Harbor district intelligent delivery management method based on Internet of Things

Abstract

The invention discloses a harbor district intelligent delivery management method based on the Internet of Things. The method comprises the following steps: S1, acquiring and preprocessing harbor district Internet of Things terminal data; s2, an improved dynamic beat rearrangement algorithm is adopted to generate a queuing sequence and port access suggested time; s3, constructing a crane position loading capacity model, and outputting an efficiency score table by using an improved pre-loading capacity perspective algorithm; s4, matching the queuing sequence with the scoring table, generating a multi-constraint scheduling matrix, and binding vehicles and crane positions through dynamic weights to generate a scheduling table; s5, an electronic waybill is generated according to the scheduling table, identity and crane position information is embedded, and verification and binding are completed; s6, after the vehicle enters and exits the electronic fence, the system identifies and compares the waybill information, and the vehicle is automatically released and a loading instruction is pushed after successful matching; and S7, continuously collecting data during task execution, and updating feedback in real time. According to the method, intelligent closed-loop scheduling of port dispatching is realized, the efficiency is improved, congestion is reduced, and resource allocation is optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent transportation and logistics scheduling, and in particular to an intelligent shipping management method for a port area based on the Internet of Things. Background Art

[0002] In modern port logistics management, especially when it comes to the transportation of bulk cargoes such as crude oil and chemicals, port shipping and dispatching systems play a crucial role. With the continued rise in transportation demand and operational density, traditional port shipping and dispatching management systems have gradually exposed various limitations, particularly in vehicle dispatching, operational coordination, safety control, and customer service response, which suffer from significant information lags, resource waste, and management blind spots. Currently, most ports still rely on manual record-keeping, on-site communication, or semi-automated single-point systems for vehicle entry and exit management, loading arrangements, and shipping dispatch. This model is not only inefficient but also highly susceptible to human interference, leading to frequent problems such as uneven loading resource utilization, vehicle congestion, long queues, and high safety risks.

[0003] Especially in short-haul transportation scenarios involving bulk raw materials, such as crude oil pickup business mainly based on truck transportation, the challenges faced by port scheduling are particularly complex. First, the layout of port terminals usually has physical bottlenecks such as limited access channels, a fixed number of loading crane positions, and narrow weighbridge circulation paths, which can easily lead to logistics congestion during peak hours; secondly, the queue scheduling before vehicles enter the port relies on manual notifications and the experience of dispatchers, and lacks a scientific basis for time allocation and resource matching. Vehicles often gather in clusters around the port area, causing increased regional traffic pressure and even causing safety hazards. In addition, most current systems cannot achieve effective linkage with driver terminals, lack accurate estimated arrival time control methods, and drivers cannot know the timing of entry and loading order in advance. They can only rely on verbal notifications or on-site observations, which greatly reduces transportation efficiency and customer satisfaction.

[0004] From a system integration perspective, existing port business systems generally suffer from module fragmentation, data silos, and inconsistent interfaces. Dispatching systems, metering systems, loading systems, and identity verification systems often rely on manual data transfer or USB flash drive import and export, lacking a unified data center and information synchronization mechanism. This decentralized management model results in a lack of real-time tracking and global optimization capabilities throughout the shipping process, making it impossible to form an end-to-end closed loop of shipping information. It is difficult for managers to grasp the overall operational status and to respond quickly and make command arrangements in emergencies. Furthermore, traditional systems are relatively rudimentary in terms of scheduling algorithms, using only static priority or first-come, first-served (FCFS) strategies. These systems ignore dynamic factors such as the vehicle's own status, differences in loading capacity, real-time port traffic, and equipment load, seriously affecting the scientific and rational nature of overall scheduling.

[0005] In recent years, with the development of emerging technologies such as the Internet of Things, artificial intelligence, and big data, some port areas have tried to introduce sensing devices and data analysis platforms in specific links. For example, license plate recognition systems are used to achieve automatic release, or video surveillance systems are used to assist in the loading confirmation operation. However, most of these improvements are local optimizations and have not yet formed an information fusion and intelligent scheduling system across links and the entire process. At the same time, most existing technologies rely on fixed rules and manual intervention and do not yet have the ability of data-driven adaptive scheduling. More importantly, there is currently a general lack of the ability to uniformly model and dynamically optimize the allocation among vehicle pick-up plans, actual operating status, and loading resources, and the advanced algorithms have not been fully utilized to accurately allocate port area resources.

[0006] At the scheduling level, traditional systems lack a mechanism for modeling and evaluating the capabilities of crane positions and cannot dynamically score and preferentially allocate crane positions according to real-time loading efficiency; in terms of queuing and sorting, a scheduling index model combining multi-dimensional factors such as time, space, equipment capabilities, and cargo demand has not been formed, resulting in a serious disconnection between vehicle sorting and operation efficiency. In addition, the system fails to effectively identify special business scenarios such as "secondary loading" and does not provide customized management capabilities such as VIP queue jumping, exception handling, and scheduling warnings. These problems make it difficult for traditional port area shipping management systems to adapt to the current increasing business volume and refined management requirements.

[0007] In summary, the existing technologies have the following main defects: First, there is a lack of an identity binding and task coordination mechanism among vehicles, drivers, and crane positions, which cannot guarantee the accuracy and safety of scheduling; second, a dynamic scheduling algorithm model has not been established, and the queuing order and crane position allocation cannot be intelligently optimized according to the real-time operating status and loading efficiency of the port area; third, the data cannot form a closed loop, and the scheduling results lack a real-time feedback and system self-learning mechanism, making it impossible to achieve continuous optimization; fourth, the adaptability to special business scenarios is insufficient, and the diversified and high-frequency loading scheduling requirements of modern ports cannot be met.

[0008] Therefore, how to provide an intelligent shipping management method for port areas based on the Internet of Things to achieve intelligent sorting of vehicle scheduling, optimal allocation of loading crane positions, accurate binding of task identities, and closed-loop feedback of the whole-process data is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0009] An object of the present invention is to propose an intelligent shipping management method for port areas based on the Internet of Things. The present invention makes full use of Internet of Things sensing technologies, big data analysis methods, and intelligent scheduling optimization algorithms, combines an improved dynamic beat rearrangement algorithm and a preloading capacity perspective algorithm, and details the whole process method for realizing intelligent vehicle sorting, dynamic gantry position scoring, multi-constraint matching scheduling, and task identity binding, and has the advantages of high shipping efficiency, high resource utilization rate, strong scheduling accuracy, full-process traceability, and strong system self-learning ability.

[0010] An intelligent shipping management method for port areas based on the Internet of Things according to an embodiment of the present invention includes the following steps: S1. Collect Internet of Things terminal data in the port area and perform preprocessing to construct a standardized shipping data set; S2. Based on the standardized shipping data set, use an improved dynamic beat rearrangement algorithm to dynamically sort all vehicles, and generate a vehicle queuing sequence and an estimated time for entering and leaving the port; S3. Based on the standardized shipping data set, construct a loading capacity model for each gantry position in the port area per unit time, and score it by an improved preloading capacity perspective algorithm to output an efficiency score table for each gantry position; S4. Cross-match the vehicle queuing sequence and the efficiency score table to form a multi-constraint matching matrix, and bind high-loading-capacity gantry positions and high-priority vehicles through a dynamic weight allocation mechanism to generate a queuing and scheduling table with priority labels; S5. Based on the queuing and scheduling table and the estimated time for entering and leaving the port, generate corresponding electronic waybills for all vehicles, embed the license plate number, the driver's ID number, and the allocated gantry position number, and perform identity verification and shipping task binding on the electronic waybills; S6. When a vehicle enters or exits the electronic fence of the port area, identify the vehicle position and compare it with the electronic waybill information. If the match is successful, trigger a release signal and push the loading instruction to the corresponding gantry position terminal; S7. During the vehicle's execution of the shipping task, continuously collect Internet of Things terminal data in the port area, update and record the feedback in real time, form a closed-loop self-learning scheduling mechanism, and complete the archiving of the entire process of shipping data.

[0011] Optionally, the Internet of Things terminal data includes the vehicle's real-time position, license plate number information, identity of the driving driver, gantry position allocation status, vehicle picking-up plan quantity, current port area traffic load, and historical loading records, and the preprocessing includes data cleaning and unified formatting.

[0012] Optionally, the S2 specifically includes: S21. Based on the standardized shipping data set, extract the estimated arrival time of each vehicle to be scheduled , picking-up plan quantity , last loading time The distance from the current location to the port area , as well as the current average loading time per beat in the port area and the real-time port area traffic load factor , and use the above data as the sorting input features; S22. Construct a dynamic beat rearrangement optimization objective function, and use an improved dynamic beat rearrangement algorithm to calculate the priority index of the vehicle: ; Among them, is the scheduling priority index of vehicle , and the higher the value, the higher the scheduling priority. is the picking-up plan quantity of vehicle , with the unit of ton. is the remaining distance of vehicle from the port area, with the unit of kilometer. is the expected arrival time of vehicle at the port area, with the unit of minute. is the last loading completion time of vehicle , with the unit of minute. is the current average beat time required for loading per vehicle in the port area, with the unit of minute / vehicle. is the current port area traffic load factor, which is a dimensionless number reflecting traffic and operation pressure. , , are the weighted weights of the picking-up quantity factor, distance factor and waiting duration factor respectively, satisfying , , are positive infinitesimal quantities introduced to avoid the denominator being zero; S23. Arrange the priority indices of all vehicles in descending order to obtain the vehicle queuing sequence , where , and at the same time, taking , , and as parameters, calculate the recommended in-out port time for each vehicle: ; Among them, is the recommended in-out port time of vehicle , is the current actual speed of vehicle , with the unit of kilometer / minute. is the empirical adjustment factor used to balance the beat time and the port area congestion level; S24. Output the vehicle queuing sequence Collection of Suggested Arrival and Departure Times 。

[0013] Optionally, S3 specifically includes: S31. Extract relevant data for each crane position based on the standardized shipping data set. The relevant data for each crane position includes the historical operation rounds of each crane position , the total loading volume per round , the corresponding operation duration , the current number of waiting vehicles , the number of abnormal events per unit time , and the equipment availability factor within the current operation cycle ; S32. Construct a port loading capacity model and define the theoretical loading capacity of each crane position per unit time: ; Among them, is the loading capacity per unit time of crane position , with the unit of tons / minute, is the loading volume of the th operation, is the actual completion duration of the th operation, is the historical operation rounds; S33. On the basis of the loading capacity model, use an improved pre-loading capacity perspective algorithm to correct the real-time efficiency of each crane position, form a multi-dimensional comprehensive scoring function, and output the efficiency score value: ; Among them, is the loading efficiency score value of crane position , is the equipment availability factor, and the value range represents the operating state of the equipment at the current crane position, is the number of abnormal events within the current cycle, is the current number of waiting vehicles, , , are the availability enhancement factor, abnormal penalty factor, and queuing penalty factor respectively, all of which are positive constants; S34. Output the set of loading efficiency scores for all crane positions 。

[0014] Optionally, S4 specifically includes: S41. Based on the vehicle queuing sequence , vehicle priority index , pick-up plan quantity and the collection of loading efficiency scores for crane positions , construct a standardized vehicle vector and a crane position vector : ; Among them, represents the remaining distance of vehicle from the port area, is a small positive number to prevent division by zero, is the crane position loading efficiency score value, is the crane position historical average single loading volume, is the number of vehicles currently waiting for operation at this crane position, is the maximum value of vehicle pick-up volume in the current batch, used as the normalization benchmark, is the maximum value function; S42. Combine the non-linear suppression function and use the weighted cosine similarity to calculate the comprehensive fitness of vehicle and crane position : ; Among them, is the comprehensive fitness score of vehicle and crane position , , are the feature values of the vehicle and the crane position in the th dimension respectively, is the feature weighting factor, satisfying , is a small value to avoid division by zero, is the loading volume deviation suppression factor, which controls the sensitivity of the difference between the vehicle and the historical capacity of the crane position; S43. To improve the overall resource utilization efficiency and fairness, introduce a dynamic weight allocation mechanism and construct a scheduling priority function: ; Among them, represents the current port area traffic load coefficient, a dimensionless number reflecting traffic and operation pressure, is the scheduling utility value of vehicle assigned to crane position , is the equipment availability coefficient of crane position , is the current port area traffic load coefficient, a dimensionless number reflecting traffic and operation pressure, is the maximum delay time in the current batch, used for normalization, is the delay time when the vehicle exceeds the recommended arrival time, is the actual entry and exit time, is the vehicle recommended entry and exit port time, , , is an adjustable coefficient to set optimization preferences; S44. Combine all scoring values to form a two-dimensional scheduling utility matrix , and calculate the vehicle-crane position binding relationship according to the maximum weight matching strategy to generate a queuing scheduling table with priority labels: ; Among them, is the scheduling table, which contains the scheduling relationship between each vehicle and the allocated crane position, is the th numbered vehicle to be scheduled, is the numbered crane position it is bound to, is the vehicle recommended entry and exit port time, is the corresponding scheduling utility value, , determined by the quantile value in the overall distribution, is the number of successfully matched vehicle-crane position pairs.

[0015] The beneficial effects of the present invention are: First of all, by constructing a standardized shipping data set and introducing real-time perception information of Internet of Things terminals, the present invention realizes the comprehensive perception and data fusion of key elements such as vehicles, drivers, crane positions, picking plans, and port area traffic, breaks the limitations of information islands and data fragmentation in traditional shipping management, provides an accurate, complete, and dynamic data foundation for the scheduling system, and effectively improves the visualization control ability of the port area over vehicle status and operation processes.

[0016] Secondly, the improved dynamic beat rearrangement algorithm and preloading capacity perspective algorithm proposed by the present invention are interlinked with each other, and a multi-constraint scheduling optimization model with "vehicle priority - crane position capacity matching - time coordination" as the core is constructed. It can not only calculate the vehicle entry order in the park in real time based on vehicle picking requirements, port area load status, and loading efficiency, but also dynamically identify and allocate the optimal crane position resources to ensure that high-priority vehicles can complete the loading operation quickly, significantly improving the utilization rate of loading beats and the accuracy of scheduling response.

[0017] Finally, the present invention constructs a dispatching feedback and waybill binding mechanism with a full-process closed loop. Through vehicle electronic fence identification, loading execution status feedback, and task data archiving, it realizes the full-process closed-loop control from dispatching generation, task binding, execution monitoring to feedback optimization. The system has the ability of self-learning, can continuously optimize algorithm parameters according to historical operation data, effectively reduce the shipping bottlenecks caused by human intervention, resource misallocation, and operation delays, and overall improve the intelligent level and operation safety of port area shipping management. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention, and do not constitute a limitation to the present invention. In the drawings: Figure 1 is a flowchart of a port area intelligent shipping management method based on the Internet of Things proposed by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] Now, the present invention will be further described in detail with reference to the drawings. These drawings are all simplified schematic diagrams, only showing the basic structure of the present invention in a schematic way, so they only show the components related to the present invention.

[0020] Refer to Figure 1 , a port area intelligent shipping management method based on the Internet of Things, includes the following steps: S1. Collect the Internet of Things terminal data in the port area and perform preprocessing to construct a standardized shipping data set; S2. Based on the standardized shipping data set, use an improved dynamic beat rearrangement algorithm to dynamically sort all vehicles, and generate a vehicle queuing sequence and the recommended time for entering and leaving the port; S3. Based on the standardized shipping data set, construct a loading capacity model of each quay position in the port area per unit time, and score it by an improved pre-loading capacity perspective algorithm to output an efficiency score table for each quay position; S4. Cross-match the vehicle queuing sequence and the efficiency score table to form a multi-constraint matching matrix, and bind high-loading-capacity quay positions and high-priority vehicles through a dynamic weight allocation mechanism to generate a queuing dispatching table with priority labels; S5. Based on the queuing dispatching table and the recommended time for entering and leaving the port, generate corresponding electronic waybills for all vehicles, embed the license plate number, driver's ID number, and the assigned quay position number, and perform identity verification and shipping task binding on the electronic waybills; S6. When the vehicle enters or exits the port area electronic fence, identify the vehicle position and compare the electronic waybill information. If the match is successful, trigger the release signal and push the loading instruction to the corresponding quay position terminal; S7. During the vehicle shipping mission, continuously collect the IoT terminal data in the port area, update and record the feedback in real time, form a closed-loop self-learning scheduling mechanism, and complete the archiving of the shipping data throughout the process.

[0021] By introducing the collection and standardized processing of IoT terminal data, this invention constructs a dynamically perceivable shipping data foundation. Combining with the improved dynamic beat rearrangement algorithm and preloading capacity perspective algorithm, it realizes the efficient matching of vehicle priority ranking and crane position resources, generates a queuing scheduling plan with timeliness and executability, and significantly improves the vehicle passing efficiency and loading resource utilization rate. Through the electronic waybill and identity information binding mechanism, it ensures the accurate correspondence between the vehicle and the task during the shipping process, effectively reducing the risks of human intervention and incorrect operations. The system has real-time feedback and self-learning capabilities during the scheduling execution process, can continuously optimize the scheduling strategy, and finally realizes the intelligent, automated and full-process closed-loop control of port area shipping management, with the beneficial effects of more accurate scheduling, more efficient operation and safer operation.

[0022] In this embodiment, the IoT terminal data includes the real-time position of the vehicle, license plate number information, identity of the driving driver, crane position allocation status, vehicle pick-up plan quantity, current port area traffic load, and historical loading records. The preprocessing includes data cleaning and unified format.

[0023] This invention clarifies that the collected data types cover key dimensions such as the real-time position of the vehicle, license plate number, driver identity, crane position status, port area traffic, and historical loading information. The preprocessing process includes data cleaning and format unification operations, ensuring the accuracy and real-time nature of the data.

[0024] In this embodiment, the specific steps of S2 are as follows: S21. Based on the standardized shipping data set, extract the expected arrival time , pick-up plan quantity , last loading time , distance from the current location to the port area of each vehicle to be scheduled, as well as the current average beat loading time and real-time port area traffic load coefficient of the port area, and use the above data as the sorting input features; S22. Construct an optimization objective function for dynamic beat rearrangement, and use the improved dynamic beat rearrangement algorithm to calculate the priority index of the vehicle: ; Wherein, is the scheduling priority index of vehicle , and the higher the value, the higher the scheduling priority. is the pick-up plan quantity of vehicle , with the unit of ton. For the vehicle The remaining distance to the port area, in kilometers For the vehicle The estimated arrival time at the port area, in minutes For the vehicle The time when the last loading of the vehicle was completed, in minutes The current average cycle time required for loading a single vehicle at the port area, in minutes / vehicle The current port area traffic load factor, a dimensionless number reflecting traffic and operation pressure and and Are the weighted weights of the pickup volume factor, distance factor, and waiting duration factor respectively, satisfying , and Are positive infinitesimal quantities introduced to avoid a zero denominator; S23. Arrange the priority indices of all vehicles in descending order to obtain the vehicle queuing sequence , where , and at the same time, using , , and as parameters, calculate the recommended in-out port time for each vehicle: ; where is the recommended in-out port time for vehicle , is the current actual speed of vehicle , in kilometers per minute is the empirical adjustment factor for balancing the cycle time and the port area congestion level; S24. Output the vehicle queuing sequence and the set of recommended in-out port times .

[0025] By introducing an improved dynamic cycle rearrangement algorithm, the present invention realizes the dynamic sorting of vehicles to be scheduled and the intelligent calculation of the recommended in-out port times. Compared with the traditional sorting logic based on first-come, first-served (FCFS), the present invention comprehensively considers multi-dimensional factors such as vehicle priority, pickup volume, location distance, and port area traffic, and can accurately estimate the most reasonable entry time and loading sequence, significantly reducing the queuing congestion phenomenon and improving the vehicle turnover efficiency and the continuity of the operation cycle.

[0026] In this embodiment, the S3 specifically includes: ​​S31. Extract the crane position-related data based on the standardized shipping data set. The crane position-related data includes, for each crane position the historical operation rounds the total loading volume per round the corresponding operation duration the current number of waiting vehicles the number of abnormal events per unit time and the equipment availability factor within the current operation cycle ; S32. Build a port loading capacity model and define the theoretical loading capacity of each crane position per unit time: ; where is the loading capacity per unit time of crane position , with the unit of tons per minute, is the loading volume of the th operation, is the actual completion duration of the th operation, is the historical operation rounds; S33. Based on the loading capacity model, use an improved pre-loading capacity perspective algorithm to correct the real-time efficiency of each crane position, form a multi-dimensional comprehensive scoring function, and output an efficiency score value: ; where is the loading efficiency score value of crane position , is the equipment availability factor, and its value range is , representing the operating state of the equipment at the current crane position, is the number of abnormal events within the current cycle, is the current number of waiting vehicles, , , are the availability enhancement factor, abnormal penalty factor, and queuing penalty factor respectively, all of which are positive constants; S34. Output the set of loading efficiency scores of all crane positions .

[0027] Based on the real-time and historical data of the port area, the present invention constructs a loading capacity model for each crane position and uses an improved pre-loading capacity perspective algorithm for dynamic scoring. This algorithm synthesizes indicators such as the loading rate, equipment availability, abnormal operation records, and current waiting pressure, and outputs a scoring table reflecting the true operation efficiency of the crane position. This effectively solves the problems of one-sided understanding and uneven distribution of loading resources in traditional scheduling, and promotes the precise matching transformation of crane position resource utilization from "average distribution" to "capacity-driven".

[0028] In this embodiment, the S4 specifically includes: S41. Based on the vehicle queuing sequence , vehicle priority index , pick-up plan quantity and the crane position loading efficiency scoring set , construct a standardized vehicle vector and a crane position vector : ; Among them, represents the remaining distance of vehicle from the port area, is a small positive number to prevent division by zero, is the loading efficiency scoring value of crane position , is the historical average single loading quantity of crane position , is the number of vehicles currently waiting for operation at this crane position, is the maximum value of the vehicle pick-up quantity in the current batch, used as the normalization benchmark, is the maximum value function; S42. Combine the non-linear suppression function and use the weighted cosine similarity to calculate the comprehensive fitness of vehicle and crane position : ; Among them, is the comprehensive fitness score of vehicle and crane position , , are the feature values of the vehicle and the crane position in the th dimension respectively, is the feature weighting factor, satisfying , is a small value to avoid division by zero, is the loading quantity deviation suppression factor, controlling the sensitivity of the historical ability difference between the vehicle and the crane position; ​​​​​S43. To improve the overall resource utilization efficiency and fairness, a dynamic weight allocation mechanism is introduced and a scheduling priority function is constructed: ; Among them, represents the current port area traffic load coefficient, which is a dimensionless number reflecting traffic and operation pressure, is for vehicle assigned to the quay position of the scheduling utility value, is the equipment availability coefficient of the quay position , is the current port area traffic load coefficient, which is a dimensionless number reflecting traffic and operation pressure, is the maximum delay time in the current batch, used for normalization, is the delay time of the vehicle exceeding the recommended arrival time, is the actual entry and exit time, is for vehicle of the recommended entry and exit time of the port, , , is an adjustable coefficient to set the optimization preference; S44. All score values form a two-dimensional scheduling utility matrix . According to the maximum weight matching strategy, the vehicle - quay position binding relationship is calculated to generate a queuing scheduling table with priority labels: ; Among them, is the scheduling table, which contains the scheduling relationship between each vehicle and the assigned quay position, is the th scheduled vehicle number, is the quay position number bound to it, is the recommended entry and exit time of vehicle , is the corresponding scheduling utility value, , determined by the quantile value in the overall distribution, is the number of successfully matched vehicle - quay position pairs.

[0029] Through the cross - matching of the vehicle queuing sequence and the quay position efficiency score, the present invention constructs a multi - constraint scheduling matrix and introduces a dynamic weight allocation mechanism. The system realizes the optimal binding logic of "high - capacity quay positions" and "high - priority vehicles". This mechanism supports real - time adjustment of the allocation strategy according to the actual operation situation, effectively improves the overall scheduling flexibility and task response ability, significantly reduces the vehicle waiting time and resource idle rate, and improves the overall operation efficiency of the loading operation in the port area.

[0030] Example 1: To verify the feasibility of the present invention in implementation, the present invention is applied to the crude oil shipping operation scenario of a large coastal port area. This port area mainly focuses on crude oil transshipment business. The number of vehicles for daily crude oil loading operations reaches more than 200. The port area is equipped with a set of import and export channel systems, two weighbridge systems, and four loading crane positions. Due to concentrated customers, high-frequency shipments, and dense vehicle arrivals, this port area has long faced problems such as difficult loading resource scheduling, vehicle queuing congestion, information asymmetry, and heavy manual scheduling load in traditional shipping management. Especially when the vehicle flow surges during peak hours, situations such as long vehicle waiting times, high crane position idle rates, and lagging information interaction often occur, seriously affecting the operation efficiency and operation safety.

[0031] In the process of implementing the present invention, first, a unified Internet of Things access platform is deployed. Existing data collection devices in the port area, such as license plate recognition systems, driver identity recognition terminals, vehicle GPS positioning systems, weighbridge sensors, and loading control units, are connected to the platform to unify the data collection frequency and format. The system automatically collects and preprocesses information such as the real-time position of the vehicle, license plate number, driver identity information, planned tonnage of the pick-up plan, operation status of the crane position, current traffic load in the port area, and historical loading efficiency, and constructs a standardized shipping data set. The data processing cycle is within 15 seconds to ensure that the basic data for scheduling calculations has high timeliness and accuracy.

[0032] In the shipping scheduling process, the system first calls the improved dynamic beat rearrangement algorithm to calculate the vehicle scheduling priority according to multiple parameters such as the vehicle pick-up plan, arrival time prediction, last operation time, and current port area load factor, and generates an executable queuing sequence and recommended entry and exit times. Subsequently, it calls the improved preloading capacity perspective algorithm to construct a unit-time loading capacity scoring model based on the loading speed, number of operation interruptions, equipment utilization rate, and current number of vehicles to be loaded at each crane position in the past 24 hours. The system cross-matches the vehicle priority index with the crane position scoring table, constructs a multi-constraint scheduling scoring matrix, and comprehensively obtains the scheduling utility score through a dynamic weight adjustment mechanism to complete the optimal binding of vehicles and crane positions, and outputs a queuing scheduling table with priority labels.

[0033] After receiving the scheduling notice through the mobile terminal applet, the driver drives into the port area within the recommended time period. The recognition system configured at the import and export channels of the port area automatically recognizes the license plate and waybill information, verifies the driver's identity, and completes the electronic waybill matching. If the information matches correctly, the system automatically triggers the barrier to release and sends the loading instruction to the corresponding crane position operation terminal. During the entire execution process, the system real-time collects the entry time, waiting time, operation duration, and actual loading data of each vehicle, and feeds the collection results back to the scheduling module in real time to automatically correct and continuously learn the scheduling parameters and algorithm models.

[0034] Taking the implementation data of a certain port area during the one-week period from March 10th to March 17th, 2025 as an example, the system dispatched 1,580 oil trucks in total, with an average of 225 trucks per day. The following is the "Comparison Table of the Implementation Results of the Intelligent Shipping Management System", showing the actual improvement of the port area in key scheduling indicators before and after the system was enabled: Table 1 Comparison Table of the Implementation Results of the Intelligent Shipping Management System ; Before the system went online, the average waiting time per truck was about 48 minutes, and the utilization rate of loading bays was 71.3%; after going online, the average waiting time per truck was shortened to 18 minutes, and the utilization rate of bays was increased to 89.6%. At the same time, the average error between the recommended arrival time calculated by the system and the actual arrival time of the vehicle was less than 3 minutes, and the overall scheduling accuracy reached 99.2%. During this period, the system automatically processed more than 130 complex scenarios such as secondary loading and priority vehicle queue jumping, and all tasks were successfully bound, the dynamic adjustment of bays was completed, and the operations were carried out smoothly, without any scheduling conflicts or failures. During the operation of the system, the number of dispatch positions was reduced from 3 per shift to 1, significantly reducing the manual dispatch intensity and personnel configuration costs.

[0035] From the implementation results, it can be seen that the application of the present invention in the port area dispatch management effectively solves the problems of long vehicle waiting, uneven resource allocation, low dispatch efficiency, and lack of information closed-loop in the traditional dispatch mode, realizes dispatch intelligence, loading precision, system linkage, and management transparency, and significantly improves the automation and operation efficiency of the port area shipping management while ensuring safe production.

[0036] The above is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. An intelligent shipping management method for port areas based on the Internet of Things, characterized in that, It includes the following steps: S1. Collect the Internet of Things (IoT) terminal data in the port area and perform preprocessing to construct a standardized shipping dataset; S2. Based on the standardized shipping dataset, use an improved dynamic beat rearrangement algorithm to dynamically sort all vehicles, generating a vehicle queuing sequence and the recommended time for entering and leaving the port; S3. Based on the standardized shipping dataset, construct a loading capacity model for each quay position in the port area per unit time, and score it using an improved preloading capacity perspective algorithm to output an efficiency score table for each quay position; S4. Cross-match the vehicle queuing sequence with the efficiency score table to form a multi-constraint matching matrix, and bind high-loading-capacity quay positions with high-priority vehicles through a dynamic weight allocation mechanism to generate a queuing scheduling table with priority labels; S5. Based on the queuing scheduling table and the recommended time for entering and leaving the port, generate corresponding electronic waybills for all vehicles, embed the license plate number, driver's ID number, and the assigned quay position number, and perform identity verification and binding with the shipping task for the electronic waybills; S6. When a vehicle enters or exits the port's electronic fence, identify the vehicle's position and compare the electronic waybill information. If the match is successful, trigger a release signal and push the loading instruction to the corresponding quay position terminal; S7. During the vehicle's shipping task execution, continuously collect the IoT terminal data in the port area, update and record the feedback in real time, and complete the archiving of the shipping data throughout the process.

2. The intelligent shipping management method for port areas based on the Internet of Things according to claim 1, wherein, The IoT terminal data includes the vehicle's real-time position, license plate number information, identity of the driving driver, quay position allocation status, vehicle pick-up planned quantity, current port traffic load, and historical loading records. The preprocessing includes data cleaning and format unification.

3. The intelligent shipping management method for port areas based on the Internet of Things according to claim 1, wherein Specifically, S2 includes: S21. Extract the estimated arrival time of each vehicle to be scheduled based on the standardized shipping data set , the planned pick-up quantity , the last loading time , the distance from the current location to the port area , and the current average beat loading time in the port area and the real-time port area traffic load factor , and use the above data as the sorting input features; S22. Construct an optimization objective function for dynamic beat rearrangement, and use an improved dynamic beat rearrangement algorithm to calculate the priority index of the vehicle; ； Among them, is the scheduling priority index of the vehicle , and the higher the value, the higher the scheduling priority. is the planned pickup volume of the vehicle , with the unit of ton. is the remaining distance of the vehicle from the port area, with the unit of kilometer. is the estimated arrival time of the vehicle at the port area, with the unit of minute. is the last loading completion time of the vehicle , with the unit of minute. is the current average beat time required for loading each vehicle in the port area, with the unit of minute / vehicle. is the current traffic load factor in the port area, which is a dimensionless number reflecting traffic and operation pressure. , , are the weighted weights of the pickup volume factor, distance factor, and waiting duration factor respectively, satisfying , , are positive infinitesimals introduced to avoid the denominator being zero. S23. Arrange the priority indices of all vehicles in descending order to obtain a vehicle queuing sequence , where . At the same time, taking , , and as parameters, calculate the recommended arrival and departure times for each vehicle: ； wherein, is the recommended time for the vehicle to enter and leave the port, is the vehicle current actual speed, in kilometers per minute, is the empirical adjustment factor for balancing the beat time and the port congestion level; S24. Output the vehicle queuing sequence and the set of recommended arrival and departure times .

4. The intelligent shipping management method for port areas based on the Internet of Things according to claim 1, wherein, Specifically, S3 includes: S31. Extract the crane position-related data based on the standardized shipping data set. The crane position-related data includes the historical operation rounds of each crane position the total loading quantity per round the corresponding operation duration the current number of waiting vehicles the number of abnormal events per unit time and the equipment availability factor within the current operation cycle ; ​ S32. Build a loading capacity model for the port area and define the theoretical loading capacity of each crane position within a unit time: ； Among them, is the loading capacity per unit time at the crane position, with the unit of ton / minute, is the loading volume of the th operation, is the actual completion duration of the th operation, and is the historical operation round; S33. On the basis of the loading capacity model, use an improved preloading capacity perspective algorithm to correct the real-time efficiency of each quay position, form a multi-dimensional comprehensive scoring function, and output the efficiency score value; ； Among them, is the loading efficiency score value of the crane position . is the equipment availability factor, and its value range is , representing the operating state of the equipment at the current crane position is the number of abnormal events during the current period is the number of waiting vehicles at present , , are the availability enhancement factor, the abnormal penalty factor and the queuing penalty factor respectively, and all are positive constants; S34. Output the loading efficiency score set for all crane positions .

5. The intelligent shipping management method for port areas based on the Internet of Things according to claim 1, wherein Specifically, S4 includes: S41. Based on the vehicle queuing sequence , the vehicle priority index , the picking plan quantity and the crane position loading efficiency score set , construct a standardized vehicle vector and a crane position vector : ； Among them, represents the remaining distance of the vehicle from the port area, is a small positive number to prevent division by zero, is the gantry position of the loading efficiency score value, is the gantry position of the historical average single loading volume, is the number of vehicles currently waiting for operation at this gantry position, is the maximum value of the vehicle pick-up volume in the current batch, used as the normalization benchmark, is the maximum value function; S42. Combine with the non-linear suppression function and calculate the comprehensive fitness of the vehicle and the crane position using the weighted cosine similarity: ； Among them, is the comprehensive adaptation degree score of the vehicle and the crane position , , are the eigenvalue of the vehicle and the crane position in the th dimension respectively, is the feature weighting factor, satisfying , is a small value to avoid division by zero, is the loading quantity deviation suppression factor, which controls the sensitivity of the historical capacity difference between the vehicle and the crane position; S43. To improve the overall resource utilization efficiency and fairness, introduce a dynamic weight allocation mechanism and construct a scheduling priority function ； Among them, represents the current port area traffic load factor, which is a dimensionless number reflecting traffic and operation pressure, is for vehicles assigned to the crane position of the scheduling utility value, is the equipment availability factor of the crane position , is the current port area traffic load factor, which is a dimensionless number reflecting traffic and operation pressure, is the maximum delay time in the current batch, used for normalization, is the delay time of the vehicle exceeding the recommended arrival time, is the actual access time, is for vehicles of the recommended access time to and from the port, , , is an adjustable coefficient to set the optimization preference; S44. All rating values form a two-dimensional scheduling utility matrix , calculate the vehicle-crane position binding relationship according to the maximum weight matching strategy, and generate a queuing scheduling table with priority labels: ； Among them, is the scheduling table, which contains the scheduling relationship between each vehicle and the allocated crane position, is the number of the th vehicle to be scheduled, is the number of the crane position bound to it, is the recommended time for vehicle to enter and leave the port, is the corresponding scheduling utility value, , which is determined by the quantile value in the overall distribution, is the number of successfully matched vehicle-crane position pairs.

Images