Port logistics intelligent simulation and decision optimization platform based on digital twinning

Abstract

The invention provides a port logistics intelligent simulation and decision optimization platform based on a digital twin technology, and the platform comprises a basic layer which provides calculation, storage, network and basic software support; the resource layer integrates data resources, model resources and algorithm resources and is used for managing port multi-source data, business process models and intelligent algorithms; the digital twinborn layer is used for constructing and maintaining a high-fidelity digital twinborn body synchronous with a physical port and realizing virtual-real mapping, dynamic data fusion and model self-evolution; the simulation layer is used for deducing based on the digital twin and comprises a simulation engine and an event driving and data interaction module; and the application layer is used for providing a user interface, analyzing decision and optimizing functions and supporting scheme simulation, evaluation and closed-loop optimization. According to the method, the high-fidelity digital twinborn body of the port can be constructed, real-time mapping, accurate simulation, prospective prediction and intelligent decision making of the whole process of port logistics are achieved, and a technical closed loop of data acquisition, digital twinborn, intelligent decision making and virtual-real linkage is formed.

Description

Technical Field

[0001] This invention belongs to the field of port logistics management and digital twin technology, and specifically relates to a port logistics intelligent simulation and decision optimization platform based on digital twin, which is suitable for port operation scheduling, resource optimization and handling of complex scenarios. Background Technology

[0002] With the continuous growth of global trade, ports, as key hubs for cargo transportation, play a crucial role in ensuring the smooth operation of the entire supply chain. Efficient port logistics can accelerate cargo turnover, reduce logistics costs, and enhance the competitiveness of enterprises.

[0003] Port logistics involves numerous complex processes and the coordinated operation of a large number of equipment and personnel. Examples include ship berthing and departure, container loading and unloading, and the scheduling of trucks and other transport vehicles. These processes are complex, interconnected, and mutually influential; a problem in any one of these processes can lead to a decrease in the overall efficiency of port logistics.

[0004] Currently, in the port logistics sector, the optimization of various operational processes and resource scheduling mainly relies on experience and traditional analytical methods. When formulating vessel berthing plans, it is often based on past experience with similar vessels and the port's current approximate berth usage, without fully considering the impact of factors such as the vessel's actual cargo capacity, loading and unloading time requirements, and subsequent operational processes on berth occupancy time and efficiency. In terms of truck scheduling, it is also mostly based on simple task allocation principles, such as first-come, first-served, lacking a comprehensive consideration of real-time port traffic conditions, the possibility of congestion on truck routes, and the priority of different tasks.

[0005] Traditional port logistics planning and management methods are increasingly revealing limitations in the face of growing business volume and complex operational demands. These limitations include high berth allocation error rates and a sharp decline in equipment utilization rates in the event of sudden failures (such as quay crane malfunctions), making it difficult to meet the requirements of efficient, intelligent, and precise operations in modern ports. Therefore, there is an urgent need to develop a digital twin platform capable of real-time mapping, accurate simulation, intelligent decision-making, and closed-loop control of the entire port logistics process.

[0006] And the following problems exist:

[0007] (1) Lack of accurate operation process simulation: Existing methods are difficult to accurately simulate the specific operation process and the coordination relationship between each operation link in port logistics. In the process of container loading and unloading, it is impossible to accurately take into account the actual operating efficiency fluctuations, failure probability of equipment such as quay cranes and gantry cranes, as well as the differences in loading and unloading of different types of containers, resulting in a large deviation in the prediction of the entire loading and unloading operation time and efficiency.

[0008] (2) Insufficient resource scheduling optimization: For various resources in ports, such as berths, equipment, and transport vehicles, existing studies mostly focus on single-objective optimization of total time or cost, ignoring individual experience factors such as ship waiting time, making it difficult to balance the interests of port management and ship owners. The scheduling model does not fully consider constraints such as port tides and equipment coordination, and cannot meet complex scheduling requirements, resulting in low resource utilization.

[0009] (3) Lack of ability to cope with complex scenarios and emergencies: The port logistics environment is complex and ever-changing, and may face emergencies such as severe weather, sudden equipment failures, and traffic congestion. Existing methods are difficult to effectively simulate and analyze these complex scenarios and emergencies, and cannot formulate response strategies in advance. Once an emergency occurs, it can easily lead to chaos and delays in port logistics.

[0010] (4) Insufficient data utilization: Ports accumulate a large amount of data in their daily operations, but traditional planning and management methods have failed to fully explore and utilize the value of this data. This data is scattered across different business systems and has not been effectively integrated and analyzed, failing to provide comprehensive and accurate data support for intelligent decision-making and optimization of port logistics, and has not formed a closed loop of "data-driven decision-making". Summary of the Invention

[0011] In view of this, the purpose of this invention is to provide a port logistics intelligent simulation and decision optimization platform based on digital twin technology, so as to solve the problems of inaccurate simulation of operation process, insufficient optimization of resource scheduling, weak ability to cope with complex scenarios and emergencies, and insufficient utilization of data in existing port logistics planning and management.

[0012] In a first aspect, embodiments of the present invention provide a port logistics intelligent simulation and decision optimization platform based on digital twin technology, comprising,

[0013] The foundational layer provides computing, storage, networking, and basic software support;

[0014] The resource layer integrates data resources, model resources, and algorithm resources for managing multi-source port data, business process models, and intelligent algorithms.

[0015] The digital twin layer constructs and maintains a high-fidelity digital twin synchronized with the physical port, enabling virtual-real mapping, dynamic data fusion, and model self-evolution.

[0016] The simulation layer performs dynamic simulations based on digital twins, including a simulation engine, event-driven mechanisms, and data interaction modules.

[0017] The application layer provides user interface, analysis, decision-making and optimization functions, and supports scheme simulation, evaluation and closed-loop optimization.

[0018] Preferably, the resource layer includes:

[0019] The data resource module is used to collect and store port logistics data from the terminal operating system, sensors, and external data sources;

[0020] The model resource module is used to construct port logistics business models, loading and unloading equipment models, and traffic environment models.

[0021] The algorithm resource module integrates scheduling algorithm libraries, prediction algorithm libraries, and optimization algorithm libraries, supporting multi-objective optimization and intelligent decision-making.

[0022] Preferably, the digital twin layer includes:

[0023] The virtual-real mapping module establishes a two-way real-time data channel between physical entities and digital mirrors;

[0024] The dynamic data fusion module cleans, correlates, and performs semantic recognition on multi-source heterogeneous data.

[0025] The model self-evolution module automatically calibrates model parameters based on the comparison between simulation and actual data.

[0026] Preferably, the simulation layer supports:

[0027] Event-driven simulation of emergencies, including equipment failure, weather changes, and traffic congestion;

[0028] Multi-port collaborative simulation supports multi-port linkage and supply chain resilience assessment along the Belt and Road Initiative.

[0029] Real-time recording and transmission of simulation process data are used for model calibration and scheme evaluation.

[0030] Preferably, the application layer provides:

[0031] A two-dimensional / three-dimensional visualization interface displays the real-time status of the port and the simulation process.

[0032] Multi-option comparative analysis and decision recommendation generation;

[0033] The optimization module calls upon algorithm resources to generate the optimal scheduling and path planning solutions.

[0034] Preferably, it supports data integration and control command issuance with external systems such as the port operating system, ERP, and WMS through RESTful APIs and standardized data interfaces.

[0035] Preferably, the scheduling algorithm library integrates genetic algorithms, particle swarm optimization algorithms, and reinforcement learning methods for the coordinated scheduling of berths, quay cranes, and container trucks.

[0036] Preferably, the optimization algorithm library integrates the NSGA-II multi-objective optimization algorithm to seek Pareto optimal solutions among multiple objectives such as cost, time, and efficiency.

[0037] The embodiments of the present invention bring the following beneficial effects:

[0038] This invention can construct a high-fidelity digital twin of a port, enabling real-time mapping, accurate simulation, forward-looking prediction, and intelligent decision-making of the entire port logistics process. It forms a closed-loop technology of "data collection - digital twin - intelligent decision-making - virtual-real linkage", providing port managers with scientific and accurate decision-making basis, thereby realizing the intelligent upgrade of port logistics, improving operational efficiency, reducing operating costs, and enhancing the port's competitiveness and resilience.

[0039] This invention improves port logistics operation efficiency: through precise simulation and resource scheduling optimization of the entire port logistics process, it assists in the rational arrangement of ship berthing, equipment operation and truck transportation, reduces operation waiting time and traffic congestion, improves the overall operational efficiency of port logistics, shortens the time goods stay in port, and accelerates the turnover speed of goods.

[0040] This invention reduces operating costs by optimizing resource allocation, improving equipment utilization, and reducing equipment idleness and energy consumption; it also rationally plans truck routes to reduce transportation costs. By simulating and responding to various unforeseen circumstances in advance, it avoids additional costs caused by operational delays and chaos.

[0041] This invention enhances the ability to cope with complex scenarios and emergencies: it can effectively simulate and analyze complex scenarios and emergencies such as severe weather, equipment failure, and traffic congestion, formulate response strategies in advance, improve the port's emergency response capabilities in the face of emergencies, and ensure the stable operation of port logistics.

[0042] This invention provides a scientific basis for decision-making: by mining and analyzing historical data and comparing and evaluating the simulation results of different schemes, it provides port managers with comprehensive and accurate decision-making information, helping them to make scientific and reasonable decisions and improve the port's management level and competitiveness.

[0043] This invention enhances supply chain collaboration and resilience: through multi-port linkage simulation, it can optimize resource allocation from a global supply chain perspective, identify cross-border logistics risks in advance, enhance the collaborative efficiency and anti-interference capabilities of the entire logistics chain, and serve the logistics infrastructure construction of the national "Belt and Road" initiative.

[0044] Other features and advantages of the invention will be set forth in the following description, and some features will be obvious from the description or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained through the structures particularly pointed out in the description, claims, and drawings.

[0045] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0046] Figure 1 This is a platform system architecture diagram according to an embodiment of the present invention;

[0047] Figure 2 This is a platform usage flow and closed-loop control diagram according to an embodiment of the present invention. Detailed Implementation

[0048] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0049] Example 1: The platform adopts a five-layer architecture design:

[0050] 1. The foundational layer provides basic hardware and software support for the entire platform, including servers, network devices, operating systems, and databases, providing computing, storage, networking, and basic software support;

[0051] 2. The resource layer is the core asset library of the platform, which integrates three major resources: data resources, model resources, and algorithm resources. It supports multi-source data access, high-precision modeling, and intelligent algorithm invocation, and is used to manage port multi-source data, business process models, and intelligent algorithms.

[0052] Specifically,

[0053] 1) Data resource module, used to collect and store port logistics data from the terminal operating system, sensors and external data sources; the logistics data acquisition module collects multi-source data from the terminal operating system (TOS), equipment management system, sensors (GPS, status sensors) and external data sources (AIS, weather forecast), and the logistics data storage module performs structured and unstructured storage.

[0054] Furthermore, the logistics data acquisition module is responsible for collecting port logistics-related data from various port business systems, such as the Terminal Operating System (TOS), equipment management system, transportation management system, sensors (e.g., vehicle positioning sensors, equipment status sensors), and external data sources (e.g., weather forecast data, ship dynamic information data). This data includes port infrastructure information (e.g., number, length, and depth of berths, yard layout), equipment information (e.g., model, quantity, and performance parameters of quay cranes, gantry cranes, container trucks, etc.), business data (e.g., ship arrival schedules, cargo loading and unloading tasks, transportation orders), and environmental data (e.g., weather conditions, tidal information). The logistics data storage module stores the collected data in a database, employing a combination of relational databases (MySQL) and non-relational databases (MongoDB). The relational database stores structured business data, ensuring data consistency and integrity; the non-relational database stores semi-structured and unstructured data, such as real-time data collected by sensors and text logs, to meet the requirements for data storage flexibility and efficiency. Simultaneously, a data backup and recovery mechanism is established to ensure data security.

[0055] 2) Model resource module, which constructs port logistics business models, loading and unloading equipment models, and traffic environment models;

[0056] Specifically,

[0057] Port logistics business models: Based on the actual business processes and rules of port logistics, various business models are constructed. For example, the ship berthing model calculates the optimal berthing position and berthing time for ships based on factors such as ship type, size, cargo capacity, and estimated loading and unloading time, combined with the port's berth resources; the container loading and unloading model simulates the container loading and unloading process and time, considering factors such as the operating efficiency, failure probability, and operation sequence of equipment such as quay cranes and gantry cranes; and the truck transportation model simulates the transportation process and time consumption of trucks within the port, based on factors such as the number of trucks, load capacity, travel speed, and route planning.

[0058] Loading and unloading equipment model module: Detailed equipment models are created for various types of equipment in ports, such as quay cranes, gantry cranes, reach stackers, and container trucks. These models include not only the physical attributes of the equipment (such as dimensions, weight, and power) but also its operational performance parameters (such as loading and unloading efficiency, cycle time, and failure rate). Through the precise construction of these equipment models, the operation of the equipment in port logistics operations can be simulated more realistically.

[0059] Port Traffic Environment Model Module: This module constructs a port traffic environment model considering factors such as road layout, traffic flow, and traffic rules within the port. Based on port transport road network data, it simulates the transport routes of trucks and other transport vehicles within the port, assisting in the analysis of the impact of traffic congestion on transport efficiency. It supports strategies such as simulating traffic light control and route adjustment optimization to alleviate traffic congestion and improve traffic flow.

[0060] 3) Algorithm resource module, integrating scheduling algorithm library, prediction algorithm library, and optimization algorithm library, supporting multi-objective optimization and intelligent decision-making. Specifically,

[0061] Scheduling Algorithm Library: Integrates multiple advanced scheduling algorithms such as genetic algorithms, particle swarm optimization, and reinforcement learning. Primarily used to solve the collaborative allocation and real-time scheduling problems of core resources such as berths, quay cranes, and container trucks, it can handle various complex constraints (such as time windows, resource capabilities, and task priorities). It is the core computing engine for the platform to achieve dynamic optimization of resource allocation, improve operational efficiency, and reduce vessel port time, supporting the platform's evolution from static planning to dynamic intelligent scheduling.

[0062] The prediction algorithm library includes prediction models such as time series analysis (e.g., ARIMA), machine learning (e.g., gradient boosting trees), and deep learning (e.g., LSTM). Its function is to accurately predict key operational indicators based on historical and real-time data, such as ship arrival times, container throughput, and equipment failure probabilities. It provides accurate initial input for simulations, enabling port logistics situational awareness and risk warning.

[0063] The optimization algorithm library's core is a multi-objective optimization solver based on the Non-Dominated Sorting Genetic Algorithm (NSGA-II), while also integrating classic algorithms such as linear programming and integer programming. It is used to solve complex optimization problems within the platform, such as finding the optimal balance (Pareto optimal solution) among multiple conflicting objectives like cost, time, and efficiency. It supports the generation of scientifically optimal decision-making schemes, enabling the platform to transition from "simulation analysis" to "decision optimization."

[0064] The resource layer gathers multi-source data from inside and outside the port through the data resource module; constructs digital models of business processes, equipment and environment through the model resource module; and integrates intelligent algorithms such as scheduling, prediction and optimization through the algorithm resource library, providing rich and standardized "ammunition" for upper-level simulation and decision-making.

[0065] 3. Digital Twin Layer: Constructs and maintains a high-fidelity digital twin synchronized with the physical port, realizing virtual-real mapping, dynamic data fusion, and model self-evolution. This layer serves as the central hub for virtual-real interaction, achieving accurate mapping of the physical port. It does not directly perform simulation calculations or store raw data, but is responsible for "activating" the raw data and models provided by the resource layer into a high-fidelity digital twin that operates synchronously with the physical port, is holographically mapped, and can make forward-looking predictions.

[0066] Specifically,

[0067] 1) The virtual-real mapping module establishes a two-way real-time data channel between the physical entity and its digital mirror; specifically,

[0068] The virtual-physical mapping module forms the foundational framework for building and maintaining digital twins. Based on the equipment model, business model, and environment model in the resource layer, it constructs digital mirrors in virtual space that correspond one-to-one with physical port elements (such as berths, quay cranes, trucks, and containers). Its core responsibility is to establish and maintain a two-way, real-time data channel: on the one hand, by accessing real-time sensor data (such as GPS location and equipment status signals) from the resource layer data acquisition module, it continuously drives the status updates of the digital mirrors, ensuring synchronization with the physical entities; on the other hand, it maps optimization instructions or predicted states generated by the simulation layer (such as new routes planned for trucks) to virtual objects, preparing them for subsequent dissemination to the physical world (such as truck navigation systems). This ensures that any changes in the physical world are reflected instantly in the virtual space, and that decisions made in the virtual space can find corresponding physical executors.

[0069] 2) The dynamic data fusion module cleans, correlates, and performs semantic recognition on multi-source heterogeneous data. Specifically,

[0070] The dynamic data fusion module is the "nerve center" of the digital twin, responsible for processing multi-source, heterogeneous, and massive amounts of real-time and historical data from the resource layer. It not only receives sensor data but also integrates external environmental data such as business order data and weather forecasts from the TOS (Transmission of System). Its core task is to extract these raw data into "events" or "states" with clear business meaning and contextual relationships through data cleaning, semantic recognition, and correlation analysis. This provides an accurate and reliable driving force for the simulation layer and a calibration basis for the model self-evolution module.

[0071] 3) The model self-evolution module automatically calibrates model parameters based on the comparison of simulation and actual data. Specifically,

[0072] The model self-evolution module is key to endowing the digital twin with the ability to "learn and evolve." It compares and analyzes the simulation results generated from multiple runs in the simulation layer with actual operational data obtained through the dynamic data fusion module. When a systematic deviation is detected between the simulation predictions and actual results (e.g., the actual loading and unloading efficiency of a certain type of quay crane consistently exceeds the model's set value), the module automatically initiates a calibration process, using machine learning algorithms to fine-tune and optimize the parameters of the business model or equipment model in the resource layer. This continuous "simulation-actuality-calibration" closed loop allows the digital twin to continuously approach and more accurately depict the real operational patterns of the physical port, thereby significantly improving the accuracy of its future predictions and the reliability of its recommended decisions.

[0073] 4. Simulation Layer: Based on digital twins, this layer performs forward-looking dynamic simulations, supporting event injection and multi-scenario simulation. It includes a simulation engine, event-driven mechanisms, and data interaction modules. Specifically,

[0074] 1) Simulation Engine Module: This is the core component of the entire simulation platform, responsible for scheduling and driving various simulation models to execute simulation tasks. It receives simulation scenarios from the application layer and initiates the simulation based on the current system state provided by the digital twin layer (such as the real-time location and status of all equipment, vehicles, and cargo). The engine sequentially calls business models (berthing models, loading / unloading models, etc.) in the resource layer for calculations according to configurable time steps, propelling the "virtual time" of the entire virtual port forward. It coordinates the execution order and data exchange of all models to ensure the continuity and logical consistency of the simulation, ultimately generating a complete "pre-simulation video" of port operations from the current state for a future period, along with all process data.

[0075] 2) Event-Driven Module: An event-driven mechanism is introduced during the simulation process. Within the framework of the simulation engine, key conditions during the simulation process are monitored. When preset trigger conditions are met (e.g., simulation time reaches a certain point, equipment utilization exceeds a threshold, or a real-world emergency is injected by the digital twin layer), this module immediately interrupts the normal process and adjusts the simulation model's operating parameters and state according to preset event handling rules to simulate the dynamic changes in actual port logistics. For example, when a quay crane malfunction is detected, the event-driven module will suspend related container loading and unloading operations and adjust subsequent operation plans and resource allocation based on the malfunction type and repair time.

[0076] 3) Data Interaction Module: Responsible for interaction between the simulation layer and external data. Before the simulation begins, it obtains the latest initial data and model parameters from the resource layer and digital twin layer. During the simulation, it records all intermediate data generated by the simulation engine in real time (such as the position of the truck at each moment and the lifting record of the container). After the simulation ends, it categorizes and stores the complete simulation process data and result data back into the database of the resource layer. This process not only provides raw materials for the analysis of the application layer, but also accumulates "simulation-reality" comparison data for model calibration for the model self-evolution module of the digital twin layer.

[0077] Furthermore, the simulation layer, based on the physical model and real-time data provided by the digital twin layer, synchronized with the physical world, performs forward-looking dynamic simulations according to set rules and time steps. It relies heavily on the precise mirror provided by the digital twin layer and provides comprehensive simulation process and result data for the application layer's analysis and decision-making. The simulation layer supports: event-driven simulation of emergencies, including equipment failure, weather changes, and traffic congestion; multi-port collaborative simulation, supporting multi-port linkage and supply chain resilience assessment along the Belt and Road Initiative; and real-time recording and transmission of simulation process data for model calibration and solution evaluation.

[0078] 5. Application Layer: Provides user interface, analysis, decision-making, and optimization functions, supporting scheme simulation, evaluation, and closed-loop optimization. This includes: a 2D / 3D visualization interface displaying the port's real-time status and simulation process; multi-scheme comparative analysis and decision suggestion generation; and an optimization module that calls upon algorithm resources to generate optimal scheduling and path planning schemes. Specifically,

[0079] 1) A 2D / 3D visualization interface transforms the simulation capabilities, data resources, and model algorithms provided by the lower layer into business functions that users can directly understand and use. It can call down to the simulation layer for inference calculations and utilize the algorithm library of the resource layer for in-depth optimization, ultimately serving users with the decision-making results or directly feeding them back to the physical world.

[0080] User Interface Module: This module provides users with an intuitive and user-friendly interface. Through 2D / 3D visualization, it presents the real-time status of the digital twin layer and the simulation process and results in a clear and intuitive manner, using graphics, charts, and animations. It also offers user-friendly interactive features, allowing users to configure simulation parameters, select simulation scenarios, trigger optimization calculations, and issue control commands.

[0081] 2) Analysis and Decision Support Module: This module deeply mines the massive amounts of output data from the simulation layer and the historical data stored in the resource layer. Through multi-dimensional and multi-indicator comparative analysis (such as comparisons of operational efficiency, cost, and resource utilization of different schemes), it generates easy-to-understand evaluation reports and decision recommendations. It transforms raw simulation data into business intelligence with direct guiding significance, assisting managers in making informed decisions.

[0082] 3) Optimization Module: After the analysis and decision support module identifies shortcomings in existing solutions and generates optimization suggestions, users can use the optimization module to set optimization objectives (such as lowest cost or shortest time) and constraints. It then calls upon scheduling, prediction, and optimization intelligent algorithms from the resource layer's algorithm library to automatically construct and generate better scheduling or planning schemes (such as new berth allocation plans or truck route planning). The generated new scheme can be immediately sent back to the simulation layer for verification and simulation to ensure its effectiveness, thus forming a rapid iterative closed loop of "analysis-optimization-verification." Ultimately, the optimal decision is recommended to the user or issued for execution through the digital twin layer.

[0083] In this embodiment of the invention, the platform supports data integration and control command issuance with external systems such as the port operating system, ERP, and WMS through RESTful APIs and standardized data interfaces.

[0084] Example 2: The specific process for implementing platform functions is as follows:

[0085] 1. Simulation of the entire port logistics process

[0086] Ship entry and exit simulation: Simulates navigation trajectory, berthing time, etc., in conjunction with hydrological and meteorological conditions. Specifically, based on factors such as ship type, size, cargo capacity, and port hydrological and meteorological conditions, it calculates parameters such as ship navigation trajectory, berthing time, and departure time, and displays the ship's real-time position and status within the port.

[0087] Container loading and unloading simulation: This simulation considers equipment efficiency and failure probability, analyzing operation time and resource utilization. Specifically, it simulates the loading, unloading, and handling process of containers between quay cranes, gantry cranes, and container yards. It considers factors such as equipment operating efficiency, failure probability, and operation sequence, analyzing indicators such as container loading and unloading time, container yard occupancy, and equipment utilization.

[0088] Truck transportation simulation: Based on route planning, this simulation model optimizes traffic flow by simulating the transportation process. Specifically, it simulates the transportation process of trucks within a port, including transport from the yard to the berth, from the berth to the yard, and between different areas within the port. Based on factors such as the number of trucks, their load capacity, travel speed, and route planning, it analyzes the transportation efficiency, traffic congestion, and transportation costs.

[0089] 2. Resource scheduling optimization

[0090] Berth allocation optimization: Combining vessel arrival schedules, the NSGA-II algorithm is used to minimize waiting time. Specifically, based on factors such as vessel arrival schedules, cargo capacity, and loading / unloading time requirements, and considering the port's berth resources, an optimization algorithm is applied to allocate optimal berths to vessels, thereby improving berth utilization and vessel loading / unloading efficiency, and reducing vessel dwell time in port.

[0091] Equipment scheduling optimization: Dynamically allocate tasks for quay cranes and gantry cranes to improve utilization. Specifically, rationally schedule equipment such as quay cranes, gantry cranes, and reach stackers in the port. Determine the operating sequence and task allocation of equipment based on its operating capacity, current operating status, and container loading and unloading requirements to improve equipment utilization and operational efficiency.

[0092] Truck scheduling optimization: Avoiding congestion and reducing transportation costs through route algorithms. Specifically, based on information such as the truck's transportation task, real-time location, and traffic conditions, route planning algorithms and scheduling strategies are used to plan the optimal driving route for the truck, rationally arrange the truck's transportation tasks, avoid traffic congestion, and improve the truck's transportation efficiency.

[0093] 3. Complex scene simulation

[0094] Severe weather simulation: Adjusting ship speed and equipment operation safety.

[0095] Equipment failure simulation: Analyze the impact of failures and develop emergency response strategies.

[0096] Traffic congestion simulation: alleviating congestion through signal control and route optimization.

[0097] 4. Data Analysis and Decision Support

[0098] Historical data mining: Analyze throughput trends to support resource planning.

[0099] Simulation result evaluation: Compare the cost and efficiency indicators of different schemes and recommend the optimal solution.

[0100] 5. Multi-port linkage simulation

[0101] Collaborative scheduling simulation: Construct a virtual supply chain with multiple ports along the "Belt and Road" and test collaborative strategies, such as resource reallocation.

[0102] Supply chain resilience assessment: Simulate disruption events, recommend contingency plans, and enhance risk resistance capabilities.

[0103] 6. Commercial software integration

[0104] Interface with TOS: Synchronize job data via RESTful API to form a closed loop.

[0105] Third-party software interface: Supports data exchange with ERP and WMS, enhancing business integration.

[0106] Example 3, see Figure 2 The platform's operation is designed with three phases:

[0107] In the planning phase of the support plan: Users design berth allocation and equipment scheduling schemes based on digital twins, and preset evaluation indicators. Specifically, users design and initialize schemes based on digital twins. First, they obtain the task schedule for logistics transportation at the designated port, synchronize real-time port data, and drive the high-fidelity digital twin to reflect the current situation. Subsequently, users allocate berths, schedule equipment, and plan routes for the logistics tasks, quickly generating one or more alternative port logistics support schemes. Each scheme is bound to a digital twin and has a preset evaluation indicator system, laying a precise initial state and evaluation benchmark for subsequent simulation and deduction.

[0108] Simulation and deduction phase: This phase involves forward-looking simulations of the logistics process, injecting unexpected events to test robustness, and presenting the results visually. Specifically, it receives plans and tasks from the planning phase, performs forward-looking simulations of port logistics operations, and simulates the entire process of ship operations, equipment operation, and truck transportation. Through the event-driven module, unexpected situations such as equipment failure and traffic congestion can be injected to test the robustness of the plans. The logistics transportation process, the status of various ships, loading and unloading facilities and equipment, and the overall port logistics situation are presented visually.

[0109] Performance evaluation phase: Comparing simulation and real-world data, calibrating the model, and generating a debriefing report. Specifically, this involves aggregating data from the port logistics simulation phase, analyzing the effectiveness of the logistics support plan based on evaluation indicators, and visually reviewing the port logistics transportation process. Furthermore, it allows for the calibration of the digital twin model's accuracy by comparing real-world port logistics process data (after completing port logistics tasks) with simulation data.

[0110] It should be noted that, in this document, the terms “comprising,” “including,” or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0111] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A port logistics intelligent simulation and decision optimization platform based on digital twins, characterized in that, include: The foundational layer provides computing, storage, networking, and basic software support; The resource layer integrates data resources, model resources, and algorithm resources for managing multi-source port data, business process models, and intelligent algorithms. The digital twin layer constructs and maintains a high-fidelity digital twin synchronized with the physical port, enabling virtual-real mapping, dynamic data fusion, and model self-evolution. The simulation layer performs dynamic simulations based on digital twins, including a simulation engine, event-driven mechanisms, and data interaction modules. The application layer provides user interface, analysis, decision-making and optimization functions, and supports scheme simulation, evaluation and closed-loop optimization.

2. The platform according to claim 1, characterized in that, The resource layer includes: The data resource module is used to collect and store port logistics data from the terminal operating system, sensors, and external data sources; The model resource module is used to construct port logistics business models, loading and unloading equipment models, and traffic environment models. The algorithm resource module integrates scheduling algorithm libraries, prediction algorithm libraries, and optimization algorithm libraries, supporting multi-objective optimization and intelligent decision-making.

3. The platform according to claim 1, characterized in that, The digital twin layer includes: The virtual-real mapping module establishes a two-way real-time data channel between physical entities and digital mirrors; The dynamic data fusion module cleans, correlates, and performs semantic recognition on multi-source heterogeneous data. The model self-evolution module automatically calibrates model parameters based on the comparison between simulation and actual data.

4. The platform according to claim 1, characterized in that, The simulation layer supports: Event-driven simulation of emergencies, including equipment failure, weather changes, and traffic congestion; Multi-port collaborative simulation supports multi-port linkage and supply chain resilience assessment along the Belt and Road Initiative. Real-time recording and transmission of simulation process data are used for model calibration and scheme evaluation.

5. The platform according to claim 1, characterized in that, The application layer provides: A two-dimensional / three-dimensional visualization interface displays the real-time status of the port and the simulation process. Multi-option comparative analysis and decision recommendation generation; The optimization module calls upon algorithm resources to generate the optimal scheduling and path planning solutions.

6. The platform according to claim 1, characterized in that, It supports data integration and control command issuance with external systems such as port operating systems, ERP, and WMS through RESTful APIs and standardized data interfaces.

7. The platform according to claim 2, characterized in that, The scheduling algorithm library integrates genetic algorithms, particle swarm optimization, and reinforcement learning methods for the coordinated scheduling of berths, quay cranes, and container trucks.

8. The platform according to claim 2, characterized in that, The optimization algorithm library integrates the NSGA-II multi-objective optimization algorithm, which is used to seek Pareto optimal solutions among multiple objectives such as cost, time and efficiency.

Images