CN118690933A - A dynamic route scheduling method based on data-driven and intelligent optimization - Google Patents

Abstract

The present invention relates to a dynamic route scheduling method based on data-driven and intelligent optimization, belonging to the field of data processing technology, including: obtaining information of stores and store visiting employees, performing data integration and processing, generating structured store information lists and employee information lists; evaluating the distance between stores and the length of time on the way between stores under specified transportation; setting employees, stores and employee initial positions as decision variables, constructing an NP-hard problem model that meets the route planning requirements, using the OptaPlanner calculation engine, obtaining the distribution of all route stores assigned to each employee within a set time period and the order of visiting stores on a single route, forming the employee's visit route; generating a visit plan after the employee obtains the assigned visit route; performing real-time monitoring and road condition prediction processing when the employee executes the visit plan, and adjusting the visit plan. The present invention realizes dynamic and flexible route scheduling.

Description

A dynamic route scheduling method based on data-driven and intelligent optimization

Technical Field

The present invention relates to the field of data processing technology, and in particular to a dynamic route scheduling method based on data drive and intelligent optimization.

Background Art

With the rapid development of information technology, route scheduling technology has shown great value in many business scenarios, especially in the fields of logistics distribution, sales visits, service scheduling, etc. However, existing technologies still have certain limitations when facing real-time traffic changes, resource optimization and allocation, and other issues. In the traditional model, for business processes that require field visits or maintenance of multiple stores, staff are often first assigned a series of stores, and then manually integrate these stores into different visit routes one by one. Under this mode of operation, the number of stores that each route should include and their specific combinations are not pre-set, resulting in inefficient resource allocation. Especially in multinational companies with a large staff team and widely distributed stores, the route allocation task is not only time-consuming, but also difficult to ensure the effectiveness and rationality of the allocation plan.

If there is no scientific route planning and employees are left to decide the order of visits, a series of new problems may arise. For example, it is difficult for employees to judge whether the selected route is efficient and whether they can successfully complete the task of visiting all stores within the specified time. At the same time, the company's management cannot achieve refined management of employee visits and it is difficult to accurately assess how many stores each employee is most suitable to visit, thus limiting the improvement of overall work efficiency and effectiveness.

Summary of the invention

In view of the above analysis, the present invention aims to disclose a dynamic route scheduling method based on data-driven and intelligent optimization; to overcome the inefficiency and imbalance of traditional manual scheduling methods, and to solve the challenge of designing multiple overall optimal routes for employees. At the same time, it can also deeply explore and optimize the number of stores allocated to employees to ensure that their work saturation is at a reasonable level.

The present invention discloses a dynamic route scheduling method based on data drive and intelligent optimization, comprising:

Step S1, obtaining information of stores and store visiting employees, integrating and processing the data, and generating a structured store information list and employee information list;

Step S2: Based on the store information list and employee information list, the distance between stores and the travel time between stores under the specified transportation are evaluated; employees, stores and employee initial positions are set as decision variables, an NP-hard problem model that meets the route planning requirements is constructed, and the OptaPlanner calculation engine is used to obtain the distribution of all route stores assigned to each employee within the set time period and the order of visiting stores on a single route, so as to form the employee's visiting route;

Step S3: After the employee obtains the assigned visiting route, he/she generates a visiting plan based on factors including the weights of the stores involved in the visiting route, the visiting frequency, and the working hours;

Step S4: When the employee executes the visiting plan, real-time monitoring and traffic condition prediction processing including location tracking and traffic condition monitoring are performed to adjust the visiting plan.

Furthermore, step S1 includes:

Step S101, extracting information of stores and store visiting employees from multiple different data sources to form a data warehouse;

Step S102: verify, clean and standardize the data in the data warehouse to obtain standardized data;

Step S103: Fill the standardized data into the store information list and the employee information list according to the formatted content of the table.

Furthermore, the store table structure includes: store ID, user/organization ID, user/organization code, user/organization name, store code, store name, store address, city, store level, calculation times/monthly visit times, store time, store latitude and longitude, creation time and tenant ID items;

The employee table structure includes: employee ID, user/organization ID, user/organization code, user/organization name, monthly working days, daily working hours, average travel time, whether to enable virtual departure point, latitude and longitude of at least one departure point, and tenant ID items.

Furthermore, in step S2, the OSRM+OptaPlanner method is used to plan the visit route;

Among them, the map route planning service OSRM is used to calculate the distance between all stores or to evaluate the travel time between stores under specified transportation tools;

Employees, stores, and their initial locations are set as decision variables, and the OptaPlanner engine is used to calculate the NP-hard problem model that meets the route planning requirements; the visit route is planned.

Furthermore, the hard constraints set on the OptaPlanner engine include:

a) Fixed visit frequency standards for stores;

b) Daily working hours limit;

c) Setting the time of staying in the store;

The soft constraints set on the OptaPlanner engine include:

a) The number of routes that employees should perform each month should not be less than the set number;

b) Each planned route should cover at least the set number of stores;

c) Minimize the total travel distance or accumulated travel time.

Furthermore, using the VRP model of the OptaPlanner calculation engine, employees, stores, and employee initial positions are respectively mapped to vehicles, stores, and warehouses in the VRP model; employees’ daily working hours correspond to the transportation capacity of the vehicles, and employees’ time in stores corresponds to the store’s demand for goods.

According to the set soft and hard constraints, the VRP model's solution algorithm is used to calculate the distribution of all route stores assigned to an employee and the order of store visits on a single route; then the store route scheduling of all employees is calculated through a recursive algorithm.

Furthermore, employees can obtain assigned visit routes through the employee task platform; set up store area maps and employee visit route display areas and visit plan editing areas on the display screen of the employee task platform;

Through the visit plan editing area, the date and time of the visit route and the stores on the route can be adjusted, the visit time can be rearranged, stores between routes can be moved and replaced, and the order of store visits within the route can be modified; and the planned employee route can be displayed in the store area map and the employee visit route display area.

Furthermore, the step S4 comprises:

Step S401: Real-time monitoring of employee location tracking, road condition monitoring, task progress feedback, store visit verification, work efficiency evaluation, and emergency response;

Step S402: predicting the road conditions based on the real-time monitored emergencies to determine the optimal driving route;

Step S403: Adjust and optimize the visit plan according to the set adjustment strategy.

Furthermore, the road condition prediction adopts a machine learning prediction model; the model captures massive traffic data streams in real time, based on historical data, real-time monitoring data and other relevant environmental factors, uses deep learning and time series analysis to simulate and predict road traffic conditions within a certain period of time in the future.

Furthermore, the step S403 includes:

1) Data collection and analysis: Real-time collection of key internal and external variables, including on-site conditions reported by employees, traffic data, changes in customer appointments, and adjustments to store business hours, to analyze the impact of these changes on the estimated arrival time and task priority of the original visit plan;

2) Set up a triggering mechanism to adjust the visit plan: When changes in key variables monitored trigger the visit plan adjustment mechanism, the route is replanned to formulate a new visit plan;

3) Re-plan the route based on real-time traffic conditions: Use OSRM to re-plan the optimal visit route;

4) Re-prioritize visit tasks: Re-prioritize visit tasks based on multiple criteria such as urgency, customer importance, business needs, etc.

5) Communicate and notify the adjusted visit plan: automatically push the adjusted visit plan to the affected employees or send it to a third-party service; and notify the customer at the same time;

6) Continuous monitoring and optimization: During the execution of the new plan, various dynamic information is continuously monitored to ensure that the plan is always consistent with the actual situation. If there is another unexpected situation, the system will immediately start a new round of adjustment and optimization cycle.

The present invention can achieve one of the following beneficial effects:

1. It realizes dynamic and flexible route scheduling, fully considers real-time factors, and effectively reduces the impact of factors such as road congestion and emergencies on the planned route.

2. The introduction of machine learning prediction improves the predictability and robustness of route planning, making the planning scheme more in line with actual environmental changes.

3. Quantitatively evaluate and reasonably allocate employees’ work saturation to ensure efficient use of human resources while avoiding excessive work pressure.

BRIEF DESCRIPTION OF THE DRAWINGS

The drawings are only for the purpose of illustrating particular embodiments and are not to be considered limiting of the present invention. Like reference symbols denote like components throughout the drawings.

FIG1 is a flow chart of a dynamic route scheduling method based on data-driven and intelligent optimization in an embodiment of the present invention;

FIG2 is a summary diagram of route results in an embodiment of the present invention;

FIG3 is a schematic diagram of an employee route list in an embodiment of the present invention;

FIG4 is a schematic diagram of a list of stores on a certain route of an employee in an embodiment of the present invention;

FIG5 is a schematic diagram of a store map distribution on a certain route of an employee in an embodiment of the present invention;

FIG6 is a schematic diagram of a list of stores on a certain route of an employee in an embodiment of the present invention;

FIG. 7 is a display screen diagram of the employee task platform in an embodiment of the present invention.

DETAILED DESCRIPTION

The preferred embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used to illustrate the principles of the present invention together with the embodiments of the present invention.

An embodiment of the present invention discloses a dynamic route scheduling method based on data-driven and intelligent optimization, as shown in FIG1 , comprising the following steps:

Step S1, obtaining information of stores and store visiting employees, integrating and processing the data, and generating a structured store information list and employee information list;

Step S2: Based on the store information list and employee information list, the distance between stores and the travel time between stores under the specified transportation are evaluated; employees, stores and employee initial positions are set as decision variables, an NP-hard problem model that meets the route planning requirements is constructed, and the OptaPlanner calculation engine is used to obtain the distribution of all route stores assigned to each employee within the set time period and the order of visiting stores on a single route, so as to form the employee's visiting route;

Step S3: After the employee obtains the assigned visiting route, he/she generates a visiting plan based on factors including the weights of the stores involved in the visiting route, the visiting frequency, and the working hours;

Step S4: When the employee executes the visiting plan, real-time monitoring and traffic condition prediction processing including location tracking and traffic condition monitoring are performed to adjust the visiting plan.

Specifically, step S1 includes:

Step S101, extracting information of stores and store visiting employees from multiple different data sources to form a data warehouse;

Specifically, data is extracted through the data interface, database or data warehouse opened by the upstream system that is directly connected, or the data interface that meets the route scheduling protocol provided for route scheduling, as well as the real-time traffic data or interface provided by a third party, and the business-related parameters required for route scheduling, including store visit frequency, employee daily working hours, and employee monthly working days, are calculated to form a data warehouse.

For example, basic store information including store location, importance level, and historical sales data can be captured from SFA (sales automation) and CRM (customer relationship management) systems;

Connect to real-time traffic data providers through API to obtain real-time traffic information;

In more selective solutions, other relevant data sources such as weather forecasts are also connected.

In order to ensure the consistency and integrity of multiple different data sources, the following processing is performed when extracting data:

1) Transaction processing: When extracting data from different data sources, encapsulate the data extraction process in a transaction to ensure the consistency of the data source before and after the extraction. If the extraction fails or a problem occurs, the transaction can be rolled back to the original state to avoid leaving inconsistent data traces.

2) Three-Phase Commit (3PC): For cross-system or cross-database data extraction, a distributed transaction processing method is adopted, in which the distributed transaction processing adopts a three-phase commit protocol to ensure that data change operations between multiple data sources can either succeed or fail completely to achieve data consistency.

3) Message queue middleware: The microservice version of route scheduling uses the message queue middleware Kafka as a buffer and intermediary. The data source first publishes the change information to the message queue, and the data center subscribes and consumes the message, and processes the message sequentially to ensure that the data arrives in order and is not missed.

4) Change Data Capture (CDC): Supports the use of CDC technology to capture change records of data sources (such as binlog and redo log of MySQL database), and then synchronizes these changes to the data center incrementally to ensure that the extracted data is always up-to-date and complete.

5) Primary key/unique constraints and indexes: At the data storage level, establish primary key constraints, unique constraints, and appropriate indexes to prevent inconsistencies caused by repeated insertion or update of data.

6) Idempotent design: The data extraction service is designed as an idempotent operation, that is, no matter how many times it is executed, as long as the input is the same, the result is the same. In this way, even if repeated extraction occurs due to network fluctuations, it will not affect the consistency of the final data.

Through the above processing, the data required for route scheduling is extracted into its own data warehouse;

Step S102: verify, clean and standardize the data in the data warehouse to obtain standardized data;

During the data loading phase after extraction, strict data verification rules are set to perform integrity and consistency checks on the extracted data. Data that does not comply with the rules will not be loaded, and abnormalities will be reported. The quality of incoming data is ensured through the data cleaning process, and data non-standardization issues are resolved through standardized processing.

Specific calibration, cleaning and standardization processes include:

1) Perform data verification including format verification, integrity verification, consistency verification, reference integrity verification and rationality verification:

in,

Format verification: Check whether the data conforms to the expected format, such as whether the date and month, latitude and longitude values, number of visits, etc. conform to the standard format. For numeric data, check whether its digital format is correct and whether there are any non-numeric characters mixed in.

Completeness check: Check whether there are missing values in the fields of the data set, determine the degree of missing values by calculating the missing value ratio, and formulate a reasonable processing strategy. Verify whether key fields are complete, such as whether necessary information such as "store latitude and longitude" and "employee departure point" is filled in.

Consistency check: Check logical contradictions in data, such as "daily working hours" should not exceed 24 hours, "monthly working days" should not exceed the number of natural days in the month, etc. Compare related data tables to ensure that foreign key constraints are met and there is no invalid data referenced. For example, employees must be assigned to stores, and the employee table information corresponding to the store table cannot be missing.

Referential integrity check: Check whether the data violates referential integrity, such as whether the "store type" enumeration value exists in the predefined enumeration list.

Reasonableness check: Verify the rationality of data based on business rules or domain knowledge, such as the longitude value of the longitude and latitude values should be in the range of 0-180, and the latitude value should be in the range of 0-90.

2) Perform data cleaning including processing of missing values, outliers and duplicate values;

in,

Handling missing values: Delete records with too many missing values and record the deleted data for later manual proofreading. Sometimes it is also necessary to fill in missing values. For example, if the latitude and longitude of the departure point of some employees are missing, an intermediate location will be calculated as a virtual departure point based on the latitude and longitude of the store they are assigned during preprocessing.

Handling outliers: Define thresholds to remove obviously abnormal data points, and use box plots to identify outliers and remove stores distributed across cities.

Handling duplicate values: Find duplicate records by unique key and decide whether to keep one copy or merge the records. For example, store data uses store code as the unique key.

3) Perform data conversion and standardization;

Convert data types, such as converting strings to dates or numbers. Standardize or normalize numerical features so that data can be easily compared and analyzed at the same scale. For example, if the longitude and latitude are using coordinate systems from different map companies, they need to be unified into the same coordinate coefficient value.

4) Clear irrelevant or erroneous records;

It is also necessary to clear irrelevant or erroneous records, including incorrectly entered test data and illegal characters.

Step S103: Fill the standardized data into the store information list and the employee information list according to the formatted content of the table.

The store table structure includes: store ID, user/organization ID, user/organization code, user/organization name, store code, store name, store address, city, store level, calculation times/monthly visits, store time, store latitude and longitude, creation time, and tenant ID items, as shown in the following table:

Column Name type describe id bigint ID org_id bigint User/Organization ID org_code varchar(50) User/Organization Code org_name varchat(200) User/Organization Name cust_code varchar(50) Store code cust_name varchar(50) Store Name addr varchar(200) Store Address city varchar(50) City cust_level varchar(5) Store level calc_count int Calculation times/monthly visits dwell_time int Length of time in store lon varchar(50) Store longitude lat varchar(50) Store Latitude created datetime Creation time tenant_id bigint Tenant ID

The employee table structure includes: employee ID, user/organization ID, user/organization code, user/organization name, monthly working days, daily working hours (unit: minutes), average travel time, whether to enable virtual departure point, latitude and longitude of at least one departure point, and tenant ID items; the details are as follows

Specifically, in step S2, the visit route is planned by using the map route planning service OSRM (Open Source Routing Machine) + OptaPlanner computing engine;

The map route planning service OSRM is used to calculate the distance between all stores or to estimate the travel time between stores under specified means of transportation. Employees, stores and their initial positions are set as decision variables to build an NP-hard (Non-deterministic Polynomial hard) problem model that meets the route planning requirements. The OptaPlanner computing engine is used to plan the visit route.

Specifically, the OptaPlanner computing engine uses heuristic optimization algorithms such as improved genetic algorithm, simulated annealing algorithm, hill climbing pedometer algorithm, flood algorithm, taboo search algorithm, delayed reception algorithm, variable neighborhood descent algorithm, etc., combined with path search algorithms such as Dijkstra algorithm and A* algorithm to build a multi-objective optimization model to solve the route planning problem.

More specifically, factors such as travel time, work saturation, and total route length are fully considered in the model to ensure that the planning results meet actual business needs.

Provide all the basic information needed for calculation, such as store information (including travel time between stores, store in-store time, coordinate longitude and latitude, number of monthly repeat visits to stores, etc.), employee information, employee departure point information, number of routes to be calculated, daily working hours, etc., to the OptaPlanner engine and start the calculation service; plan multiple different visit routes for employees every day for a whole month at one time. For example, in a month with 20 working days, calculate a total of 20 reasonable travel routes for each employee.

In order to formulate planning rules that meet business needs in a targeted manner and generate the optimal route plan that not only meets customer needs but also conforms to actual operations, this embodiment sets soft and hard constraints on the OptaPlanner engine;

Among them, hard constraints are the basic rules that are indispensable and must be strictly followed in route scheduling calculations, and soft constraints are the rules for achieving ideal goals as much as possible under the premise of fully satisfying all hard constraints;

Specific hard constraints include:

a) Fixed visit frequency standards for stores;

For example, a Class A store needs to be visited four times a month, while a Class B store needs to be visited twice.

b) Daily working hours limit;

Ensure that employees' daily working hours do not exceed the legal limit of 8 hours.

c) Setting the time of staying in the store;

For example, the requirement for each visit to a Class A store is 28 minutes, while for a Class B store it is 22 minutes.

Specific soft constraints include:

a) A certain type of employee should not perform less than a set number of routes per month; for example, 15.

b) Each planned route should cover at least a set number of stores; for example, 8 stores.

c) Minimize the total travel mileage or accumulated travel time to minimize the overall route length.

It is worth noting that the soft constraints will be adjusted flexibly as the number of stores an employee is responsible for changes. If an employee is only responsible for one store, some of the above soft constraints may be moderately relaxed in certain circumstances to ensure the rationality and feasibility of the overall solution.

Preferably, the VRP model of the OptaPlanner calculation engine is used to correspond employees, stores, and employee initial positions to vehicles, stores, and warehouses in the VRP model, respectively; the employee's daily working hours correspond to the vehicle's transportation capacity, and the employee's in-store time corresponds to the store's demand for goods.

According to the set soft and hard constraints, the VRP model's solution algorithm is used to calculate the distribution of all route stores assigned to an employee and the order of store visits on a single route; then the store route scheduling of all employees is calculated through a recursive algorithm.

In a specific simulation environment, the route results calculated using the VRP model are summarized in Figure 2; the employee route list is shown in Figure 3; the store list on a certain employee route is shown in Figure 4; the store map distribution on a certain employee route is shown in Figure 5, and the store list on a certain employee route is shown in Figure 6.

Specifically, in step S3, the employee can obtain the assigned visit route through the employee task platform;

On the employee task platform, a visit plan is generated based on factors including the weight of the stores involved in the visit route, visit frequency, and working hours;

Visit planning mainly involves arranging the date and time for employees to use visit routes and manually fine-tuning the route stores.

Preferably, a store area map and an employee visit route display area and a visit plan editing area are set on the display screen of the employee task platform;

Through the visit plan editing area, you can adjust the date and time of the visit route and the route stores, reschedule the visit time, move and replace stores between routes, and modify the order of store visits within the route. The planned employee route is displayed in the store area map and employee visit route display area.

As shown in Figure 7, this is the display screen of the employee task platform. In the figure, the left side is the store area map and the employee visit route display area; the area map is displayed and the employee visit route is superimposed on the map; the right side is the visit plan editing area, which includes input controls, drop-down selection controls, etc. By modifying or selecting the control content, the visit time can be rearranged, stores between routes can be moved and replaced, and the order of store visits within the route can be modified to adjust the visit plan.

Specifically, in step S4, it includes:

Step S401: Real-time monitoring of employee location tracking, road condition monitoring, task progress feedback, store visit verification, work efficiency evaluation, and emergency response;

in,

1) Employee location tracking: Using mobile communication device positioning technology and mobile applications, employees’ current location information is obtained in real time, compared with the planned visit route, and monitored to see whether the employees follow the established route.

2) Traffic condition monitoring: Integrate real-time traffic information from map service providers to monitor traffic congestion, accidents, construction and other situations that affect travel on the road. When encountering emergencies, employees can be reminded to adjust routes or re-plan in a timely manner.

3) Task progress feedback: After completing a store visit, employees can submit visit results through the mobile application, including completed task items, time spent, performance achieved and other information. The system receives and updates the task completion status in real time.

4) Store visit verification: Through the store’s sign-in and sign-out, verify whether the employee has arrived on time and completed the store visit task.

5) Work efficiency evaluation: Monitor performance indicators such as the number of visits actually completed by employees each day, average stay time, customer satisfaction, etc. to evaluate employee work efficiency and effectiveness.

6) Emergency response: When emergencies occur, such as employees being unable to complete tasks on time or customer stores requiring emergency visits, the system can quickly receive relevant feedback and assist in re-planning task priorities and routes.

The above real-time monitoring function can not only help managers fully understand the team's execution status, but also dynamically adjust plans according to actual conditions to improve the team's overall operational efficiency and service quality. At the same time, it also ensures that employees can get effective support and guidance when facing unforeseen road conditions and work changes.

Step S402: predicting the road conditions based on the real-time monitored emergencies to determine the optimal driving route;

Specifically, the road condition prediction adopts a machine learning prediction model; the model captures massive traffic data streams in real time, based on historical data, real-time monitoring data and other relevant environmental factors, uses deep learning and time series analysis to simulate and predict road traffic conditions within a certain period of time in the future.

Once congestion, accidents, etc. that may affect the efficiency of existing routes are predicted, the system can respond quickly, re-evaluate and automatically adjust the optimal driving route based on the prediction results, ensuring the smoothness and efficiency of the entire delivery or service process.

Step S403: Adjust and optimize the visit plan according to the set adjustment strategy.

Specifically, they include:

1) Data collection and analysis: Real-time collection of key internal and external variables, including on-site conditions reported by employees, traffic data, changes in customer appointments, and adjustments to store business hours, to analyze the impact of these changes on the estimated arrival time and task priority of the original visit plan;

2) Set up a triggering mechanism to adjust the visit plan: When changes in key variables monitored trigger the visit plan adjustment mechanism, the route is replanned to formulate a new visit plan;

If a certain road section is severely congested and it is impossible to reach the next store on time, the system will automatically trigger the adjustment mechanism. According to the preset rules and weights, it is determined whether the visit plan needs to be re-optimized immediately or whether it can be adjusted after a certain threshold is reached.

3) Re-plan the route based on real-time traffic conditions: Use OSRM to re-plan the optimal visit route;

Use OSRM route planning to re-plan the optimal visit route based on real-time traffic conditions, try to avoid blocked sections, and shorten the total driving time. For stores that cannot be visited for some reason, the system will find alternatives, such as replacing them with other visit plan routes and postponing the visit to the next available time period.

4) Re-prioritize visit tasks: Re-prioritize visit tasks based on multiple criteria such as urgency, customer importance, business needs, etc.

If there are new stores with urgent tasks, they will be inserted into the appropriate positions in the existing plan to ensure that high-priority tasks are completed first.

5) Communicate and notify the adjusted visit plan: automatically push the adjusted visit plan to the affected employees or send it to a third-party service; and notify the customer at the same time;

Third-party services include SFA or CRM to ensure they are aware of new routes and schedules. At the same time, the system can also send notifications to customers to inform them of changes in visiting times to maintain a good customer service experience.

6) Continuous monitoring and optimization: During the execution of the new plan, the system will continue to monitor various dynamic information to ensure that the plan is always consistent with the actual situation. If there is another change, the system will immediately start a new round of adjustment and optimization cycle.

In a more preferred solution, an employee mobile app is set up;

The employee mobile application is used to display the latest visit task list and specific routes; allow employees to record visits, upload on-site photos, and fill out visit reports; and provide feedback on real-time geographic location and visit progress to facilitate real-time adjustment and scheduling.

In summary, the data-driven and intelligently optimized dynamic route scheduling method of this embodiment realizes dynamic and flexible route scheduling, fully considers real-time factors, and effectively reduces the impact of factors such as road congestion and emergencies on the planned route. The introduction of machine learning prediction improves the predictability and robustness of route planning, making the planning scheme more in line with actual environmental changes. Quantitative evaluation and reasonable allocation of employee work saturation ensures efficient use of human resources while avoiding excessive work pressure.

The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.

Claims (10)

1. A dynamic route scheduling method based on data-driven and intelligent optimization, characterized by comprising: Step S1, obtaining information of stores and store visiting employees, integrating and processing the data, and generating a structured store information list and employee information list; Step S2: Based on the store information list and employee information list, the distance between stores and the travel time between stores under the specified transportation are evaluated; employees, stores and employee initial positions are set as decision variables, an NP-hard problem model that meets the route planning requirements is constructed, and the OptaPlanner calculation engine is used to obtain the distribution of all route stores assigned to each employee within the set time period and the order of visiting stores on a single route, so as to form the employee's visiting route; Step S3: After the employee obtains the assigned visiting route, he/she generates a visiting plan based on factors including the weights of the stores involved in the visiting route, the visiting frequency, and the working hours; Step S4: When the employee executes the visiting plan, real-time monitoring and traffic condition prediction processing including location tracking and traffic condition monitoring are performed to adjust the visiting plan. 2. The dynamic route scheduling method based on data-driven and intelligent optimization according to claim 1, characterized in that: Step S1 includes: Step S101, extracting information of stores and store visiting employees from multiple different data sources to form a data warehouse; Step S102: verify, clean and standardize the data in the data warehouse to obtain standardized data; Step S103: Fill the standardized data into the store information list and the employee information list according to the formatted content of the table. 3. The dynamic route scheduling method based on data-driven and intelligent optimization according to claim 2 is characterized in that: The store table structure includes: store ID, user/organization ID, user/organization code, user/organization name, store code, store name, store address, city, store level, calculation times/monthly visits, store time, store latitude and longitude, creation time, and tenant ID items; The employee table structure includes: employee ID, user/organization ID, user/organization code, user/organization name, monthly working days, daily working hours, average travel time, whether to enable virtual departure point, latitude and longitude of at least one departure point, and tenant ID items. 4. The dynamic route scheduling method based on data-driven and intelligent optimization according to claim 2, characterized in that: In step S2, the OSRM+OptaPlanner method is used to plan the visit route; Among them, the map route planning service OSRM is used to calculate the distance between all stores or to evaluate the travel time between stores under specified transportation tools; Employees, stores, and their initial locations are set as decision variables, and the OptaPlanner engine is used to calculate the NP-hard problem model that meets the route planning requirements; the visit route is planned. 5. The dynamic route scheduling method based on data-driven and intelligent optimization according to claim 4 is characterized in that: The hard constraints set on the OptaPlanner engine include: a) Fixed visit frequency standards for stores; b) Daily working hours limit; c) Setting the time of staying in the store; The soft constraints set on the OptaPlanner engine include: a) The number of routes that employees should perform each month should not be less than the set number; b) Each planned route should cover at least the set number of stores; c) Minimize the total travel distance or accumulated travel time. 6. The dynamic route scheduling method based on data-driven and intelligent optimization according to claim 5, characterized in that: Using the VRP model of the OptaPlanner calculation engine, employees, stores, and their initial positions correspond to vehicles, stores, and warehouses in the VRP model respectively; employees’ daily working hours correspond to the transportation capacity of vehicles, and employees’ time in stores corresponds to the demand for goods in stores; According to the set soft and hard constraints, the VRP model's solution algorithm is used to calculate the distribution of all route stores assigned to an employee and the order of store visits on a single route; then the store route scheduling of all employees is calculated through a recursive algorithm. 7. The dynamic route scheduling method based on data-driven and intelligent optimization according to claim 5, characterized in that: Employees can obtain assigned visit routes through the employee task platform; set up store area maps and employee visit route display areas and visit plan editing areas on the display screen of the employee task platform; Through the visit plan editing area, the date and time of the visit route and the stores on the route can be adjusted, the visit time can be rearranged, stores between routes can be moved and replaced, and the order of store visits within the route can be modified; and the planned employee route can be displayed in the store area map and the employee visit route display area. 8. The dynamic route scheduling method based on data-driven and intelligent optimization according to claim 5, characterized in that: The step S4 comprises: Step S401: Real-time monitoring of employee location tracking, road condition monitoring, task progress feedback, store visit verification, work efficiency evaluation, and emergency response; Step S402: predicting the road conditions based on the real-time monitored emergencies to determine the optimal driving route; Step S403: Adjust and optimize the visit plan according to the set adjustment strategy. 9. The dynamic route scheduling method based on data-driven and intelligent optimization according to claim 5, characterized in that: Traffic condition prediction uses a machine learning prediction model; the model captures massive amounts of traffic data streams in real time, based on historical data, real-time monitoring data and other relevant environmental factors, uses deep learning and time series analysis to simulate and predict road traffic conditions within a certain period of time in the future. 10. The dynamic route scheduling method based on data-driven and intelligent optimization according to claim 9, characterized in that: The step S403 includes: 1) Data collection and analysis: Real-time collection of key internal and external variables, including on-site conditions reported by employees, traffic data, changes in customer appointments, and adjustments to store business hours, to analyze the impact of these changes on the estimated arrival time and task priority of the original visit plan; 2) Set up a triggering mechanism to adjust the visit plan: When the changes in the monitored key variables trigger the adjustment mechanism of the visit plan, the route is replanned to formulate a new visit plan; 3) Re-plan the route based on real-time traffic conditions: Use OSRM to re-plan the optimal visit route; 4) Re-prioritize visit tasks: Re-prioritize visit tasks based on multiple criteria such as urgency, customer importance, business needs, etc. 5) Communicate and notify the adjusted visit plan: automatically push the adjusted visit plan to the affected employees or send it to a third-party service; and notify the customer at the same time; 6) Continuous monitoring and optimization: During the implementation of the new plan, various dynamic information will continue to be monitored to ensure that the plan is always consistent with the actual situation; if any unexpected events occur again, a new round of adjustment and optimization cycle will be initiated.