A port cargo intelligent scheduling system and method based on an internet of things

Abstract

The application discloses a port cargo intelligent scheduling system and method based on Internet of Things, and relates to the technical field of port logistics intelligent scheduling.The whole-chain linkage mechanism of real-time sensing, intelligent analysis, dynamic decision-making, precise execution and closed-loop feedback is constructed to realize dynamic self-adaptive adjustment of the port scheduling strategy.The system can automatically expand or shrink the buffer zone boundary of the yard according to the congestion index of the core operation area and the ship tide prediction, and synchronously adjust the AGV and yard bridge scheduling rules.In the peak tide period, the AGV preferentially guarantees the rapid transfer of the core area cargo, avoids the efficiency loss caused by long-distance transportation, and actively pre-migrates the flexible storage area cargo to the core reserve area in the trough return period to layout in advance for the subsequent operation peak.This dynamic scheduling mode can effectively respond to the sharp fluctuation of the port operation quantity, keeps each operation link highly efficient and coordinated at all times, and greatly improves the adaptability and response capability of the port in the complex operation environment.

Description

Technical Field

[0001] This invention relates to the field of intelligent scheduling technology for port logistics, specifically to an intelligent port cargo scheduling system and method based on the Internet of Things. Background Technology

[0002] As a key hub for sea and land transportation, the efficiency of port operations and the quality of its services have become one of the core elements of trade competitiveness. With the continuous expansion of trade, the throughput of port cargo has shown a rapid growth trend. However, upon examining the current state of port operations, the traditional port operation model has shortcomings in the face of the ever-increasing cargo throughput and complex and ever-changing market demands.

[0003] In terms of yard management, most ports still use the traditional fixed zoning model, which makes it difficult to make flexible adjustments based on fluctuations in real-time workload. This rigid zoning strategy leads to severe congestion and cargo backlog in some areas during peak hours, while other areas remain largely idle during off-peak hours, resulting in a significant waste of yard resources. Yard space utilization has remained at a low level for a long time. In terms of task scheduling, traditional methods often rely on manual experience and lack accurate prediction and dynamic scheduling mechanisms for peak ship arrival / departure times. This directly leads to low operating efficiency of key equipment such as AGVs and yard cranes, high empty running rates, and long waiting times, which cannot match the port's growing business scale. The extended ship stay time in port not only increases operating costs but also reduces the port's overall competitiveness.

[0004] In summary, the existing port cargo scheduling model has drawbacks in terms of yard zoning and task scheduling, which has become a bottleneck restricting the efficient development of ports. It is of great significance to develop a system and method that can perceive the port operation status in real time and dynamically adjust the yard zoning and task scheduling strategies based on the prediction of peak operation periods to realize intelligent port cargo scheduling. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and provide an intelligent port cargo scheduling system and method based on the Internet of Things. This system can achieve dynamic adaptive adjustment of port scheduling strategies by constructing a full-chain linkage mechanism of real-time perception, intelligent analysis, dynamic decision-making, precise execution, and closed-loop feedback. The system can automatically expand or shrink the boundaries of the yard buffer zone based on the congestion index of the core operating area and ship tidal predictions, and simultaneously adjust the scheduling rules of AGVs and yard cranes. During peak tidal periods, AGVs prioritize the rapid transfer of cargo in the core area to avoid efficiency losses due to long-distance transportation. During off-peak return periods, cargo in the flexible storage area is proactively moved to the core reserve area to prepare for subsequent peak operations. This dynamic scheduling mode can effectively cope with drastic fluctuations in port operations, ensuring efficient collaboration among all operational links and significantly improving the port's adaptability and responsiveness in complex operating environments.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: On the one hand, a port cargo intelligent scheduling system based on the Internet of Things, the system comprising: an Internet of Things sensing module, an intelligent analysis module, a dynamic decision-making module, a tidal scheduling execution module, and an execution feedback module;

[0007] The IoT sensing module is used to collect multi-source data from the yard in real time, including container location and type, yard crane working status, AGV location and load, and ship arrival and departure plans, and synchronize the data to the intelligent analysis module.

[0008] The intelligent analysis module calculates the real-time congestion index of the core work area based on the data input from the IoT sensing module, predicts the peak periods of work tidal flow, and outputs the results to the dynamic decision-making module.

[0009] The dynamic decision-making module dynamically adjusts the virtual partition boundaries of the storage yard based on the congestion index and tidal prediction results output by the intelligent analysis module, and generates corresponding tidal scheduling rules.

[0010] The tidal scheduling execution module receives the partition boundary and scheduling instructions issued by the dynamic decision module, generates dynamic task packages for the yard crane / AGV, and distributes them to the loading and unloading equipment for execution.

[0011] The execution feedback module collects real-time data on changes in yard occupancy status after the yard crane / AGV performs its operation, and feeds this data back to the IoT sensing module, triggering a new round of coordinated adjustments to the zoning and scheduling strategies.

[0012] Furthermore, the IoT sensing module includes a container monitoring unit, a yard crane monitoring unit, an AGV monitoring unit, and a ship monitoring unit;

[0013] The container monitoring unit acquires the location, type, and weight data of containers by using RFID tag readers and weight sensors deployed in various areas of the yard.

[0014] The yard crane monitoring unit uses position sensors and pressure sensors installed on the yard crane to monitor the yard crane's position, running speed, load weight, and utilization rate data in real time.

[0015] The AGV monitoring unit uses GPS positioning, lidar and weighing sensors on the AGV vehicle to collect the AGV's position coordinates, driving trajectory, load status and AGV density data.

[0016] The vessel monitoring unit acquires data on the arrival plans, vessel positions, and cargo loading / unloading volumes of vessels waiting to be operated through the AIS system, radar equipment, and shore-based sensors at the port waters. The data collected by the above units is then aggregated and synchronized to the intelligent analysis module.

[0017] Furthermore, the intelligent analysis module preprocesses the received data, removes abnormal data, and calculates the congestion index of the core operating area: ,in, This is the congestion index for the core work area, with a value ranging from [0, 1]. A higher value indicates a higher degree of congestion. The density coefficient of the container and , This represents the current actual number of containers in the work area. This represents a safe threshold for the number of containers in the operating area. This refers to the actual area occupied by the container. The total area of ​​the core operating area AGV saturation coefficient and , This represents the current number of AGVs operating in the work area. The maximum number of AGVs that can be carried in the core work area. For the first The current load rate of the AGV. This is the upper limit of the rated load of the AGV. The load factor of the yard bridge and , The number of yard bridges in the core operating area. For the first The current load weight of the bridge. This refers to the rated load weight of the yard crane. During peak operating hours for the bridge, The total time of the statistical period. Let be the weighting coefficient, satisfying , The container density weight has a value range of [0.4, 0.6]. This represents the AGV saturation weight, with a value range of [0.2, 0.3]. This represents the load weight of the field bridge, with a value range of [0.2, 0.3].

[0018] Furthermore, the intelligent analysis module constructs a peak operation tide prediction model based on an LSTM neural network. Combining historical ship arrival and departure data, meteorological and hydrological data, and real-time traffic flow data, it predicts peak operation tide periods for 1-6 hours. The prediction process is as follows:

[0019] Real-time data acquisition: Real-time acquisition of ship arrival and departure data, meteorological and hydrological data, and traffic flow data for the previous 24 hours, and normalization of the raw data so that its values ​​are distributed between 0 and 1;

[0020] Data partitioning: The processed data is divided into a training set and a test set, where the training set is used for model training and the test set is used to evaluate model performance;

[0021] Input model prediction: Input the preprocessed data sequence into the LSTM neural network, and calculate and output the probability value of the operation peak at each moment within 6 hours based on the learned time series features and patterns, forming a probability sequence;

[0022] Probability threshold judgment: Set the probability threshold to 0.7. Compare the probability value of each moment output by the model with the threshold. When the probability value of a moment is greater than 0.7, it is determined that the moment is the peak period of the operation tide. Otherwise, it is determined to be a non-peak period and a peak warning signal is sent to the dynamic decision module.

[0023] Dynamic update forecast: Repeat the above data acquisition, segmentation and forecasting process to update the forecast of future peak operation times in real time to ensure the timeliness and accuracy of the forecast results.

[0024] Furthermore, the virtual zoning boundary of the yard in the dynamic decision-making module includes a work area, a buffer zone, and a flexible storage area. The work area is a high-efficiency loading and unloading area adjacent to the port shore. The buffer zone is a transition area between the work area and the flexible storage area, used for temporary storage and transshipment. The flexible storage area is used for long-distance, long-term cargo storage. Based on the peak warning signal, congestion index, and tidal peak prediction results output by the intelligent analysis module, the dynamic decision-making module adjusts the zoning boundary and formulates a scheduling strategy according to the following tidal scheduling rules, and sends scheduling instructions to the tidal scheduling execution module. The tidal scheduling rules are as follows:

[0025] when When the value is ≥0.8 and the predicted peak duration is >2 hours: Due to the high congestion in the core operation area and the long peak duration, in order to avoid ships waiting due to insufficient space in the shore loading and unloading area, 1 / 3 of the area in the flexible storage area will be incorporated into the buffer zone. This operation can quickly expand the temporary storage capacity of the core operation area, so that after the AGV unloads the ship, the cargo can be stacked in the expanded buffer zone nearby, reducing the efficiency loss caused by the long-distance transportation of AGVs. At the same time, the cargo unloaded by the AGV will be prioritized to be stacked in the buffer zone nearby. After the buffer zone is saturated, it will be diverted to the remaining flexible storage area, prioritizing the high-speed turnover of cargo in the operation area and reducing the risk of ships being stranded in port.

[0026] When 0.6≤ When the value is <0.8 and the predicted peak duration is >1: When moderate congestion occurs in the work area and the peak will last for a period of time, the 1 / 5 area of ​​the elastic storage area closest to the core work area will be included in the buffer zone. While appropriately increasing the buffer space, avoid excessive occupation of elastic storage area resources, ensure that AGVs give priority to using the newly added buffer area, and rationally allocate yard resources to balance loading and unloading operation pressure while ensuring the efficiency of the core work area.

[0027] Furthermore, the generation process of the field bridge / AGV dynamic task package in the tidal scheduling execution module is as follows:

[0028] Task parsing and preprocessing: Receive partition boundaries and scheduling instructions from the dynamic decision module, and parse task type, job location and time constraints;

[0029] Real-time status acquisition: The current position, load status, remaining power, and queuing status of the work point of the yard crane / AGV are obtained through the Internet of Things sensing module;

[0030] Time efficiency calculation: For each task to be assigned, calculate the estimated completion time based on the current device status and path planning. and task urgency coefficient The The estimated completion time for task p and , This refers to the travel time of the AGV from its current location to the work point. The standard time for the yard crane / AGV to complete loading and unloading operations. The queuing time at the work site, the Let be the urgency coefficient of task p and , For the time of departure of the vessel, For the current time, For the preset time threshold, Priority coefficient and ;

[0031] Priority ranking: Calculate the priority score for each task. Sort the tasks by score from highest to lowest;

[0032] Equipment matching: Assign tasks with higher priority scores to the field bridges that will complete the task in order of priority.

[0033] Dynamic task package generation: The task allocation results are encapsulated into dynamic task packages, which include task details, execution devices, path planning, and time windows.

[0034] Furthermore, the execution feedback module utilizes inductive loops deployed on the yard ground to acquire the real-time operating trajectories of AGVs and yard cranes, and combines this with image data collected by cameras in the yard to determine the number of containers in the images. Simultaneously, based on the AGV and yard crane operation trajectory data, the dwell time and cargo handling situation in each area of ​​the yard are determined, and the dynamic occupied area of ​​each area is calculated. This allows for the calculation of the occupancy rate of each area of ​​the storage yard. ,in, This refers to the average floor space occupied by a single shipping container. Given the total area of ​​the storage yard, calculate the occupancy rate of each area. Container quantity and dynamic occupied area The system performs integrated analysis, generates a report on changes in yard occupancy, and feeds it back to the database of the IoT sensing module. This triggers the IoT sensing module to update its data collection cycle and key monitoring areas, thereby initiating a new round of coordinated adjustments to zoning and scheduling strategies. When the area occupancy rate... > At that time, an early warning message is sent to the dynamic decision-making module to adjust the scheduling strategy.

[0035] On the other hand, a port cargo intelligent scheduling method based on the Internet of Things (IoT) includes the following specific steps:

[0036] S100 Real-time data acquisition: Through various sensors in the IoT sensing module, continuously collect multi-source data on the number and type of containers in each area of ​​the yard, the location and working status of yard cranes, the location and load of AGVs, and the arrival and departure plans of ships.

[0037] S200, Data Analysis and Prediction: The received data is preprocessed, outliers are removed, the real-time congestion index of the work area is calculated, and the peak work period of the next 1-6 hours is predicted. The results are then transmitted to the dynamic decision-making module.

[0038] S300, Dynamic Scheduling Decision: Based on the output congestion index and tidal prediction results, dynamically adjust the virtual partition boundary of the yard and bind and generate the corresponding tidal scheduling rules.

[0039] S400, Task Execution: Generate dynamic task packages for yard cranes / AGVs based on time efficiency and distribute them to loading and unloading equipment. The equipment then executes the cargo handling tasks according to the instructions.

[0040] S500, closed-loop feedback and strategy adjustment: Real-time acquisition of yard occupancy status change data after equipment execution, generation of yard occupancy status change report, and feedback to IoT sensing module to trigger a new round of data acquisition and analysis.

[0041] Compared with existing technologies, this Internet of Things-based intelligent port cargo scheduling system and method has the following advantages:

[0042] I. This invention achieves dynamic adaptive adjustment of port scheduling strategies by constructing a full-chain linkage mechanism of real-time perception, intelligent analysis, dynamic decision-making, precise execution, and closed-loop feedback. The system can automatically expand or shrink the boundary of the yard buffer zone based on the congestion index of the core operation area and ship tidal predictions, and simultaneously adjust the scheduling rules of AGVs and yard cranes. During peak tidal periods, AGVs prioritize the rapid transfer of goods in the core area to avoid efficiency losses caused by long-distance transportation. During off-peak return periods, goods in the flexible storage area are proactively moved to the core reserve area in advance to prepare for subsequent peak operations. This dynamic scheduling mode can effectively cope with the drastic fluctuations in port operations, ensuring that all operational links maintain efficient collaboration and significantly improving the port's adaptability and responsiveness in complex operating environments.

[0043] Second, based on multi-dimensional data fusion analysis, this invention realizes the refined management and optimized allocation of port resources. By collecting data on containers, equipment and ships through the Internet of Things sensing module and combining it with the congestion index calculation of the intelligent analysis module, the system can accurately identify areas of idle or overloaded resources. Through the dynamic decision-making module, the system can optimize the yard zoning and equipment tasks in real time, and achieve precise connection of loading and unloading actions. This resource optimization allocation method effectively reduces equipment empty running and waiting time, improves the utilization rate of yard space and equipment use efficiency, and realizes the maximum value conversion of port resources.

[0044] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0046] Figure 1 This is an operation flowchart for an IoT-based intelligent port cargo scheduling system.

[0047] Figure 2 This is a flowchart illustrating the steps of an IoT-based intelligent cargo scheduling method for ports. Detailed Implementation

[0048] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0049] Example 1

[0050] This embodiment provides a working principle of a port cargo intelligent scheduling system based on the Internet of Things (IoT). By constructing a full-chain linkage system of IoT sensing, intelligent analysis, dynamic decision-making, tidal scheduling execution and execution feedback, intelligent and dynamic management of port cargo scheduling is achieved.

[0051] The aforementioned IoT sensing module achieves real-time monitoring of all elements of port operations through the coordinated deployment of multiple types of sensors. In the yard area, RFID tag readers and weight sensors are distributed in a grid pattern. Each container is affixed with a unique RFID tag. When a container enters the yard, the reader reads its location coordinates and type information through radio frequency signals, while the weight sensor simultaneously collects cargo weight data to form a basic information file for the container. The yard crane monitoring unit obtains the three-dimensional coordinates, running speed, and load weight of the yard crane in real time through encoders installed on the traveling mechanism and pressure sensors on the lifting mechanism. The pressure sensor monitors the load changes of the spreader in real time and then calculates the equipment utilization rate. The AGV monitoring unit integrates GPS positioning, LiDAR, and weighing sensor data. The GPS provides positioning information, the LiDAR builds an environmental map and updates the AGV's travel trajectory in real time to avoid collisions, and the weighing sensor is integrated into the AGV platform to monitor cargo weight in real time and calculate the load rate. The ship monitoring unit receives ship automatic identification information (including arrival and departure plans, cargo type, etc.) through the AIS system. Combined with the real-time tracking data of the radar equipment and the loading and unloading monitoring of shore-based sensors, a ship operation dataset is formed. All sensor data is wirelessly transmitted and synchronized to the intelligent analysis module.

[0052] The intelligent analysis module preprocesses the received multi-source data, removes outliers, and then uses a congestion index calculation model and an LSTM neural network to assess the congestion status of the core work area and predict peak work hours. The formula for calculating the congestion index is as follows: ,in This is the congestion index for the core work area, with a value ranging from [0, 1]. The higher the value, the greater the congestion. For container density coefficient, This is the current actual number of containers in the work area. This represents a safe threshold for the number of containers in the operating area. This refers to the actual area occupied by the container. This coefficient represents the total area of ​​the core operating area and reflects the density of containers within that area. This is the saturation coefficient of the AGV. This indicates the current number of AGVs operating in the work area. The maximum number of AGVs that can be carried in the core work area. For the first The current load rate of the AGV. This is the rated load limit for the AGV. This coefficient reflects the saturation level and load distribution of the AGV in the work area. For the load factor of the yard bridge The number of yard bridges in the core operating area. For the first The current load weight of the bridge. This refers to the rated load weight of the yard crane. During peak operating hours for the bridge, To calculate the total time of the statistical cycle, this coefficient comprehensively considers the load weight of the yard crane and the duration of peak traffic. β and γ are weighting coefficients, satisfying ,in The value range is [0.4, 0.6], representing the container density weight. The value range is [0.2, 0.3], which represents the AGV saturation weight. The value range is [0.2, 0.3], representing the load weight of the yard crane. For predicting peak operating tides, a prediction model is built based on an LSTM neural network. First, real-time data on ship arrivals and departures, meteorological and hydrological data, and traffic flow data for the 24 hours preceding the current moment are acquired. The raw data is then normalized to ensure its values ​​are distributed between 0 and 1, eliminating the influence of data units. The processed data is then divided into training and testing sets. The training set is used for model training, enabling the LSTM neural network to learn time-series characteristics and patterns. The testing set is used to evaluate model performance. Next, the preprocessed data sequence is input into the LSTM neural network. The model calculates and outputs the probability value of the peak operating tide at each moment within 6 hours based on the learned patterns, forming a probability sequence. A probability threshold of 0.7 is set. The probability value at each moment is compared with the threshold; if it is greater than 0.7, it is considered a peak operating tide period; otherwise, it is considered a non-peak period. A peak warning signal is sent to the dynamic decision module. Finally, by repeating the data acquisition, division, and prediction process, the prediction is updated in real time to ensure the timeliness and accuracy of the results.

[0053] The dynamic decision-making module adjusts the virtual zoning boundaries of the yard and formulates scheduling strategies based on the peak warning signals, congestion index, and tidal peak prediction results output by the intelligent analysis module. The virtual zoning boundaries of the yard include the operating area, buffer zone, and flexible storage area. The operating area is a high-efficiency loading and unloading area adjacent to the port shore. The buffer zone is a transition area between the operating area and the flexible storage area, used for temporary storage and transshipment. The flexible storage area is used for long-distance, long-term cargo storage. When CI ≥ 0.8 and the predicted peak duration > 2, the core operating area is in a state of high congestion and the peak duration is long. To avoid insufficient space in the shore loading and unloading area causing ships to wait, 1 / 3 of the area in the flexible storage area is incorporated into the buffer zone. This can quickly expand the temporary storage capacity of the core operating area, allowing AGVs to store cargo in the expanded buffer zone after unloading, reducing efficiency losses caused by long-distance transportation. Simultaneously, it stipulates that AGV-unloaded cargo should be prioritized for storage in the nearest buffer zone. Once the buffer zone is saturated, cargo will be diverted to the remaining flexible storage area. This prioritizes ensuring high-speed turnover of cargo in the operating area and reduces the risk of ships being held up in port. When 0.6 ≤ CI < 0.8 and the predicted peak duration is > 1, moderate congestion occurs in the operating area, and the peak will continue for some time. In this case, the 1 / 5 of the flexible storage area closest to the core operating area will be included in the buffer zone. While appropriately increasing the buffer space, this avoids excessive occupation of flexible storage area resources, ensuring that AGVs prioritize the use of the newly added buffer area. Under the premise of ensuring the efficiency of the core operating area, yard resources are rationally allocated to balance loading and unloading pressure. Through this dynamic adjustment of zoning boundaries, yard resources can be optimized according to fluctuations in workload, improving overall operational efficiency.

[0054] After receiving the partition boundaries and scheduling instructions from the dynamic decision-making module, the tidal scheduling execution module generates and distributes dynamic task packages for the yard cranes / AGVs. This generation process includes task parsing and preprocessing, real-time status acquisition, time efficiency calculation, priority sorting, equipment matching, and dynamic task package generation. Specifically: First, task parsing and preprocessing are performed. Upon receiving instructions, the task type, work location, and time constraints are analyzed to clarify task requirements. Then, the IoT sensing module acquires the current location, load status, remaining power, and queuing status of the yard cranes / AGVs, providing real-time equipment status information for task allocation. Regarding time efficiency calculation, for each task to be assigned, the estimated completion time is calculated based on the current equipment status and path planning. and task urgency coefficient The ,in This refers to the travel time of the AGV from its current location to the work point. The standard time for the yard crane / AGV to complete loading and unloading operations. The formula, which takes into account the queuing time at the work site, comprehensively considers the travel, loading / unloading, and queuing times during the task execution process, and accurately estimates the task completion time. , For the time of departure of the vessel, For the current time, For the preset time threshold, Priority coefficient and This coefficient reflects the urgency of the task; the closer the departure time, the more urgent the task. Next, a priority score is calculated for each task. ,in The priority score is used as a weighting factor to sort tasks from highest to lowest. The priority score takes into account the estimated completion time and task urgency, ensuring that urgent and short-duration tasks are processed first. Then, tasks with high priority scores are assigned to suitable yard cranes in order to complete the task. Factors such as the current status and load capacity of the equipment are taken into account to achieve the best match between tasks and equipment. Finally, the task allocation results are encapsulated into a dynamic task package, which includes task details, execution equipment, path planning and time window, and distributed to loading and unloading equipment for execution, so that the equipment can complete the cargo handling task efficiently according to plan.

[0055] The execution feedback module collects real-time data on changes in yard occupancy after equipment execution, generates reports, and feeds them back to the IoT sensing module, triggering a new round of data collection and analysis to form a closed-loop feedback mechanism. It utilizes ground-penetrating inductive loops deployed on the yard floor to acquire the real-time operating trajectories of AGVs and yard cranes, and combines this with image data collected by yard cameras to determine the number of containers in the images. Simultaneously, based on the AGV and yard crane operation trajectory data, the dwell time and cargo handling situation in each area of ​​the yard are determined, and the dynamic occupied area of ​​each area is calculated. Through formula Calculate the occupancy rate of each area of ​​the storage yard, where This refers to the average floor space occupied by a single shipping container. This formula, representing the total area of ​​the yard, comprehensively considers the impact of container quantity and equipment operating area on the area occupancy rate, accurately reflecting the actual occupancy status of each area and calculating the occupancy rate of each area. Container quantity and dynamic occupied area The system performs integrated analysis to generate a report on changes in yard occupancy, which is then fed back to the database of the IoT sensing module. Upon receiving the report, the IoT sensing module is triggered to update its data collection cycle and key monitoring areas. When the area occupancy rate... In this case, an early warning message is sent to the dynamic decision-making module. The dynamic decision-making module adjusts the scheduling strategy based on the early warning message to ensure the rational use of yard resources and the efficient operation. This closed-loop feedback mechanism enables the system to adjust the strategy in real time based on the actual execution effect and continuously optimize the scheduling process.

[0056] In summary, the IoT-based intelligent port cargo scheduling system of this embodiment achieves real-time collection of multi-source data through the IoT sensing module, completes congestion index calculation and tidal peak prediction through the intelligent analysis module, adjusts zoning boundaries and scheduling strategies based on analysis results through the dynamic decision-making module, generates and distributes dynamic task packages through the tidal scheduling execution module, and forms a closed-loop feedback through the execution feedback module. All modules work together to build a full-chain linkage mechanism, realizing dynamic adaptive adjustment of port scheduling strategies. The system can flexibly adjust the yard buffer zone boundaries and optimize AGV and yard crane scheduling rules based on the congestion situation in the core operating area and ship tidal predictions, effectively cope with fluctuations in port operations, reduce equipment idle time and waiting time, improve yard space utilization and equipment efficiency, enable the port to maintain efficient collaborative operation in complex operating environments, and significantly enhance the port's adaptability and responsiveness.

[0057] Example 2

[0058] like Figure 1 As shown in Example 1, this example elaborates on the specific steps of a port cargo intelligent scheduling system based on the Internet of Things (IoT) in performing transfer operations. The specific steps are as follows:

[0059] (1) Real-time data acquisition

[0060] By utilizing various sensors in the IoT sensing module, data on the quantity and type of containers in each area of ​​the yard are continuously collected;

[0061] Collect data on the location and operational status of the field bridge;

[0062] Collect AGV location and load data;

[0063] Collect ship arrival and departure schedule data;

[0064] (2) Data preprocessing and analysis prediction

[0065] The collected data is preprocessed to remove outliers;

[0066] Based on the congestion index calculation model, the real-time congestion index of the core work area is calculated.

[0067] Using an LSTM neural network, predict the peak periods of work activity in the next 1-6 hours;

[0068] The calculated congestion index and the predicted peak hours are transmitted to the dynamic decision-making module.

[0069] (3) Dynamic scheduling decision

[0070] The dynamic decision-making module dynamically adjusts the virtual partition boundaries of the yard based on the received congestion index and tidal prediction results, including the boundary adjustment of the work area, buffer zone and elastic storage area;

[0071] Based on the adjusted partition boundaries, the corresponding tidal scheduling rules are generated.

[0072] (4) Task generation and execution

[0073] The tidal scheduling execution module receives the partition boundaries and scheduling instructions issued by the dynamic decision-making module, parses and preprocesses the tasks, and clarifies the task type, job location and time constraints.

[0074] The IoT sensing module acquires real-time status information such as the current location of the yard crane / AGV, load status, remaining power, and queuing status at work points.

[0075] For each task to be assigned, calculate the estimated completion time and task urgency coefficient based on the current device status and path planning;

[0076] Calculate the priority score for each task and sort the tasks from highest to lowest score;

[0077] Assign tasks with high priority scores to the appropriate yard bridges / AGVs in sequence;

[0078] The task allocation results are encapsulated into dynamic task packages, which include information such as task details, execution equipment, route planning, and time windows, and then distributed to loading and unloading equipment. The equipment executes the cargo handling tasks according to the instructions.

[0079] (5) Closed-loop feedback and strategy adjustment

[0080] The execution feedback module uses inductive loops deployed on the yard ground to obtain the running trajectory of AGVs and yard cranes in real time, and combines it with image data collected by cameras to obtain the number of containers in the image;

[0081] Based on the AGV and yard crane operation trajectory data, determine their dwell time and cargo handling in each area of ​​the yard, calculate the dynamic occupied area of ​​each area, and then calculate the occupancy rate of each area of ​​the yard.

[0082] The occupancy rate, number of containers, and dynamic occupied area of ​​each area are integrated and analyzed to generate a report on changes in yard occupancy status;

[0083] The report is fed back to the database of the IoT sensing module, triggering the IoT sensing module to update the data collection cycle and key monitoring areas, thereby initiating a new round of linkage adjustment of partitioning and scheduling strategies;

[0084] When the area occupancy rate is greater than 90%, an early warning message is sent to the dynamic decision-making module to adjust the scheduling strategy.

[0085] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A port cargo intelligent scheduling system based on the Internet of Things, characterized in that, The system consists of: an IoT sensing module, an intelligent analysis module, a dynamic decision-making module, a tidal scheduling execution module, and an execution feedback module; The IoT sensing module is used to collect multi-source data from the yard in real time, including container location and type, yard crane working status, AGV location and load, and ship arrival and departure plans, and synchronize the data to the intelligent analysis module. The intelligent analysis module calculates the real-time congestion index of the work area based on the data input from the IoT sensing module, predicts the peak periods of work tidal flow, and outputs the results to the dynamic decision-making module. The intelligent analysis module preprocesses the received data, removes abnormal data, and calculates the congestion index of the core operating area: ,in, The congestion index for the core work area ranges from [0, 1]. The density coefficient of the container and , This represents the current actual number of containers in the work area. This represents a safe threshold for the number of containers in the operating area. This refers to the actual area occupied by the container. The total area of ​​the core operating area AGV saturation coefficient and , This represents the current number of AGVs operating in the work area. The maximum number of AGVs that can be carried in the core work area. For the first The current load rate of the AGV. This is the upper limit of the rated load of the AGV. The load factor of the yard bridge and , The number of yard bridges in the core operating area. For the first The current load weight of the bridge. This refers to the rated load weight of the yard crane. During peak operating hours for the bridge, The total time of the statistical period. Let be the weighting coefficient, satisfying , The container density weight has a value range of [0.4, 0.6]. This represents the AGV saturation weight, with a value range of [0.2, 0.3]. This represents the load weight of the yard bridge, with a value range of [0.2, 0.3]. The intelligent analysis module constructs a peak operation tide prediction model based on an LSTM neural network. Combining historical ship arrival and departure data, meteorological and hydrological data, and real-time traffic flow data, it predicts the peak operation tide periods for 1-6 hours. The prediction process is as follows: Real-time data acquisition: Real-time acquisition of ship arrival and departure data, meteorological and hydrological data, and traffic flow data for the previous 24 hours, and normalization of the raw data so that its values ​​are distributed between 0 and 1; Data partitioning: The processed data is divided into a training set and a test set, where the training set is used for model training and the test set is used to evaluate model performance; Input model prediction: Input the preprocessed data sequence into the LSTM neural network, and calculate and output the probability value of the operation peak at each moment within 6 hours based on the learned time series features and patterns, forming a probability sequence; Probability threshold judgment: Set the probability threshold to 0.

7. Compare the probability value of each moment output by the model with the threshold. When the probability value of a moment is greater than 0.7, it is determined that the moment is the peak period of the operation tide. Otherwise, it is determined to be a non-peak period and a peak warning signal is sent to the dynamic decision module. Dynamic update forecast: Repeat the above data acquisition, segmentation and forecasting process to update the forecast of future peak operation times in real time; The dynamic decision-making module dynamically adjusts the virtual partition boundaries of the storage yard based on the congestion index and tidal prediction results output by the intelligent analysis module, and generates corresponding tidal scheduling rules. The tidal scheduling execution module receives the partition boundary and scheduling instructions issued by the dynamic decision module, generates dynamic task packages for the yard crane / AGV, and distributes them to the loading and unloading equipment for execution. The execution feedback module collects real-time data on changes in yard occupancy status after the yard crane / AGV performs its operation, and feeds this data back to the IoT sensing module, triggering a new round of coordinated adjustments to the zoning and scheduling strategies.

2. The port cargo intelligent scheduling system based on the Internet of Things according to claim 1, characterized in that, The IoT sensing module includes a container monitoring unit, a yard crane monitoring unit, an AGV monitoring unit, and a ship monitoring unit; The container monitoring unit acquires the location, type, and weight data of containers by using RFID tag readers and weight sensors deployed in various areas of the yard. The yard crane monitoring unit uses position sensors and pressure sensors installed on the yard crane to monitor the yard crane's position, running speed, load weight, and utilization rate data in real time. The AGV monitoring unit uses GPS positioning, lidar and weighing sensors on the AGV vehicle to collect the AGV's position coordinates, driving trajectory, load status and AGV density data. The vessel monitoring unit acquires data on the arrival plans, vessel positions, and cargo loading / unloading volumes of vessels waiting to be operated through the AIS system, radar equipment, and shore-based sensors at the port waters. The data collected by the above units is then aggregated and synchronized to the intelligent analysis module.

3. The port cargo intelligent scheduling system based on the Internet of Things according to claim 1, characterized in that, The dynamic decision-making module defines the virtual zoning boundaries of the storage yard, including a work area, a buffer zone, and a flexible storage area. The work area is a high-efficiency loading and unloading area adjacent to the port shore. The buffer zone is a transition area between the work area and the flexible storage area, used for temporary storage and transshipment. The flexible storage area is used for long-distance, long-term cargo storage. Based on the peak warning signal, congestion index, and tidal peak prediction results output by the intelligent analysis module, the dynamic decision-making module adjusts the zoning boundaries and formulates a scheduling strategy according to the following tidal scheduling rules, and sends scheduling instructions to the tidal scheduling execution module. The tidal scheduling rules are as follows: when When the value is ≥0.8 and the predicted peak duration is >2: 1 / 3 of the area in the elastic storage area is included in the buffer zone, so that the goods are stacked in the expanded buffer zone after the AGV unloads the ship; When 0.6≤ When the value is <0.8 and the predicted peak duration is >1: include the 1 / 5 of the elastic storage area closest to the core work area in the buffer zone.

4. The port cargo intelligent scheduling system based on the Internet of Things according to claim 1, characterized in that, The generation process of the dynamic task package for the field bridge / AGV in the tidal scheduling execution module is as follows: Task parsing and preprocessing: Receive partition boundaries and scheduling instructions from the dynamic decision module, and parse task type, job location and time constraints; Real-time status acquisition: The current position, load status, remaining power, and queuing status of the work point of the yard crane / AGV are obtained through the Internet of Things sensing module; Time efficiency calculation: For each task to be assigned, calculate the estimated completion time based on the current device status and path planning. and task urgency coefficient The The estimated completion time for task p and , This refers to the travel time of the AGV from its current location to the work point. The standard time for the yard crane / AGV to complete loading and unloading operations. The queuing time at the work site, the Let be the urgency coefficient of task p and , For the time of departure of the vessel, For the current time, For the preset time threshold, Priority coefficient and ; Priority ranking: Calculate the priority score for each task. Sort the tasks by score from highest to lowest; Equipment matching: Assign tasks with higher priority scores to the field bridges that will complete the task in order of priority. Dynamic task package generation: The task allocation results are encapsulated into dynamic task packages, which include task details, execution devices, path planning, and time windows.

5. A port cargo intelligent scheduling system based on the Internet of Things according to claim 1, characterized in that, The execution feedback module uses inductive loops deployed on the yard ground to acquire the running trajectories of AGVs and yard cranes in real time, and combines this with image data collected by cameras in the yard to obtain the number of containers in the images. Simultaneously, based on the AGV and yard crane operation trajectory data, the dwell time and cargo handling situation in each area of ​​the yard are determined, and the dynamic occupied area of ​​each area is calculated. This allows for the calculation of the occupancy rate of each area of ​​the storage yard. ,in, This refers to the average floor space occupied by a single shipping container. Given the total area of ​​the storage yard, calculate the occupancy rate of each area. Container quantity and dynamic occupied area The system performs integrated analysis, generates a report on changes in yard occupancy, and feeds it back to the database of the IoT sensing module. This triggers the IoT sensing module to update its data collection cycle and key monitoring areas, thereby initiating a new round of coordinated adjustments to zoning and scheduling strategies. When the area occupancy rate... > At that time, an early warning message is sent to the dynamic decision-making module to adjust the scheduling strategy.

6. A port cargo intelligent scheduling method based on the Internet of Things (IoT), applicable to the port cargo intelligent scheduling system based on the Internet of Things as described in any one of claims 1-5, characterized in that, The specific steps of this method are as follows: S100 Real-time data acquisition: Through various sensors in the IoT sensing module, continuously collect multi-source data on the number and type of containers in each area of ​​the yard, the location and working status of yard cranes, the location and load of AGVs, and the arrival and departure plans of ships. S200, Data Analysis and Prediction: The received data is preprocessed, outliers are removed, the real-time congestion index of the work area is calculated, and the peak work period of the next 1-6 hours is predicted. The results are then transmitted to the dynamic decision-making module. S300, Dynamic Scheduling Decision: Based on the output congestion index and tidal prediction results, dynamically adjust the virtual partition boundaries of the yard and bind the corresponding tidal scheduling rules; S400, Task Execution: Generate dynamic task packages for yard cranes / AGVs based on time efficiency and distribute them to loading and unloading equipment. The equipment then executes the cargo handling tasks according to the instructions. S500, closed-loop feedback and strategy adjustment: Real-time acquisition of yard occupancy status change data after equipment execution, generation of yard occupancy status change report, and feedback to IoT sensing module to trigger a new round of data acquisition and analysis.

Images