Vehicle scheduling management method and system based on logistics order

Abstract

The invention provides a logistics order-based vehicle scheduling management method and system, and relates to the technical field of data processing, and the method comprises the steps: determining a preliminary vehicle scheduling scheme through employing a genetic algorithm; in the transportation process, data streams from all vehicles are received and processed in real time; dynamically calculating a corresponding vehicle path by using the data stream and the preliminary vehicle scheduling scheme; according to the vehicle path, updating a vehicle scheduling scheme, and determining a corresponding key logistics node; based on the vehicle path and the key logistics node, adopting an ant colony algorithm to optimize the deployment position of the Internet of Things sensor so as to obtain an optimized sensor deployment scheme; through an optimized sensor deployment scheme, key parameters are monitored in real time, and the key parameters comprise temperature, humidity and vibration. According to the invention, more efficient and more intelligent vehicle scheduling management can be realized.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a vehicle dispatching management method and system based on logistics orders. Background Art

[0002] Vehicle dispatch management is the core link in logistics operations, which directly affects logistics efficiency, cost control and customer satisfaction. However, some traditional vehicle dispatch management methods have many defects and cannot meet the high efficiency, flexibility and intelligence requirements of modern logistics industry.

[0003] For example, some traditional methods are based on fixed transportation routes and plans for scheduling, lacking data-driven insights. This approach ignores the information contained in historical transportation data, such as transportation patterns, peak periods, and areas with frequent delays, which are crucial for optimizing scheduling solutions.

[0004] Secondly, some traditional methods have some shortcomings in processing real-time data. During the transportation process, key factors such as vehicle location, changes in cargo demand and real-time road conditions will continue to change, and some traditional methods are difficult to respond to these changes in real time, resulting in rigid scheduling plans and failure to maximize transportation efficiency. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a vehicle dispatching management method and system based on logistics orders, which can achieve more efficient and intelligent vehicle dispatching management.

[0006] In order to solve the above technical problems, the technical solution of the present invention is as follows: In a first aspect, a vehicle dispatching management method based on logistics orders comprises: Step 11: Analyze historical transportation data to identify transportation patterns, peak periods, and areas with frequent delays hidden in the data; Step 12, using the transportation mode, peak period and frequently delayed area, combined with the number of vehicles, load, time window and cost, a genetic algorithm is used to determine the preliminary vehicle dispatch plan; Step 13, during the transportation process, real-time receiving and processing of data streams from each vehicle, including GPS location, changes in cargo demand, and real-time road condition information; using the data stream and the preliminary vehicle scheduling plan, dynamically calculating the corresponding vehicle path; based on the vehicle path, updating the vehicle scheduling plan and determining the corresponding key logistics nodes; Step 14: Based on the vehicle path and key logistics nodes, the deployment location of the IoT sensors is optimized using the ant colony algorithm to obtain an optimized sensor deployment plan; through the optimized sensor deployment plan, key parameters are monitored in real time, including temperature, humidity and vibration; Step 15: Perform anomaly detection on key parameters from IoT sensors. If an abnormality is found, an early warning is triggered.

[0007] Furthermore, historical transportation data is analyzed to identify transportation patterns, peak periods, and areas with frequent delays hidden in the data, including: Obtain historical transportation data, including transportation time, route, vehicle information, cargo information and weather conditions; clean and preprocess the historical transportation data to obtain preprocessed historical transportation data; Using cluster analysis, the pre-processed historical transportation data is used to identify the routing and time window patterns of cargo transportation; The time dimension in the pre-processed historical transportation data is analyzed by peak detection method to identify the peak period of transportation activities; The delay records in the preprocessed historical transportation data are analyzed by hot spot analysis to obtain the areas where delays occur.

[0008] Furthermore, using the transportation mode, peak hours and frequently delayed areas, combined with the number of vehicles, loading capacity, time window and cost, a genetic algorithm is used to determine the preliminary vehicle scheduling plan, including: Randomly generate multiple initial solutions as the initial population, each solution represents a vehicle scheduling plan; Calculate a fitness value for each solution based on the transportation cost, time window constraint, and load constraint; The roulette wheel selection method is used to select the corresponding individuals according to the fitness value to enter the next generation, and the partial matching crossover method is used to generate new solutions; the positions of genes are randomly exchanged to generate a new population, and the selection, crossover and mutation are repeated until the preset number of iterations is reached to obtain the final individual, which is used as the preliminary vehicle scheduling plan.

[0009] Furthermore, the calculation formula of the fitness value is: ; in, , and is the weight coefficient; It represents the transportation cost per unit distance; Indicates The distance travelled by the vehicle; is the number of vehicles; Indicates that the vehicle has arrived at The actual time of each node; and Respectively represent The earliest and latest arrival times allowed for each node; is the number of logistics nodes; Indicates The actual load of the vehicle; Indicates The maximum load of the vehicle; is the number of vehicles used; and; , and is the index value.

[0010] Furthermore, the corresponding vehicle paths are dynamically calculated using the data stream and the preliminary vehicle scheduling scheme, including: Extract the GPS location information of all current vehicles from the data stream; obtain real-time traffic information; The entire logistics network is abstracted into a graph model, where nodes represent logistics nodes and edges represent driving paths. Logistics nodes include warehouses, distribution centers, and customer locations. Using real-time traffic information, dynamically update the weights of each edge in the graph model; For each vehicle, use its current GPS position as the starting point and initialize the Dijkstra algorithm in the graph model; create a distance array to store the estimated shortest distance from the starting point to each node; during initialization, set the distance of the starting point to 0 and the distances of the remaining nodes to infinity, indicating that no path has been found; create a set of visited nodes, initialized to an empty set; Select the unvisited node corresponding to the starting point from the distance array as the current node; traverse all neighbor nodes of the current node and update the estimated distance from the current node to the neighbor node; if the distance to the neighbor node through the current node is shorter, update the distance of the neighbor node; mark the current node as visited and add it to the set of visited nodes until all nodes are visited to get the final path; According to the final path, obtain detailed path information from the starting point to the target node; According to the detailed path information from the starting point to the target node, combined with the time window and vehicle load in the vehicle scheduling plan, the final path is screened to obtain the updated path information.

[0011] Furthermore, based on the vehicle path, the vehicle dispatching plan is updated and the corresponding key logistics nodes are determined, including: Analyze the updated path information to obtain analysis results, including the advantages, disadvantages and potential risk points of each path; Adjust the preliminary vehicle dispatch plan based on the strengths, weaknesses and potential risk points of each route, including determining the specific departure and arrival times of each vehicle, assigning corresponding cargo loading and unloading tasks, and updating the driver arrangements and vehicle maintenance plans in the dispatch plan; Obtain real-time and historical traffic flow data for each section of the road on the updated route information; Analyze real-time and historical traffic flow data to identify the time periods and locations where traffic peaks occur; Calculate the frequency of each path used by vehicles and identify frequently used paths; Identify the intersection areas between different paths and evaluate the connectivity of each intersection area in the network, that is, the number of different paths connected to each intersection area; Based on the time periods and locations where traffic peaks occur, frequently used routes, and the number of different routes connected to each intersection area, a comprehensive assessment is conducted to identify intersection points, which are key logistics nodes.

[0012] Furthermore, based on vehicle paths and key logistics nodes, the ant colony algorithm is used to optimize the deployment locations of IoT sensors to obtain an optimized sensor deployment solution, including: Analyze the driving paths of all vehicles, including the starting point, end point, key logistics nodes passed through, and the length and driving time of each path segment; Obtain the location of key logistics nodes and cargo flow volume; Define the objective function, and randomly generate corresponding ants, i.e. candidate solutions, based on the vehicle path and key logistics nodes. Each ant represents a sensor deployment solution. Define the movement rules of ants during the search process. Ants choose the next deployment location based on the pheromone concentration and heuristic information at the current location. After each iteration, the pheromone is updated according to the ant's search results and the evaluation value of the objective function. Repeat the ant search and pheromone update until the preset number of iterations is reached to obtain the optimized sensor deployment plan.

[0013] Furthermore, real-time traffic information includes road congestion, traffic accidents and road maintenance.

[0014] In a second aspect, a vehicle dispatching management system based on logistics orders includes: An analytics module that analyzes historical transportation data to identify transportation patterns, peak periods, and areas of frequent delays hidden in the data; A determination module is used to determine the preliminary vehicle dispatch plan using genetic algorithm by taking into account the transportation mode, peak period and frequently delayed area, the number of vehicles, load, time window and cost; The processing module is used to receive and process data streams from each vehicle in real time during the transportation process, including GPS location, changes in cargo demand and real-time road condition information; dynamically calculate the corresponding vehicle path using the data stream and the preliminary vehicle scheduling plan; update the vehicle scheduling plan based on the vehicle path and determine the corresponding key logistics nodes; The optimization module is used to optimize the deployment location of IoT sensors based on vehicle paths and key logistics nodes using an ant colony algorithm to obtain an optimized sensor deployment plan. Through the optimized sensor deployment plan, key parameters are monitored in real time, including temperature, humidity, and vibration. The early warning module is used to detect anomalies in key parameters from IoT sensors and trigger an early warning if an abnormality is found.

[0015] According to a third aspect, a computing device includes: one or more processors; The storage device is used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the method described.

[0016] In a fourth aspect, a computer-readable storage medium stores a program, and when the program is executed by a processor, the method described is implemented.

[0017] The above scheme of the present invention includes at least the following beneficial effects.

[0018] Analyze historical transportation data to identify transportation patterns, peak periods, and areas with frequent delays hidden in the data. This step helps to understand various patterns and potential problems in the transportation process, such as which periods are peak periods and which areas are prone to delays, so as to better plan transportation routes and times, reduce delays, and improve efficiency.

[0019] By comprehensively considering multiple factors and using genetic algorithms, this step can generate a preliminary but efficient vehicle scheduling plan, which helps enterprises maximize transportation efficiency, reduce costs, and meet time window requirements under limited resources. Real-time data processing makes vehicle scheduling more flexible and accurate, and can adjust vehicle routes and scheduling plans in time according to actual conditions to cope with traffic congestion or changes in cargo demand, greatly improving the reliability and response speed of transportation.

[0020] By optimizing the deployment location of sensors, environmental parameters such as temperature, humidity and vibration at key logistics nodes can be more effectively monitored, which not only helps to ensure the safety and quality of goods, but also can detect and solve problems in a timely manner and improve customer satisfaction. The anomaly detection mechanism can detect potential problems in a timely manner, such as damaged goods or environmental changes. Once an early warning is triggered, measures can be taken quickly to prevent the problem from escalating. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a flow chart of a vehicle dispatching management method based on logistics orders provided by an embodiment of the present invention.

[0022] Figure 2 It is a schematic diagram of a vehicle dispatching management system based on logistics orders provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0024] like Figure 1 As shown, an embodiment of the present invention proposes a vehicle dispatching management method based on logistics orders, and the method comprises the following steps: Step 11: Analyze historical transportation data to identify transportation patterns, peak periods, and areas with frequent delays hidden in the data; Step 12, using the transportation mode, peak period and frequently delayed area, combined with the number of vehicles, load, time window and cost, a genetic algorithm is used to determine the preliminary vehicle dispatch plan; Step 13, during the transportation process, real-time receiving and processing of data streams from each vehicle, including GPS location, changes in cargo demand, and real-time road condition information; using the data stream and the preliminary vehicle scheduling plan, dynamically calculating the corresponding vehicle path; based on the vehicle path, updating the vehicle scheduling plan and determining the corresponding key logistics nodes; Step 14: Based on the vehicle path and key logistics nodes, the deployment location of the IoT sensors is optimized using the ant colony algorithm to obtain an optimized sensor deployment plan; through the optimized sensor deployment plan, key parameters are monitored in real time, including temperature, humidity and vibration; Step 15: Perform anomaly detection on key parameters from IoT sensors. If an abnormality is found, an early warning is triggered.

[0025] In an embodiment of the present invention, step 11, by analyzing historical data, regular patterns and potential problem areas in the transportation process can be identified, which helps to predict future transportation needs, plan resources in advance, and reduce transportation risks during peak hours or areas with frequent delays, thereby improving overall transportation efficiency and reliability. Step 12, the application of genetic algorithms can comprehensively consider multiple complex factors and generate preliminary but efficient vehicle scheduling plans. This helps enterprises maximize the utilization efficiency of transportation resources while meeting time windows and cost control, and achieve more optimized transportation planning. Step 13, real-time data processing and dynamic path planning enable enterprises to respond immediately to changes in the transportation process, such as traffic conditions, cargo demand adjustments, etc. This enhances the flexibility and accuracy of vehicle scheduling, helps reduce delays, and improves customer satisfaction. Step 14, by optimizing sensor deployment through ant colony algorithms, enterprises can more effectively monitor the environmental conditions of key logistics nodes. This helps to ensure the safety and quality of goods during transportation, promptly discover potential problems, and take measures to prevent losses, thereby improving the overall quality of logistics services. Step 15, the anomaly detection mechanism can monitor and identify abnormal conditions in the transportation process in real time, such as temperature fluctuations, humidity changes, or vibration anomalies. Once an anomaly is detected, an early warning is triggered in a timely manner, enabling enterprises to respond quickly to ensure the integrity and transportation safety of goods while reducing potential losses and risks.

[0026] In an embodiment of the present invention, the above step 11, analyzing the historical transportation data to identify the transportation mode, peak period and frequently delayed area hidden in the data, may include: Step 111, obtain historical transportation data, including transportation time, route, vehicle information, cargo information and weather conditions; clean and pre-process the historical transportation data to obtain pre-processed historical transportation data, specifically including: extracting historical transportation data from the enterprise's transportation management system, GPS tracking system, cargo management system and other databases. These data usually include the start time and end time of the transportation task, the transportation route (starting point, end point and passing points), the vehicle information used (model, number, etc.), the cargo information (type, weight, volume, etc.) and the weather conditions during transportation; the cleaning process includes removing duplicate records, processing missing values ​​(for example, filling missing data by interpolation, deletion or estimation), correcting outliers (identifying and processing those obviously unreasonable or erroneous data points), data formatting (ensuring that all data follows a unified format and standard), and data type conversion (such as converting text data into numerical data for numerical analysis); after cleaning, some pre-processing operations are required, such as normalization (scaling the data to a uniform range, such as between 0 and 1) and standardization (converting the data to have zero mean and unit variance).

[0027] Step 112, using cluster analysis to identify the pre-processed historical transportation data to identify the path and time window patterns of cargo transportation, specifically including: extracting key features from the cleaned data, such as the coordinate points of the transportation route, the start and end time of transportation, etc.; defining a suitable distance metric (such as Euclidean distance) to measure the similarity between different transportation tasks; using a clustering algorithm (such as K-means) to group the transportation tasks, and these groups should be based on the similarity of transportation paths and / or time windows; analyzing the clustering results to identify common transportation paths and time window patterns, which can be used to optimize future transportation plans and resource allocation.

[0028] Step 113, analyzing the time dimension in the pre-processed historical transportation data by a peak detection method to identify the peak period of transportation activities, specifically including: arranging the historical transportation data in chronological order to form a time series data set; applying a peak detection algorithm (such as a threshold-based method) to identify the peak periods in the time series, which usually correspond to the time periods with the most frequent transportation activities.

[0029] Step 114, analyzing the delay records in the pre-processed historical transportation data by a hot spot analysis method to obtain the areas where delays occur, specifically including: extracting all records related to delays from the cleaned historical transportation data, including information such as the location, time and cause of the delay; using spatial statistical methods (such as kernel density estimation, Getis-Ord Gi* statistics, etc.) to analyze the spatial distribution of delay records, so as to identify hot spots where delays occur, which are usually places where problems or congestion are prone to occur during transportation.

[0030] In an embodiment of the present invention, step 111, cleaning and preprocessing data can ensure the accuracy and reliability of the analysis results. By removing invalid data, correcting erroneous data and standardizing the data format, the enterprise can obtain a clean and consistent data set. Step 112, cluster analysis can help enterprises discover common paths and time window patterns in cargo transportation. This helps enterprises optimize transportation routes, avoid congested areas, and choose to transport during periods with smoother traffic, thereby improving transportation efficiency and punctuality. Step 113, identifying the peak period of transportation activities can help enterprises reasonably allocate transportation resources and avoid a large number of transportation activities during peak hours, thereby reducing congestion and delays. In addition, enterprises can also use peak period data to adjust transportation plans to improve transportation efficiency and customer satisfaction. Step 114, hot spot analysis can help enterprises accurately identify areas where delays often occur. This enables enterprises to provide early warnings and take corresponding measures, such as optimizing routes, adding alternative transportation plans, etc., to reduce the risk of delays in these areas.

[0031] In the embodiment of the present invention, the above step 12 uses the transportation mode, peak period and frequently delayed area, combined with the number of vehicles, loading capacity, time window and cost, to determine the preliminary vehicle scheduling plan using a genetic algorithm, including: Step 121, randomly generate multiple initial solutions as the initial population, each solution represents a vehicle scheduling solution, specifically including: determining the population size (i.e., the number of initial solutions), which is usually selected based on the complexity of the problem and the availability of computing resources; for each initial solution (i.e., each vehicle scheduling solution), randomly assigning vehicles to different transportation tasks, ensuring that each task has at least one vehicle assigned and the load of each vehicle does not exceed its maximum capacity; recording the specific vehicle allocation and task allocation of each initial solution, which will serve as the starting point of the genetic algorithm.

[0032] Step 122, according to the transportation cost, time window constraint and load constraint, a fitness value is calculated for each solution, specifically including: for each solution, the total transportation cost is calculated, which usually includes vehicle operation cost, driver cost, etc.; check whether each solution meets the time window constraints of all tasks, that is, whether the vehicle completes the loading and unloading of the task within the specified time. If not, a corresponding penalty cost is given; verify whether the vehicle load in each solution exceeds its maximum capacity. If it exceeds, a corresponding penalty cost is also given; combine the above costs to calculate a total fitness value for each solution.

[0033] Step 123, using the roulette wheel selection method, select the corresponding individuals according to the fitness value to enter the next generation, and use the partial matching crossover method to generate a new solution; randomly exchange the positions of genes to generate a new population, repeat the selection, crossover and mutation until the preset number of iterations is reached to obtain the final individual, and use the final individual as the preliminary vehicle scheduling plan, which specifically includes: According to the fitness value of each solution, the probability of its selection is calculated. The probability of each individual being selected is proportional to its fitness value, which means that individuals with higher fitness values ​​will have a higher probability of selection; when calculating specifically, the normalized fitness value of each individual can be used as its probability of selection. For example, if the normalized fitness value of an individual is 0.2, then its probability of selection can be set to 0.2, and the lower the fitness value, the higher the probability of selection. Use the roulette wheel selection method, that is, randomly select solutions to enter the next generation based on the above probability. This means that better solutions (solutions with lower fitness values) have a greater chance of being selected. For each pair of solutions (i.e., parents) selected, a new solution (i.e., offspring) is generated using the partial matching crossover (PMX) method. In PMX, two cutting points are randomly selected, and the genes of the corresponding fragments between the parents are exchanged, and then the possible duplicate genes are repaired through the mapping relationship to ensure the legitimacy of the new solution; random mutation operations are performed on the newly generated solutions. This usually involves randomly selecting a portion of the genes in the solution (i.e., the allocation of vehicles or tasks) and randomly adjusting or exchanging them; combining the newly generated solution with some of the original excellent solutions into a new population for the next round of iteration; repeating the above selection, crossover, and mutation process until the preset number of iterations is reached or other stopping conditions are met (such as finding the optimal solution that satisfies all constraints). During the iteration process, the optimal solution found is recorded and updated. After the iteration termination condition is reached, the optimal individual in the current population is selected as the preliminary vehicle scheduling solution.

[0034] In an embodiment of the present invention, by comprehensively considering a variety of transportation parameters and constraints, the genetic algorithm can generate an efficient vehicle scheduling solution, thereby optimizing resource allocation, reducing costs, and improving the reliability and punctuality of transportation. By randomly generating an initial population, the diversity of the search space can be increased to avoid falling into a local optimal solution. This method helps to find a better vehicle scheduling solution in a global scope. By calculating the fitness value of each solution, the pros and cons of each vehicle scheduling solution can be quantitatively evaluated. This helps to select better individuals for reproduction according to the fitness value in the subsequent genetic algorithm process, thereby gradually approaching the optimal solution. The roulette selection method can perform probabilistic selection according to the fitness value of the individual, ensuring that excellent individuals have a greater chance of being selected and passed on to the next generation. The partial matching crossover method and the operation of randomly exchanging gene positions can increase the diversity of the population and avoid the algorithm from converging to the local optimal solution too early. By repeating the process of selection, crossover and mutation, the genetic algorithm can continuously approach the global optimal solution and finally obtain an efficient vehicle scheduling solution. This method can significantly improve transportation efficiency, reduce costs, and meet various constraints.

[0035] In the embodiment of the present invention, in the above step 122, the calculation formula of the fitness value is: ; in, , and is the weight coefficient; It represents the transportation cost per unit distance; Indicates The distance travelled by the vehicle; is the number of vehicles; Indicates that the vehicle has arrived at The actual time of each node; and Respectively represent The earliest and latest arrival times allowed for each node; is the number of logistics nodes; Indicates The actual load of the vehicle; Indicates The maximum load of the vehicle; is the number of vehicles used; and; , and is the index value.

[0036] In the embodiment of the present invention, the formula achieves comprehensive optimization of the vehicle scheduling solution by combining multiple key factors (transportation cost, time window and load). This comprehensive evaluation method helps to find solutions that perform well in multiple dimensions. By adjusting the weight coefficient , and , you can flexibly adjust the importance of the optimization goal according to actual needs. For example, if cost is the primary consideration, you can increase If arriving on time is more important, you can increase The formula takes into account in detail the driving distance of each vehicle, the arrival time of each node, and the load of each vehicle, all of which are key factors affecting the efficiency of vehicle scheduling. By accurately calculating these factors, a more accurate fitness value can be obtained, thereby more accurately evaluating the quality of each solution. By minimizing the fitness value, the algorithm can automatically find scheduling solutions that perform well in terms of cost, time, and load, thereby improving the overall efficiency of corporate logistics and transportation. Because the formula takes into account multiple constraints, the generated scheduling solution may have better robustness and adaptability when facing uncertainties and changes in actual situations (such as traffic congestion, weather changes, etc.).

[0037] In the embodiment of the present invention, in the above step 13, during the transportation process, data streams from each vehicle are received and processed in real time, including GPS locations, changes in cargo demand, and real-time road condition information, including: During the transportation process, each transport vehicle is equipped with corresponding equipment (such as GPS trackers, sensors, on-board computers, etc.), which can collect and send various data of the vehicle in real time. The management system maintains connection with these vehicles through wireless communication (such as 4G / 5G network, satellite communication, etc.), so that it can receive data streams from each vehicle in real time.

[0038] Data processing content: GPS location data: GPS (Global Positioning System) data provides the precise current location of vehicles. By analyzing GPS data, management systems can track the paths of vehicles and ensure that they follow the intended routes. GPS data can also be used to estimate vehicle arrival times, thereby better coordinating loading and unloading operations and other logistics activities.

[0039] Changes in cargo demand: During transportation, there may be temporary changes in cargo demand, for example, customers may need to increase or decrease the amount of cargo. Sensors and management systems on the vehicle can monitor the status of the cargo in real time, including quantity, temperature, and humidity, to ensure that the cargo is safe and meets customer needs. By processing these changes in demand in real time, transportation plans can be adjusted in a timely manner to improve logistics efficiency and customer satisfaction.

[0040] Real-time traffic information: Traffic information includes factors that may affect transportation efficiency, such as traffic congestion, road construction, and accidents. By receiving and analyzing this information in real time, the management system can provide drivers with the best route suggestions to avoid congested areas and reduce transportation time and costs. In some cases, real-time traffic information can also be used to predict potential delays and notify customers in advance.

[0041] Real-time data processing enables logistics management systems to respond quickly, optimize transportation processes, and improve overall efficiency. It helps reduce unnecessary delays and costs while improving customer service and satisfaction.

[0042] In the embodiment of the present invention, in the above step 13, the corresponding vehicle path is dynamically calculated using the data stream and the preliminary vehicle scheduling plan, including: Extract the GPS location information of all current vehicles from the data stream; obtain real-time traffic information, specifically including: receiving the location data sent by the vehicle in real time through the vehicle-mounted GPS device. These data usually include longitude, latitude and timestamp; collect these GPS data to the central server or cloud platform through the data interface or network transmission; the server parses the received data and extracts the current location information of each vehicle. By establishing a connection with the traffic information center, map service provider or other relevant data sources, obtain real-time road congestion, traffic accidents, road maintenance and other road condition information, which is provided in the form of data streams, and the system needs to be able to parse and process these data in real time.

[0043] The entire logistics network is abstracted into a graph model, where nodes represent logistics nodes and edges represent driving paths. Logistics nodes include warehouses, distribution centers, and customer locations. Specifically, the key nodes in the logistics network are determined. These nodes usually include warehouses, distribution centers, and customer locations. These nodes are hubs or destinations for the flow of goods. Each key node is abstracted as a vertex in the graph. In graph theory, a vertex represents an object or entity, so in this scenario, each warehouse, distribution center, or customer location is considered an independent vertex. Each vertex can be represented by a unique identifier, such as a number, name, or coordinates, to facilitate subsequent calculations and identification. Analyze the roads or transportation routes in the logistics network, which are the actual paths for goods to move from one node to another. Each road or transportation route is abstracted as an edge in the graph. In graph theory, an edge is used to connect two vertices, indicating that there is a certain relationship between them or that they can be directly connected; each edge can be represented by the identifiers of the two vertices it connects, and also contains some additional attributes, such as distance, travel time, cost, etc.

[0044] Using real-time traffic information, the weights of each edge in the graph model are dynamically updated. Real-time traffic information includes road congestion, traffic accidents, and road repairs. Specifically, it includes: establishing real-time connections with data sources such as traffic information centers and map service providers to obtain the latest traffic information, which includes road congestion, traffic accident reports, road repair notices, etc.; parsing the acquired traffic information into a format that the system can understand, and extracting key information, such as the affected roads, congestion level, and expected recovery time. Matching the parsed traffic information with the edges in the graph model. This usually involves matching the actual roads with the edges in the graph to ensure that each edge reflects the real-time status of the road it represents. Based on the matching results and the specific content of the traffic information, the weights of each edge in the graph model are dynamically adjusted. The weight usually represents the cost required to travel the edge, which can be time, distance, cost, etc. For example, if a road is congested, the weight of the edge corresponding to the road (such as travel time) will increase to reflect the delays that may be encountered in actual driving. Similarly, if a road is undergoing maintenance or a traffic accident occurs, its weight will be adjusted accordingly. The weights are adjusted in real time to ensure that the path planning algorithm is always calculated based on the latest traffic information.

[0045] For each vehicle, use its current GPS position as the starting point and initialize the Dijkstra algorithm in the graph model. Create a distance array to store the estimated shortest distance from the starting point to each node. When initializing, set the distance of the starting point to 0 and the distances of the remaining nodes to infinity, indicating that no path has been found. Create a set of visited nodes, initialized to an empty set, including: The real-time location information of each vehicle is obtained through the GPS system, including longitude and latitude. These location information are converted into corresponding nodes in the logistics network graph model, and the node closest to the current location of the vehicle is selected as the starting point. For each vehicle, the node corresponding to its current location is used as the starting point to initialize the Dijkstra algorithm in the graph model. The Dijkstra algorithm is an algorithm for finding the shortest path from a single source in a weighted graph. Create a distance array (or distance vector) to store the shortest distance estimates from the starting point to each node in the graph model. The size of the array should be equal to the number of nodes in the graph model. Set the distance of the starting point to 0, indicating that the distance from the starting point to itself is 0. Set the distance of the remaining nodes to infinity (or use a sufficiently large number to represent it), indicating that the path from the starting point to these nodes has not yet been found. Create an empty set to record the nodes that have been visited by the Dijkstra algorithm. During the execution of the algorithm, this set will be gradually filled until it contains all the nodes in the graph model.

[0046] Select the unvisited node corresponding to the starting point from the distance array as the current node; traverse all neighbor nodes of the current node and update the estimated distance from the current node to the neighbor node; if the distance to the neighbor node through the current node is shorter, update the distance of the neighbor node; mark the current node as visited and add it to the set of visited nodes until all nodes are visited to get the final path, including: Select a node from the distance array that is closest to the starting point and has not been visited as the current node. This is usually achieved by traversing the distance array and finding the minimum value; getting all neighbor nodes of the current node (that is, nodes directly connected to the current node) and traversing these neighbor nodes; for each neighbor node, calculate the distance from the starting point through the current node to the neighbor node. If this distance is shorter than the distance to the neighbor node currently stored in the distance array, update the value in the distance array; mark the current node as visited, indicating that the shortest path from the starting point to the node has been found. This is usually achieved by adding the current node to the visited node set; repeat the steps until all nodes in the graph model have been visited. This can be determined by checking whether the size of the visited node set is equal to the total number of nodes in the graph model. When all nodes have been visited, the distance array stores the shortest distance from the starting point to each node. Based on this distance array and the intermediate results of the algorithm (such as parent node pointers or path records), the specific path from the starting point to the target node can be backtracked.

[0047] According to the final path, get the detailed path information from the starting point to the target node, including: after the Dijkstra algorithm is executed, the distance array stores the shortest distance from the starting point to each node; in order to obtain detailed path information, it is necessary to backtrack from the target node. If the predecessor node of each node (that is, from which node to the current node) is recorded during the execution of the algorithm, then this predecessor relationship can be used to backtrack to the starting point to obtain the complete path. In the process of backtracking, record the nodes and edges passed, and this information constitutes the detailed path from the starting point to the target node. It can be saved as a sequence of nodes and edges, or as a series of coordinate points (if the node has geographic location information). Format the extracted path information for subsequent processing and display.

[0048] According to the detailed path information from the starting point to the target node, combined with the time window and vehicle load in the vehicle scheduling plan, the final path is screened to obtain the updated path information, including: Reading the data in the vehicle scheduling plan, including the time window of each vehicle (i.e., the time period during which the vehicle is available) and the vehicle load limit. For each path, check whether it meets the time window of the vehicle. If the estimated travel time of the path exceeds the time window of the vehicle, then this path needs to be screened out. According to the vehicle load limit, analyze whether the total weight of the goods to be delivered on each path exceeds the standard. If it exceeds the standard, then this path also needs to be screened out. After screening out the paths that meet the time window and load limit, they can be further optimized based on other indicators (such as total travel time, total distance, cost, etc.). This involves reordering the nodes on the path or choosing a more efficient route. Update the screened and optimized path information to the system. These paths are now not only the shortest, but also take into account the actual operating conditions and restrictions of the vehicle. Output the updated path information to the vehicle scheduling system or the driver for actual logistics distribution operations.

[0049] In the embodiment of the present invention, by receiving and processing vehicle data streams and road condition information in real time, the system can adjust the vehicle path in real time to ensure that the vehicle always travels on the optimal route. This dynamic adjustment capability greatly improves the flexibility and efficiency of logistics transportation. By updating road condition information in real time and recalculating the shortest path, the system can avoid congestion and accident areas, reduce transportation time and fuel consumption, and thus reduce transportation costs. Accurate path planning and real-time adjustment help ensure that goods are delivered on time, reduce delays, and thus improve customer satisfaction. At the same time, the system can also filter the path according to the time window to further meet the customer's timeliness requirements. Path filtering combined with the time window and vehicle load in the vehicle scheduling plan can ensure that each vehicle is effectively utilized and avoid resource waste, such as empty driving or overloading. By abstracting the entire logistics network into a graph model and using algorithms to calculate the path, managers can more intuitively understand the operating status of the entire network. Real-time data processing and dynamic path planning enable the system to quickly adjust in the face of traffic accidents and road repairs, enhancing the robustness and response capabilities of the system.

[0050] In the embodiment of the present invention, in the above step 13, the vehicle dispatching plan is updated according to the vehicle path, and the corresponding key logistics nodes are determined, including: Analyze the updated path information to obtain analysis results, including the advantages, disadvantages and potential risk points of each path, including: collect updated path information, which includes detailed data such as the starting point, end point, nodes along the way, total distance, and estimated driving time of each path; for each path, use the collected data to evaluate its advantages (such as short distance, good traffic conditions, etc.) and disadvantages (such as congested areas, complex road conditions, etc.); combine the geographic information system (GIS) and historical data to identify possible risk points on the path, such as accident-prone areas, construction sections, weather-affected areas, etc.

[0051] According to the advantages, disadvantages and potential risk points of each route, adjust the preliminary vehicle dispatch plan, including determining the specific departure and arrival time of each vehicle, assigning corresponding cargo loading and unloading tasks, and updating the driver arrangement and vehicle maintenance plan in the dispatch plan. Specifically, it includes: determining the preliminary vehicle dispatch plan, understanding the basic conditions such as vehicle allocation, departure time, arrival time, etc. in the existing plan; adjusting the specific departure time and estimated arrival time of each vehicle according to the results of the route analysis to avoid congested periods and take advantage of periods with good traffic conditions; assigning reasonable cargo loading and unloading tasks to each vehicle according to the distribution needs of the goods and the characteristics of the route to ensure the optimization of vehicle load and distribution efficiency; updating the driver scheduling plan in the dispatch plan to ensure that the driver has sufficient rest time; at the same time, arranging a vehicle maintenance plan to prevent vehicle failures during driving; Obtain real-time and historical traffic flow data for each section of the updated path information; analyze real-time and historical traffic flow data to find out the time period and location of traffic peaks; calculate the frequency of each path used by vehicles and identify frequently used paths, including: selecting reliable real-time and historical traffic flow data sources, such as the data platform of the traffic management department, third-party traffic information service providers, etc.; using API interfaces, data crawlers or purchasing services to obtain real-time and historical traffic flow data for each section of the updated path information; cleaning, format conversion and standardization of the obtained data; analyzing the time series characteristics of historical traffic flow data to find out the time period when traffic peaks occur (such as rush hour); combining with geographic information systems, analyzing the spatial distribution of traffic flow and identifying locations with concentrated traffic (such as transportation hubs, commercial centers, etc.); combining the results of time series and spatial distribution analysis to determine the time period and location of traffic peaks; summarizing the driving records of all vehicles, including vehicle ID, driving route, driving time and other information; counting the number of times each path is used by vehicles and calculating the frequency of use of each path; identifying frequently used paths based on the frequency of use.

[0052] Identify the intersection areas between different paths and evaluate the connectivity of each intersection area in the network, that is, the number of different paths connected to each intersection area, including: Determine all updated path information, the starting point, end point, and passing nodes of each path. By comparing the nodes of different paths, find the nodes that two or more paths pass through together. These nodes are intersections. For example, path A passes through nodes 1, 2, 3, and 5, and path B passes through nodes 2, 4, and 6, then node 2 is an intersection. Preliminary assessment of the importance of intersections is based on the number of paths connected to the intersection and the busyness of the paths (such as cargo volume, vehicle frequency, etc.). The more paths connected and the busier the paths, the higher the importance of the intersection. Centered on each intersection, a reasonable area range is defined as the intersection area based on the actual geographical, traffic, and logistics conditions. This range can be a fixed geographical distance (such as a circular area with a radius of 1 km) or a dynamic range based on the actual road network and logistics activities. This allows you to intuitively see the location and relationship of each intersection area. For each intersection area, count the number of different paths directly connected to it. This includes paths starting from the area and paths entering the area. Connectivity can be quantified in a variety of ways. For example, the number of connected paths can be used directly as an indicator of connectivity, or the connectivity can be weighted by considering factors such as the busyness and capacity of the paths. Based on the connectivity data and other relevant indicators (such as cargo throughput, vehicle stop frequency, etc.), set standards for evaluating the logistics importance of the intersection area. Conduct a comprehensive assessment of each intersection area to determine its logistics importance level. This can be a qualitative assessment (such as high, medium, and low importance) or a quantitative scoring system. Based on the results of the logistics importance assessment of the intersection area, support is provided for decisions such as logistics network optimization, resource allocation, and risk management. For example, for intersection areas with high logistics importance, priority can be given to adding storage facilities, optimizing traffic flow lines, or strengthening safety management measures.

[0053] Based on the time period and location of traffic peaks, frequently used routes, and the number of different routes connected to each intersection area, a comprehensive assessment is conducted to identify intersection points, which are key logistics nodes, including: Collect and record the time periods and locations of peak traffic on each route. This data can be obtained from traffic management departments, road sensors, or third-party data providers. Count the frequency of use of each route. This can be done by analyzing vehicle driving records or using GPS tracking data. Identify each intersection and count the number of different routes connected to each intersection; Sort the traffic peaks on each path and mark the time periods and locations with the highest traffic. High traffic peaks mean that the location carries a large number of vehicles in a specific time period, so these locations are more important in the logistics network. Sort by the frequency of use of the paths and identify frequently used paths. Frequently used paths are usually the main channels in logistics activities and have an important impact on overall logistics efficiency. Count the number of different paths connected to each intersection area and sort them. The more paths connected to the intersection area, the better the connectivity of the area and the more intensive the logistics activities.

[0054] Combining the evaluation results of the above three aspects, a comprehensive score is assigned to each intersection. For example, different weights can be given according to the ranking of traffic peak, path usage frequency and intersection area connectivity, and the total score can be calculated.

[0055] Set threshold or select top N: Set a comprehensive score threshold, and the intersection points exceeding this threshold are regarded as key logistics nodes. Or directly select the top N intersection points with the highest comprehensive scores as key logistics nodes.

[0056] In the embodiments of the present invention, by optimizing the route selection and vehicle scheduling, the transportation time and cost are reduced, thereby improving the overall logistics efficiency. By identifying the disadvantages and risk points on the route, potential problems can be prevented in advance and the possibility of accidents can be reduced. Identifying and optimizing key logistics nodes can improve the response capacity of the entire logistics network and ensure that the network can still operate efficiently when facing challenges. Through the meticulous management of vehicles, drivers and maintenance plans, the optimal allocation of resources is achieved and the efficiency of resource utilization is improved.

[0057] In the embodiment of the present invention, in the above step 14, based on the vehicle path and the key logistics nodes, the deployment location of the IoT sensor is optimized by using the ant colony algorithm to obtain an optimized sensor deployment solution, including: Analyze the driving paths of all vehicles, including the starting point, end point, key logistics nodes passed through, and the length and travel time of each path segment. Specifically, it includes: collecting the driving data of all vehicles from the vehicle tracking system, GPS records or other relevant data sources; using the map matching algorithm to map the GPS points to the actual road, so as to extract the specific driving path of each vehicle; recording the starting point and end point of each path, as well as all the key logistics nodes passed on the path (such as transportation hubs, warehouses, etc.); calculating the length and travel time of each path segment, which can be achieved through GIS software.

[0058] Obtain the location and cargo flow of key logistics nodes, including: Use the previously extracted vehicle travel path data, which includes information about each node that the vehicle passes through; ensure that the data contains sufficient time span and vehicle samples. For each key logistics node, use GIS system, map API or on-site survey to obtain its precise geographic location information, such as longitude and latitude. Record the collected geographic location information in the key logistics node list; collect relevant information for each key logistics node, such as floor area, building scale, facilities and equipment, etc.; evaluate the scale of each node based on the collected data and on-site survey, and use a unified evaluation standard or scoring system to compare different nodes; determine the data source for statistical cargo turnover, including logistics companies' operational data, warehouse management system records, etc. Collect cargo turnover data for each key logistics node within a period of time (such as a month, a quarter or a year) from the determined data source, clean and organize the collected data to ensure the accuracy and completeness of the data, and count the total cargo turnover of each key logistics node within a specified time period based on the cleaned data.

[0059] Define the objective function. Based on the vehicle path and key logistics nodes, randomly generate corresponding ants, namely candidate solutions. Each ant represents a sensor deployment solution, which includes: based on the identification and analysis of key logistics nodes, determine the node set where sensors may be deployed; for each ant (candidate solution), randomly select a part of the nodes as the deployment location of the sensor. At the same time, randomly assign the number and type of sensors to each selected node; for each ant (candidate solution), calculate the corresponding objective function value according to the objective function calculation formula defined above. The calculation formula of the objective function is: ; in, is the number of key logistics nodes; is the number of sensor types that can be selected; is the unit cost of deploying a sensor of type j at node i; is the number of sensors of type j deployed on node i in deployment scheme x; is the number of segments of all vehicle paths; is the length of path segment k; is an indicator function that if path segment k is covered by a sensor in deployment scenario x, then ,otherwise ; and Represents the coefficient.

[0060] Define the movement rules of ants during the search process. Ants select the next deployment location based on the pheromone concentration and heuristic information at the current location. After each iteration, the pheromone is updated based on the ant's search results and the evaluation value of the objective function. Repeat the ant search and pheromone update until the preset number of iterations is reached to obtain the optimized sensor deployment plan, which includes: Set an initial pheromone concentration value for all possible paths (i.e., connections between key logistics nodes), usually a small positive number; initialize a pheromone concentration matrix, where each element represents the pheromone concentration on the corresponding path. Create a certain number of ants (candidate solutions), each ant represents a potential sensor deployment solution, and assign a starting node to each ant, which can be randomly selected or determined according to specific rules (such as the node with the largest cargo flow).

[0061] During the search process, each ant chooses the next deployment location according to the following rules: For the node where the current ant is located, calculate its transition probability to all adjacent nodes. The transition probability consists of two parts: pheromone concentration and heuristic information (such as the inverse of the node's cargo turnover, path length, etc.).

[0062] The calculation formula for the transition probability can be: ; in, and is a parameter that regulates the importance of pheromone and heuristic information; is the transition probability from node i to node j; is the pheromone concentration on path (i, j); is the value of the heuristic information from node i to node j, which is inversely proportional to factors such as path length and cargo turnover; represents the set of all adjacent nodes that the ant can choose at node i; It is an identifier used to refer to a specific ant in the algorithm. In the ant colony algorithm, there will be multiple ants searching at the same time, and each ant will independently choose a path according to the algorithm rules.

[0063] According to the calculated transfer probability, the roulette wheel selection method is used to determine the next node to be visited. Move to the selected node and update the current position of the ant. At each visited node, decide whether to deploy sensors and the type and number of sensors according to certain rules (such as random, fixed number, or node characteristic-based strategies). After each ant completes a path search, the local pheromone update is performed on the path it has passed to reduce the probability of other ants choosing the same path and increase the diversity of the search.

[0064] After the ant completes a path search, it performs a local pheromone update on the path it has passed. The formula for the local update can be: ; in, It is the coefficient of pheromone volatilization, usually ranging from 0 to 1; is the initial pheromone concentration value; in this formula, the first (on the right side of the equation) represents the pheromone concentration on the current path, that is, the concentration before the update. (On the left side of the equation) represents the new pheromone concentration on the path after local update. This formula means that the pheromone on the path will retain a part The original concentration, plus a portion Initial concentration of pheromone. This is done to simulate the volatilization and replenishment of pheromones to maintain the diversity of the search. After all ants complete a round of search, the best few ants are selected for global pheromone update based on the deployment plan of each ant and the evaluation value of the objective function.

[0065] The update formula is: ; in, is the volatility coefficient of the global update, is the pheromone increment calculated based on the best ant’s path and the objective function value; in this formula, the first represents the original pheromone concentration on the current path (before updating). represents the pheromone concentration after global update. This formula means that during the global update process, a portion of the pheromone on the path will be retained. Original concentration, and increase The pheromone increment is calculated based on the optimal path. In this way, the algorithm can gradually emphasize the paths that constitute the optimal solution during the iteration process. Repeat the ant search and pheromone update process until the preset number of iterations is reached or other stop conditions are met (such as objective function value convergence, reaching the maximum running time, etc.). After the search is completed, the best ant (that is, the deployment plan with the best objective function value) is selected as the final sensor deployment plan, and the final deployment plan and its corresponding objective function value are output.

[0066] In an embodiment of the present invention, by optimizing the deployment location of sensors, the flow of goods at key logistics nodes can be monitored more accurately, thereby adjusting logistics strategies in a timely manner, reducing unnecessary waiting and delays, and improving overall logistics efficiency. Reasonable sensor deployment can avoid installing sensors in unnecessary locations, thereby saving hardware costs and maintenance costs. At the same time, by reducing redundant and ineffective sensor deployment, energy consumption and network communication costs can also be reduced. Deploying sensors at key logistics nodes can ensure that the collected data is more representative and accurate, because these nodes are important hubs for the flow of goods. High-quality data helps companies make more informed decisions, improve operational efficiency and customer satisfaction. The sensor deployment scheme optimized by the ant colony algorithm can ensure full coverage of key logistics nodes, reduce monitoring blind spots, and thus improve the reliability and stability of the entire logistics monitoring system. The ant colony algorithm can find the optimal solution among many possible deployment schemes, so that sensor resources are reasonably allocated and utilized, avoiding waste or shortage of resources. The optimized sensor deployment scheme provides logistics companies with more comprehensive operational data, which helps companies achieve intelligent management and improve response speed and decision-making efficiency. By deploying sensors at key logistics nodes, the status of goods and the transportation environment can be monitored in real time, potential risks can be discovered in a timely manner, and measures can be taken to prevent and control them, ensuring the safety and reliability of the logistics process.

[0067] In the embodiment of the present invention, in the above step 14, key parameters are monitored in real time through an optimized sensor deployment scheme, and the key parameters include temperature, humidity and vibration, including: According to the sensor deployment scheme previously optimized by the ant colony algorithm, temperature, humidity and vibration sensors are installed at key locations. The data collected by the sensors will be transmitted to the central data processing system in real time via wireless or wired means. The received raw data will be pre-processed, including data cleaning, format conversion and standardization, so as to carry out subsequent key parameter analysis and anomaly detection.

[0068] In the embodiment of the present invention, the above step 15 performs abnormal detection on the key parameters from the IoT sensor, and triggers an early warning if an abnormality is found, including: For the three key parameters of temperature, humidity and vibration, normal range thresholds are set. For example, the temperature may be considered normal between 20°C and 25°C, the humidity may be between 40% and 60%, and the vibration amplitude may be within a certain range. IoT sensors collect temperature, humidity and vibration data in real time and send this data to the central processing system. After the central processing system receives the data, it immediately compares it with the preset thresholds. If the data of any parameter exceeds its normal range, it is considered abnormal. Once abnormal data is detected, the system automatically triggers an early warning. The early warning can be sent in a variety of ways, such as SMS, email or APP push notification, to ensure that relevant personnel can be informed of the abnormal situation immediately. The early warning information should include the name of the abnormal parameter (temperature, humidity or vibration), the abnormal value, the time when the abnormality occurred, and the recommended initial response measures.

[0069] like Figure 2 As shown, an embodiment of the present invention further provides a vehicle dispatching management system based on logistics orders, comprising: An analytics module that analyzes historical transportation data to identify transportation patterns, peak periods, and areas of frequent delays hidden in the data; A determination module is used to determine the preliminary vehicle dispatch plan using genetic algorithm by taking into account the transportation mode, peak period and frequently delayed area, the number of vehicles, load, time window and cost; The processing module is used to receive and process data streams from each vehicle in real time during the transportation process, including GPS location, changes in cargo demand and real-time road condition information; dynamically calculate the corresponding vehicle path using the data stream and the preliminary vehicle scheduling plan; update the vehicle scheduling plan based on the vehicle path and determine the corresponding key logistics nodes; The optimization module is used to optimize the deployment location of IoT sensors based on vehicle paths and key logistics nodes using an ant colony algorithm to obtain an optimized sensor deployment plan. Through the optimized sensor deployment plan, key parameters are monitored in real time, including temperature, humidity, and vibration. The early warning module is used to detect anomalies in key parameters from IoT sensors and trigger an early warning if an abnormality is found.

[0070] It should be noted that the system is a system corresponding to the above method, and all implementation methods in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0071] The embodiment of the present invention further provides a computing device, comprising: a processor, a memory storing a computer program, wherein when the computer program is executed by the processor, the method described above is executed. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0072] The embodiment of the present invention also provides a computer-readable storage medium storing instructions, which, when executed on a computer, enable the computer to execute the method described above. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0073] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A vehicle dispatching management method based on logistics orders, characterized in that: The method comprises: Step 11: Analyze historical transportation data to identify transportation patterns, peak periods, and areas with frequent delays hidden in the data; Step 12, using the transportation mode, peak period and frequently delayed area, combined with the number of vehicles, load, time window and cost, a genetic algorithm is used to determine the preliminary vehicle dispatch plan; Step 13, during the transportation process, real-time receiving and processing of data streams from each vehicle, including GPS location, changes in cargo demand, and real-time road condition information; using the data stream and the preliminary vehicle scheduling plan, dynamically calculating the corresponding vehicle path; based on the vehicle path, updating the vehicle scheduling plan and determining the corresponding key logistics nodes; Step 14: Based on the vehicle path and key logistics nodes, the deployment location of the IoT sensors is optimized using the ant colony algorithm to obtain an optimized sensor deployment plan; through the optimized sensor deployment plan, key parameters are monitored in real time, including temperature, humidity and vibration; Step 15: Perform anomaly detection on key parameters from IoT sensors. If an abnormality is found, an early warning is triggered.

2. A vehicle dispatching management method based on logistics orders according to claim 1, characterized in that: Analyze historical transportation data to identify transportation patterns, peak times, and areas of frequent delays hidden in the data, including: Obtain historical transportation data, including transportation time, route, vehicle information, cargo information and weather conditions; clean and preprocess the historical transportation data to obtain preprocessed historical transportation data; Using cluster analysis, the pre-processed historical transportation data is used to identify the routing and time window patterns of cargo transportation; The time dimension in the pre-processed historical transportation data is analyzed by peak detection method to identify the peak period of transportation activities; The delay records in the preprocessed historical transportation data are analyzed by hot spot analysis to obtain the areas where delays occur.

3. A vehicle dispatching management method based on logistics orders according to claim 2, characterized in that: Using transportation modes, peak hours and frequently delayed areas, combined with vehicle numbers, loads, time windows and costs, a genetic algorithm is used to determine a preliminary vehicle dispatching solution, including: Randomly generate multiple initial solutions as the initial population, each solution represents a vehicle scheduling plan; Calculate a fitness value for each solution based on the transportation cost, time window constraint, and load constraint; The roulette wheel selection method is used to select the corresponding individuals according to the fitness value to enter the next generation, and the partial matching crossover method is used to generate new solutions; the positions of genes are randomly exchanged to generate a new population, and the selection, crossover and mutation are repeated until the preset number of iterations is reached to obtain the final individual, which is used as the preliminary vehicle scheduling plan.

4. A vehicle dispatching management method based on logistics orders according to claim 3, characterized in that: The calculation formula of fitness value is: ； in, , and is the weight coefficient; It represents the transportation cost per unit distance; Indicates The distance travelled by the vehicle; is the number of vehicles; Indicates that the vehicle has arrived at The actual time of each node; and Respectively represent The earliest and latest arrival times allowed for each node; is the number of logistics nodes; Indicates The actual load of the vehicle; Indicates The maximum load of the vehicle; is the number of vehicles used; and; , and is the index value.

5. A vehicle dispatching management method based on logistics orders according to claim 4, characterized in that: Using data streams and preliminary vehicle scheduling solutions, dynamically calculate the corresponding vehicle paths, including: Extract the GPS location information of all current vehicles from the data stream; obtain real-time traffic information; The entire logistics network is abstracted into a graph model, where nodes represent logistics nodes and edges represent driving paths. Logistics nodes include warehouses, distribution centers, and customer locations. Using real-time traffic information, dynamically update the weights of each edge in the graph model; For each vehicle, use its current GPS position as the starting point and initialize the Dijkstra algorithm in the graph model; create a distance array to store the estimated shortest distance from the starting point to each node; during initialization, set the distance of the starting point to 0 and the distances of the remaining nodes to infinity, indicating that no path has been found; create a set of visited nodes, initialized to an empty set; Select the unvisited node corresponding to the starting point from the distance array as the current node; traverse all neighbor nodes of the current node and update the estimated distance from the current node to the neighbor node; if the distance to the neighbor node through the current node is shorter, update the distance of the neighbor node; mark the current node as visited and add it to the set of visited nodes until all nodes are visited to get the final path; According to the final path, obtain detailed path information from the starting point to the target node; According to the detailed path information from the starting point to the target node, combined with the time window and vehicle load in the vehicle scheduling plan, the final path is screened to obtain the updated path information.

6. A vehicle dispatching management method based on logistics orders according to claim 5, characterized in that: Update the vehicle dispatch plan based on the vehicle path and determine the corresponding key logistics nodes, including: Analyze the updated path information to obtain analysis results, including the advantages, disadvantages and potential risk points of each path; Adjust the preliminary vehicle dispatch plan based on the strengths, weaknesses and potential risk points of each route, including determining the specific departure and arrival times of each vehicle, assigning corresponding cargo loading and unloading tasks, and updating the driver arrangements and vehicle maintenance plans in the dispatch plan; Obtain real-time and historical traffic flow data for each section of the road on the updated route information; Analyze real-time and historical traffic flow data to identify the time periods and locations where traffic peaks occur; Calculate the frequency of each path used by vehicles and identify frequently used paths; Identify the intersection areas between different paths and evaluate the connectivity of each intersection area in the network, that is, the number of different paths connected to each intersection area; Based on the time periods and locations where traffic peaks occur, frequently used routes, and the number of different routes connected to each intersection area, a comprehensive assessment is conducted to identify intersection points, which are key logistics nodes.

7. A vehicle dispatching management method based on logistics orders according to claim 6, characterized in that: Based on vehicle paths and key logistics nodes, the ant colony algorithm is used to optimize the deployment locations of IoT sensors to obtain an optimized sensor deployment solution, including: Analyze the driving paths of all vehicles, including the starting point, end point, key logistics nodes passed through, and the length and driving time of each path segment; Obtain the location of key logistics nodes and cargo flow volume; Define the objective function, and randomly generate corresponding ants, i.e. candidate solutions, based on the vehicle path and key logistics nodes. Each ant represents a sensor deployment solution. Define the movement rules of ants during the search process. Ants choose the next deployment location based on the pheromone concentration and heuristic information at the current location. After each iteration, the pheromone is updated according to the ant's search results and the evaluation value of the objective function. Repeat the ant search and pheromone update until the preset number of iterations is reached to obtain the optimized sensor deployment plan.

8. A vehicle dispatching management method based on logistics orders according to claim 7, characterized in that: Real-time traffic information includes road congestion, traffic accidents and road repairs.

9. A vehicle dispatching management system based on logistics orders, characterized in that: Applied to the method according to any one of claims 1 to 8, comprising: An analytics module that analyzes historical transportation data to identify transportation patterns, peak periods, and areas of frequent delays hidden in the data; A determination module is used to determine the preliminary vehicle dispatch plan using genetic algorithm by taking into account the transportation mode, peak period and frequently delayed area, the number of vehicles, load, time window and cost; The processing module is used to receive and process data streams from each vehicle in real time during the transportation process, including GPS location, changes in cargo demand and real-time road condition information; dynamically calculate the corresponding vehicle path using the data stream and the preliminary vehicle scheduling plan; update the vehicle scheduling plan based on the vehicle path and determine the corresponding key logistics nodes; The optimization module is used to optimize the deployment location of IoT sensors based on vehicle paths and key logistics nodes using an ant colony algorithm to obtain an optimized sensor deployment plan. Through the optimized sensor deployment plan, key parameters are monitored in real time, including temperature, humidity, and vibration. The early warning module is used to detect anomalies in key parameters from IoT sensors and trigger an early warning if an abnormality is found.

10. A computing device, characterized in that include: one or more processors; A storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the method as claimed in any one of claims 1 to 8.

Images