table of the database, and the final pickup scheduling learning record is generated.

[0048] S4, after the business personnel place an order through the platform computer or handheld terminal, the newly added order data is pushed to the Kafka server through Oracle GoldenGate. The newly added order data is regarded as change data, and then the data is pulled from the Kafka server to query the province, city, district, and detailed address corresponding to the order. The province, city, district, and detailed address of the order are used to query the learning data stored in the database in real time to obtain the corresponding pickup outlets. The outlets are the predicted pickup outlets obtained after learning, and the obtained outlet data are pushed to the platform computer or handheld terminal. There is a network connection on the platform computer, and the platform computer is used to log in to the operation platform website to place orders and inquire, and obtain the predicted recommended outlet data.

[0049] Oracle GoldenGate is a log-based structured data replication software used to achieve real-time capture, transformation, and delivery of transaction data in a heterogeneous environment. Oracle database can achieve high success rate and high data consistency during data synchronization through GoldenGate. New data is pushed to the Kafka server through Oracle GoldenGate, and then data is pulled from the Kafka server to query the province, city, district, and detailed address corresponding to the order. Kafka was originally developed by LinkedIn and later became an open source project of the Apache Software Foundation. Its main design goal is to provide a high-throughput, persistent, distributed messaging system. The Kafka server is suitable for offline and online message consumption, such as regular message collection, website activity tracking, aggregated statistical system operation data (monitoring data), log collection, and other data collection scenarios for Internet services with large amounts of data.

[0050] Example 2

[0051] As attached Figure 2 As shown, this embodiment is an order automatic learning scheduling system, which includes a pickup information collection module, a pickup quantity statistics module, a scheduling learning module and an order automatic scheduling module. Specifically,

[0052] The pickup information collection module is used to collect the pickup information of the order in real time, including the pickup outlet and the pickup clerk, and save it in the pickup scheduling learning details table;

[0053] The pickup volume statistics module is used to count the network point information with the largest pickup volume corresponding to each province, city, district, and detailed address in the past 15 days, and generate it in the RocketMQ server;

[0054] The scheduling learning module is used to pull the data in the RabbitMQ server and update the pickup quantity data of the learning table;

[0055] The automatic order scheduling module pushes the order data to the Kafka server in real time through Oracle GoldenGate, and queries the province, city, district, and detailed address corresponding to the order. It queries the learning data stored in the database in real time through the province, city, district, and detailed address of the order to obtain the corresponding pickup point.

[0056] Example 3

[0057] like Figure 3As shown, this embodiment is a set of automatic order learning and scheduling equipment, which includes a server side 1, a computer platform terminal 2 and a handheld terminal 3. The computer platform terminal 2 is communicatively connected to the server side 1, and the handheld terminal 3 is communicatively connected to the server side 1. The above-mentioned automatic order learning and scheduling system is stored in the server side 1. The scheduling equipment executes the above-mentioned automatic order learning and scheduling method, and realizes data input and entry and storage of pickup point data through the computer platform terminal 2 or the handheld terminal 3, that is, completes the automatic order learning and auxiliary utilization, and then saves the obtained point data into the order table of the Oracle database, and the order table data is used for data utilization and processing in subsequent links.

[0058] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for automatic learning and scheduling of orders, characterized in that: The method comprises the following steps: The first step is to add a pickup scheduling learning details table and a pickup scheduling learning table to the database of the local server, and create a unique index for the pickup scheduling learning details table according to province, city, district, detailed address, outlet and date; The second step is to obtain the dispatching network point information through the order pickup information, store the obtained dispatching network point information in the pickup scheduling learning details table in the database on the local server, and accumulate the daily network point pickup volume; The third step is to obtain each province, city, district and detailed address through scheduled task calls, so as to determine the outlets with the largest pickup volume in the past 15 days, send the called data to the RocketMQ server for transfer, and generate the final scheduling learning data from the RocketMQ server, pull the scheduling learning data and update and save it in the pickup scheduling learning table in the database on the local server; The fourth step is to push the new order data to the Kafka server after the order is placed, and pull the province, city, district, and detailed address corresponding to the order data from the Kafka server to match the pickup scheduling learning table saved in the Oracle database. The province, city, district, and detailed address of the new order can be used to query the learning data saved in the Oracle database in real time to obtain the corresponding pickup point. The obtained pickup point is the predicted pickup point obtained after learning, and the obtained point data is saved in the order table of the Oracle database.

2. The method and system for automatic order learning and scheduling according to claim 1, characterized in that: In the first step, indexes are created in the database of the local server, including the pickup scheduling learning details table and the pickup scheduling learning table, and UNIQUE constraints are used to ensure that the data in one or more columns is unique and no duplicate values ​​appear.

3. The method for automatic learning and scheduling of orders according to claim 1, characterized in that: The step of generating the pickup scheduling learning table includes: By creating a pickup scheduling learning details table in the database, and adding a unique index based on province, city, district, detailed address, pickup point, and date to the table; Obtain the actual pickup point information through the pickup information data; Count the actual pickup volume data of each outlet in the system on a daily basis; The acquired dispatching network point information is stored in the pickup dispatching learning detail table in the database, and the daily pickup volume of the network points is accumulated to obtain the daily pickup dispatching learning detail data.

4. The method for automatic learning and scheduling of orders according to claim 1, characterized in that: In the third step, by counting the outlets with the largest pickup volume in the past 15 days for each province, city, district, and detailed address in the pickup scheduling learning details table every day, the province, city, district, detailed address and corresponding pickup outlet information data are sent to the RocketMQ server; data is pulled from the RocketMQ server to compare the pickup volume in the past 15 days, and updated to the pickup scheduling learning table, thus generating the final pickup scheduling learning record.

5. The method for automatic learning and scheduling of orders according to claim 1 is characterized by: In the fourth step, after the operator places the order, the new order data is pushed to the Kafka server through Oracle GoldenGate, and the province, city, district, and detailed address corresponding to the order are pulled from the Kafka server, matched with the data in the pickup scheduling learning table saved in the Oracle database, and the pickup point corresponding to the province, city, district, and detailed address is obtained, which is the pickup point corresponding to the new order.

6. An automatic order learning and scheduling system, characterized in that: The system consists of a pickup information collection module, a pickup quantity statistics module, a scheduling learning module and an order automatic scheduling module which are connected in sequence, wherein: The pickup information collection module is used to collect the pickup information of the order in real time, the pickup information includes the pickup outlet and the pickup clerk, and save the collected pickup information into the pickup scheduling learning details table; The pickup volume statistics module obtains pickup information from the pickup scheduling learning details table, and counts the network point information with the largest pickup volume corresponding to each province, city, district, and detailed address in the past 15 days, and sends the statistical results to the RocketMQ server; The scheduling learning module pulls the data in the RabbitMQ server, updates the pickup quantity data in the pickup scheduling learning table, and updates the data to the pickup scheduling learning table to form learning data; The order automatic scheduling module pushes the order data to the Kafka server in real time through Oracle GoldenGate, and queries the learning data stored in the pickup scheduling learning table in the Oracle database in real time through the order province, city, district, and detailed address in the Kafka server, thereby obtaining the pickup point corresponding to the order data.

7. The automatic order learning and scheduling system according to claim 6 is characterized in that: In the order automatic scheduling module, the Kafka server has complete order data, and automatic scheduling is achieved by obtaining the order province, city, district, and detailed address from the Kafka server to match the pickup scheduling learning table data in the Oracle database.

8. An automatic order learning and scheduling device, characterized in that: The device includes a server side, a computer platform terminal and a handheld terminal. The computer terminal is communicatively connected to the server side, and the handheld terminal is communicatively connected to the server side. The server side stores the order automatic learning and scheduling system described in claim 6, executes the order automatic learning and scheduling method described in claims 1-5, and realizes the entry of pickup point data through the computer platform terminal or the handheld terminal, and then saves the obtained point data into the order table of the Oracle database.

Images