A port production operation deviation analysis and adjustment system based on port digitization

Abstract

The present invention discloses a port production operation deviation analysis and adjustment system based on port digitization, which relates to the field of operation control technology, including a data acquisition module, a deviation analysis module, a planning and adjustment module and a feedback optimization module. It solves the problems of weak spatiotemporal data analysis capabilities, port congestion and lack of dynamic diversion solutions in port operations. The data acquisition module uniformly extracts port production operation information to achieve cross-system data sharing and business collaboration, and constructs a port spatiotemporal knowledge graph through the deviation analysis module to analyze the production operation deviation of port equipment, obtain the propagation law and influencing factors of the port equipment congestion status in time and space, and then dynamically generates a port equipment planning and adjustment solution through the planning and adjustment module, and regulates ship scheduling, resource coordination and path planning respectively. Then, the feedback optimization module evaluates the application effect of port dynamic diversion and performs adaptive growth iterative optimization.

Description

Technical Field

[0001] The present invention relates to the technical field of operation control, and in particular to a port production operation deviation analysis and adjustment system based on port digitization. Background Art

[0002] The port digital management and control platform provides digital management tools for port managers at all levels through the construction of digital management and control scenarios such as production operation deviation analysis and guidance of port equipment, preventive equipment early warning and collaborative management, and preventive safety and environmental operation early warning and collaborative management. It enables situational awareness, interactive management, and cross-domain collaboration of port operations.

[0003] However, existing port digital platforms have weak spatiotemporal data analysis capabilities during port production operations, port congestion caused by equipment operation deviations, and a lack of dynamic diversion solutions. Traditional port business systems, such as production, equipment, and safety systems, operate independently, and data sharing relies on manual export and import, resulting in low collaboration efficiency and delayed decision-making. They struggle to support real-time scheduling and cross-domain collaboration, and after abnormal critical events, ships may return to port in large numbers, making it difficult to combine spatiotemporal data for early diversion, leading to port congestion. Insufficient path planning capabilities also make it difficult to divert loading and unloading equipment.

[0004] In view of the above technical defects, a solution is now proposed. Summary of the Invention

[0005] The purpose of the present invention is to solve the application problems of the existing port digital platform, such as weak spatiotemporal data analysis capabilities during port production operations, port congestion caused by equipment operation deviations, and lack of dynamic diversion solutions.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A port production operation deviation analysis and adjustment system based on port digitalization, including a data acquisition module, a deviation analysis module, a planning and adjustment module, and a feedback optimization module;

[0008] The data acquisition module is used to connect to the port business system and collect port production operation information: port production operation information includes spatiotemporal data, environmental data and business data;

[0009] The deviation analysis module includes edge computing nodes and cloud processors. The edge computing nodes preprocess port production operation information to construct a port spatiotemporal knowledge graph. The cloud processor then builds a port congestion propagation model, divides the port into dynamic grid units, and constructs a spatiotemporal graph convolutional network structure to analyze the production operation deviations of port equipment, obtain the propagation patterns of port equipment congestion status in time and space, and analyze the factors affecting port equipment congestion.

[0010] The planning and adjustment module is used to dynamically generate port equipment planning and adjustment plans. By considering the factors affecting port equipment congestion, the module constructs an action space and a state space, and establishes a reward function to obtain the optimal plan. The port equipment planning and adjustment plan includes ship scheduling instructions, resource coordination instructions, and path planning instructions.

[0011] The feedback optimization module is used to receive port equipment planning and adjustment plans, analyze the impact of the plans on port congestion relief and resource utilization after actual application, evaluate the application effect of port dynamic diversion, and perform adaptive growth iterative optimization.

[0012] Furthermore, the process of collecting port production operation information is as follows:

[0013] Spatiotemporal data include vessel arrival time, berth occupancy status, yard container density, and equipment operating time;

[0014] Environmental data includes weather parameters, tidal cycles, and sudden fault alarms. Weather parameters include wind speed, rainfall, and visibility.

[0015] Business data includes vessel priorities, operational planning constraints, equipment operating efficiency, and resource consumption levels;

[0016] Among them, the databases of various business systems are synchronized to the intermediate library, and the data collection module uniformly extracts data from the intermediate library.

[0017] Furthermore, the specific process of constructing the port spatiotemporal knowledge graph is as follows:

[0018] Mark the N0 entity types in the port scene and mark any entity type as node V;

[0019] The entity types are classified by attributes based on the port production operation information, the attribute information of the node V is obtained, and the vector form of the node V is constructed: , where node V contains n attributes;

[0020] Data preprocessing is performed through edge computing nodes: the raw data of port production operation information is cleaned and filtered using the arithmetic mean method. Missing values ​​are then processed using minimum-maximum normalization and linear interpolation. The preprocessed port production operation information is obtained and marked as a preprocessed dataset. The preprocessed dataset is then spatially aligned in time.

[0021] Mark the relationship types between N0 entity types and mark any relationship type as edge E;

[0022] Construct the port spatiotemporal knowledge graph G by integrating nodes V and edges E: .

[0023] Furthermore, the specific process of building a port congestion propagation model is as follows:

[0024] The port area is dynamically divided into m0 grid cells, and any grid cell is marked as i. The state of grid cell i at time t is marked as ;

[0025] Construct a spatiotemporal graph convolutional network structure to analyze the production operation deviation of port equipment, obtain the temporal and spatial correlation of port equipment congestion, and analyze the influencing factors of port equipment congestion;

[0026] Then, by performing convolution operations on time and space, a state update formula is generated;

[0027] Mark the nodes of key events of port operations and construct a causal relationship network: Based on the historical data of key events, mark the nodes of any two key events as and , the conditional probability between statistical nodes , quantitatively analyze the degree of causal images between nodes, and thus build a causal network.

[0028] Furthermore, the specific process of the port equipment planning and adjustment plan is as follows:

[0029] Ship dispatch instructions include optimizing ship docking sequence and berth allocation. The action space A1 is established through ship dispatch instructions, and the state space S1 is established through the causal relationship network related to ship dispatch. The state space S1 includes the ship arrival time, cargo priority and berth status.

[0030] Resource coordination instructions include dynamically adjusting equipment resources, including cranes and / or trailers. The resource coordination instructions establish an action space A2, and then a state space S2 is established through a causal network related to resource coordination. The state space S2 includes the idle rate of loading and unloading equipment and the cargo yard load.

[0031] Path planning instructions include generating the shortest path between the yard and the berth, avoiding congested areas. Action space A3 is established through path planning instructions, and then state space S3 is established through the causal relationship network related to path planning. State space S3 includes the location of the yard and the berth, and the congestion status of the network unit.

[0032] Through the dynamic adjustment of state space and action space, the synchronous control of ship scheduling, resource coordination and path planning is achieved, and the ship waiting time, yard load balance and resource utilization rate during the adjustment process are monitored to obtain the optimal port equipment planning and adjustment plan.

[0033] Furthermore, the specific process of obtaining the optimal port equipment planning and adjustment plan is as follows:

[0034] The reward function R is established by taking a weighted sum of the ship waiting time, yard load balance, and resource utilization.

[0035] The port equipment planning adjustment plan corresponding to the maximum value of the reward function R is marked as the optimal plan.

[0036] Furthermore, the specific process of iterative optimization is as follows:

[0037] The congestion relief analysis process is to compare the congestion-related data before and after the application of the port equipment planning adjustment plan;

[0038] The resource utilization analysis process is to calculate the changes in equipment utilization before and after applying the port equipment planning adjustment plan;

[0039] Thus, quantitative index data is obtained and through weighted fusion, the comprehensive score of the application effect of the port equipment planning and adjustment plan is calculated;

[0040] The application effect of port dynamic diversion is comprehensively evaluated through the comprehensive score of application effect, so as to adjust the model parameters involved in port equipment planning and adjustment, and thus update and optimize them.

[0041] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0042] The present invention uses a data acquisition module to uniformly extract port production operation information, achieving cross-system data sharing and business collaboration. It also uses a deviation analysis module to construct a port spatiotemporal knowledge graph, analyze the production operation deviations of port equipment, and obtain the propagation patterns and influencing factors of port equipment congestion in time and space. It then uses a planning and adjustment module to dynamically generate a port equipment planning and adjustment plan, and separately regulates ship scheduling, resource coordination, and path planning. Furthermore, it uses a feedback optimization module to evaluate the application effect of port dynamic diversion and perform adaptive growth iterative optimization.

[0043] This invention connects port production, operations, equipment and facilities, safety and environmental protection, and other links to achieve data integration and collaborative management, improve data consistency, and support refined operations management. By combining edge computing nodes (real-time data cleaning, multi-source spatiotemporal alignment) with cloud processors (spatiotemporal graph convolutional network, causal reasoning conditional probability), it constructs a port spatiotemporal knowledge graph and a port congestion propagation model, dynamically switches diversion strategies, improves data processing and complex scenario analysis efficiency, and enhances real-time and intelligent analysis capabilities.

[0044] The present invention generates optimal ship, resource, and route plans and updates optimization model parameters based on actual application results, forming a "monitoring-decision-execution-feedback" closed loop to adaptively respond to dynamic changes in ports such as sudden failures and critical events. It establishes a reward function based on ship waiting time, yard balance, and resource utilization, optimizes scheduling strategies, and reduces equipment idleness and energy consumption, thereby improving resource utilization and alleviating congestion. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 shows a schematic diagram of the connection of the system modules of the present invention;

[0046] Figure 2 A schematic diagram showing the steps of the workflow of the present invention is shown. DETAILED DESCRIPTION

[0047] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0048] Example 1:

[0049] like Figure 1-2 As shown, a port production operation deviation analysis and adjustment system based on port digitalization includes a data acquisition module, a deviation analysis module, a planning and adjustment module, and a feedback optimization module, wherein the data acquisition module, the deviation analysis module, the planning and adjustment module, and the feedback optimization module are communicatively connected;

[0050] S1, the data collection module is used to connect to the port business system and collect port production operation information: port production operation information includes spatiotemporal data, environmental data and business data; the collection process of port production operation information is as follows:

[0051] Spatiotemporal data includes vessel arrival time, berth occupancy status, container density in the yard, and equipment operating hours; environmental data includes weather parameters, tidal cycles, and sudden fault alerts. Weather parameters include wind speed, rainfall, and visibility; operational data includes vessel priority, operation plan constraints, equipment operating efficiency, and resource consumption.

[0052] The databases of various business systems are synchronized to the intermediate database. The data collection module uniformly extracts data from the intermediate database, and the analyzed data results are then returned to each business system. Each business system then transforms its own system based on the feedback results. The list of business systems involved includes: business and financial integration system, equipment and energy system, integrated pipe network system, production and business integration system, container business system, roll-on / roll-off logistics production system, production cost accounting system, comprehensive safety management, logistics information network, and Internet of Things meteorological monitoring system.

[0053] Through functions such as data integration, data standardization, data development, data storage, metadata management, data asset management, data openness, and data security management, we provide an integrated and highly integrated platform tool system to provide data collection services for digital management and control scenarios, thereby achieving data sharing, business collaboration, and refined management of port operations by connecting port production, operations, equipment and facilities, safety and environmental protection, and other links.

[0054] S2, the deviation analysis module includes edge computing nodes and cloud processors:

[0055] S2-1, pre-processing the port production operation information through edge computing nodes to build the port spatiotemporal knowledge graph;

[0056] The specific process of constructing the port spatiotemporal knowledge graph is as follows:

[0057] Mark the N0 entity types in the port scene and mark any entity type as node V. The entity types include ships, berths, and loading and unloading equipment.

[0058] The entity types are classified by attributes based on the port production operation information, the attribute information of the node V is obtained, and the vector form of the node V is constructed: , where node V contains n attributes;

[0059] Among them, the ship node includes attributes such as arrival time and cargo priority; the berth node includes attributes such as occupancy status (free / occupied), maximum carrying capacity, and applicable ship type; the loading and unloading equipment node includes attributes such as operation efficiency, operation status (operating / faulty), and region.

[0060] Data preprocessing is performed through edge computing nodes: the raw data of port production operation information is cleaned and filtered using the arithmetic mean method. Missing values ​​are then processed using minimum-maximum normalization and linear interpolation. The preprocessed port production operation information is obtained and marked as a preprocessed dataset. The preprocessed dataset is then spatially and temporally aligned to unify the data format and ensure the consistency and accuracy of multi-source data.

[0061] Mark the relationship types between N0 entity types and mark any relationship type as edge E. The relationship types include the docking relationship between ships and berths, the operation relationship between loading and unloading equipment and berths, and the scheduling relationship between ships and loading and unloading equipment.

[0062] Construct the port spatiotemporal knowledge graph G by integrating nodes V and edges E: .

[0063] S2-2: A port congestion propagation model is then built using a cloud processor. This model divides the port into dynamic grid units and constructs a spatiotemporal graph convolutional network structure to analyze the production operation deviations of port equipment, obtain the propagation patterns of port equipment congestion in time and space, and analyze the factors affecting port equipment congestion.

[0064] The specific process of building a port congestion propagation model is as follows:

[0065] The port area is dynamically divided into m0 grid cells, and any grid cell is marked as i, including berths, yards, and loading and unloading equipment;

[0066] Mark the state of grid cell i at time t as ,state Includes attribute information of each entity type within the area where grid unit i is located;

[0067] Among them, custom quantification is performed on state-related attributes. For example, the occupancy status of berth nodes is quantitatively marked, with idle = 0 and occupied = 1. Standardized quantification is performed on data-related attributes. The operating efficiency of loading and unloading equipment nodes is marked by the standardized value of the processing volume per unit time.

[0068] Construct a spatiotemporal graph convolutional network structure to analyze the production operation deviation of port equipment, obtain the temporal and spatial correlation of port equipment congestion, and analyze the influencing factors of port equipment congestion;

[0069] By performing convolution operations on time and space, the state update formula is generated:

[0070] ;

[0071] Where Ni is the set of spatially adjacent grids; K is the time step; Ws and Wt aggregate the convolution weights of the adjacent network j and the current network i, respectively. Through historical data training, the spatial and temporal congestion correlation is captured to analyze the production operation deviation of port equipment. As the activation function, this scheme uses the ReLU function to introduce nonlinear mapping characteristics;

[0072] Mark the nodes of key events in port operations and construct a causal network;

[0073] The nodes of key events of port operations include loading and unloading equipment failure a1 (yes 1 no 0), concentrated arrival of ships at the port a1 (yes 1 no 0), and loading and unloading equipment operation delay a3 (yes 1 no 0);

[0074] Through the historical data of key events, any two key event nodes are marked as and , the conditional probability between statistical nodes , quantitatively analyze the degree of causal image between nodes;

[0075] Among them, when the conditional probability The higher the value, the more An event node occurs under the condition The higher the probability of an event and The stronger the causal relationship between them;

[0076] In this way, a causal network is built to conduct a comprehensive analysis of the causes of port operation deviations, thereby realizing dynamic diversion decisions.

[0077] S3, the planning and adjustment module dynamically generates a port equipment planning and adjustment plan: by considering the factors affecting port equipment congestion, it constructs an action space and a state space, and establishes a reward function to obtain the optimal plan. The port equipment planning and adjustment plan includes ship scheduling instructions, resource coordination instructions, and path planning instructions.

[0078] The specific process of port equipment planning and adjustment plan is as follows:

[0079] S3-1, ship dispatch instructions include optimizing ship docking sequence and berth allocation; the action space A1 is established through the ship dispatch instructions, and then the state space S1 is established through the causal relationship network related to ship dispatch; the state space S1 includes the ship arrival time, cargo priority and berth status;

[0080] In state space S1, the ship arrival time reflects the order of ship arrival and is used to determine whether backlogged ships need to be processed first. For example, ships carrying cold chain cargo have a higher priority and need to be allocated berths first to meet special operation timeliness or safety requirements. Berth status indicates whether a berth is free to avoid duplicate allocation. Action space A1 generates a berth allocation order list. For example, ships with high priority and early arrival time are given priority to be allocated to free berths, thereby optimizing the overall docking process and reducing ship waiting time.

[0081] S3-2, resource coordination instructions include dynamically adjusting equipment resources such as cranes and trailers; through resource coordination instructions, action space A2 is established, and then through the causal relationship network related to resource coordination, state space S2 is established; state space S2 includes the idle rate of loading and unloading equipment and the cargo yard load;

[0082] In state space S2, the idle rate of loading and unloading equipment refers to the ratio of idle equipment to the total number of equipment. A higher value indicates that resources are underutilized. The cargo yard load refers to the amount or weight of containers stacked at each yard, reflecting the operational pressure in the area. Action space A2 dynamically adjusts the equipment allocation strategy. For example, if a yard has a high load and a high equipment idle rate, idle equipment is dispatched to that yard to balance resource utilization and improve overall operational efficiency.

[0083] S3-3, path planning instructions include generating the shortest path from the yard to the berth and avoiding congested areas; the path planning instructions establish the action space A3, and then the state space S3 is established through the causal relationship network related to the path planning; the state space S3 includes the location of the yard and the berth and the congestion status of the network unit;

[0084] In state space S3, the grid congestion state is the traffic volume passing through the intersection within a certain period of time. The grid unit congestion level is output and quantitatively determined, such as high 3, medium 2, and low 1. The yard-berth location is determined by determining the starting point and end point and searching for a path between them based on map information. Action space A3 generates the shortest path that avoids highly congested grids, reducing transportation time and congestion impact.

[0085] S3-4, dynamically adjust the action space based on the state space, through the simultaneous adjustment of ship scheduling, resource coordination and path planning, and monitor the ship waiting time, yard load balance and resource utilization during the adjustment process, to obtain the optimal port equipment planning and adjustment plan;

[0086] Among them, ship waiting time is used to measure scheduling efficiency, yard load balance is used to measure resource allocation balance, and resource utilization is used to measure equipment idle rate and energy consumption.

[0087] S3-5: A reward function R is established by performing a weighted summation of the ship waiting time, the yard load balance, and the resource utilization rate. The port equipment planning adjustment plan corresponding to the maximum value of the reward function R is marked as the optimal plan.

[0088] The reward function formula is designed based on the principle that the shorter the ship waiting time, the more balanced the yard load, and the higher the resource utilization rate, the higher the reward function R. By dynamically generating port equipment planning and adjustment plans, this reduces ship waiting time, balances yard load, and reduces equipment idleness and resource consumption.

[0089] Therefore, it can autonomously learn and execute optimal diversion decisions for complex dynamic scenarios in ports, achieve efficient collaboration in ship scheduling, resource coordination and route planning, and solve the current problems of weak spatiotemporal data analysis capabilities of port digital platforms, port congestion caused by equipment operation deviations, and lack of dynamic diversion solutions.

[0090] S4: The feedback optimization module receives the port equipment planning and adjustment plan, analyzes its impact on port congestion relief and resource utilization after its actual implementation, evaluates the effectiveness of the port dynamic diversion application, and performs adaptive growth and iterative optimization.

[0091] S4-1, the congestion relief analysis process involves comparing congestion-related data before and after the implementation of the solution. For example, comparing the average waiting time for ships to berths. If this time is significantly reduced after the implementation of the solution, it indicates that the solution has a positive effect in alleviating ship waiting congestion. By analyzing the changing trend of cargo backlog in the yard, if the backlog decreases and remains at a reasonable level, it indicates that yard congestion has been effectively alleviated.

[0092] S4-2, the resource utilization analysis process evaluates the impact of the plan on resource utilization by calculating changes in equipment utilization. For example, the ratio of the crane's actual operating time to the total operating time is calculated. If this ratio increases after the plan is implemented, it means that the equipment is more fully utilized. By analyzing the collaborative operation of different equipment, if the waiting time between equipment for coordination is reduced, it means that resource allocation is more reasonable and overall utilization is improved.

[0093] S4-3, based on the above impact analysis, obtain quantitative indicator data and conduct a comprehensive evaluation of the application effect of port dynamic diversion by weighted calculation of the comprehensive application effect score; the higher the comprehensive application effect score, the better the application effect of port dynamic diversion;

[0094] Based on the evaluation results of the application effect of dynamic diversion in ports, the model parameters involved in port equipment planning and adjustment are adjusted. For example, if it is found that the judgment of ship priority in the ship scheduling instructions is unreasonable and causes congestion, the back-end management personnel will adjust and update the parameters of ship priority based on the actual situation, thereby changing the priority judgment rules, realizing adaptive growth and iterative optimization, and continuously improving the port's operational efficiency.

[0095] The application effects of the present invention are as follows:

[0096] This invention connects port production, operations, equipment and facilities, safety and environmental protection, and other links to achieve data integration and collaborative management, improve data consistency, and support refined operations management. It then combines edge computing nodes (real-time data cleaning, multi-source spatiotemporal alignment) with cloud processors (spatiotemporal graph convolutional networks, causal reasoning conditional probability) to construct a port spatiotemporal knowledge graph and a port congestion propagation model, dynamically switching diversion strategies, improving data processing and complex scenario analysis efficiency, and enhancing real-time and intelligent analysis capabilities.

[0097] This invention generates optimal ship, resource, and routing solutions and updates optimization model parameters based on actual application results, forming a "monitoring-decision-execution-feedback" closed loop to adaptively respond to dynamic port changes such as sudden failures and critical events. It also establishes a reward function based on ship waiting time, yard balance, and resource utilization, optimizes scheduling strategies, and reduces equipment idleness and energy consumption, thereby improving resource utilization and alleviating congestion.

[0098] In summary, the present invention uniformly extracts port production operation information through the data acquisition module to achieve cross-system data sharing and business collaboration, and constructs a port spatiotemporal knowledge graph through the deviation analysis module to analyze the production operation deviation of port equipment, obtain the propagation law and influencing factors of the port equipment congestion status in time and space, and then dynamically generates a port equipment planning and adjustment plan through the planning and adjustment module, and regulates ship scheduling, resource coordination and path planning respectively, and then evaluates the application effect of port dynamic diversion through the feedback optimization module, and performs adaptive growth iterative optimization.

[0099] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art will appreciate that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution.

[0100] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0101] The above data processing is to remove the dimension and obtain its numerical value for calculation. The setting of the interval and threshold size is for the convenience of comparison. The size of the threshold depends on the amount of sample data and the cardinality set by technical personnel in this field for each group of sample data. As long as it does not affect the proportional relationship between the parameter and the quantized value, the preset parameters are set by technical personnel in this field according to actual conditions.

[0102] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A port production operation deviation analysis and adjustment system based on port digitization, characterized by: It includes data acquisition module, deviation analysis module, planning adjustment module and feedback optimization module; The data acquisition module is used to connect to the port business system and collect port production operation information: port production operation information includes spatiotemporal data, environmental data and business data; The deviation analysis module includes edge computing nodes and cloud processors. The edge computing nodes preprocess port production operation information to construct a port spatiotemporal knowledge graph. The cloud processor then builds a port congestion propagation model, divides the port into dynamic grid units, and constructs a spatiotemporal graph convolutional network structure to analyze the production operation deviations of port equipment, obtain the propagation patterns of port equipment congestion status in time and space, and analyze the factors affecting port equipment congestion. The planning and adjustment module is used to dynamically generate port equipment planning and adjustment plans. By considering the factors affecting port equipment congestion, the module constructs an action space and a state space, and establishes a reward function to obtain the optimal plan. The port equipment planning and adjustment plan includes ship scheduling instructions, resource coordination instructions, and path planning instructions. The feedback optimization module is used to receive port equipment planning and adjustment plans, analyze the impact of the plans on port congestion relief and resource utilization after actual implementation, evaluate the effectiveness of port dynamic diversion applications, and perform adaptive growth and iterative optimization; The specific process of constructing the port spatiotemporal knowledge graph is as follows: Mark the N0 entity types in the port scene and mark any entity type as node V; The entity types are classified by attributes based on the port production operation information, the attribute information of the node V is obtained, and the vector form of the node V is constructed: , where node V contains n attributes; Data preprocessing is performed through edge computing nodes: the raw data of port production operation information is cleaned and filtered using the arithmetic mean method. Missing values ​​are then processed using minimum-maximum normalization and linear interpolation. The preprocessed port production operation information is obtained and marked as a preprocessed dataset. The preprocessed dataset is then spatially aligned in time. Mark the relationship types between N0 entity types and mark any relationship type as edge E; Construct the port spatiotemporal knowledge graph G by integrating nodes V and edges E: ; The specific process of building a port congestion propagation model is as follows: The port area is dynamically divided into m0 grid cells, and any grid cell is marked as i. The state of grid cell i at time t is marked as ; Construct a spatiotemporal graph convolutional network structure to analyze the production operation deviation of port equipment, obtain the temporal and spatial correlation of port equipment congestion, and analyze the influencing factors of port equipment congestion; Then, by performing convolution operations on time and space, a state update formula is generated; Mark the nodes of key events of port operations and construct a causal relationship network: Based on the historical data of key events, mark the nodes of any two key events as and , the conditional probability between statistical nodes , quantitatively analyze the degree of causal images between nodes, and thus build a causal network.

2. A port production operation deviation analysis and adjustment system based on port digitization according to claim 1, characterized in that: The process of collecting port production operation information is as follows: Spatiotemporal data include vessel arrival time, berth occupancy status, yard container density, and equipment operating hours; Environmental data includes weather parameters, tidal cycles, and sudden fault alarms. Weather parameters include wind speed, rainfall, and visibility. Business data includes vessel priorities, operational planning constraints, equipment operating efficiency, and resource consumption levels; Among them, the databases of various business systems are synchronized to the intermediate library, and the data collection module uniformly extracts data from the intermediate library.

3. The port production operation deviation analysis and adjustment system based on port digitization according to claim 2 is characterized by: The specific process of generating the port equipment planning adjustment plan is as follows: Ship dispatch instructions include optimizing ship docking sequence and berth allocation. The action space A1 is established through ship dispatch instructions, and the state space S1 is established through the causal relationship network related to ship dispatch. The state space S1 includes the ship arrival time, cargo priority and berth status. Resource coordination instructions include dynamically adjusting equipment resources, including cranes and trailers. The resource coordination instructions establish an action space A2, and then a state space S2 is established through a causal network related to resource coordination. State space S2 includes the idle rate of loading and unloading equipment and the cargo yard load. Path planning instructions include generating the shortest path between the yard and the berth, avoiding congested areas. Action space A3 is established through path planning instructions, and then state space S3 is established through the causal relationship network related to path planning. State space S3 includes the location of the yard and the berth, and the congestion status of the network unit. Through the dynamic adjustment of state space and action space, the synchronous control of ship scheduling, resource coordination and path planning is achieved, and the ship waiting time, yard load balance and resource utilization rate during the adjustment process are monitored to obtain the optimal port equipment planning and adjustment plan.

4. The port production operation deviation analysis and adjustment system based on port digitization according to claim 3 is characterized by: The specific process of obtaining the optimal port equipment planning and adjustment plan is as follows: The reward function R is established by taking a weighted sum of the ship waiting time, yard load balance, and resource utilization. The port equipment planning adjustment plan corresponding to the maximum value of the reward function R is marked as the optimal plan.

5. The port production operation deviation analysis and adjustment system based on port digitization according to claim 4 is characterized in that: The specific process of iterative optimization is: The congestion relief analysis process is to compare the congestion-related data before and after the application of the port equipment planning adjustment plan; The resource utilization analysis process is to calculate the changes in equipment utilization before and after applying the port equipment planning adjustment plan; Thus, quantitative index data is obtained and through weighted fusion, the comprehensive score of the application effect of the port equipment planning and adjustment plan is calculated; The application effect of port dynamic diversion is comprehensively evaluated through the comprehensive score of application effect, so as to adjust the model parameters involved in port equipment planning and adjustment, and thus update and optimize them.

Images