Ship dispatching and carrying method based on space-time constraint scheduling simulation and conflict optimization

Abstract

The invention relates to the technical field of port ship scheduling, and discloses a ship scheduling carrying method based on space-time constraint scheduling simulation and conflict optimization. Obtaining ship scheduling tasks and port space-time resource data, and generating structured data in a classified manner; constructing a multi-dimensional space-time constraint model, and determining an initial constraint condition; scheduling simulation is carried out on the basis, and the robustness of the scheme is improved by introducing real-time data; detecting conflicts and carrying out iterative optimization by adopting a multi-objective optimization algorithm; and synchronously correcting model parameters with real-time data, generating a scheduling instruction set, and pushing and executing the scheduling instruction set. And meanwhile, path planning is performed by using a deep reinforcement learning algorithm, and a visual interface is constructed to assist scheduling. The method can effectively integrate data, optimize a scheduling scheme, improve the utilization rate of port resources, enhance the ability to deal with emergencies, and realize efficient and intelligent management of port ship scheduling.

Description

Technical Field

[0001] The present invention relates to the technical field of port ship scheduling, and in particular to a ship scheduling and embarkation method based on time-space constraint scheduling simulation and conflict optimization. Background Art

[0002] Ship scheduling is a complex and critical task in port operations. With the booming global trade, port cargo throughput continues to climb, the number of ships is increasing, and the difficulty of ship scheduling is increasing. Traditional ship scheduling methods are no longer able to meet the requirements of efficient modern port operations and have exposed numerous problems.

[0003] Early ship scheduling relied primarily on manual experience, with dispatchers making arrangements based on basic information such as the order in which ships arrive at the port and the type of cargo. This approach lacked systematicity and scientific integrity, relying entirely on personal judgment and experience. Different dispatchers made widely varying decisions, making it difficult to achieve optimal resource allocation. When ports were small and ship traffic was low, manual scheduling might have been able to keep things running smoothly. However, as port business volume grew, the inefficiency and error-prone nature of manual scheduling became increasingly apparent, often resulting in ships staying in port for extended periods, leaving berths and loading and unloading equipment idle or overused, severely impacting the port's overall operational efficiency.

[0004] With the development of information technology, some simple computer-aided dispatching systems have begun to emerge. These systems utilize databases to store vessel and port resource information and perform preliminary scheduling arrangements using simple algorithms. However, most of these systems only consider a single factor, such as allocating berths based solely on the vessel's estimated arrival time, ignoring the vessel type, cargo attributes, dynamic changes in port resources, and various complex spatiotemporal constraints. Different vessel types require different berthing conditions, loading and unloading equipment, and time; differences in cargo attributes affect loading and unloading priorities and operational processes; and the occupancy of port resources such as berths, waterways, and loading and unloading equipment is dynamic and subject to multiple constraints. Due to their failure to fully consider these factors, simple computer-aided dispatching systems have been ineffective in practical applications, frequently leading to problems such as long vessel waiting times and resource conflicts.

[0005] Existing technologies also have significant shortcomings in the management of space and time resources. Ports' time resources are not properly and precisely divided and effectively utilized. The scheduling of ship arrivals and departures lacks flexibility and scientific rationality, which can easily lead to wasted time slots or overcrowding. In terms of spatial resource management, the geographical layout and functional zoning of ports are not fully utilized, and allocation is not optimized based on the operational needs of ships and the distribution of resources. For example, different types of ships operate at different efficiencies in different areas, but existing scheduling methods often ignore this fact, resulting in low spatial resource utilization.

[0006] Furthermore, traditional scheduling methods and existing systems have limited adaptability to emergencies such as severe weather and equipment failures. Once these unforeseen circumstances occur, the entire scheduling plan becomes chaotic, making timely and effective adjustments impossible, further exacerbating port operations. Inclement weather can affect ship speed and berthing safety, while equipment failures can disrupt loading and unloading operations. However, existing scheduling mechanisms struggle to quickly respond and replan scheduling plans, posing significant challenges to port operations. Summary of the Invention

[0007] The purpose of the present invention is to provide a ship dispatching and embarking method based on time-space constraint scheduling simulation and conflict optimization to solve the problems raised in the above background technology.

[0008] To achieve the above objectives, the present invention provides the following technical solution: a ship dispatching and embarking method based on spatiotemporal constraint scheduling simulation and conflict optimization, the method comprising: Obtaining ship dispatch task information and port spatiotemporal resource data, and classifying the task information according to preset dimensions to generate structured dispatch task data; the preset dimensions include ship type, cargo attributes, berthing window, and navigation path; Construct a multi-dimensional spatiotemporal constraint model and define a spatiotemporal coordinate system based on the classification dimension of the scheduling task data. The spatiotemporal coordinate system includes a time axis, a space axis, and a resource axis. Determining initial spatiotemporal constraints based on the ship type and cargo attributes, including berthing time thresholds, channel occupancy conflict rules, and resource allocation priorities; Perform scheduling simulation on ship dispatching tasks based on the spatiotemporal constraint model, simulate the dynamic behavior of ships in the port through the discrete event simulation engine, and generate preliminary dispatching plans; Conduct conflict detection on the preliminary scheduling plan, identify ship task nodes with overlapping time and space resources or conflicting rules, and mark the conflict type and conflict intensity; Iteratively optimize conflicting nodes using a multi-objective optimization algorithm, adjust the spatiotemporal parameters of ship missions or resource allocation strategies, and generate conflict-free optimized scheduling solutions. Synchronize the optimized scheduling plan with the port's real-time operation data to dynamically modify the parameters of the spatiotemporal constraint model; Generate a final ship scheduling instruction set based on the revised spatiotemporal constraint model, the instruction set including ship berthing timing, path planning, and resource allocation information; The ship dispatch instruction set is pushed to the port dispatch terminal and the ship control system to trigger the automated dispatch execution process.

[0009] Preferably, the constructing of a multidimensional spatiotemporal constraint model includes: Define gridded zones of spatial axes based on the geographical layout of the port, with each zone associated with channel capacity, number of berths, and distribution of loading and unloading equipment; Divide the time axis into continuous or discrete time slots, and associate each time slot with a priority weight for ships entering or leaving the port; The dynamic occupation status of port resources is defined based on the resource axis, including energy supply, loading and unloading equipment availability, and human resource allocation.

[0010] Preferably, the multi-objective optimization algorithm adopts an improved genetic algorithm, including: The spatiotemporal parameters of conflicting nodes are encoded as chromosomes, and a fitness function is defined to evaluate the conflict intensity, resource utilization, and task delay cost. A population of new scheduling solutions is generated through crossover and mutation operations, and the Pareto front screening strategy is used to retain the non-inferior solution set.

[0011] Preferably, the scheduling simulation includes: Introducing a real-time data stream processing module to dynamically receive port weather data, ship positioning information, and equipment failure alarms; The input parameters of the simulation model are adjusted according to real-time data to generate a robust scheduling solution.

[0012] Preferably, the conflict detection includes: Establish a time-space resource occupancy matrix and detect the overlapping areas of ship tasks in time slots and space partitions through matrix operations; Based on the rule engine matching the preset conflict judgment logic, it identifies conflict events that violate channel priority or safety distance.

[0013] Preferably, the path planning adopts a deep reinforcement learning algorithm, including: The port channel topology is modeled as a graph network, and the ship movement actions are defined as state transitions; The Q-learning strategy is used to train the agent to select the optimal path to minimize navigation time and avoid conflicts.

[0014] Preferably, the definition of the resource axis includes: Integrate port environmental factors into resource allocation strategies, including tidal cycles, wind speed limits, and visibility conditions; Dynamically adjust the start and stop sequence of loading and unloading equipment and the speed thresholds of ships entering and leaving the port based on environmental factors.

[0015] Preferably, the method further comprises: Build a ship mission pattern library based on historical scheduling data to extract high-frequency mission combinations and typical conflict scenarios; The pattern matching algorithm is used to predict potential conflicts of newly added scheduling tasks and pre-generate optimization strategies.

[0016] Preferably, the resource allocation strategy includes: Use dynamic priority queues to manage ship tasks, adjusting queue order based on cargo urgency and ship fuel remaining; Reserve buffers for critical resources to ensure that high-priority tasks can preempt resources when conflicts arise.

[0017] Preferably, the method further comprises: Build a visual scheduling monitoring interface to map the spatiotemporal constraint model, conflict detection results, and optimized scheduling plan into a three-dimensional dynamic view; Receive manual intervention instructions through the interactive operation interface and update the scheduling instruction set in real time.

[0018] Compared with the prior art, the present invention has the following beneficial effects: The present invention has many significant beneficial effects: The ship dispatching and embarkation method based on time-space constraint scheduling simulation and conflict optimization proposed in this invention has brought significant positive impacts in many aspects to port ship dispatching, and has effectively promoted the improvement of port operation efficiency and management level.

[0019] During data processing and task classification, structured scheduling task data is generated by acquiring ship scheduling task information and port spatiotemporal resource data, and categorizing them according to pre-set dimensions such as ship type, cargo attributes, berthing window, and navigation path. This process streamlines previously complex and disorganized data, providing a solid foundation for subsequent precise scheduling. Compared to traditional, general recording methods, structured data facilitates rapid processing and analysis by computer systems, enabling precise identification of the characteristics and requirements of different ship tasks, avoiding scheduling errors caused by information disorganization, and significantly improving the accuracy and targeted nature of scheduling.

[0020] One of the core innovations of this invention is the construction of a multidimensional spatiotemporal constraint model. By defining a spatiotemporal coordinate system consisting of a time axis, a space axis, and a resource axis, the spatial axis is divided into grid zones based on the port's geographical layout, and information such as waterway capacity, berth number, and the distribution of loading and unloading equipment is associated with this model. This enables refined management of port spatial resources. Before entering the port, ships can rationally plan their berthing locations based on their needs and the availability of resources in each zone, avoiding waste or competition for berth resources. By dividing the time axis into continuous or discrete time slots and assigning a priority weight to each time slot for ship entry and exit, the system streamlines the scheduling of ship arrivals and departures. For example, ships transporting fresh produce can be assigned a higher priority in their time slots to ensure timely loading and unloading, minimizing losses. The dynamic occupancy of port resources is defined based on the resource axis, taking into account factors such as energy supply, loading and unloading equipment availability, and human resource allocation, further improving the efficiency and rationality of resource utilization. In the event of a loading and unloading equipment failure, the system can promptly adjust the scheduling plan based on the resource axis information, assigning tasks to other available equipment and ensuring operational continuity.

[0021] When determining the initial spatiotemporal constraints, the system prioritizes berthing time thresholds, channel occupancy conflicts, and resource allocation based on vessel type and cargo attributes, fully considering the unique characteristics of different vessels and cargoes. Large tanker loading and unloading operations are complex, so a longer berthing time threshold is set to prevent operations from being delayed due to tight schedules. Vessels carrying flammable and explosive cargoes are given higher priority in channel access and resource allocation to ensure safe port operations.

[0022] During the scheduling simulation phase, a discrete event simulation engine is used to simulate the dynamic behavior of ships within the port, and a real-time data stream processing module is introduced. This module dynamically receives real-time data such as port weather data, ship positioning information, and equipment failure alarms. Based on this data, the module adjusts the input parameters of the simulation model to generate a robust scheduling plan. In the event of inclement weather, the system can promptly adjust the ship's speed and berthing plan to ensure ship safety while minimizing the impact on the overall scheduling. This real-time adjustment capability greatly enhances the adaptability of the scheduling plan, effectively responding to various emergencies in port operations.

[0023] In the conflict detection phase, by establishing a spatiotemporal resource occupancy matrix and matching conflict determination logic based on a rule engine, we can quickly and accurately identify ship mission nodes with overlapping spatiotemporal resources or conflicting rules, and mark the conflict type and intensity. This enables dispatchers to promptly identify potential issues, providing a clear basis for subsequent conflict optimization and preventing further escalation of conflicts that could impact port operations.

[0024] A multi-objective optimization algorithm is used to iteratively optimize conflicting nodes, adjusting the spatiotemporal parameters of ship missions or resource allocation strategies. Using an improved genetic algorithm as an example, the spatiotemporal parameters of conflicting nodes are encoded as chromosomes. A fitness function is defined to evaluate conflict intensity, resource utilization, and mission delay costs. A population of new scheduling solutions is generated through crossover and mutation operations, and a Pareto frontier screening strategy is used to retain the set of non-inferior solutions. This optimization approach balances multiple objectives, generating conflict-free optimized scheduling solutions, effectively improving resource utilization, reducing ship waiting time, and mission delay costs.

[0025] By synchronizing the optimized scheduling plan with real-time port operational data, the parameters of the spatiotemporal constraint model are dynamically modified, enabling the model to reflect the actual port operations in real time. The final ship scheduling instruction set generated based on the modified model includes information on ship berthing timing, path planning, and resource allocation, more closely aligned with actual needs. Path planning utilizes a deep reinforcement learning algorithm, modeling the port's waterway topology as a graph network. Using a Q-learning strategy, the intelligent agent is trained to select the optimal path, minimize navigation time, and avoid conflicts, further improving the efficiency and safety of ship navigation within the port.

[0026] Finally, the ship dispatch instruction set is pushed to the port dispatch terminal and ship control system, triggering the automated dispatch execution process, achieving automated and intelligent dispatch. Simultaneously, a visual dispatch monitoring interface is constructed, mapping the spatiotemporal constraint model, conflict detection results, and optimized dispatch plans into a three-dimensional dynamic view, allowing dispatchers to intuitively understand port operations. Manual intervention instructions are received through an interactive operation interface, and the dispatch instruction set is updated in real time. This ensures automation while retaining the flexibility of manual intervention, further improving the accuracy and reliability of dispatch. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a working principle diagram of the ship dispatching and embarkation method based on spatiotemporal constraint scheduling simulation and conflict optimization according to the present invention; Figure 2 Schematic diagram of the process built for the multidimensional spatiotemporal constraint model; Figure 3 Schematic diagram of the process for ship scheduling conflict detection; Figure 4 A flowchart for resource axis definition and resource allocation considering environmental factors. DETAILED DESCRIPTION

[0028] 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 creative efforts are within the scope of protection of the present invention.

[0029] See also Figures 1-4 The present invention provides a technical solution: a ship dispatching and embarking method based on time-space constraint scheduling simulation and conflict optimization. The specific implementation scheme is as follows: Data Acquisition and Processing: This data is collected from a wide range of sources, including ship dispatch information provided by shipping companies, including vessel type, cargo attributes, projected berthing windows, and planned sailing routes. Port management systems also provide data on port spatiotemporal resources, covering the port's geographic layout, waterway capacity, number of berths, and distribution of loading and unloading equipment. This information is categorized according to pre-set dimensions, including vessel type, cargo attributes, berthing windows, and sailing routes. This categorization generates structured dispatch data for subsequent processing.

[0030] Constructing a multidimensional spatiotemporal constraint model: Based on the classification dimensions of scheduling task data, a spatiotemporal coordinate system is defined. This coordinate system consists of a time axis, a spatial axis, and a resource axis. The time axis can be divided into continuous or discrete time slots based on actual scheduling needs to facilitate the management of ship arrival and departure times. The spatial axis is gridded based on the port's geographical layout. Each zone is associated with information such as channel capacity, number of berths, and distribution of loading and unloading equipment, thereby clarifying the resource status of different areas. The resource axis is used to define the dynamic occupancy status of port resources, including energy supply, loading and unloading equipment availability, and human resource allocation.

[0031] Determine initial spatiotemporal constraints: Initial spatiotemporal constraints are determined based on vessel type and cargo attributes. For example, different ship types have different berthing time thresholds. Large oil tankers have longer berthing time thresholds due to the complexities of loading and unloading operations. Different cargo attributes also influence channel occupancy conflict rules and resource allocation priorities. For example, vessels transporting flammable and explosive cargoes enjoy higher priority when navigating the channel, and resources required for safe loading and unloading are prioritized in resource allocation.

[0032] Scheduling Simulation: Based on a spatiotemporal constraint model, scheduling simulation is performed on ship dispatch tasks. A discrete event simulation engine is used to simulate the dynamic behavior of ships within the port, including entry and exit, berthing, loading and unloading, and other operations, to generate preliminary scheduling plans. During the simulation process, a real-time data stream processing module can be introduced to dynamically receive real-time data such as port weather data, ship positioning information, and equipment failure alarms. This data is used to adjust the simulation model's input parameters, making the generated scheduling plan more robust.

[0033] Conflict Detection: Conflict detection is performed on the preliminary scheduling plan, identifying ship mission nodes with overlapping time and space resources or conflicting rules, and marking the conflict type and intensity. For example, by establishing a time and space resource occupancy matrix, matrix operations are used to detect overlapping areas of ship missions in time slots and spatial partitions. Based on the rule engine, the preset conflict determination logic is matched to identify conflict events that violate channel priority or safety distance.

[0034] Conflict Optimization: A multi-objective optimization algorithm is used to iteratively optimize conflicting nodes, adjust the spatiotemporal parameters of ship missions, or adjust resource allocation strategies, to generate conflict-free optimal scheduling solutions. This multi-objective optimization algorithm can employ an improved genetic algorithm, encoding the spatiotemporal parameters of conflicting nodes as chromosomes and defining a fitness function to assess conflict intensity, resource utilization, and task delay costs. A population of new scheduling solutions is generated through crossover and mutation operations, and a Pareto frontier screening strategy is employed to retain the set of non-inferior solutions.

[0035] Model parameter correction: Synchronize the optimized scheduling plan with real-time port operation data to dynamically adjust the parameters of the spatiotemporal constraint model based on actual operation. For example, if the actual utilization frequency of a berth in a specific time period differs significantly from the model preset frequency, the berth's occupancy parameters for the corresponding time slot can be adjusted.

[0036] Generate a dispatching instruction set: Based on the modified spatiotemporal constraint model, a final dispatching instruction set is generated. This instruction set includes information on ship berthing timing, path planning, and resource allocation. Path planning utilizes a deep reinforcement learning algorithm, modeling the port channel topology as a graph network and defining ship movement actions as state transitions. A Q-learning strategy is then used to train the agent to select the optimal path, minimizing travel time and avoiding conflicts.

[0037] Instruction push and execution: Push the ship dispatch instruction set to the port dispatch terminal and ship control system, triggering the automated dispatch execution process to achieve efficient dispatch and boarding of ships in the port.

[0038] The present invention will be further described below in conjunction with Examples 1 to 6: Example 1:

[0039] In the construction of multi-dimensional space-time constraint model, the grid partition of space axis is divided into several rectangular grid areas according to the geographical layout of the port. Taking the center of each area as the reference point, the channel capacity, number of berths and distribution of handling equipment information in the area are associated. Assuming that one of the grid areas A, the channel capacity can be determined by historical ship flow data and channel width, depth and other physical parameters, for example, the channel in this area allows 5 medium-sized ships to pass simultaneously at a certain time period; the number of berths is 8, marked as B1-B8, each berth has different length and depth, suitable for different types of ships to dock; the distribution of handling equipment, the area is equipped with 3 large container cranes, 2 bulk cargo handling machines, etc., the location and working capacity of each equipment are recorded in detail.

[0040] The time axis is divided into discrete time slots considering the different operation conditions of the port day and night, each time slot is 1 hour. During the busy period of the day (6:00-20:00), the priority weight of the ship entering and leaving the port is determined according to the urgency of the goods and the type of the ship. For ships transporting fresh goods, the priority weight of entering and leaving the port is high during this period; while for some ordinary dry cargo transport ships, the priority weight is relatively low. During the night period (20:00-6:00), due to the maintenance of part of the handling equipment and the arrangement of manpower, the priority weight of the ship entering and leaving the port is adjusted as a whole, for large ships and special handling equipment, it will be arranged in the appropriate time slot after the maintenance of the equipment at night, and the corresponding priority weight is given.

[0041] The definition of resource axis not only includes energy supply, handling equipment availability and human resource allocation, but also integrates port environmental factors into resource allocation strategy. Taking the tidal cycle as an example, during the tidal flood period, the water depth of part of the channel increases, allowing ships with deeper draft to pass. At this time, according to the type of the ship and the task arrangement, the speed threshold of the ship entering and leaving the port is adjusted appropriately. For large oil tankers, the speed of entering and leaving the port can be increased under the premise of safety to make full use of tidal resources. When the wind speed of the port exceeds a certain threshold, part of the handling operation will be suspended, and the start and stop timing of the handling equipment will be adjusted to ensure the safety of the operation. The visibility condition has a great influence on the navigation of the ship, when the visibility is low, the ship entering and leaving the port will be limited, and additional lighting and guiding resources will be allocated to the ships in the port to ensure the safe docking and operation of the ships in the port. By integrating these environmental factors, the dynamic and reasonable allocation of port resources is realized.

[0042] Example 2:

[0043] An improved genetic algorithm is used in the conflict optimization phase. For example, consider a scheduling scenario at a port during a specific time period. Assume that multiple ships conflict over berth use and channel access. The spatiotemporal parameters of the conflicting nodes are encoded to construct a chromosome. For example, ship A has an estimated arrival time of t1, an estimated berthing duration of t2, and a planned berth of b1. Ship B has an estimated arrival time of t3, an estimated berthing duration of t4, and a planned berth of b2. These parameters are combined according to specific rules to form a chromosome encoding. Here, t1, t2, b1, t3, t4, and b2 are treated as gene segments within the chromosome, and together they form a chromosome describing the ship scheduling parameters for the current conflict scenario.

[0044] The fitness function is defined to evaluate the conflict intensity, resource utilization and task delay cost. Taking berth conflict as an example, if ship A and ship B plan to use the same berth in the same time interval, assuming that the berth can only serve one ship per unit time, ship A plans to occupy the berth for Ta, ship B plans to occupy the berth for Tb, and the overlap between the two times is To, then the conflict intensity calculation formula is: The larger the C value, the more serious the conflict. Taking berth resources as an example, assuming that the theoretical usable time of a berth in a statistical period T is Ttotal and the actual usage time is Tused, the resource utilization formula is: , the larger the R value, the more fully the resources are utilized. To calculate the task delay cost, assume that the urgency coefficient of the cargo transported by the ship is k (the larger the k value, the more urgent the cargo), the unit time delay cost is c, and the expected delay time is t, then the task delay cost formula is Taking these three factors into consideration, the fitness function , where α, β, and γ are weight coefficients, which are adjusted based on the actual needs and priorities of port operations. For example, during periods of high urgency, the value of γ can be appropriately increased; if the port prioritizes efficient resource utilization, the weight of β can be increased.

[0045] A population of new scheduling solutions is generated through crossover and mutation operations. During the crossover operation, two chromosomes, such as chromosome A and chromosome B, are randomly selected and their gene segments are swapped at certain positions. For example, the gene segment for ship arrival time is swapped. The gene for ship A's arrival time in chromosome A is swapped with the gene for ship B's arrival time in chromosome B, generating two new chromosomes representing new possible scheduling solutions. The mutation operation randomly changes certain genes within the chromosomes, for example, randomly changing the estimated ship's berthing time. Ship A's original estimated berthing time is t2, but through mutation, it may become t2'. This introduces new scheduling variables and increases the diversity of the population. A Pareto front screening strategy is used to retain the non-inferior solution set. From the numerous new scheduling solutions generated, solutions that are relatively superior across multiple objectives, such as conflict intensity, resource utilization, and task delay cost, are selected. This set of solutions is the non-inferior solution set. This set of solutions is then iteratively optimized to ultimately obtain an optimal scheduling solution that is conflict-free and has good overall performance.

[0046] Example 3:

[0047] A real-time data stream processing module is introduced in the scheduling simulation link. Taking a large port as an example, the port is equipped with complete real-time data monitoring equipment, including meteorological monitoring stations, ship positioning systems, and equipment status monitoring sensors. The real-time data stream processing module obtains port meteorological data in real time, such as wind speed, wind direction, precipitation, visibility and other information. When the wind speed reaches a certain threshold, the simulation model will adjust the ship's sailing speed and berthing plan according to the type and performance parameters of the ship. For example, the normal sailing speed of a small container ship is v0. When the wind speed is detected to reach v1 (v1 exceeds the ship's safe sailing wind speed threshold), the simulation model automatically adjusts its sailing speed to v2 (v2 according to the ship's navigation safety standards). <v0),并相应延长其预计到达时间和靠泊时间。同时,重新评估该船舶在新的到达时间下对泊位和其他港口资源的占用情况,避免与其他船舶产生冲突。

[0048] Ship positioning information is transmitted in real time to the real-time data stream processing module via the Automatic Identification System (AIS). If a ship's actual navigation trajectory deviates significantly from its planned trajectory, the module will immediately feed this information back to the simulation model. Suppose a bulk carrier deviates from its course due to a sudden mechanical failure. After receiving the deviation information, the simulation model will replan the ship's navigation path based on the ship's current position, the distribution of surrounding ships, and the real-time conditions of the port's waterway. The ship is guided to a safe area, such as a nearby emergency anchorage, while the entire port's scheduling plan is dynamically adjusted. For example, the navigation plans of other ships that originally planned to use a certain section of the waterway at the same time as the bulk carrier will be adjusted to avoid collisions and ensure the safety and smooth navigation of ships in the port.

[0049] Equipment failure alarms are also received in real time by the real-time data stream processing module. When a critical piece of loading and unloading equipment, such as a large crane, fails, the simulation model responds quickly. For ships originally scheduled to use that crane for loading and unloading, the model reallocates available loading and unloading equipment. If no other suitable equipment is available, the loading and unloading order of these ships is adjusted, prioritizing those with access to other functioning equipment. By processing this data in real time and continuously adjusting the simulation model's input parameters, the resulting scheduling solution is more adaptable to various emergencies in port operations, offering greater robustness and ensuring efficient port operations.

[0050] Example 4:

[0051] During the conflict detection process, a spatiotemporal resource occupancy matrix is ​​established. For example, a simplified port model has four time slots (T1, T2, T3, and T4) and five spatial partitions (S1, S2, S3, S4, and S5). For each ship's mission, if ship 1 occupies spatial partition S3 for loading or unloading during time slot T2, the corresponding element (T2, S3) in the spatiotemporal resource occupancy matrix is ​​marked as 1, indicating that the spatiotemporal region is occupied. If no ship occupies the region, the corresponding element (T2, S3) is marked as 0. In this way, the mission information of all ships in the port is mapped into the matrix.

[0052] Matrix operations are used to detect overlapping areas of ship missions across time slots and spatial partitions. For example, to detect conflicts in the occupancy of various spatial partitions within a specific time period, the matrix columns corresponding to the corresponding time slots are summed. If the sum of the column elements corresponding to spatial partition S2 in time slot T3 is 2, this indicates that two ships are scheduled to operate in area S2 during time slot T3, resulting in a spatial resource overlap conflict. This matrix operation method can quickly and intuitively identify conflicts between ship missions in time and space.

[0053] The preset conflict judgment logic includes channel priority rules and safety distance rules. For the channel priority rules, certain special vessels, such as vessels performing emergency rescue tasks, harbor pilot vessels, etc., have absolute priority on the channel, and other vessels must give way. When a vessel enters the channel, the rule engine checks whether it violates the channel priority according to the type and task information of the vessel. For the safety distance rules, different minimum safety distance standards are set according to the type and size of the vessel. For example, the minimum safety distance between large oil tankers is d1, and the minimum safety distance between container ships and oil tankers is d2. The position information of the vessels is obtained in real time through the vessel positioning system, and the rule engine calculates the actual distance between the vessels and compares it with the preset minimum safety distance. If the actual distance is less than the minimum safety distance, it is judged as a conflict event that violates the safety distance, and the conflict type and conflict intensity are marked. The conflict intensity can be quantified according to the difference between the actual distance and the minimum safety distance, the larger the difference, the higher the conflict intensity, providing detailed information for subsequent conflict optimization.

[0054] Embodiment 5:

[0055] The path planning adopts a deep reinforcement learning algorithm, taking a complex channel network of a certain port as the background. The topological structure of the port channel is modeled as a graph network, where the nodes represent key positions of the channel, such as channel junctions, berths, anchorages, etc., and the edges represent the channels connecting these nodes. The movement of the vessel is defined as a state transition, for example, the vessel moves from node n1 to adjacent node n2, which is considered as a state transition.

[0056] The agent is trained to choose the optimal path through the Q-learning strategy. During the training process, the agent selects an action (i.e., moves to an adjacent node) according to the state information of the current node of the vessel, including the positions of surrounding vessels, the congestion status of the channel, the occupancy status of the berth, etc. After executing the action, the agent will obtain a reward value. The setting of the reward value is related to the two goals of minimizing the sailing time and avoiding conflicts. If the vessel successfully avoids other vessels and has a shorter sailing time, the agent will obtain a higher reward value; otherwise, if a conflict occurs or the sailing time is too long, the agent will obtain a lower reward value.

[0057] The agent updates the Q-value table according to the reward values obtained by continuously trying different actions. The Q-value table records the expected cumulative reward value of selecting each action in each state. Suppose the agent selects action a1 in state s1, executes the action and obtains reward value r1, and moves to new state s2. According to the update formula of Q-learning, , where α is the learning rate, which controls the magnitude of each update and typically ranges from 0 to 1; γ is the discount factor, which reflects the importance the agent places on future rewards and also ranges from 0 to 1. As the number of training sessions increases, the agent gradually learns to select the action with the highest Q-value according to the Q-value table, thereby achieving the goals of minimizing navigation time and avoiding conflicts. In practical applications, when a new ship requires path planning, the agent uses the trained model to find the optimal path based on the ship's initial and target positions. This path planning information is then provided to the ship's control system, guiding the ship to navigate safely and efficiently within the port.

[0058] Example 6:

[0059] A ship mission pattern library is constructed based on historical dispatch data. Ship dispatch data from a specific port over the past several years is collected, including detailed information such as vessel type, cargo attributes, expected berthing windows, navigation routes, and conflicts that occurred during the dispatch process. This massive data set is then analyzed in depth, and data mining algorithms are used to extract high-frequency task combinations and typical conflict scenarios. For example, data analysis revealed that during the peak summer tourist season, a large number of passenger ships carrying tourists arrive at the port in the morning hours. These ships often clash when using specific berths and boarding / disembarking lanes, creating a typical conflict scenario.

[0060] A pattern matching algorithm is used to predict potential conflicts for newly added scheduling tasks. When a new ship scheduling task is added, the relevant information about the task is compared and matched with the data in the ship task pattern library. If the new task is similar to a typical conflict scenario, the system can pre-generate an optimization strategy. For example, for the aforementioned scenario of concentrated passenger ship arrivals during the summer tourist peak season, a pre-generated optimization strategy might include coordinating berth allocation plans in advance to reserve dedicated berths for passenger ships; or adjusting the arrival times of passenger ships to spread the arrival times of some passenger ships to other time periods to avoid conflicts caused by concentrated arrivals.

[0061] In terms of resource allocation, a dynamic priority queue is used to manage ship tasks. The queue order is adjusted based on the urgency of the cargo and the ship's fuel remaining. Vessels transporting urgently needed supplies, such as emergency medical supplies and disaster relief supplies, are given higher priority. Vessels with lower fuel reserves are also given a higher priority, taking into account potential safety risks. This ensures that these vessels can refuel in a timely manner and avoid safety issues in the port. For example, when a ship transporting emergency medical supplies enters the port, the system automatically increases its mission priority, prioritizing berths and loading and unloading equipment, and expediting its loading and unloading operations.

[0062] Reserve a buffer zone for critical resources to ensure that high-priority tasks can preempt them in the event of a conflict. For example, for a port's limited and critical loading and unloading equipment, such as large gantry cranes, a certain amount of operating time is reserved as a buffer zone. When a high-priority vessel arrives, if the equipment is currently serving other vessels, the current operation can be suspended to prioritize the high-priority vessel, while ensuring safety and not affecting overall operational efficiency.

[0063] A visual scheduling and monitoring interface is constructed to present the spatiotemporal constraint model, conflict detection results, and optimized scheduling solutions in a three-dimensional dynamic view. Different colors and shapes are used to distinguish different ships, berths, and waterways on the visual interface, and dynamic graphics are used to display the ship's position, navigation trajectory, and resource occupancy in real time. Users can perform manual intervention through the interactive operation interface, such as adjusting a ship's berthing time or reallocating berths. When the system receives a manual intervention instruction, it will update the scheduling instruction set in real time and promptly push the new instruction to the port scheduling terminal and ship control system, enabling flexible and efficient management of port ship scheduling and improving the overall efficiency and safety of port operations.

[0064] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0065] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A ship dispatching and embarkation method based on time-space constraint scheduling simulation and conflict optimization, characterized in that: include: Obtaining ship dispatch task information and port spatiotemporal resource data, and classifying the task information according to preset dimensions to generate structured dispatch task data; the preset dimensions include ship type, cargo attributes, berthing window, and navigation path; Construct a multi-dimensional spatiotemporal constraint model and define a spatiotemporal coordinate system based on the classification dimension of the scheduling task data. The spatiotemporal coordinate system includes a time axis, a space axis, and a resource axis. Determining initial spatiotemporal constraints based on the ship type and cargo attributes, including berthing time thresholds, channel occupancy conflict rules, and resource allocation priorities; Perform scheduling simulation on ship dispatching tasks based on the spatiotemporal constraint model, simulate the dynamic behavior of ships in the port through the discrete event simulation engine, and generate preliminary dispatching plans; Conduct conflict detection on the preliminary scheduling plan, identify ship task nodes with overlapping time and space resources or conflicting rules, and mark the conflict type and conflict intensity; Iteratively optimize conflicting nodes using a multi-objective optimization algorithm, adjust the spatiotemporal parameters of ship missions or resource allocation strategies, and generate conflict-free optimized scheduling solutions. Synchronize the optimized scheduling plan with the port's real-time operation data to dynamically modify the parameters of the spatiotemporal constraint model; Generate a final ship scheduling instruction set based on the revised spatiotemporal constraint model, the instruction set including ship berthing timing, path planning, and resource allocation information; Pushing the ship dispatch instruction set to the port dispatch terminal and the ship control system to trigger the automated dispatch execution process; The constructing of the multidimensional spatiotemporal constraint model comprises: Define gridded zones of spatial axes based on the geographical layout of the port, with each zone associated with channel capacity, number of berths, and distribution of loading and unloading equipment; Divide the time axis into continuous or discrete time slots, and associate each time slot with a priority weight for ships entering or leaving the port; The dynamic occupation status of port resources is defined based on the resource axis, including energy supply, loading and unloading equipment availability, and human resource allocation.

2. The ship dispatching and embarking method based on time-space constraint scheduling simulation and conflict optimization according to claim 1, characterized in that: The multi-objective optimization algorithm adopts an improved genetic algorithm, including: The spatiotemporal parameters of conflicting nodes are encoded as chromosomes, and a fitness function is defined to evaluate the conflict intensity, resource utilization, and task delay cost. A population of new scheduling solutions is generated through crossover and mutation operations, and the Pareto front screening strategy is used to retain the non-inferior solution set.

3. The ship dispatching and embarking method based on time-space constraint scheduling simulation and conflict optimization according to claim 1, characterized in that: The scheduling simulation includes: Introducing a real-time data stream processing module to dynamically receive port weather data, ship positioning information, and equipment failure alarms; The input parameters of the simulation model are adjusted according to real-time data to generate a robust scheduling solution.

4. The ship dispatching and embarking method based on time-space constraint scheduling simulation and conflict optimization according to claim 1, characterized in that: The conflict detection includes: Establish a time-space resource occupancy matrix and detect the overlapping areas of ship tasks in time slots and space partitions through matrix operations; Based on the rule engine matching the preset conflict judgment logic, it identifies conflict events that violate channel priority or safety distance.

5. The ship dispatching and embarking method based on time-space constraint scheduling simulation and conflict optimization according to claim 1, characterized in that: The path planning uses a deep reinforcement learning algorithm, including: The port channel topology is modeled as a graph network, and the ship movement actions are defined as state transitions; The Q-learning strategy is used to train the agent to select the optimal path to minimize navigation time and avoid conflicts.

6. The ship dispatching and embarking method based on time-space constraint scheduling simulation and conflict optimization according to claim 1, characterized in that: The definition of the resource axis includes: Integrate port environmental factors into resource allocation strategies, including tidal cycles, wind speed limits, and visibility conditions; Dynamically adjust the start and stop sequence of loading and unloading equipment and the speed thresholds of ships entering and leaving the port based on environmental factors.

7. The ship dispatching and embarking method based on time-space constraint scheduling simulation and conflict optimization according to claim 1, characterized in that: The method further comprises: Build a ship mission pattern library based on historical scheduling data to extract high-frequency mission combinations and typical conflict scenarios; The pattern matching algorithm is used to predict potential conflicts of newly added scheduling tasks and pre-generate optimization strategies.

8. The ship dispatching and embarking method based on time-space constraint scheduling simulation and conflict optimization according to claim 1, characterized in that: The resource allocation strategy includes: Use dynamic priority queues to manage ship tasks, adjusting queue order based on cargo urgency and ship fuel remaining; Reserve buffers for critical resources to ensure that high-priority tasks can preempt resources when conflicts arise.

9. The ship dispatching and embarking method based on time-space constraint scheduling simulation and conflict optimization according to claim 1, characterized in that: The method further comprises: Build a visual scheduling monitoring interface to map the spatiotemporal constraint model, conflict detection results, and optimized scheduling plan into a three-dimensional dynamic view; Receive manual intervention instructions through the interactive operation interface and update the scheduling instruction set in real time.

Images