Logistics management system based on intelligent scheduling

Abstract

The invention relates to the technical field of logistics transportation scheduling, in particular to a logistics management system based on intelligent scheduling, which comprises a data acquisition module, a demand analysis module, an intelligent scheduling module, a path planning module, a monitoring and feedback module, an optimization module and the like. The data acquisition module is used for collecting information such as goods, vehicles, storage, customer demands and traffic road conditions; the demand analysis module predicts logistics demands by using a time sequence analysis or machine learning algorithm; the intelligent scheduling module adopts a greedy and dynamic programming mixed algorithm to allocate resources; the path planning module calculates an optimal path according to road conditions and starting and ending points; the monitoring and feedback module monitors logistics parameters in real time and feeds back abnormity; and the optimization module optimizes the scheduling scheme according to the fitness function based on the genetic algorithm. According to the invention, data and resources of all links of logistics are effectively integrated, accurate demand prediction, efficient resource configuration, transportation path optimization, real-time monitoring feedback and overall performance improvement of the system are realized, the logistics cost is significantly reduced, and the transportation efficiency and the service quality are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of logistics transportation scheduling, and in particular to a logistics management system based on intelligent scheduling. Background Art

[0002] As the logistics industry is booming, traditional logistics management methods are gradually unable to cope with the growing business complexity and high customer demands. With the rise of e-commerce, the volume of logistics orders has exploded, the variety of goods has increased, and the distribution range has continued to expand, which has brought huge challenges to logistics transportation scheduling.

[0003] Traditional logistics scheduling often relies on manual experience, making it difficult to quickly and accurately analyze massive amounts of data, resulting in irrational resource allocation. For example, in the arrangement of transport vehicles, factors such as vehicle location, load, transport time and traffic conditions may not be comprehensively considered, resulting in long empty mileage and low transport efficiency; in terms of warehouse management, inventory layout cannot be optimized according to cargo demand forecasts, resulting in low utilization of warehouse space, frequent backlogs or out-of-stock phenomena.

[0004] At the same time, customers have higher and higher expectations for the quality of logistics services, requiring goods to be delivered on time and the transportation process to be traceable. However, the traditional model lacks real-time and flexibility in route planning, cannot effectively deal with emergencies such as traffic congestion, and is difficult to meet customer needs, resulting in a decrease in customer satisfaction and an increase in the operating costs of logistics companies. Therefore, a logistics management system based on intelligent scheduling is proposed to address the above problems. Summary of the invention

[0005] The purpose of the present invention is to provide a logistics management system based on intelligent scheduling to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A logistics management system based on intelligent scheduling, comprising:

[0008] Data collection module: used to collect historical data and real-time data in logistics business in real time. The historical data and real-time data involve cargo information, transportation vehicle information, warehouse inventory information, customer demand information and traffic condition information;

[0009] Demand analysis module: connected to the data collection module, based on the collected historical data and real-time data, uses data analysis algorithms to predict and analyze logistics demand, and determines the logistics service demand in different time periods and regions;

[0010] Intelligent scheduling module: connected to the demand analysis module, it uses the preset intelligent scheduling algorithm to dynamically plan transportation tasks, allocate transportation resources, and plan storage use strategies based on the logistics demand output by the demand analysis module and the real-time vehicle, storage and traffic resource conditions, so as to achieve the optimal configuration of logistics resources;

[0011] Path planning module: connected to the intelligent scheduling module, for the transportation tasks assigned by the intelligent scheduling module, the optimal transportation path is calculated through the path planning algorithm based on the real-time traffic information and the transportation starting and ending locations;

[0012] Monitoring and feedback module: connected with the intelligent scheduling module, path planning module and data acquisition module, it monitors the parameters in the process of logistics task execution in real time. The parameters in the process of logistics task execution include vehicle location, transportation progress, and changes in warehouse inventory. When an abnormal situation deviates from the preset plan, it will provide feedback to the intelligent scheduling module.

[0013] Optimization module: connected to the intelligent scheduling module, it performs global optimization on the scheduling scheme output by the intelligent scheduling module based on the genetic algorithm. The genetic algorithm conducts a comprehensive evaluation of the scheduling scheme in terms of cost, efficiency and service quality according to the set fitness function, so as to continuously optimize the scheduling scheme and optimize the overall performance of the logistics system.

[0014] As a preferred solution, the data acquisition module includes on-board sensors installed on transport vehicles, inventory monitoring equipment in warehouses, data interfaces in logistics business systems, and communication units for data connection with traffic management departments. The data collected by the on-board sensors include vehicle speed. Mileage , the average speed of the vehicle is calculated by the following formula: ,in, Indicates the mileage of each section. Indicates the driving time corresponding to each segment of mileage. To calculate the number of mileage segments, the inventory monitoring device uses the inventory quantity calculation formula To monitor warehouse inventory in real time, is the current inventory quantity, is the initial inventory quantity, is the incoming quantity, and O is the outgoing quantity.

[0015] As a preferred solution, the data analysis algorithm used by the demand analysis module includes at least one of a time series analysis algorithm and a machine learning prediction algorithm to accurately predict logistics demand at different time scales; if the autoregressive integrated moving average (ARIMA) model is used for time series analysis and prediction, the formula is: ;

[0016] in, For current needs, To predict demand, are the autoregressive model parameters, is the autoregressive order, is the sliding average order, is the mean, is the sliding average model parameter, is the white noise error, Indicates at a point in time Push forward Time series The value of Indicates at a point in time Push forward White noise error value per time unit;

[0017] If the multivariate linear regression algorithm in machine learning is used for prediction, the prediction model formula is: ;

[0018] in, For the predicted logistics demand, is the intercept term, For each characteristic variable The corresponding coefficients are, is the number of characteristic variables, and the value of each coefficient is determined by fitting the training data to achieve demand forecasting.

[0019] As a preferred solution, the intelligent scheduling algorithm adopts a hybrid algorithm that combines the greedy algorithm with the dynamic programming algorithm. While meeting the priority requirements of transportation tasks, it realizes the efficient allocation of overall logistics resources among different tasks. In the greedy algorithm part, the task priority The calculation formula is: ;

[0020] in, For the task The demand time, For the task The transportation time is shortened and resources are allocated in descending order according to the task priority;

[0021] In the dynamic programming algorithm part, suppose Indicates in front After the tasks are assigned resources, the resource status is The optimal value when (the value here can take into account factors such as cost and efficiency), its state transfer equation is: ;

[0022] in, Indicates execution of The amount of resources required for each task, Indicates execution of The value increment brought by each task is calculated through continuous iteration. To determine the optimal resource allocation plan.

[0023] As a preferred solution, when the path planning algorithm calculates the optimal transportation path, its heuristic function is: ,in, Represents the actual cost from the starting point to the current node, Represents the estimated cost from the current node to the target node, Represents the total cost of the current node; if distance is used as the cost measurement standard, By Euclidean distance formula The cumulative calculation is as follows: and are the coordinates of the two points, Using the Manhattan distance formula Make an estimate, among which, Front node coordinates, is the target node coordinate, by continuously comparing the value to determine the optimal path.

[0024] As a preferred solution, the monitoring and feedback module exchanges data with the transport vehicles, warehouse management systems and terminal devices of logistics business-related personnel through wireless communication networks to ensure that abnormal situations are promptly communicated to the relevant responsible parties and the dispatch center. The monitoring and feedback module monitors the vehicle position through GPS positioning and locates the coordinates. Every time interval Update, the update formula is: ;

[0025] in, for The coordinates of the moment, for The coordinates of the moment, and They are Always Axis and The velocity component in the axial direction is used to determine whether there is a deviation by comparing the actual coordinates with the coordinates corresponding to the preset planned route. If the deviation exceeds the set threshold, it is judged as an abnormality and feedback is given to the intelligent scheduling module.

[0026] As a preferred solution, in the genetic algorithm of the optimization module, the fitness function is set as: ;

[0027] in, are the weight coefficients corresponding to logistics operation cost, average transportation time, on-time delivery rate of goods, and resource idle rate, respectively. Quality(x), Utilization Respectively represent the scheduling scheme The corresponding indicator value under

[0028] In the crossover operation of the genetic algorithm, a single-point crossover method is used, and the parent individual and The intersection point is , then the generated offspring individuals and The gene fragment combination is as follows: ; ;

[0029] In the mutation operation, let the length of the individual gene sequence be , the mutation probability is , for each gene position , according to probability Perform mutation operation. If the gene bit represents the number of resource allocations, the mutation is performed according to the uniform distribution. Randomly generate new values ​​to replace the original gene position values, where and To set the lower and upper limits of the variation range, through continuous iterative genetic operations, the individual pros and cons are evaluated according to the fitness function, and the optimal scheduling plan is screened out.

[0030] It can be seen from the technical solution provided by the present invention that the logistics management system based on intelligent scheduling provided by the present invention has the following beneficial effects:

[0031] Improve resource utilization efficiency: The intelligent scheduling module uses a greedy algorithm combined with a dynamic programming algorithm to reasonably allocate logistics resources such as transport vehicles and storage space based on the results of the demand analysis module and real-time resource conditions. For example, when allocating vehicles, it fully considers vehicle load, location, and availability to avoid idle or overloaded vehicles and improve vehicle utilization. Warehouse resource allocation optimizes the storage and outbound order based on the characteristics of the goods and inventory layout, reduces storage space waste, effectively reduces logistics costs, and achieves efficient resource allocation.

[0032] Enhanced transportation timeliness: The route planning module uses real-time traffic conditions and heuristic algorithms to calculate the optimal transportation route, which can timely avoid unfavorable factors such as congested roads and construction areas. For example, in urban distribution, the route can be dynamically adjusted according to real-time traffic conditions to enable transportation vehicles to pass quickly, significantly shortening transportation time, improving the punctuality of goods delivery, and enhancing customer satisfaction.

[0033] Optimize demand forecast accuracy: The demand analysis module uses time series analysis and machine learning algorithms to conduct in-depth mining of large amounts of historical and real-time data. It provides precise guidance for subsequent scheduling by making detailed forecasts of logistics demand in different time periods and regions. It can also continuously update the model to adapt to business changes, such as preparing resources in advance during e-commerce promotions to ensure the timeliness and stability of logistics services.

[0034] Ensure the reliability of the logistics process: The monitoring and feedback module tracks key parameters of vehicles, transportation and warehousing in real time, and uses GPS positioning and data comparison to promptly detect abnormalities. Once problems such as vehicle failure, inventory shortage or transportation delay occur, feedback is quickly given to the intelligent scheduling module so that strategies can be adjusted in a timely manner, tasks and resource allocation can be replanned to ensure smooth operation of the logistics process and reduce operational risks.

[0035] Achieve optimal overall system performance: The optimization module is based on a genetic algorithm. It comprehensively evaluates and optimizes scheduling plans based on a fitness function that covers cost, time, service quality, and resource idle rate. After multiple iterations, the optimal plan is selected to achieve a balance in the logistics system in many aspects, improve overall operational efficiency and service quality, enhance the company's competitiveness in the logistics market, and promote efficient and sustainable development of the logistics business. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 The figure is a schematic diagram of the overall structure of a logistics management system based on intelligent scheduling according to the present invention. DETAILED DESCRIPTION

[0037] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0038] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0039] like Figure 1 As shown, an embodiment of the present invention provides a logistics management system based on intelligent scheduling, including a data acquisition module, a demand analysis module, an intelligent scheduling module, a path planning module, a monitoring and feedback module and an optimization module.

[0040] In this embodiment, the data collection module is used to collect historical data and real-time data in the logistics business in real time. The historical data and real-time data involve cargo information, transportation vehicle information, storage inventory information, customer demand information, and traffic condition information;

[0041] Furthermore, the data collection module includes on-board sensors installed on transport vehicles, inventory monitoring equipment in warehouses, data interfaces in logistics business systems, and communication units for data connection with traffic management departments. The data collected by on-board sensors include vehicle speed. Mileage , the average speed of the vehicle is calculated by the following formula: ,in, Indicates the mileage of each section. Indicates the driving time corresponding to each segment of mileage. To calculate the number of mileage segments, the inventory monitoring device uses the inventory quantity calculation formula To monitor warehouse inventory in real time, is the current inventory quantity, is the initial inventory quantity, is the incoming quantity, O is the outgoing quantity;

[0042] The specific steps for the data acquisition module to collect various types of information are as follows:

[0043] Cargo information collection:

[0044] The basic attributes of goods, such as name, weight, volume, category, etc., are manually input through the data entry interface in the logistics business system at the initial stage of goods entering the logistics system. At the same time, this information will be associated with the unique identification of the goods (such as barcode, QR code or goods number) and stored in the database for subsequent query and update; during the transportation of goods, if there is a change in the status of the goods (such as damage, deterioration, etc.), the transportation personnel can supplement the entry of the goods status information through handheld terminal devices (such as PDA) and synchronize it to the logistics business system;

[0045] For the transportation track information of the goods, after the goods are loaded onto the transport vehicle, the on-board sensors (such as GPS positioning devices) are integrated with the cargo information system. When the vehicle is driving, the GPS positioning device automatically records the latitude and longitude information of the location of the goods at a set time interval (for example, every 10 minutes), and transmits this information back to the data collection module in the logistics business system in real time through a wireless communication network (such as 4G / 5G network), realizing the dynamic collection of the cargo transportation track;

[0046] Transport vehicle information collection:

[0047] The speed sensor and mileage sensor in the vehicle sensor continuously monitor the vehicle's driving speed and mileage The speed sensor senses the rotation speed of the vehicle wheels and calculates the real-time speed of the vehicle based on parameters such as the wheel circumference. The mileage sensor accumulates the distance traveled by the vehicle. At a certain time period (for example, every hour), the data acquisition module reads this data from the on-board sensor and calculates the speed of the vehicle based on the formula. Calculate the average speed of the vehicle during this time period, where Indicates the mileage of each section. Indicates the driving time corresponding to each segment of mileage. The number of mileage segments to be counted; these speed and mileage data will be associated and stored with the vehicle number, transport task number and other information to facilitate subsequent analysis of the vehicle's driving performance and transport efficiency;

[0048] The vehicle's load information is collected through a pressure sensor installed on the vehicle chassis when the cargo is loaded; when the cargo is loaded onto the vehicle, the pressure sensor measures the pressure change on the vehicle and converts it into the weight information of the cargo; before the vehicle departs, the data acquisition module obtains and records the load information and stores it together with other vehicle information. During transportation, if the load changes due to cargo loading and unloading, the pressure sensor will monitor and update the data in real time, and the data acquisition module will also obtain new load data in a timely manner;

[0049] The vehicle maintenance information is obtained from the vehicle management system. After each vehicle repair, maintenance, annual inspection, etc., the repair shop or relevant staff will enter the maintenance items, time, cost and other detailed information into the vehicle management system. The data acquisition module is connected to the vehicle management system through the data interface, and queries and obtains the latest vehicle maintenance information at a certain frequency (such as once a week) to evaluate the reliability and availability of the vehicle and provide a reference for transportation task arrangement.

[0050] Warehouse inventory information collection:

[0051] Deploy inventory monitoring equipment in the warehouse, such as infrared sensors, RFID readers, etc.; for goods stored on shelves, install infrared sensors on each shelf. When goods are placed on or taken out of the shelf, the infrared sensors detect the presence or absence of goods and transmit the signal to the inventory management system; the inventory management system calculates the inventory quantity according to the signal changes of the sensor according to the formula To monitor warehouse inventory in real time, is the current inventory quantity, is the initial inventory quantity, is the incoming quantity, and O is the outgoing quantity; the data acquisition module obtains the updated information of the inventory quantity in real time through the data interface with the inventory management system, and stores it in the database of the logistics business system;

[0052] For goods stored in pallets or containers, RFID tags are attached to the pallets or containers, and RFID readers are installed at the warehouse entrances and exits or on handling equipment. When goods with RFID tags pass through the reader, the reader reads the tag information and transmits the goods' in and out status to the inventory management system. The inventory management system also updates the inventory quantity according to the above inventory calculation formula. The data acquisition module then obtains the updated inventory data from the inventory management system to ensure the accuracy and timeliness of the inventory information.

[0053] Customer demand information collection:

[0054] Set up a customer demand input interface in the e-commerce platform or customer service system of the logistics enterprise; when placing an order, the customer needs to fill in detailed information such as the delivery address, recipient information, expected delivery time, special requirements for the goods (such as fragile goods, refrigerated goods, etc.); after the customer submits the order, this information will be transmitted to the data collection module of the logistics business system through the data processing process of the system background, and stored in the customer demand database; the data collection module will conduct preliminary verification and sorting of customer demand information to ensure the integrity and accuracy of the information;

[0055] Logistics companies can also communicate with customers through customer service personnel to further understand their special needs or supplement their order information. Customer service personnel will record the results of communication with customers in the customer service system. The data collection module regularly obtains this supplementary information from the customer service system and integrates it with the customer's original order information to better meet customer needs during the logistics transportation process.

[0056] Traffic information collection:

[0057] Establish a data connection mechanism with the traffic management department, and obtain traffic condition information at certain time intervals (such as every 15 minutes) through the traffic data interface provided by the traffic management department; this information includes road congestion (such as congested road section, congestion degree, expected congestion time, etc.), traffic accident information (accident location, accident type, impact range, etc.), road construction information (construction road section, construction time, impact of construction on traffic, etc.); the data acquisition module will parse and organize the acquired traffic condition information, extract information related to the logistics transportation route, and store it in the local database to provide real-time road condition data support for the path planning module;

[0058] Utilize the traffic information feedback function of the vehicle navigation system or mobile map application; when the transport vehicle is driving on the road, the vehicle navigation system or the mobile map application used by the driver will receive real-time traffic conditions information on the surrounding roads, and feed this information back to the data collection module of the logistics business system through the wireless communication network (such as 4G / 5G network); the data collection module summarizes and analyzes the traffic information from the vehicle side, and integrates it with the information obtained from the traffic management department to further improve the accuracy and comprehensiveness of traffic road condition information, so as to more effectively plan transportation routes, avoid congested sections, and improve transportation efficiency.

[0059] In this embodiment, the demand analysis module is connected to the data collection module. Based on the collected historical data and real-time data, the data analysis algorithm is used to predict and analyze the logistics demand to determine the logistics service demand in different time periods and different regions. The data analysis algorithm used by the demand analysis module includes at least one of a time series analysis algorithm and a machine learning prediction algorithm to accurately predict the logistics demand under different time scales. If the autoregressive integrated moving average (ARIMA) model is used for time series analysis and prediction, the formula is: ;

[0060] in, For current needs, To predict demand, are the autoregressive model parameters, is the autoregressive order, is the sliding average order, is the mean, is the sliding average model parameter, is the white noise error, Indicates at a point in time Push forward Time series The value of Indicates at a point in time Push forward White noise error value per time unit;

[0061] If the multivariate linear regression algorithm in machine learning is used for prediction, the prediction model formula is: ;

[0062] in, For the predicted logistics demand, is the intercept term, For each characteristic variable The corresponding coefficients are, is the number of characteristic variables, and the value of each coefficient is determined by fitting the training data to achieve demand forecasting;

[0063] Further, the operation steps of the demand analysis module are as follows:

[0064] Data reception and preprocessing:

[0065] The demand analysis module receives the collected historical and real-time data of logistics business through the connection with the data acquisition module; these data include cargo information, transportation vehicle information, storage inventory information, customer demand information, traffic condition information, etc.

[0066] Clean and organize the received data; remove noise and erroneous values ​​in the data, for example, correct or delete obviously unreasonable cargo weight or transportation mileage data; convert data in different formats into a format suitable for analysis, such as unifying time data into a standard time format, encoding text-based cargo category information, etc.;

[0067] Classify and archive data, and store them by cargo type, transportation area, time period, etc., so that relevant data can be quickly retrieved and called for subsequent analysis;

[0068] Feature engineering (optional step, depending on the algorithm used):

[0069] If machine learning algorithms are used for predictive analysis, features related to logistics demand can be extracted from preprocessed data; for example, for logistics demand forecasting, features such as cargo transportation volume in historical time periods, number of orders in different regions, seasonal factors (such as fluctuations in logistics demand in specific months or during holidays), and economic data (such as the impact of local GDP growth trends on logistics demand) can be extracted;

[0070] Standardize or normalize the extracted features to make the numerical ranges of different features comparable; for example, use Min-Max normalization to map feature values ​​to the interval [0,1], or use Z-score normalization to convert feature values ​​to a distribution with a mean of 0 and a standard deviation of 1;

[0071] Model selection and training (if it is the first time to use or the model needs to be updated):

[0072] Select appropriate data analysis algorithms based on the characteristics and data of logistics demand forecasting. If the data presents obvious time series characteristics and the historical data is relatively abundant, the autoregressive integrated moving average (ARIMA) model can be given priority. If the data has multiple influencing factors and the relationship is relatively complex, machine learning forecasting algorithms such as multivariate linear regression and neural networks can be selected.

[0073] For the selected model, a portion of the historical data (usually 70%-80%) is used for training; when training the ARIMA model, the parameters of the model are determined by analyzing the autocorrelation function (ACF) and partial autocorrelation function (PACF) of the time series data. (autoregressive order), (difference order) and (moving average order), and then use methods such as maximum likelihood estimation to estimate other parameters of the model, such as the autoregressive coefficients Moving average coefficient For machine learning models, such as multivariate linear regression, the least squares method is used to fit the model parameters, i.e., to determine the coefficients The value of makes the error between the predicted value and the true value minimum;

[0074] Use a validation set (usually 10%-15% of the historical data) to validate the trained model; calculate the model's evaluation indicators on the validation set, such as mean square error (MSE), mean absolute error (MAE), etc., and adjust and optimize the model based on the evaluation results; if the model performs poorly on the validation set, you may need to adjust the model's parameters, increase the amount of training data, or try other models;

[0075] Logistics demand forecasting and analysis:

[0076] Input the real-time data and pre-processed historical data into the trained model for prediction; if the ARIMA model is used, according to the model formula Calculate forecast demand ,in, For current needs, are the autoregressive model parameters, is the autoregressive order, is the sliding average order, is the mean, is the sliding average model parameter, is the white noise error, Indicates at a point in time Push forward Time series The value of Indicates at a point in time Push forward The white noise error value at time units; if it is a multivariate linear regression model, according to the formula Calculate forecasted logistics demand ,in, is the characteristic variable, is the corresponding coefficient;

[0077] Analyze the forecast results, combine the division of different time periods (such as hours, days, weeks, months, etc.) and different regions (such as cities, provinces, countries, etc.), and count and analyze the distribution of logistics demand; for example, draw logistics demand trend charts for different time periods, analyze the hot spots and cold spots of logistics demand in different regions, so as to determine the key areas and times for resource allocation;

[0078] Based on the prediction and analysis results, a logistics demand report is generated. The report should include logistics demand forecast values ​​for different time periods and regions, demand change trends, factors that may affect demand, and other information. The report is then output to the intelligent scheduling module to provide a basis for its decision-making. At the same time, the demand analysis module will continue to monitor new data inflows, and regularly (such as daily or weekly) retrain and update the model to adapt to the dynamic changes in the logistics business and improve the accuracy of the prediction.

[0079] In this embodiment, the intelligent scheduling module is connected to the demand analysis module. According to the logistics demand output by the demand analysis module and the real-time vehicle, storage and traffic resource conditions, the intelligent scheduling module uses a preset intelligent scheduling algorithm to dynamically plan transportation tasks, allocate transportation resources and plan storage use strategies to achieve optimal allocation of logistics resources. The intelligent scheduling algorithm uses a hybrid algorithm that combines a greedy algorithm with a dynamic programming algorithm to achieve efficient allocation of overall logistics resources among different tasks while meeting the priority requirements of transportation tasks. In the greedy algorithm part, the task priority The calculation formula is: ;

[0080] in, For the task The demand time, For the task The transportation time is shortened and resources are allocated in descending order according to the task priority;

[0081] In the dynamic programming algorithm part, suppose Indicates in front After the tasks are assigned resources, the resource status is The optimal value when (the value here can take into account factors such as cost and efficiency), its state transfer equation is: ;

[0082] in, Indicates execution of The amount of resources required for each task, Indicates execution of The value increment brought by each task is calculated through continuous iteration. To determine the optimal resource allocation plan;

[0083] Furthermore, the specific operation steps of the intelligent scheduling module are as follows:

[0084] Receive demand and resource information:

[0085] The intelligent scheduling module establishes a real-time connection with the demand analysis module and receives the logistics demand situation report of different time periods and different regions output by it; the report contains detailed information such as the predicted cargo transportation volume, delivery time requirements, cargo categories and special needs;

[0086] At the same time, the intelligent dispatching module exchanges data with the vehicle management subsystem, warehouse management subsystem and traffic information system in the logistics business system to obtain real-time vehicle location, vehicle load capacity, vehicle availability (whether it is under repair or maintenance), warehouse inventory capacity, warehouse layout, inventory distribution, current traffic conditions (road congestion, traffic restriction information, etc.) and other resource status information;

[0087] Task priority determination (based on the greedy algorithm part):

[0088] For each transportation task, according to the demand time provided by the demand analysis module and shipping time , according to the task priority calculation formula Calculating task priorities ; Requirement time Can be obtained from customer demand information, transportation time The transportation distance (calculated based on the coordinates of the origin and destination), traffic conditions (congested roads will increase transportation time), and the average vehicle speed (estimated based on the vehicle's historical driving data or real-time speed monitoring data) are considered to determine the transportation time.

[0089] Sort all transportation tasks by priority from high to low so that high-priority tasks are given priority in subsequent resource allocation;

[0090] Resource allocation and task planning (combined with dynamic programming algorithm):

[0091] set up Indicates in front After the tasks are assigned resources, the resource status is The optimal value when (the value here takes into account factors such as cost and efficiency); initialization is the value of the initial resource state (usually 0);

[0092] For each task , traverse all possible resource allocation schemes; in terms of vehicle allocation, select the appropriate vehicle to perform the transportation task based on the vehicle's load capacity, location and availability; for example, if a task requires the transportation of 5 tons of goods, and there are two available vehicles, one with a load capacity of 8 tons and the other with a load capacity of 3 tons, and other conditions are the same, the vehicle with a load capacity of 8 tons is given priority; at the same time, considering the distance between the current location of the vehicle and the starting point of the task, selecting a vehicle with a shorter distance can reduce empty mileage and improve transportation efficiency; in terms of storage resource allocation, determine the storage location and outbound order of the goods based on the category, inventory distribution and storage capacity of the goods; for goods that need to be delivered urgently, give priority to allocating cargo spaces close to the outbound port with sufficient inventory, and arrange for priority outbound delivery;

[0093] Calculate the incremental value under each resource allocation scheme and the amount of resources required to perform the task ; The incremental value can be determined by comprehensive evaluation of factors such as reducing transportation costs (such as choosing a more economical route, making rational use of vehicle load), improving customer satisfaction (delivering goods on time), etc.; The amount of resources required to perform the task includes vehicle usage time, fuel consumption, storage space occupation, etc.;

[0094] According to the state transfer equation renew value, find the optimal resource allocation plan and task execution order;

[0095] Dynamic adjustment and optimization:

[0096] During the execution of logistics tasks, the intelligent scheduling module continuously monitors the real-time information of vehicle location, transportation progress, warehouse inventory changes, and traffic conditions; if there are abnormal situations such as vehicle failure, traffic congestion seriously affecting transportation time, new emergency orders, etc., the task priority and resource allocation plan will be re-evaluated;

[0097] According to the new situation, it may be necessary to adjust the execution order of transportation tasks, such as bringing forward urgent tasks originally planned to be executed later, and reallocating vehicles and storage resources; for vehicles affected by traffic congestion, re-plan routes and choose alternative routes to reduce delays; at the same time, timely communicate and coordinate with relevant departments (such as vehicle maintenance teams, warehouse managers, and transportation personnel) to ensure that the adjusted plan can be smoothly implemented;

[0098] Generate a schedule and output:

[0099] The intelligent scheduling module generates a detailed logistics scheduling plan based on the finalized resource allocation and task planning results; the plan includes the specific execution vehicle, departure time, driving route, estimated arrival time, storage resource usage plan (warehousing, outbound and storage location arrangement of goods) for each transportation task, etc.

[0100] The scheduling plan is output to relevant modules and departments such as the route planning module, transportation execution team and warehousing management team to ensure that all links work together according to the scheduling plan to achieve optimal allocation and efficient operation of logistics resources.

[0101] In this embodiment, the path planning module is connected to the intelligent scheduling module. For the transportation task assigned by the intelligent scheduling module, the optimal transportation path is calculated by the path planning algorithm based on the real-time traffic information and the transportation starting and ending locations. When the path planning algorithm calculates the optimal transportation path, its heuristic function is: ,in, Represents the actual cost from the starting point to the current node, Represents the estimated cost from the current node to the target node, Represents the total cost of the current node; if distance is used as the cost measurement standard, By Euclidean distance formula The cumulative calculation is as follows: and are the coordinates of the two points, Using the Manhattan distance formula Make an estimate, among which, Front node coordinates, is the target node coordinate, by continuously comparing the value to determine the optimal path;

[0102] Further, the specific operation steps of the path planning module are as follows:

[0103] Task reception and information extraction:

[0104] The route planning module receives the transport task information assigned by the intelligent scheduling module, including the detailed addresses or geographic coordinates of the transport start and end points, as well as the relevant attributes of the goods (such as weight, volume, whether it is fragile, etc.); at the same time, it obtains real-time traffic information from the traffic information system. The traffic information includes the congestion level of each section (expressed as a congestion coefficient, such as a value between 0-1, 0 means smooth traffic and 1 means severe congestion), road construction conditions (the location of the construction section and the estimated construction time), traffic control information (such as restricted sections and time), etc.

[0105] Map data preparation:

[0106] Call the Geographic Information System (GIS) database to obtain map data including road networks, intersection information, and road attributes (such as road grade, speed limit, number of lanes, etc.); locate the transportation starting point and end point on the map, and filter out the passable road network based on the cargo attributes and vehicle type (such as large trucks, small passenger cars, etc.); for example, for large trucks, exclude roads with height and weight restrictions;

[0107] The cost function is determined as:

[0108] When distance is used as the main cost metric, for each section of the path , using the Euclidean distance formula Calculate its actual distance cost ,in, and are the coordinates of the two end points of the road respectively; starting from the starting point, as the path expands, the The value of

[0109] For the current node Estimated cost to the target node , using the Manhattan distance formula Calculate, where is the current node coordinate, Target node coordinates; at the same time, according to real-time traffic information Make corrections; for example, if a road section is severely congested, appropriately increase the estimated cost of that section; if a new expressway is opened, reduce the estimated cost of the corresponding section;

[0110] Calculate the total cost of each node ;

[0111] Path Search and Extension:

[0112] The starting point is added as the initial node to the priority queue (usually using a binary heap or other data structure), and the priority queue is based on the total cost of the node. Sort by

[0113] Take the total cost from the priority queue The smallest node is expanded as the current node; check the adjacent nodes of the current node, and calculate the adjacent nodes that are accessible and have not been visited. , and value and add it to the priority queue; during the expansion process, the parent node information of each node is recorded so that the complete path can be generated by subsequent backtracking;

[0114] Repeat the above expansion process until the target node is added to the priority queue and becomes the current node;

[0115] Path optimization and determination:

[0116] When the target node becomes the current node, an initial path is generated by backtracking the parent node chain from the target node to the starting point;

[0117] Check the initial path for optimization; consider the actual conditions of the road, such as whether there are one-way streets, turning restrictions at intersections, etc.; if the initial path violates these rules, make local adjustments; for example, when encountering a one-way street, find an alternative path to bypass the road section;

[0118] Comprehensively consider factors such as transportation time, transportation cost and cargo safety, and evaluate the optimized path; if there are multiple feasible paths, select the path with the lowest total cost as the final optimal transportation path;

[0119] The optimal transportation route information (including the coordinates of each node on the route, driving direction, estimated driving time, etc.) is output to the transportation execution department and the intelligent scheduling module so that the vehicle can travel according to the planned route, and the intelligent scheduling module is provided with route-related feedback information, such as the deviation between the actual driving time and the estimated driving time, for reference in subsequent scheduling decisions.

[0120] In this embodiment, the monitoring and feedback module is connected to the intelligent scheduling module, the path planning module and the data acquisition module to monitor the parameters in the process of executing the logistics task in real time. The parameters in the process of executing the logistics task include the vehicle position, transportation progress, and changes in warehouse inventory. When an abnormal situation deviates from the preset plan, feedback is given to the intelligent scheduling module. The monitoring and feedback module exchanges data with the transport vehicle, the warehouse management system and the terminal equipment of the logistics business-related staff through the wireless communication network to ensure that the abnormal situation is promptly communicated to the relevant responsible parties and the scheduling center. The monitoring and feedback module monitors the vehicle position through GPS positioning, and the positioning coordinates Every time interval Update, the update formula is: ;

[0121] in, for The coordinates of the moment, for The coordinates of the moment, and They are Always Axis and The velocity component in the axial direction is used to determine whether there is a deviation by comparing the actual coordinates with the coordinates corresponding to the preset planned route. If the deviation exceeds the set threshold, it is judged as abnormal and feedback is given to the intelligent scheduling module;

[0122] Further, the specific operation steps of the monitoring and feedback module are as follows:

[0123] Establish data connection and initialization:

[0124] Establish stable data interaction connection with transport vehicles, warehouse management systems, terminal equipment of logistics business-related staff, intelligent scheduling modules, path planning modules and data acquisition modules through wireless communication networks;

[0125] Set normal threshold ranges for vehicle location, transportation progress, warehouse inventory changes, and time interval parameters for data collection and updating;

[0126] Data collection and real-time monitoring:

[0127] The vehicle location information is obtained from the GPS positioning system of the transport vehicle and the vehicle location information is obtained at a set time interval (e.g. Time) Update positioning coordinates , the update formula is: ;

[0128] in, for Time coordinates, and They are Always Axis and Axial velocity component;

[0129] Obtain transportation progress information from the logistics business system, such as the time nodes and completion status data of cargo loading, transportation, unloading, etc., and update and record them in real time according to the business process nodes;

[0130] Connect with the warehouse management system to obtain warehouse inventory changes, and calculate the inventory quantity according to the formula ( is the current inventory quantity, is the initial inventory quantity, The system monitors inventory dynamics in real time and updates data promptly after each inbound and outbound operation.

[0131] Data comparison and abnormal judgment:

[0132] Compare the actual position coordinates of the vehicle with the corresponding coordinates of the preset planned route and calculate the deviation value; if the deviation value exceeds the set position threshold, it is determined that the vehicle position is abnormal;

[0133] Compare the transportation progress with the planned progress to determine whether it is delayed or advanced; if the transportation progress deviates from the planned progress by more than the set time threshold, the transportation progress is considered abnormal;

[0134] Analyze whether the changes in warehouse inventory are in line with the planned inventory change curve or the established inventory strategy; if the inventory quantity, in and out time, etc. are inconsistent with the plan and exceed the allowable error range, the warehouse inventory is considered abnormal;

[0135] Abnormal feedback and information transmission:

[0136] Once any abnormal situation is detected, feedback information including the abnormality type (vehicle position abnormality, transportation progress abnormality or warehouse inventory abnormality), detailed data (such as the actual vehicle position coordinates, deviation value, current transportation progress, planned progress, actual inventory quantity, planned inventory quantity, etc.) and occurrence time is immediately sent to the intelligent scheduling module through the wireless communication network, so that the intelligent scheduling module can take timely adjustment measures.

[0137] In this embodiment, the optimization module is connected to the intelligent scheduling module, and the scheduling scheme output by the intelligent scheduling module is globally optimized based on the genetic algorithm. The genetic algorithm comprehensively evaluates the scheduling scheme in terms of cost, efficiency and service quality according to the set fitness function, so as to continuously optimize the scheduling scheme and optimize the overall performance of the logistics system. In the genetic algorithm of the optimization module, the set fitness function is: ;

[0138] in, are the weight coefficients corresponding to logistics operation cost, average transportation time, on-time delivery rate of goods, and resource idle rate, respectively. Quality(x), Utilization Respectively represent the scheduling scheme The corresponding indicator value under

[0139] In the crossover operation of the genetic algorithm, a single-point crossover method is used, and the parent individual and The intersection point is , then the generated offspring individuals and The gene fragment combination is as follows: ; ;

[0140] In the mutation operation, let the length of the individual gene sequence be , the mutation probability is , for each gene position , according to probability Perform mutation operation. If the gene bit represents the number of resource allocations, the mutation is performed according to the uniform distribution. Randomly generate new values ​​to replace the original gene position values, where and To set the lower and upper limits of the variation range, through continuous iterative genetic operations, the individual advantages and disadvantages are evaluated according to the fitness function, and the optimal scheduling plan is selected;

[0141] Further, the specific operation steps of the optimization module are as follows:

[0142] Fitness function setting and parameter initialization:

[0143] Determine the fitness function of the genetic algorithm as , and set according to the focus and goals of logistics business The weight coefficients correspond to logistics operation cost, average transportation time, on-time delivery rate of goods, and resource idle rate respectively; at the same time, the genetic algorithm parameters such as population size, crossover probability, mutation probability, and number of iterations are initialized;

[0144] Initial population generation:

[0145] Obtain the scheduling plan from the intelligent scheduling module as the initial population individuals; each individual represents a combination of resource allocation and task arrangement, and its gene code corresponds to decision variables such as transportation vehicle allocation, storage resource usage, and transportation route selection; for example, binary code is used to represent vehicle selection (00 represents vehicle 1, 01 represents vehicle 2, etc.), and integer code is used to represent storage cargo space allocation, etc., and multiple individuals of the initial population are generated according to certain rules;

[0146] Fitness evaluation:

[0147] For each individual (scheduling plan) in the population, its fitness value is calculated according to the set fitness function; first, the corresponding scheduling plan details are parsed according to the individual's genetic code, such as vehicle allocation, transportation routes, storage resource usage plan, etc.; then, the logistics operation cost under the plan is calculated based on the logistics business data and actual operation conditions. , average transportation time , Quality(x), Utilization And other index values, and finally substitute them into the fitness function to calculate the fitness value ;

[0148] Select an action:

[0149] Use appropriate selection methods (such as roulette selection, tournament selection, etc.) to select some individuals from the current population as parents according to their fitness values ​​for subsequent crossover and mutation operations; individuals with higher fitness values ​​have a greater probability of being selected to ensure that good genes have a greater chance of being passed on to the next generation; for example, in roulette selection, calculate the selection probability of each individual ( is the population size), and then randomly select according to probability;

[0150] Crossover operation:

[0151] Perform crossover operation on the selected parent individuals, using single-point crossover method; randomly select a pair of parent individuals and , determine the intersection , generating offspring individuals and , whose gene fragment combination is and , through the crossover operation, the good genes of the parents are combined to produce new individuals and increase the diversity of the population;

[0152] Mutation operation:

[0153] Perform mutation operation on the offspring individuals after crossover; suppose the length of individual gene sequence is , the mutation probability is For each gene position , according to probability Perform mutation operation; if the gene bit represents the number of resource allocations, the mutation is performed according to the uniform distribution Randomly generate new values ​​to replace the original gene position values ​​( and The lower and upper limits of the set variation range are introduced to introduce new gene features to prevent the algorithm from falling into the local optimum;

[0154] Population update and iteration:

[0155] Replace some individuals in the original population with offspring individuals generated through crossover and mutation operations to form a new population; determine whether the set number of iterations has been reached. If not, return to the fitness evaluation step and continue a new round of selection, crossover, and mutation operations to continuously iterate and optimize the population so that the individuals in the population (scheduling scheme) continue to approach the optimal solution;

[0156] Optimal solution output:

[0157] When the number of iterations is reached, the individual with the highest fitness value is selected from the final population as the optimal scheduling plan; the plan is output to the intelligent scheduling module, and the intelligent scheduling module re-allocates logistics resources and replans tasks according to the optimized plan to improve the overall performance of the logistics system.

[0158] The present invention will be further described in detail below with reference to specific embodiments:

[0159] 1. Background and Objectives

[0160] A large e-commerce logistics company, whose business covers many regions across the country, handles a large number of cargo transportation orders every day. Faced with problems such as high logistics costs, transportation efficiency to be improved, and on-time delivery rate of goods to be optimized, it decided to introduce the logistics management system based on intelligent scheduling of the present invention to improve the overall performance of logistics operations;

[0161] 2. System Implementation Process

[0162] Data acquisition module:

[0163] Install high-precision on-board sensors on transport vehicles, including speed sensors, mileage sensors and GPS positioning devices; for example, on a truck responsible for urban distribution, the speed sensor collects vehicle speed data once a second, the mileage sensor accurately records the mileage, and the GPS positioning device records the vehicle location information every 5 minutes;

[0164] Advanced inventory monitoring equipment, such as infrared sensors and RFID readers, is deployed in the warehouse. In a large storage center, infrared sensors are installed on each shelf. For electronic products stored on pallets, RFID tags are affixed to the pallets. RFID readers are installed at the warehouse entrances and exits and on handling equipment to achieve real-time and accurate monitoring of the inventory of goods.

[0165] Establish stable data connection with the data interface in the logistics business system and the traffic management department; obtain detailed information of the goods through the data interface of the logistics business system, such as the name, weight (0.5 kg per piece, 1,000 pieces in total), volume (0.05 cubic meters per package, 100 packages in total), category (fashion clothing), etc. of a batch of clothing goods, and store them in association with the unique identification of the goods; at the same time, obtain traffic information from the traffic management department every 15 minutes, including congestion conditions of major urban roads (such as a congestion coefficient of 0.7 on a certain road section), traffic accident information (such as a minor scratch accident at a certain intersection, affecting the surrounding 200-meter roads) and road construction information (construction on a certain road section, with an estimated construction time of 3 hours);

[0166] Requirements Analysis Module:

[0167] Receive various data from the data acquisition module in real time, and clean and organize them. For example, if the speed data from a certain vehicle fluctuates abnormally (the instantaneous speed reaches 200 km / h, which is obviously unreasonable), it will be adjusted to a reasonable range after data verification and correction. At the same time, convert data in different formats into a unified format, such as unifying time data into the format of "YYYY-MM-DDHH:MM:SS" and encoding the cargo category information (clothing is coded as 01).

[0168] The time series analysis algorithm (ARIMA model) and machine learning prediction algorithm (multivariate linear regression algorithm) are used to predict and analyze logistics demand. Taking the logistics demand prediction of a certain region as an example, for the ARIMA model, the time series data of the cargo transportation volume in the region in the past year are analyzed, and the model parameters are determined through the autocorrelation function and partial autocorrelation function. , and then use methods such as maximum likelihood estimation to estimate other parameters and predict the logistics demand of the region in different time periods in the next week; for the multivariate linear regression algorithm, extract the characteristics of cargo transportation volume in the historical time period, local GDP growth trend, seasonal factors (such as increased logistics demand for clothing in summer), etc., use the least squares method to fit the model parameters, and predict the logistics demand of different regions;

[0169] Generate a logistics demand report based on the forecast results. The report shows that the logistics demand peak in the region will be from 9:00 to 11:00 a.m. from Monday to Friday next week. The cargo transportation volume is expected to increase by 30% compared with usual times. The main demand goods are clothing and electronic products, etc. The report is then output to the intelligent scheduling module.

[0170] Intelligent scheduling module:

[0171] Receive the logistics demand report from the demand analysis module and the resource status information from the logistics business system; for example, if you know that a transportation task requires the transportation of 5 tons of electronic products from City A to City B within 48 hours, and you also know that there are currently 3 available vehicles, Vehicle 1 has a load capacity of 8 tons and is located near a warehouse in the suburbs of City A; Vehicle 2 has a load capacity of 6 tons and is located in the center of City A; Vehicle 3 has a load capacity of 10 tons and is located in another warehouse 100 kilometers away from City A; in terms of warehousing, the warehouse in City A has sufficient inventory of electronic products, and the cargo space near the exit has enough space to store and retrieve goods;

[0172] Calculate the task priority according to the formula Calculate the task priority; assuming that based on the transportation distance, traffic conditions and average vehicle speed, it is estimated that the transportation time for vehicle 1 is 10 hours, the transportation time for vehicle 2 is 8 hours, and the transportation time for vehicle 3 is 12 hours, and the task requirement time is 48 hours, then vehicle 2 has the highest task priority, followed by vehicle 1, and vehicle 3 last; sort the transportation tasks according to priority;

[0173] Perform resource allocation and task planning; select vehicle 2 to perform the transportation task, and arrange to load the goods from the cargo space near the exit according to the storage resource situation; calculate the value increment under the resource allocation plan, such as selecting vehicle 2 can reduce transportation costs (because of its shorter distance and lower fuel consumption), improve customer satisfaction (high probability of on-time delivery), determine the amount of resources required to perform the task (vehicle 2 usage time, fuel consumption estimation, storage space occupancy, etc.), and calculate the value increment according to the state transition equation. Update the optimal value and determine the final resource allocation and task execution order;

[0174] During the execution of logistics tasks, the vehicle location, transportation progress, changes in warehouse inventory, and traffic conditions are continuously monitored; if it is found that vehicle 2 encounters a sudden traffic accident during transportation, resulting in serious road congestion and an estimated delay of 6 hours, the task priority and resource allocation plan will be re-evaluated at this time; another emergency transportation task originally planned to be executed later will be brought forward, and vehicle 1 will be reallocated to perform the task, and the route of vehicle 2 will be re-planned to choose an alternative route to bypass the accident section, and timely communication and coordination will be carried out with the vehicle maintenance team, warehouse management personnel, and transportation personnel;

[0175] Generate a detailed logistics scheduling plan, including the departure time (adjusted time), driving route (newly planned alternative route), estimated arrival time (after delay), and storage resource usage plan (cargo outbound and storage location arrangement) of vehicle 2, and output it to the route planning module, transportation execution team, warehouse management team and other relevant departments;

[0176] Path planning module:

[0177] Receive the transportation task information assigned by the intelligent scheduling module, such as the task of vehicle 2 to transport 5 tons of electronic products from the central warehouse of city A to city B, as well as the relevant attributes of the goods (electronic products, fragile, need to be transported with care); at the same time, obtain real-time traffic conditions information from the traffic information system, including the congestion level of each section along the way from city A to city B (such as the congestion coefficient of a certain section is 0.8), road construction conditions (none), and traffic control information (none);

[0178] Call the Geographic Information System (GIS) database to obtain map data including road networks, intersection information, and road attributes (such as road grade, speed limit, number of lanes, etc.); locate the transportation starting point and end point on the map, and screen the passable road network based on the cargo attributes and vehicle type, excluding roads with height restrictions, weight restrictions, and poor road conditions;

[0179] Using distance as the main cost metric, the Euclidean distance formula is used to calculate the actual distance cost of each road segment in the path. , and gradually accumulate from the starting point; for the estimated cost from the current node to the target node , calculated using the Manhattan distance formula, and based on real-time traffic information Make corrections; if a road section is severely congested, appropriately increase the estimated cost of that section; calculate the total cost of each node ;

[0180] Add the starting point as the initial node to the priority queue, according to the total cost of the node Sort; take the node with the smallest total cost from the priority queue as the current node for expansion, check the adjacent nodes, and calculate their , and value, and add it to the priority queue, recording the parent node information of each node; repeat the expansion process until the target node is added to the priority queue and becomes the current node;

[0181] When the target node becomes the current node, an initial path is generated by backtracking the parent node chain from the target node to the starting point; the initial path is optimized and checked. If a section is found to be a one-way street, an alternative path is found to bypass the section; the optimized path is evaluated by comprehensively considering factors such as transportation time, transportation cost and cargo safety, and the path with the smallest total cost is selected as the final optimal transportation path, including the coordinates of each node on the path, driving direction, estimated driving time and other information, and output to the transportation execution department and intelligent scheduling module;

[0182] Monitoring and feedback module:

[0183] Real-time data interaction is maintained through the wireless communication network with the transport vehicles, warehouse management system and the terminal devices of the logistics business related staff; for example, the location information of vehicle 2 is received every 10 minutes, and the update formula is used Update the vehicle position coordinates, where minute, and For vehicles in Axis and The velocity component in the axial direction;

[0184] At the same time, the transportation progress and warehouse inventory changes are monitored; if it is found that the transportation progress of vehicle 2 is 30 minutes later than the planned progress, or the inventory of a certain type of goods in the warehouse is lower than the safety inventory threshold (for example, there are only 100 pieces of electronic products left in stock, and the safety inventory threshold is 200 pieces), by comparing the actual data with the preset plan data, it is determined that there is an abnormal situation that deviates from the preset plan;

[0185] When an abnormal situation is detected, the intelligent scheduling module is immediately fed back with feedback information including the abnormality type (such as delayed vehicle transportation progress, insufficient warehouse inventory), detailed data (such as the vehicle's current location coordinates, delay time, actual inventory quantity, etc.) and the time of occurrence;

[0186] Optimization module:

[0187] Based on genetic algorithm, the scheduling scheme output by the intelligent scheduling module is globally optimized; the fitness function is set ,in, , , , , corresponding to logistics operation cost, average transportation time, on-time delivery rate of goods, and idle resource rate;

[0188] The initial scheduling plan is obtained from the intelligent scheduling module as the initial population individuals. The gene code of each individual represents a different combination of resource allocation and task arrangement. For example, binary code is used to represent vehicle selection, integer code is used to represent storage space allocation, etc., and an initial population size of 50 individuals is generated.

[0189] For each individual in the population, calculate its fitness value according to the set fitness function; parse the individual gene code to obtain the details of the scheduling plan, calculate the logistics operation cost, average transportation time, on-time delivery rate, resource idle rate and other indicators under the plan, and substitute them into the fitness function to calculate the fitness value;

[0190] The roulette wheel selection method is used to select some individuals as parents according to their individual fitness values, and single-point crossover and mutation operations are performed; the crossover probability is set to 0.7 and the mutation probability is set to 0.1; for the mutation operation, if the gene position represents the number of vehicles to be allocated, the mutation is performed according to the uniform distribution. (like ) Randomly generate new values ​​to replace the values ​​of the original gene positions;

[0191] Replace some individuals in the original population with offspring individuals generated by crossover and mutation operations to form a new population. After 100 iterations, select the individual with the highest fitness value from the final population as the optimal scheduling solution and output it to the intelligent scheduling module. The intelligent scheduling module re-allocates logistics resources and replans tasks based on the optimized solution.

[0192] 3. Implementation Effect

[0193] After a period of operation, the logistics company's logistics costs have been reduced by 15%, mainly due to the rational allocation of vehicle resources and the optimization of transportation routes, which have reduced empty mileage and fuel consumption; the average transportation time has been shortened by 20%, and through intelligent scheduling and route planning, congested sections have been effectively avoided, improving transportation efficiency; the on-time delivery rate of goods has increased by 25%, ensuring that customers can receive the goods on time, improving customer satisfaction, and enhancing the company's competitiveness in the market.

[0194] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A logistics management system based on intelligent scheduling, characterized by: include: Data collection module: used to collect historical data and real-time data in logistics business in real time. The historical data and real-time data involve cargo information, transportation vehicle information, warehouse inventory information, customer demand information and traffic condition information; Demand analysis module: connected to the data collection module, based on the collected historical data and real-time data, uses data analysis algorithms to predict and analyze logistics demand, and determines the logistics service demand in different time periods and different regions; Intelligent scheduling module: connected to the demand analysis module, it uses a preset intelligent scheduling algorithm to dynamically plan transportation tasks, allocate transportation resources, and plan storage use strategies based on the logistics demand output by the demand analysis module and the real-time vehicle, storage and traffic resource conditions, so as to achieve optimal configuration of logistics resources; Path planning module: connected to the intelligent scheduling module, for the transportation task assigned by the intelligent scheduling module, based on the real-time traffic information and the transportation starting and ending locations, the optimal transportation path is calculated through the path planning algorithm; Monitoring and feedback module: connected to the intelligent scheduling module, path planning module and data acquisition module, real-time monitoring of parameters in the process of logistics task execution, including vehicle location, transportation progress, and changes in warehouse inventory. When an abnormal situation deviates from the preset plan, feedback is given to the intelligent scheduling module; Optimization module: connected to the intelligent scheduling module, it performs global optimization on the scheduling scheme output by the intelligent scheduling module based on the genetic algorithm. The genetic algorithm conducts a comprehensive evaluation of the scheduling scheme in terms of cost, efficiency and service quality according to the set fitness function, so as to continuously optimize the scheduling scheme and optimize the overall performance of the logistics system.

2. A logistics management system based on intelligent scheduling according to claim 1, characterized in that: The data acquisition module includes on-board sensors installed on transport vehicles, inventory monitoring equipment in warehouses, data interfaces in logistics business systems, and communication units for data connection with traffic management departments. The data collected by the on-board sensors include vehicle speed. Mileage , the average speed of the vehicle is calculated by the following formula: ,in, Indicates the mileage of each section. Indicates the driving time corresponding to each segment of mileage. To calculate the number of mileage segments, the inventory monitoring device uses the inventory quantity calculation formula To monitor warehouse inventory in real time, is the current inventory quantity, is the initial inventory quantity, is the incoming quantity, and O is the outgoing quantity.

3. According to the logistics management system based on intelligent scheduling according to claim 1, it is characterized by: The data analysis algorithm used by the demand analysis module includes at least one of a time series analysis algorithm and a machine learning prediction algorithm to accurately predict logistics demand at different time scales; if the autoregressive integrated moving average (ARIMA) model is used for time series analysis and prediction, the formula is: ； in, For current needs, To predict demand, are the autoregressive model parameters, is the autoregressive order, is the sliding average order, is the mean, is the sliding average model parameter, is the white noise error, Indicates at a point in time Push forward Time series The value of Indicates at a point in time Push forward White noise error value per time unit; If the multivariate linear regression algorithm in machine learning is used for prediction, the prediction model formula is: ； in, For the predicted logistics demand, is the intercept term, For each characteristic variable The corresponding coefficients are, is the number of characteristic variables, and the value of each coefficient is determined by fitting the training data to achieve demand forecasting.

4. The logistics management system based on intelligent scheduling according to claim 1 is characterized by: The intelligent scheduling algorithm adopts a hybrid algorithm of greedy algorithm and dynamic programming algorithm to achieve efficient allocation of overall logistics resources among different tasks while meeting the priority requirements of transportation tasks; in the greedy algorithm part, the task priority The calculation formula is: ； in, For the task The demand time, For the task The transportation time is shortened and resources are allocated in descending order according to the task priority; In the dynamic programming algorithm part, suppose Indicates in front After the tasks are assigned resources, the resource status is The optimal value when , its state transfer equation is: ； in, Indicates execution of The amount of resources required for each task, Indicates execution of The value increment brought by each task is calculated through continuous iteration. To determine the optimal resource allocation plan.

5. The logistics management system based on intelligent scheduling according to claim 1 is characterized by: When the path planning algorithm calculates the optimal transportation path, its heuristic function is: ,in, Represents the actual cost from the starting point to the current node, Represents the estimated cost from the current node to the target node, Represents the total cost of the current node; if distance is used as the cost measurement standard, By Euclidean distance formula The cumulative calculation is as follows: and are the coordinates of the two points, Using the Manhattan distance formula Make an estimate, among which, Front node coordinates, is the target node coordinate, by continuously comparing the value to determine the optimal path.

6. The logistics management system based on intelligent scheduling according to claim 1 is characterized by: The monitoring and feedback module exchanges data with the transport vehicles, warehouse management systems and terminal devices of logistics business-related personnel through the wireless communication network to ensure that abnormal situations are promptly communicated to the relevant responsible parties and the dispatch center. The monitoring and feedback module monitors the vehicle position through GPS positioning and locates the coordinates. Every time interval Update, the update formula is: ； in, for The coordinates of the moment, for The coordinates of the moment, and They are Always Axis and The velocity component in the axial direction is used to determine whether there is a deviation by comparing the actual coordinates with the coordinates corresponding to the preset planned route. If the deviation exceeds the set threshold, it is judged as an abnormality and feedback is given to the intelligent scheduling module.

7. The logistics management system based on intelligent scheduling according to claim 1 is characterized by: In the genetic algorithm of the optimization module, the fitness function is set as: ； in, are the weight coefficients corresponding to logistics operation cost, average transportation time, on-time delivery rate of goods, and resource idle rate, respectively. Quality(x), Utilization Respectively represent the scheduling scheme The corresponding indicator value under In the crossover operation of the genetic algorithm, a single-point crossover method is used, and the parent individual and The intersection point is , then the generated offspring individuals and The gene fragment combination is as follows: ； ； In the mutation operation, let the length of the individual gene sequence be , the mutation probability is , for each gene position , according to probability Perform mutation operation. If the gene bit represents the number of resource allocations, the mutation is performed according to the uniform distribution. Randomly generate new values ​​to replace the original gene position values, where and To set the lower and upper limits of the variation range, through continuous iterative genetic operations, the individual pros and cons are evaluated according to the fitness function, and the optimal scheduling plan is screened out.

Images