Roll-on roll-off ship stowage optimization method, device and electronic equipment

Abstract

Embodiments of the present invention provide a method, apparatus, and electronic device for optimizing the loading of roll-on / roll-off (Ro-Ro) vessels. The Ro-Ro loading optimization method includes: acquiring basic inherent data of the target vessel and vehicle data of multiple transport vehicles to be loaded into the cargo hold of the target vessel; determining battery load distribution data corresponding to each deck area within the cargo hold; determining loading constraint characteristics based on the battery load distribution data, a first loading zone, a second loading zone, and a preset isolation distance; determining multiple initial loading schemes corresponding to the target vessel based on the loading constraint characteristics and the basic inherent data; acquiring a multi-objective optimization function; and iteratively optimizing the multiple initial loading schemes based on the multi-objective optimization function to determine at least one loading layout result. The solution of the present invention can reduce the risk of large-scale fires in Ro-Ro vessel mixed-transportation scenarios of new energy vehicles, ensuring the safety of transportation operations.

Description

Technical Field

[0001] This invention relates to the field of maritime management technology, and more specifically, to a method, apparatus, and electronic equipment for optimizing the stowage of roll-on / roll-off (Ro-Ro) vessels. Background Technology

[0002] In high-frequency roll-on / roll-off (Ro-Ro) shipping operations, including ocean-going, near-sea, and river-sea combined transport, the explosive growth in demand for new energy vehicles by sea has led to the common practice of transporting various types of new energy vehicles, such as pure electric and plug-in hybrid vehicles. However, the power batteries in new energy vehicles have an inherent risk of thermal runaway in the enclosed environment of a ship's cabin. Once a single battery explodes, it can easily trigger a chain reaction, leading to a catastrophic fire throughout the ship.

[0003] When a new energy vehicle in a ship's hold experiences thermal runaway, due to the lack of effective physical isolation and density control, the flames will quickly break through the single vehicle position and spread rapidly to adjacent vehicles along the deck. In addition, traditional loading does not take into account the operating radius and passageway reservation for fire rescue, making firefighting more difficult exponentially. The transportation safety of new energy vehicles in mixed-model transportation scenarios is insufficient, and once thermal runaway occurs, the risk of fire spreading is extremely high. Summary of the Invention

[0004] The purpose of this invention is to provide a method, apparatus, and electronic device for optimizing the loading of roll-on / roll-off (Ro-Ro) vessels, which can solve the problem of insufficient transportation safety in mixed-model transportation scenarios of new energy vehicles.

[0005] In view of this, an embodiment of the first aspect of the present invention provides a method for optimizing the loading of roll-on / roll-off ships.

[0006] A second aspect of the present invention provides a roll-on / roll-off (Ro-Ro) ship loading optimization device.

[0007] An embodiment of the third aspect of the present invention provides an electronic device.

[0008] To achieve the above objectives, an embodiment of the first aspect of the present invention provides a method for optimizing the loading of roll-on / roll-off (Ro-Ro) vessels, executed by a loading optimization device, comprising: acquiring basic inherent data of the target vessel and vehicle data of multiple transport vehicles to be loaded into the cargo hold of the target vessel, wherein the transport vehicles include a first type of vehicle and a second type of vehicle, the first type of vehicle being a new energy vehicle equipped with a battery pack and the second type of vehicle being a fuel vehicle; the basic inherent data including deck load strength data of each deck and fire compartment planning data; and the vehicle data including battery pack parameters of the first type of vehicle; determining battery load distribution data corresponding to each deck area in the cargo hold based on the battery pack parameters and deck load strength data; and determining a first loading compartment corresponding to the first type of vehicle, a second loading compartment corresponding to the second type of vehicle, and a preset separation between the first type of vehicle and the second type of vehicle based on the fire compartment planning data. The loading constraint features are determined based on battery load distribution data, the first loading zone, the second loading zone, and the preset isolation distance. These features include mandatory constraints that must be met, including at least battery load symmetry constraints, deck load strength constraints, and fire compartment isolation constraints. Multiple initial loading schemes corresponding to the target vessel are determined based on the loading constraint features and basic inherent data. Each initial loading scheme includes loading position data for multiple transport vehicles. A multi-objective optimization function is obtained, with cargo capacity utilization, navigation stability, and loading / unloading efficiency as sub-objectives, and weighting coefficients assigned to each sub-objective. Under the premise of meeting mandatory constraints, the multiple initial loading schemes are iteratively optimized according to the multi-objective optimization function to determine at least one loading layout result. This result indicates the loading layout of multiple transport vehicles within the cargo hold of the target vessel.

[0009] The roll-on / roll-off (Ro-Ro) ship loading optimization method provided by this invention is executed by loading optimization equipment. First, it acquires basic inherent data of the target ship, such as deck load intensity data and fire compartment planning data, as well as data on various types of transport vehicles, including parameters of new energy vehicle battery packs and parameters of fuel vehicles. Based on the battery pack parameters and deck load intensity data, it calculates the battery load distribution data for each deck area of ​​the cargo hold. Then, according to the fire compartment planning data, it delineates the first loading compartment corresponding to new energy vehicles and the second loading compartment corresponding to fuel vehicles, clarifying the preset isolation distance between the two types of vehicles, thereby constructing loading constraint features. Under the premise of meeting the mandatory constraints, it generates multiple initial loading schemes containing vehicle loading position data. Combining a multi-objective optimization function, iteratively optimizes the initial loading schemes, finally outputting the loading layout results used to indicate the parking layout of all transport vehicles in the cargo hold.

[0010] Understandably, the roll-on / roll-off (Ro-Ro) ship loading optimization method takes battery load distribution control and fire compartment physical isolation as mandatory pre-constraints for loading. By limiting the battery energy density per unit cargo hold space, it physically blocks the spread path of thermal runaway from a single vehicle to surrounding vehicles, reduces the risk of large-scale fires and the difficulty of fire fighting and rescue in the mixed transportation scenario of new energy vehicles on Ro-Ro ships, and ensures the safety and continuity of transportation operations.

[0011] In some technical solutions, optionally, load constraint characteristics are determined based on battery load distribution data, a first loading zone, a second loading zone, and a preset isolation distance. These include: determining an upper limit for deck load based on battery load distribution data, where the upper limit does not exceed a preset proportion of the deck load strength data for the corresponding deck; determining an upper limit for the port and starboard weight difference in the concentrated battery load area based on battery load distribution data; and determining a center of gravity height constraint based on the battery pack installation location and the vehicle's center of gravity height in the vehicle data, where the center of gravity height constraint includes the proportion of transport vehicles whose single-vehicle center of gravity height exceeds a preset height threshold on their respective decks. Limits and the upper limit of the overall center of gravity height of the target vessel; determine the minimum interval between the first type of vehicles and the second type of vehicles based on the preset isolation distance, and limit the first type of vehicles to be parked in an independent fire compartment; determine the stability reserve constraint based on the stability limit threshold of the target vessel, including the lower limit of the initial stability height and the lower limit of the complete stability criterion number; determine the area range in which each type of transport vehicle can be parked according to the first loading zone and the second loading zone; determine the loading constraint characteristics based on the upper limit of deck load, the upper limit of the weight difference between port and starboard sides, the center of gravity height constraint, the minimum interval, and the area range.

[0012] In this scheme, the inherent legal attributes of the anchored target vessel and the unique risk characteristics of the first type of vehicle are defined, clarifying the quantitative sources and applicable boundaries of various constraints, and ultimately combining them into loading constraint characteristics.

[0013] In some technical solutions, optionally, multiple initial loading schemes corresponding to the target vessel are determined based on loading constraint characteristics and basic inherent data, including: determining available parking spaces in the cargo hold based on basic inherent data; allocating candidate parking spaces that conform to the corresponding loading zone for each transport vehicle based on the area range corresponding to each transport vehicle in the loading constraint characteristics; randomly generating multiple sets of correspondences between transport vehicles and parking spaces within the range of candidate parking spaces to obtain multiple initial loading schemes; and performing initial screening on the multiple initial loading schemes based on mandatory constraints to eliminate initial loading schemes that do not meet the mandatory constraints.

[0014] In this scheme, based on the basic inherent data of the target ship, parking spaces occupied by fixed facilities or with insufficient load margin are eliminated to determine the available parking spaces in the cargo hold. Then, according to the parking area range of vehicle type defined in the loading constraint features, multiple sets of vehicle-parking space correspondences are randomly generated within the candidate parking spaces to obtain multiple sets of initial loading schemes.

[0015] In some technical solutions, optionally, after obtaining the multi-objective optimization function, the roll-on / roll-off (Ro-Ro) ship loading optimization method further includes: obtaining the target vessel's route sea state data and port call operation duration data; determining the forecast wind and wave level based on the route sea state data; determining the operational urgency based on the port call operation duration data; when the forecast wind and wave level is not less than a preset sea state level threshold, increasing the weight coefficient corresponding to navigation stability and correspondingly decreasing the weight coefficients corresponding to the other sub-objectives; when the port call operation duration data is not greater than a preset duration threshold, increasing the weight coefficient corresponding to loading and unloading efficiency and correspondingly decreasing the weight coefficients corresponding to the other sub-objectives; updating the multi-objective optimization function based on the adjusted weight coefficients; wherein, the sum of the weight coefficients corresponding to each sub-objective is a constant.

[0016] In this scheme, after the loading optimization equipment obtains the multi-objective optimization function, it collects the route and sea state data and port call operation duration data of the target vessel, derives the forecast wind and wave level and the urgency of the operation respectively, adjusts the target weights, and updates the multi-objective optimization function according to the adjusted weights.

[0017] In some technical solutions, optionally, the loading constraint features also include priority constraints. Under the premise of satisfying the mandatory constraints, multiple initial loading schemes are iteratively optimized according to a multi-objective optimization function to determine at least one loading layout result. This includes: using the function value of the multi-objective optimization function as the fitness, iteratively optimizing multiple initial loading schemes according to an improved adaptive genetic algorithm; in each iteration, loading schemes that do not satisfy the mandatory constraints are eliminated, and the fitness of the retained loading schemes is corrected according to the priority constraints; adaptively adjusting the crossover probability according to the fitness of each loading scheme to make the fitness higher than the preset fitness. The crossover probability corresponding to the loading scheme with a fitness threshold is less than the crossover probability corresponding to the loading scheme with a fitness lower than the preset fitness threshold; the mutation probability is adaptively adjusted according to the iteration process, so that the mutation probability in the early stage of the iteration is greater than the mutation probability in the later stage of the iteration; a dynamic correction operator is introduced into the fitness calculation, which includes at least one of the fuel consumption center of gravity shift correction factor and the mid-port loading and unloading load change correction factor; when the preset termination condition is met, the loading scheme with the highest fitness is output as the loading layout result. The preset termination condition includes no improvement in fitness for a consecutive preset number of generations or the number of iterations reaching the preset number of times limit.

[0018] In this scheme, based on the priority constraints included in the load allocation constraints, an improved adaptive genetic algorithm is used to iteratively optimize the load allocation by using the output value of the multi-objective optimization function as the fitness. In each iteration, schemes that do not meet the mandatory constraints are first eliminated, and then the fitness of the retained schemes is adjusted according to the priority constraints. The optimal load allocation scheme is output after the termination condition of no improvement in fitness for several consecutive generations or the maximum number of iterations is reached.

[0019] In some technical solutions, optionally, the roll-on / roll-off (Ro-Ro) vessel loading optimization method further includes: acquiring real-time navigation data at time intervals not exceeding a preset period during the target vessel's navigation, the real-time navigation data including at least one of real-time draft, real-time center of gravity, fuel consumption, ballast water state, and real-time sea state; determining dynamic optimization triggering features based on the real-time navigation data and loading layout results, the dynamic optimization triggering features being defined based on battery load distribution data and loading constraint features; when the dynamic optimization triggering features meet preset triggering conditions, performing local iterative correction on the loading layout results based on the real-time navigation data and loading constraint features to obtain iterative loading layout results; updating the loading layout results based on the iterative loading layout results; wherein, the preset triggering conditions include at least one of the following: the target vessel's center of gravity offset exceeds a preset offset threshold, the real-time initial stability height is lower than a preset stability threshold, and the wind and wave level corresponding to the real-time sea state is higher than the predicted wind and wave level by no less than a preset difference.

[0020] In this scheme, the load optimization equipment collects real-time navigation data at a preset cycle. Based on the battery load distribution data and load constraint characteristics in the output load layout results, dynamic optimization trigger characteristics are determined. When the trigger characteristics meet the conditions such as the center of gravity offset exceeding the preset threshold, the initial stability height being lower than the stability threshold, or the sea state rising suddenly exceeding the preset level difference, local iterative correction is carried out. After correction, the load results are updated and synchronized to the ship's bridge.

[0021] In some technical solutions, the roll-on / roll-off (Ro-Ro) ship loading optimization method may optionally include: responding to berthing instructions at intermediate ports, obtaining the current ship status characteristics, the corresponding unloading list and loading list at the intermediate port; determining the stability prediction characteristics after unloading based on the current ship status characteristics and the unloading list, combined with preset isolation distances; iteratively solving the vehicle parking positions in the remaining cargo hold space based on the stability prediction characteristics, the loading list and loading constraint characteristics, to determine the loading layout results at the intermediate port; generating loading and unloading operation instructions corresponding to the intermediate port based on the loading layout results at the intermediate port, and synchronizing the loading and unloading operation instructions to the ship's bridge terminal, deck work team terminal and port dispatch center terminal through the ship-shore communication link.

[0022] In this scheme, the loading optimization equipment responds to the berthing command of the intermediate port, retrieves the current status characteristics of the target vessel within a preset time window before the target vessel arrives at the port, deducts the corresponding loads in combination with the unloading list, and strictly predicts the stability parameters after unloading according to the loading constraint characteristics. Under the forced constraint framework, iteratively solves the optimal loading layout of the intermediate port, generates loading and unloading operation commands, and synchronizes them to the ship's bridge terminal, deck operation team terminal and port dispatch center terminal through the ship-shore communication link.

[0023] In some technical solutions, optionally, after iteratively optimizing multiple initial loading schemes according to a multi-objective optimization function to determine at least one loading layout result, the ro-ro ship loading optimization method further includes: determining three-dimensional visualization data corresponding to the loading layout result through a digital twin model, the three-dimensional visualization data including vehicle parking positions, deck load distribution cloud map, ship center of gravity coordinates, fire compartment division, and loading / unloading path planning; performing static compliance verification on the loading layout result based on the three-dimensional visualization data to obtain the static compliance verification result; under the condition that the static compliance verification result indicates compliance, performing full-condition dynamic simulation verification on the loading layout result, the full-condition dynamic simulation verification includes at least extreme sea state verification, the verification parameters of the extreme sea state verification include a roll angle not less than a preset roll angle threshold and a wind and wave level not less than a preset wind and wave level threshold; when the loading layout result passes the full-condition dynamic simulation verification, determining the stability calculation result and fire compartment verification data based on the loading layout result.

[0024] In this scheme, after the load layout results are obtained through iterative optimization, the load optimization equipment imports them into the digital twin model, performs a digital twin 3D visualization of the optimal scheme, and sequentially passes static compliance verification and full-condition dynamic simulation verification. After all verifications are passed, the stability calculation results and fire compartment verification data are output.

[0025] An embodiment of the second aspect of the present invention provides a roll-on / roll-off (Ro-Ro) ship loading optimization device, comprising: a data acquisition module, configured to acquire basic inherent data of the target ship and vehicle data of multiple transport vehicles to be loaded into the cargo hold of the target ship, the transport vehicles including a first type of vehicle and a second type of vehicle, the first type of vehicle being a new energy vehicle equipped with a battery pack, and the second type of vehicle being a fuel vehicle; the basic inherent data including deck load strength data of each deck and fire compartment planning data, and the vehicle data including battery pack parameters of the first type of vehicle; a battery load module, configured to determine battery load distribution data corresponding to each deck area in the cargo hold based on the battery pack parameters and deck load strength data; a compartment planning module, configured to determine a first loading compartment corresponding to the first type of vehicle, a second loading compartment corresponding to the second type of vehicle, and a preset isolation distance between the first type of vehicle and the second type of vehicle based on the fire compartment planning data; and a loading constraint module, configured to... Based on battery load distribution data, the first loading zone, the second loading zone, and the preset isolation distance, the loading constraint characteristics are determined. These characteristics include mandatory constraints that must be met, at least including battery load symmetry constraints, deck load strength constraints, and fire compartment isolation constraints. A scheme planning module is used to determine multiple initial loading schemes corresponding to the target vessel based on the loading constraint characteristics and basic inherent data. Each initial loading scheme includes loading position data for multiple transport vehicles. A function construction module is used to obtain a multi-objective optimization function, which uses cargo capacity utilization, navigation stability, and loading / unloading efficiency as sub-objectives, and assigns weight coefficients to each sub-objective. A loading optimization module is used to iteratively optimize multiple initial loading schemes according to the multi-objective optimization function, under the premise of meeting mandatory constraints, to determine at least one loading layout result. This result indicates the loading layout of multiple transport vehicles within the cargo hold of the target vessel.

[0026] An embodiment of the third aspect of this application provides an electronic device, including a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the roll-on / roll-off ship loading optimization method of the first aspect.

[0027] Additional aspects and advantages of the technical solutions of the present invention will become apparent in the following description or may be learned by practice of the invention. Attached Figure Description

[0028] Figure 1 One of the flowcharts of the roll-on / roll-off ship loading optimization method according to this application is shown;

[0029] Figure 2 A second schematic flowchart of the roll-on / roll-off ship loading optimization method according to this application is shown;

[0030] Figure 3The third flowchart of the roll-on / roll-off ship loading optimization method according to this application is shown;

[0031] Figure 4 The fourth flowchart of the roll-on / roll-off ship loading optimization method according to this application is shown;

[0032] Figure 5 Fifth of the flowcharts for the roll-on / roll-off ship loading optimization method according to this application is shown;

[0033] Figure 6 A flowchart of the roll-on / roll-off ship loading optimization method according to this application is shown in diagram six.

[0034] Figure 7 The seventh flowchart of the roll-on / roll-off ship loading optimization method according to this application is shown;

[0035] Figure 8 The eighth flowchart of the roll-on / roll-off ship loading optimization method according to this application is shown;

[0036] Figure 9 A schematic block diagram of the structure of the roll-on / roll-off ship loading optimization device according to this application is shown;

[0037] Figure 10 A schematic block diagram of the structure of an electronic device according to this application is shown.

[0038] Among them, 900: Ro-Ro ship loading optimization device; 902: data acquisition module; 904: battery load module; 906: zoning planning module; 908: loading constraint module; 910: scheme planning module; 912: function construction module; 914: loading optimization module;

[0039] 1000: Electronic device; 1109: Memory; 1110: Processor. Detailed Implementation

[0040] To better understand the above-described objectives, features, and advantages of the embodiments of the present invention, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0041] Many specific details are set forth in the following description in order to provide a full understanding of this application. However, embodiments of the invention may also be implemented in other ways different from those described herein. Therefore, the scope of protection of this application is not limited to the specific embodiments disclosed below.

[0042] With the continued development of the global new energy vehicle industry, China's new energy vehicle exports have maintained high growth. As the core carrier for ocean-going vehicle transportation, car carriers are experiencing continuous market demand expansion. The capacity of a single vessel has increased from 3,000 vehicles to 7,000-10,000 vehicles, and the types of vehicles transported have gradually evolved from traditional fuel vehicles to a structure primarily consisting of pure electric, plug-in hybrid, and fuel-powered vehicles. The safety, compliance, and operational efficiency of cargo loading schemes have an increasingly significant impact on operating costs and navigation safety.

[0043] The loading technology is mainly divided into three stages: the first is pure manual experience loading, which relies on the chief mate to calculate manually. It is inefficient and has large errors, and is only seen in a small number of small inland waterway roll-on / roll-off ships; the second is basic loading software-assisted loading, which generates static schemes based on ship statics formulas and is the current industry mainstream; the third is preliminary digital twin loading, which realizes scheme visualization and basic stability simulation and is in the pilot stage of leading shipping companies.

[0044] Among them, the relevant loading technologies are not adapted to the load characteristics of new energy vehicle batteries: the weight of the battery pack of new energy vehicles accounts for 25% to 35% of the total vehicle weight, the center of gravity is concentrated and the load density is high. The existing loading technologies do not set specific load constraints for the battery pack, which can easily lead to problems such as excessive local load on the deck, imbalance of the port and starboard sides of the ship and excessive center of gravity. Under the condition of rolling in the open ocean, there is a risk of insufficient stability.

[0045] Treating vehicles merely as ordinary loads without incorporating fire compartmentation, explosion-proof spacing, and hazardous vehicle management into the constraints results in compliance deficiencies. In the event of thermal runaway, the risk of fire spreading is extremely high.

[0046] Furthermore, once a fixed plan is generated before loading, it cannot respond to dynamic changes during the voyage, such as fuel consumption, ballast water adjustment, loading and unloading at intermediate ports, and sudden changes in sea conditions. This is especially true for river-sea intermodal transport routes with many intermediate ports and frequent loading and unloading, where the static plan is seriously inadequate in terms of practicality.

[0047] The following detailed description of the roll-on / roll-off ship loading optimization method, apparatus, and electronic equipment provided in this application, through specific embodiments and application scenarios, is provided in detail.

[0048] This embodiment provides a method for optimizing the loading of roll-on / roll-off (Ro-Ro) vessels, which is executed by loading optimization equipment, such as... Figure 1 As shown, the optimized loading methods for roll-on / roll-off (Ro-Ro) ships include:

[0049] Step S100: Obtain the basic inherent data of the target vessel and the vehicle data of multiple transport vehicles to be loaded into the cargo hold of the target vessel;

[0050] Step S102: Determine the battery load distribution data for each deck area in the cargo hold based on the battery pack parameters and deck load strength data;

[0051] Step S104: Determine the first loading zone corresponding to the first type of vehicle, the second loading zone corresponding to the second type of vehicle, and the preset isolation distance between the first type of vehicle and the second type of vehicle based on the fire compartment planning data;

[0052] Step S106: Determine the load constraint characteristics based on the battery load distribution data, the first loading zone, the second loading zone, and the preset isolation distance;

[0053] Step S108: Determine multiple initial loading schemes corresponding to the target vessel based on the loading constraint characteristics and basic inherent data;

[0054] Step S110: Obtain the multi-objective optimization function, which takes cabin capacity utilization, navigation stability and loading and unloading efficiency as sub-objectives, and assigns weight coefficients to each sub-objective.

[0055] Step S112: Under the premise of satisfying the mandatory constraints, iteratively optimize multiple initial loading schemes according to the multi-objective optimization function to determine at least one loading layout result;

[0056] The transport vehicles include Type 1 vehicles and Type 2 vehicles. Type 1 vehicles are new energy vehicles equipped with battery packs, and Type 2 vehicles are fuel vehicles. The basic inherent data includes deck load strength data and fire compartment planning data for each deck. The vehicle data includes battery pack parameters for Type 1 vehicles. The loading layout results are used to indicate the loading layout of multiple transport vehicles in the cargo hold of the target ship. The loading constraint features include mandatory constraints that must be met. The mandatory constraints include at least battery load symmetry constraints, deck load strength constraints, and fire compartment isolation constraints. Each initial loading scheme includes loading position data for each of the multiple transport vehicles.

[0057] The roll-on / roll-off (Ro-Ro) ship loading optimization method provided by this invention is executed by loading optimization equipment. First, it acquires the target ship's basic inherent data, as well as data on various types of transport vehicles, including parameters of new energy vehicle battery packs and parameters of fuel-powered vehicles. Based on the battery pack parameters and deck load intensity data, it calculates the battery load distribution data for each deck area of ​​the cargo hold. Then, based on fire protection zoning planning data, it delineates the first and second loading zones, clarifying the preset isolation distance between the two types of vehicles, thereby constructing loading constraint features. Under the premise of meeting the mandatory constraints, it generates multiple initial loading schemes containing vehicle loading position data. Combining a multi-objective optimization function, iteratively optimizes the initial loading schemes, finally outputting the loading layout results used to indicate the parking arrangement of all transport vehicles in the cargo hold.

[0058] Understandably, the roll-on / roll-off (Ro-Ro) ship loading optimization method takes battery load distribution control and fire compartment physical isolation as mandatory pre-constraints for loading. By limiting the battery energy density per unit cargo hold space, it physically blocks the spread path of thermal runaway from a single vehicle to surrounding vehicles, reduces the risk of large-scale fires and the difficulty of fire fighting and rescue in the mixed transportation scenario of new energy vehicles on Ro-Ro ships, and ensures the safety and continuity of transportation operations.

[0059] The loading optimization equipment for implementing the roll-on / roll-off ship loading optimization method can be implemented by any one of the following: a ship shore-based loading center server, a shipboard loading workstation, an edge computing unit, or a cloud service node, and includes at least a processor, a memory, a data input interface, and a data output interface.

[0060] The data input interface is used to connect to the ship design drawing database, classification society verification report database, vehicle factory inspection report database, port operation system and maritime reporting system to receive the basic inherent data of the target ship and the vehicle data of the transport vehicles.

[0061] The memory stores deck load strength data, fire compartment planning data, battery pack parameters, multi-objective optimization functions, and a population of loading schemes generated during the iteration process. The processor executes each loading step.

[0062] The data output interface is used to send the loading layout results to the ship's bridge terminal, deck operation team terminal and port dispatch center terminal. It supports at least one of 5G link, satellite communication link or ship-shore wired link. It can be deployed independently in the port loading center or integrated into the ship's existing loading system. It can also adopt a cloud-edge collaborative architecture of shore-based training and ship-based inference.

[0063] The target vessel is a roll-on / roll-off ship carrying out mixed transportation of new energy vehicles. Its basic inherent data is a set of legal attributes verified by the classification society upon completion of the ship's construction and does not change with each voyage, including deck load strength data and fire protection zone planning data for each deck.

[0064] The deck load strength data are the maximum uniformly distributed load value and the maximum concentrated load value specified in the design stage of each deck structure. They are recorded in the ship's hydrostatic calculation report and deck structure drawings and are the legal benchmark for deck load-bearing capacity.

[0065] Fire compartment planning data refers to the physical boundaries, explosion-proof ratings, maximum number of vehicles, and fire resistance ratings of the partitions between adjacent compartments as defined during the ship's fire protection design. It serves as the legal basis for fire isolation.

[0066] Transport vehicles are the collection of all vehicles to be loaded into the cargo hold of the target vessel, and are divided into Class I vehicles and Class II vehicles according to their power type.

[0067] Type 1 vehicles are new energy vehicles equipped with rechargeable power batteries, including pure electric vehicles, plug-in hybrid electric vehicles, and range-extended electric vehicles. Their common feature is that a high-voltage battery pack is arranged under the vehicle or inside the vehicle. The risk of thermal runaway and the energy density are much higher than those of Type 2 vehicles. The battery pack parameters of Type 1 vehicles included in the vehicle data are battery pack weight, battery pack dimensions, coordinates of the battery pack installation position in the vehicle body coordinate system, battery pack pressure per unit area, and battery pack energy density. The above parameters are obtained from the vehicle factory inspection report, shipping declaration form, or port pre-entry system.

[0068] The second type of vehicle is a gasoline-powered vehicle powered by a gasoline engine. The fire risk mainly comes from fuel leakage from the fuel tank. Its energy density is lower than that of the first type of vehicle. Vehicle data also includes the vehicle's weight, overall dimensions, wheelbase, center of gravity height, location of fastening points, and fuel type.

[0069] After completing the above data collection, the load optimization equipment determines the battery load distribution data corresponding to each deck area in the cargo hold based on the battery pack parameters and deck load intensity data.

[0070] The determination process involves projecting the battery pack of each Type I vehicle onto the corresponding deck's coordinate system according to its installation position on the vehicle body. Using a 0.5-meter by 0.5-meter grid as the smallest unit of measurement, the total weight of the battery packs within each grid is calculated by summing them up. This results in a battery load distribution heat map covering the entire deck area. The values ​​in the battery load distribution heat map are the battery load distribution data, which can accurately reflect the degree of battery energy accumulation in any area of ​​the cargo hold.

[0071] Traditional load planning schemes only focus on the total load capacity and number of vehicles of a ship, and cannot identify the spatial distribution characteristics of batteries. If batteries are concentrated on one side of the deck or in a local area, once a vehicle experiences thermal runaway, the high-density battery energy will quickly fuel combustion, leading to a chain reaction of fires. However, the quantitative processing of battery load distribution data in this invention can identify high-energy accumulation areas from a physical perspective, providing a precise basis for subsequent constraint settings and fundamentally avoiding the risk of fire spread caused by local energy overload.

[0072] Subsequently, the load optimization equipment determines the first loading zone corresponding to the first type of vehicle, the second loading zone corresponding to the second type of vehicle, and the preset isolation distance between the first type of vehicle and the second type of vehicle based on the fire protection zone planning data.

[0073] The first loading zone is a fire protection unit designated in the fire protection zone planning data, equipped with independent fire sprinkler, explosion-proof ventilation and temperature monitoring facilities, and is specifically used for parking Type I vehicles.

[0074] The second loading zone is a non-explosion-proof fireproof unit defined in the fire protection zone planning data. It is used to park the second type of vehicles and can be arranged on the same floor or on different floors as the first loading zone.

[0075] The preset isolation distance is the minimum net distance between the first loading zone and the second loading zone, or between the first type of vehicle and the second type of vehicle within the same fire compartment, which is determined by the fire resistance rating of the fire-resistant partition between adjacent zones.

[0076] For example, when the fire separation between adjacent zones is Class A60, the preset separation distance is three meters, which complies with the mandatory requirements of the International Maritime Organization's Fire Safety Systems Code.

[0077] Traditional loading and unloading often allows two types of vehicles to be parked together or reduces the isolation distance in order to improve loading and unloading efficiency, leaving hidden dangers for the spread of fire. The setup of this step can completely physically isolate the two types of risk sources. Even if the first type of vehicle experiences thermal runaway, the fireproof partition and the preset isolation distance can effectively prevent the spread of flames, high temperature and toxic fumes to the area of ​​the second type of vehicle, buy time for fire rescue, and meet the compliance requirements of maritime supervision.

[0078] Based on the above distribution data and boundary rules, the load optimization equipment determines the load constraint characteristics according to the battery load distribution data, the first loading zone, the second loading zone, and the preset isolation distance.

[0079] Loading constraint features are the complete set of rules for judging the legality of loading schemes. They are divided into two categories: mandatory constraints and priority constraints. Mandatory constraints are constraints that the loading scheme must meet, otherwise it will be directly rejected. They include at least battery load symmetry constraints, deck load strength constraints, and fire compartment isolation constraints.

[0080] The battery load symmetry constraint is that the total weight deviation of the first type of vehicles on both the port and starboard sides of the cargo hold shall not exceed three percent, so as to avoid the ship listing due to uneven battery weight and affecting navigation stability.

[0081] The deck load strength constraint is that the battery load distribution data in each deck area shall not exceed 85% of the corresponding deck load strength data, leaving sufficient margin for additional loads caused by fuel consumption during navigation, sea conditions, and loading and unloading at intermediate ports. The fire compartment isolation constraint is that all Type I vehicles shall be parked in the first loading compartment, and the distance between the parking positions of Type I vehicles and Type II vehicles shall not be less than the preset isolation distance. Vehicles of different explosion-proof levels are strictly prohibited from being parked together in the same compartment.

[0082] In addition to the three points mentioned above, mandatory constraints also include stability reserve constraints. Stability reserve constraints are determined by the ship stability limit thresholds in the basic inherent data, including at least the lower limit of the initial stability height and the lower limit of the complete stability criterion number. Both values ​​are obtained from the ship stability calculation report and are used to ensure the ship's ability not to capsize under extreme sea conditions.

[0083] Traditional load planning constraint verification is mostly done manually after the fact, which is prone to oversights. This step embeds constraints into the entire load planning process from the very beginning. Schemes that do not meet the mandatory constraints will be directly eliminated, thus eliminating the possibility of compromising safety standards to improve cabin capacity utilization and ensuring that all schemes meet safety and compliance requirements from the very beginning.

[0084] The loading optimization equipment then determines multiple initial loading schemes corresponding to the target vessel based on the loading constraint characteristics and basic inherent data. Each initial loading scheme includes loading position data for multiple transport vehicles.

[0085] The initial loading plan is generated by first identifying all available parking spaces in the cargo hold based on existing data. The selection criteria are that the remaining load margin on the deck where the parking space is located can meet the tare weight requirements of the corresponding vehicle, and the parking space is located within the loading zone of the corresponding vehicle type. Then, each transport vehicle is randomly assigned a parking position within the range of candidate parking spaces that meet the zone requirements. Multiple sets of vehicle-parking-space correspondences are randomly generated, and each set of correspondences constitutes an initial loading plan. The loading position data includes the deck layer number, row number, column number, and lashing method corresponding to each transport vehicle.

[0086] After all initial load configuration schemes are generated, a preliminary screening is performed using mandatory constraints as the criterion. Schemes that do not meet any constraint are eliminated, and only compliant schemes are retained to enter the subsequent optimization process. By setting up multiple schemes, sufficient search space is provided for subsequent multi-objective optimization, avoiding the optimization process from getting stuck in local optima.

[0087] The load optimization equipment then obtains a multi-objective optimization function, which takes cargo capacity utilization, navigation stability and loading / unloading efficiency as sub-objectives, and assigns weight coefficients to each sub-objective. The sum of the weight coefficients of each sub-objective is a constant value of one.

[0088] The multi-objective optimization function is a mathematical function used to evaluate the merits of each loading scheme. Its output value is a scalar evaluation value, and the higher the evaluation value, the better the overall performance of the scheme.

[0089] The cargo hold utilization rate is quantified as the ratio of the number of vehicles currently loaded to the theoretical maximum number of vehicles that can be loaded in the cargo hold, reflecting the operational benefits of the voyage.

[0090] The quantification of navigation stability is the normalized difference between the initial stability altitude of the current scheme and the minimum stability value specified in the standard. The larger the difference, the better the stability.

[0091] The loading and unloading efficiency is quantified by the reciprocal normalized value of the total path length of all vehicle loading and unloading operations under the current scheme. The shorter the path, the higher the efficiency.

[0092] The sum of the weight coefficients of each sub-objective is a fixed value of one. The weight coefficients can be dynamically adjusted according to the actual needs of the voyage. If the forecast sea state level of the route is high, the weight coefficient of navigation stability will be increased to prioritize navigation safety. If the port operation time is short, the weight coefficient of loading and unloading efficiency will be increased to prioritize operation timeliness and adapt to diverse transportation scenarios.

[0093] Under the premise of satisfying the mandatory constraints, multiple initial loading schemes are iteratively optimized according to the multi-objective optimization function to determine at least one loading layout result.

[0094] The iterative optimization process employs an improved adaptive genetic algorithm, using the output of a multi-objective optimization function as the evaluation value to iteratively optimize the retained initial loading schemes. In each iteration, schemes that do not meet the mandatory constraints are first eliminated. Then, the crossover probability is adaptively adjusted based on the evaluation value of each scheme, with schemes with higher evaluation values ​​having a lower crossover probability than those with lower evaluation values. Simultaneously, the mutation probability is adaptively adjusted according to the iteration progress, with the mutation probability in the early stages of the iteration being higher than that in later stages. Dynamic correction parameters are introduced into the calculation of the evaluation value, including a first correction parameter characterizing the center of gravity shift caused by fuel consumption and a second correction parameter characterizing load changes caused by loading and unloading at intermediate ports, allowing the optimization process to simulate the dynamic changes during actual navigation.

[0095] When the change in evaluation value over multiple iterations is less than a preset threshold, or when the number of iterations reaches a preset upper limit, the device outputs the scheme with the highest evaluation value as the loading layout result. The loading layout result is used to indicate the loading layout of multiple transport vehicles in the cargo hold of the target ship, including the specific parking location, lashing requirements, and fire compartment affiliation information of each vehicle.

[0096] The stowage layout result is the final solution output from the iterative optimization process, used to directly guide on-site loading and unloading operations. The stowage layout result includes at least the specific parking location of each transport vehicle, its fire compartment, lashing requirements, and loading / unloading sequence. It may also include a deck load distribution cloud map, the ship's center of gravity coordinates, and stability calculation results. The stowage layout result can be further input into a digital twin model for 3D visualization and full-condition dynamic simulation verification.

[0097] In some embodiments, the first loading zone, the second loading zone, and the preset isolation distance can be adaptively adjusted according to the maritime regulatory requirements corresponding to the voyage, the ship's temporary modification records, and the real-time operating status of the fire protection systems of adjacent zones. When the fire sprinkler system of a certain fire protection zone is in a fault state, the zone corresponding to the fault state is automatically removed from the first loading zone, or the preset isolation distance between the zone corresponding to the fault state and other zones is increased to cover loading scenarios with abnormal ship fire protection systems and temporary modifications, thereby improving the robustness of the solution.

[0098] In some embodiments, the iterative optimization process can optionally cover the entire stage of ship berthing and loading / unloading, access the real-time operation status data of port loading and unloading equipment and the real-time sequence data of vehicles boarding and disembarking, dynamically iteratively correct the parking positions of vehicles that have not yet been loaded, and when a fire warning is detected in the cargo hold, a temporary vehicle transfer layout plan can be quickly output with the goal of minimizing the spread of the fire as a new optimization objective, covering dynamic adjustments and emergency scenarios during the loading and unloading process, and improving the emergency response capability of the plan.

[0099] In some embodiments, optionally, loading optimization can also take into account the loading zoning margin, isolation spacing margin, and emergency rescue resource distribution of multiple target vessels departing from the same port, generate a convoy-level loading coordination scheme, optimize the efficiency of cross-ship rescue dispatch when the convoy encounters a sudden fire, cover the scenario of multi-ship joint transportation, and improve the guarantee capability of large-scale new energy vehicle maritime transport.

[0100] In some embodiments, the process of determining battery load distribution data may also include access to real-time health status data of the battery pack. When the health status of the battery pack of a certain type of vehicle is lower than a preset threshold, the corresponding battery energy density calculation weight is automatically increased, or a larger surrounding safety distance is reserved for it in the loading layout to cover the special risk scenarios of transporting old new energy vehicles and further reduce the probability of thermal runaway causing a chain reaction.

[0101] In some embodiments, optionally, such as Figure 2 As shown, step S106: Determine the load constraint characteristics based on the battery load distribution data, the first loading zone, the second loading zone, and the preset isolation distance, including:

[0102] Step S1062: Determine the upper limit of the deck load based on the battery load distribution data. The upper limit of the deck load shall not exceed the preset ratio of the deck load strength data of the corresponding deck.

[0103] Step S1064: Determine the upper limit of the weight difference between the port and starboard sides in the battery load concentration area based on the battery load distribution data;

[0104] Step S1066: Determine the center of gravity height constraint based on the battery pack installation location and the vehicle center of gravity height in the vehicle data;

[0105] Step S1068: Determine the minimum distance between the first type of vehicle and the second type of vehicle according to the preset isolation distance, and restrict the first type of vehicle to be parked in an independent fire compartment;

[0106] Step S1070: Determine the stability reserve constraint based on the stability limit threshold of the target vessel;

[0107] Step S1072: Determine the permitted parking area for each type of transport vehicle based on the first loading zone and the second loading zone;

[0108] Step S1074: Determine the loading constraint characteristics based on the upper limit of deck load, the upper limit of the weight difference between port and starboard sides, the center of gravity height constraint, the minimum interval value, and the area range;

[0109] Among them, the center of gravity height constraint includes the upper limit of the proportion of transport vehicles whose center of gravity height exceeds the preset height threshold on their respective decks, as well as the upper limit of the overall center of gravity height of the target ship. The stability reserve constraint includes the lower limit of the initial stability height and the lower limit of the complete stability criterion number.

[0110] In this embodiment, the inherent legal attributes of the anchored target vessel and the unique risk characteristics of the first type of vehicle are used to clarify the quantitative sources and applicable boundaries of various constraints, which are ultimately combined into loading constraint characteristics.

[0111] Understandably, solidifying the scattered ship design parameters and stability requirements into insurmountable mandatory constraints provides a clear compliance boundary for the iterative optimization of subsequent loading schemes, significantly reduces the risk of fire and overturning in the mixed transportation of new energy vehicles, and ensures the safety and continuity of intelligent shipping operations.

[0112] Specifically, the upper limit of the deck load is determined based on the battery load distribution data, and the upper limit of the deck load is used as the core indicator for deck load strength constraint.

[0113] The battery load distribution data reflects the degree of battery energy accumulation in each deck area of ​​the cargo hold. The upper limit of the deck load is taken as a preset ratio of the corresponding deck load intensity data. The value of the preset ratio must simultaneously cover the sea state turbulence load during the ship's navigation, the load changes caused by fuel consumption, and the temporary load impacts of loading and unloading at intermediate ports, to ensure that the deck will not suffer structural damage due to overload in actual operation.

[0114] For example, the preset ratio can be set to 85%, that is, the upper limit of the deck load is 85% of the corresponding deck load strength data, leaving sufficient margin for various dynamic loads.

[0115] Based on the battery load distribution data, the upper limit of the weight difference between the port and starboard sides in the concentrated battery load area is further determined, serving as the core indicator for symmetrical battery load constraints.

[0116] For Type I vehicles, the battery pack can account for 30% to 50% of the total vehicle weight. If the battery load is concentrated on the port or starboard side of the cargo hold, it can cause the ship to list, affecting its navigational stability. The upper limit of the weight difference between the port and starboard sides is determined by working backward from the allowable list threshold in the target ship's stability calculation report, ensuring that the port and starboard distribution deviation of the battery load does not exceed the ship's list tolerance limit.

[0117] The center of gravity height constraint is determined based on the battery pack installation location and the vehicle's center of gravity height in the vehicle data. The center of gravity height constraint includes two layers of control requirements.

[0118] The first layer is the control requirements at the single vehicle level, which is the upper limit of the proportion of transport vehicles with a center of gravity height exceeding the preset height threshold on their respective decks. For models with battery packs installed on the roof, the overall center of gravity height is significantly higher than that of models with battery packs installed on the bottom. If the proportion of these high-center-of-gravity vehicles is too high, it will raise the overall center of gravity of the ship and reduce stability. Therefore, it is necessary to control the upper limit of their proportion.

[0119] The second layer is the control requirements at the ship-wide level, namely the upper limit of the overall center of gravity height of the target ship. The upper limit of the overall center of gravity height of the target ship is the core indicator in the stability limit threshold of the target ship. It is verified by the inclining test during the ship construction stage and is the legal red line for ship navigation safety.

[0120] The minimum distance between the first type of vehicle and the second type of vehicle is determined based on the preset isolation distance, and the first type of vehicle is restricted to parking in an independent fire compartment, which is the core content of the fire compartment isolation constraint.

[0121] The value of the preset isolation distance is determined by the fire resistance rating of the fire-resistant partition between the first loading zone and the second loading zone. The higher the fire resistance rating, the smaller the minimum allowable distance, and vice versa.

[0122] The independent fire compartment for parking Type 1 vehicles is the first loading compartment. The first loading compartment is equipped with an independent fire sprinkler system, explosion-proof ventilation system and temperature monitoring system. The fire compartment adjacent to the first loading compartment is the second loading compartment, which only allows Type 2 vehicles to be parked. The separation of the two types of vehicles can prevent the spread of fire caused by thermal runaway of Type 1 vehicles to the area of ​​Type 2 vehicles from the source.

[0123] Stability reserve constraints are determined based on the target vessel's stability limit thresholds. These constraints include the lower limit of the initial metacentric height and the lower limit of the intact stability criterion number. The initial metacentric height is a core indicator of vessel stability; a higher initial metacentric height indicates a stronger ability to right itself after encountering wind and waves. The lower limit of the initial metacentric height is the minimum permissible value specified in the stability limit thresholds, and no loading scheme can be lower than this minimum permissible value. The intact stability criterion number is a comprehensive indicator of a vessel's ability to withstand wind and waves. The lower limit of the intact stability criterion number is also a legal requirement within the stability limit thresholds, used to ensure the vessel's navigational safety under predicted wind and wave levels.

[0124] The permitted parking area for each type of transport vehicle is determined based on the first and second loading zones. Type 1 vehicles can only select available parking spaces within the first loading zone, and Type 2 vehicles can only select available parking spaces within the second loading zone. Parking across zones is strictly prohibited. The area is clearly defined by the fire protection zone planning data and is fully compatible with the ship's fire protection system design.

[0125] The determined upper limit of deck load, upper limit of weight difference between port and starboard sides, center of gravity height constraint, minimum interval, area range and stability reserve constraint are combined to determine the loading constraint characteristics. The loading constraint characteristics are the mandatory verification benchmark for all subsequent loading schemes. Any loading scheme that violates any one of these constraints will be directly eliminated, thereby ensuring that all loading schemes meet safety and compliance requirements from the beginning.

[0126] In some embodiments, the threshold value of the center of gravity height of a single vehicle and the upper limit value of the proportion of the number of decks in the center of gravity height constraint can be adjusted in combination with the battery health status data of the first type of vehicle. When the battery health of a certain first type of vehicle is lower than the preset health threshold, it is treated as a high center of gravity vehicle for management, or the upper limit value of the proportion of the number of decks in which it belongs is directly reduced to cover the special risk scenarios of transporting old new energy vehicles and further reduce the probability of chain reaction caused by thermal runaway of low-health batteries.

[0127] In some embodiments, optionally, such as Figure 3 As shown, step S108: Determine multiple initial loading schemes corresponding to the target vessel based on loading constraint characteristics and basic inherent data, including:

[0128] Step S1082: Determine the available parking spaces in the cargo hold based on the basic inherent data;

[0129] Step S1084: Based on the area range corresponding to each transport vehicle in the loading constraint features, assign a candidate parking space that conforms to the corresponding loading zone to each transport vehicle;

[0130] Step S1086: Within the range of candidate parking spaces, randomly generate multiple sets of correspondences between transport vehicles and parking spaces to obtain multiple initial loading schemes;

[0131] Step S1088: Perform initial screening on multiple initial loading schemes based on mandatory constraints, and eliminate initial loading schemes that do not meet the mandatory constraints.

[0132] In this embodiment, based on the basic inherent data of the target ship, parking spaces occupied by fixed facilities or with insufficient load margin are eliminated to determine the available parking spaces in the cargo hold. Then, according to the parking area range of vehicle type defined in the loading constraint features, multiple sets of vehicle-parking space correspondences are randomly generated within the candidate parking spaces to obtain multiple sets of initial loading schemes.

[0133] Understandably, the rules for matching candidate parking spaces by zone eliminate the risk of cross-zone parking from the generation stage. Multiple randomly generated schemes avoid path dependence caused by human experience and prevent the optimization results from getting stuck in local optima.

[0134] Furthermore, the initial screening process eliminates all non-compliant schemes in advance, significantly reducing the computational load for subsequent iterations and optimizations, thereby improving loading efficiency while ensuring the safety baseline of loading.

[0135] Specifically, after completing the construction of the load allocation constraint features, the load allocation optimization equipment begins to generate an initial set of load allocation schemes that meet the safety baseline, providing sufficient search space for subsequent iterative optimization.

[0136] The process of determining available parking spaces relies entirely on the target vessel's basic inherent data. This basic inherent data is a set of legal attributes verified by the classification society during the ship's construction phase. It includes deck load strength data for each deck, fire compartment planning data, lashing arrangement diagrams, coordinates, dimensions, load limits, and reserved boundaries for fire lanes and loading / unloading operation lanes in the cargo hold. The system first excludes parking spaces occupied by fixed facilities such as fire pipes and lashing posts, then eliminates parking spaces where the remaining deck load margin cannot meet the corresponding vehicle's weight. At the same time, it reserves space for fire lanes and operation lanes that meet the regulatory requirements. Finally, all unoccupied compliant parking spaces are the available parking spaces in the cargo hold.

[0137] When allocating candidate parking spaces, the permitted parking areas for each type of transport vehicle are strictly defined in the loading constraint characteristics. The area is defined by the legal boundary between the first loading zone and the second loading zone. Type 1 vehicles can only select available parking spaces as candidates within the first loading zone, and Type 2 vehicles can only select available parking spaces as candidates within the second loading zone. This avoids violations of cross-zone parking from the source of allocation.

[0138] If the number of vehicles of a certain type to be loaded exceeds the total number of available parking spaces in the corresponding zone, the system will immediately trigger an alert, prompting an adjustment to the loading plan for the voyage, rather than forcibly allocating parking spaces that do not meet the requirements of the area.

[0139] Within the candidate parking space range, the system randomly generates multiple sets of correspondences between transport vehicles and parking spaces, and each set of correspondences is an independent initial loading scheme.

[0140] The random generation process automatically verifies the uniqueness of parking spaces, ensuring that a single parking space in the same plan corresponds to only one transport vehicle. At the same time, it avoids blind spots in the binding and ensures that the parking position of each vehicle is within the fixed coverage area of ​​the binding.

[0141] The initial screening process involves a full verification of all initial loading schemes. The verification criteria are the explicit mandatory constraints in the loading constraint characteristics, including deck load strength constraints, battery load symmetry constraints, center of gravity height constraints, fire compartment isolation constraints, and stability reserve constraints. Any scheme that violates any of these constraints is immediately eliminated. Only schemes that pass all mandatory constraint verifications are retained and enter the subsequent iterative optimization process.

[0142] This logic for generating and selecting initial loading schemes avoids the optimization results from getting stuck in local optima. At the same time, the initial screening eliminates all unqualified schemes in advance, which greatly reduces the amount of computation in subsequent iterations, improves loading efficiency, and lays a solid foundation for the final output of a loading layout that balances safety and efficiency.

[0143] In some embodiments, optionally, during the process of allocating candidate parking spaces for each transport vehicle, the allocation rules can also be dynamically adjusted in conjunction with the priority tags of the transport vehicles. Priority tags include high-value vehicle tags, low-risk vehicle tags with battery health below a threshold, and priority loading and unloading vehicle tags designated by the lessee. Candidate parking spaces that are close to fire lanes, away from heat sources, and convenient for loading and unloading operations are allocated to high-priority vehicles first. Under the premise of meeting the loading zoning requirements, the transportation needs of special vehicles are also taken into account, thereby improving the commercial adaptability of the loading scheme.

[0144] In some embodiments, optionally, when randomly generating multiple sets of correspondences between transport vehicles and parking spaces, an experience-guided mechanism based on historical compliant loading schemes can be introduced. The system pre-stores high-frequency parking space and vehicle mapping relationships in historical high-quality loading schemes for the same type of target vessel. During the random generation process, high-frequency mapping relationships are matched first, and the remaining vehicles are then allocated purely randomly. This not only preserves the search breadth brought by random generation and avoids path dependence caused by human experience, but also improves the overall compliance rate of the initial loading scheme and reduces the computational load of subsequent iterative optimization.

[0145] In some embodiments, optionally, when the total number of candidate parking spaces in the corresponding loading zone is less than the number of vehicles of the same type to be loaded, the system automatically triggers a loading plan adjustment warning and generates a temporary compatible allocation scheme. The temporary compatible allocation scheme includes temporarily adjusting the use of parking spaces in non-core areas under the premise of compliance, prompting a reduction in the loading quantity of vehicles of the corresponding type, and suggesting adjustments to the voyage arrangements of some vehicles, so as to avoid violating the area range requirements in order to forcibly complete the loading and improve the robustness of the system in extreme scenarios.

[0146] In some embodiments, optionally, such as Figure 4 As shown, after obtaining the multi-objective optimization function in step S110, the roll-on / roll-off ship loading optimization method further includes:

[0147] Step S1102: Obtain the target vessel's route and sea state data and port call duration data;

[0148] Step S1104: Determine the forecast wind and wave level based on the sea state data of the shipping route;

[0149] Step S1106: Determine the urgency of the operation based on the port call operation duration data;

[0150] Step S1108: When the predicted wind and wave level is not less than the preset sea state level threshold, increase the weight coefficient corresponding to navigation stability and decrease the weight coefficient corresponding to the other sub-targets accordingly.

[0151] Step S1110: When the port call operation time data is not greater than the preset time threshold, increase the weight coefficient corresponding to the loading and unloading efficiency, and correspondingly decrease the weight coefficient corresponding to the other sub-targets.

[0152] Step S1112: Update the multi-objective optimization function according to the adjusted weight coefficients;

[0153] The sum of the weight coefficients corresponding to each sub-objective is a constant.

[0154] In this embodiment, after the loading optimization equipment obtains the multi-objective optimization function, it collects the route and sea state data and port call operation duration data of the target vessel, derives the forecast wind and wave level and the urgency of the operation respectively, adjusts the target weights, and updates the multi-objective optimization function according to the adjusted weights.

[0155] Understandably, by dynamically updating the loading weights to adapt to dynamic scenarios, the system automatically tilts towards navigation stability in adverse sea conditions, thus avoiding the risk of capsizing from the source. When port operation windows are tight, it prioritizes loading and unloading efficiency, reduces port demurrage costs, achieves a dynamic balance between safety and efficiency, adapts to the differentiated needs of different voyages, and significantly improves the practicality of the loading scheme.

[0156] Specifically, after acquiring the multi-objective optimization function, the load optimization equipment introduces the actual operating environment parameters of the voyage and dynamically calibrates the weight coefficients of each sub-objective to adapt the optimization direction to the specific risks and needs of the current voyage.

[0157] The process of obtaining route sea state data relies on real-time retrieval of marine environmental forecast information via ship-to-shore communication links. The route sea state data is a comprehensive meteorological and sea state report covering the entire navigation area of ​​the target vessel, including key indicators such as wind speed, wave height, swell direction, ocean current speed, and visibility. This data is provided by maritime meteorological service departments or ocean meteorological navigation systems and is the core input for assessing navigation risks.

[0158] The forecast wave level is determined by referring to the Beaufort scale or Dow scale standards set by the International Meteorological Organization based on the significant wave height and wind speed parameters in the sea state data of the shipping route. The forecast wave level quantifies the severity of the external environment that a ship will face during navigation. The higher the level, the greater the risk of the ship capsizing, deck waves, and cargo displacement.

[0159] At the same time, the loading optimization equipment acquires port call operation time data, which includes the expected berthing time window, total tonnage requirements for loading and unloading operations, and port congestion index for each port that the target vessel plans to call at during this voyage. This information is synchronized to the loading optimization equipment by the port scheduling system or shipping planning management system.

[0160] The urgency level of an operation is determined based on the maximum allowed berthing time in the port call operation duration data. When the port congestion index is high or subsequent voyages are closely connected, the urgency level of the operation is judged as high, which means that a large number of vehicles must be loaded and unloaded within a shorter time limit, otherwise high demurrage fees or delays in subsequent voyages will be faced.

[0161] Based on the above two evaluation results, the device executes a dynamic adjustment logic for the weight coefficients.

[0162] When the predicted wind and wave level is not less than the preset sea state level threshold, it indicates that the ship is about to face severe sea conditions. At this time, the equipment automatically increases the weight coefficient corresponding to navigation stability. The preset sea state level threshold is a pre-set safety warning line, usually set to Beaufort scale 6 or Dow scale 4. Once the preset sea state level threshold is exceeded, the priority of navigation safety will be raised to the highest level. By increasing the corresponding weight, the optimization algorithm is forced to prioritize eliminating those schemes with high loading rates but low stability margins in subsequent iterations, ensuring that the ship has sufficient wind and wave resistance. The weight coefficients of cargo hold utilization and loading and unloading efficiency are correspondingly reduced. This means that under extreme sea conditions, it is permissible to sacrifice some loading space or reduce considerations of loading and unloading convenience in exchange for higher stability reserves.

[0163] Conversely, when the port call operation time data is not greater than the preset time threshold, it indicates that the port operation window is extremely tight. The loading optimization equipment automatically increases the weight coefficient corresponding to loading and unloading efficiency. The preset time threshold is the economic warning line for port operations, usually 24 hours. Once it falls below the preset time threshold, the priority of operational efficiency is raised to the highest level. By increasing the corresponding weight, the optimization algorithm is guided to prioritize generating layout schemes with concentrated vehicle distribution that facilitates mechanical operations. At this time, the loading optimization equipment correspondingly reduces the weight coefficient of the navigation stability and cargo capacity utilization sub-objectives, maximizing the loading and unloading speed without exceeding the safety limit.

[0164] After completing the above adjustments, the device updates the multi-objective optimization function based on the adjusted weight coefficients. The update process is not a simple numerical replacement, but rather the new weight coefficients are written into the mathematical expression of the multi-objective optimization function, and the weighted contribution of each sub-objective is recalculated, so that each subsequent fitness evaluation is based on the new weight system.

[0165] Throughout the entire weight adjustment and update process, the device strictly maintains the sum of the weight coefficients corresponding to each sub-target as a constant value, which is usually set to one. This normalization process ensures that the evaluation values ​​under different weight configurations are comparable and avoids algorithm convergence anomalies caused by fluctuations in the total weight.

[0166] The dynamic weight adjustment mechanism transforms abstract shipping experience into precise algorithm parameters, achieving an intelligent balance between safety and efficiency. By responding in real time to changes in the external environment and internal production pressures, it can consistently output the optimal solution that best meets current interests under different navigation stages and port conditions. This ensures navigation safety under extreme sea conditions while improving the economic efficiency of port operations, fully demonstrating the decision-making advantages of intelligent cargo loading.

[0167] In some embodiments, the route sea state data can optionally be integrated with real-time marine environmental observation data, which includes real-time ocean current speed, swell cycle, underwater obstacle distribution along the target vessel's navigation path, as well as extreme weather warnings such as typhoons and cold waves in the sea areas it passes through. The system combines real-time observation data to dynamically correct the forecast wind and wave levels, avoiding the lag in weight adjustment caused by relying solely on previous forecasts, and covering scenarios of sudden weather changes during navigation.

[0168] In some embodiments, the preset sea state level threshold and the preset duration threshold can be dynamically adjusted in conjunction with the age, type, and priority of the target vessel. The preset sea state level threshold for older vessels over 15 years old is lowered by half a level to level one, while the preset duration threshold for priority voyages transporting high-value Type I vehicles is increased by 10% to 20%, adapting to the differentiated risk tolerance of different vessels and voyages and avoiding insufficient adaptability caused by fixed thresholds.

[0169] In some embodiments, the weight adjustment can optionally be performed in segments according to the navigation phase. The preset sea state level threshold corresponding to the long-distance navigation phase across the ocean is set strictly, while the preset sea state level threshold corresponding to the short-distance navigation phase near the coast is appropriately relaxed. When the target vessel approaches an intermediate port, the system retrieves the operation time data of the next port of call 24 hours in advance. If the operation time is tight, the weight adjustment process of loading and unloading efficiency is started in advance to avoid operation delays caused by adjusting the weight after berthing, covering the scenario of continuous operation in multiple segments.

[0170] In some embodiments, the loading constraint features may optionally include priority constraints, such as... Figure 5 As shown, step S112: Under the premise of satisfying the mandatory constraints, iteratively optimize multiple initial loading schemes according to the multi-objective optimization function to determine at least one loading layout result, including:

[0171] Step S1120: Using the function value of the multi-objective optimization function as the fitness, iteratively optimize multiple initial loading schemes according to the improved adaptive genetic algorithm;

[0172] Step S1122: In each iteration, load planning schemes that do not meet the mandatory constraints are eliminated, and the fitness of the retained load planning schemes is modified according to the priority constraints.

[0173] Step S1124: Adaptively adjust the crossover probability according to the fitness of each loading scheme, so that the crossover probability corresponding to the loading scheme with fitness higher than the preset fitness threshold is less than the crossover probability corresponding to the loading scheme with fitness lower than the preset fitness threshold.

[0174] Step S1126: Adaptively adjust the mutation probability according to the iteration process, so that the mutation probability in the early stage of the iteration is greater than the mutation probability in the later stage of the iteration.

[0175] Step S1128: Introduce a dynamic correction operator into the fitness calculation. The dynamic correction operator includes at least one of the fuel consumption center of gravity offset correction factor and the intermediate port loading and unloading load change correction factor.

[0176] Step S1130: When the preset termination condition is met, output the loading scheme with the highest fitness as the loading layout result;

[0177] The preset termination conditions include no improvement in fitness for consecutive preset generations or reaching the preset maximum number of iterations.

[0178] In this embodiment, based on the load allocation constraint features including priority constraints, an improved adaptive genetic algorithm is used to perform iterative optimization, with the output value of the multi-objective optimization function as the fitness. In each iteration, schemes that do not meet the mandatory constraints are first eliminated, and then the fitness of the retained schemes is adjusted according to the priority constraints. The optimal load allocation scheme is output after the termination condition of no improvement in fitness for several consecutive generations or the maximum number of iterations is reached.

[0179] Understandably, by using an adaptive adjustment mechanism that balances global search capabilities and convergence efficiency, while adhering strictly to safety standards, it also considers operational convenience and efficiency, adapting to the complex loading scenarios of multi-objective roll-on / roll-off ships and ensuring that the output solution remains compliant throughout the entire voyage cycle.

[0180] After obtaining the multi-objective optimization function and multiple initial loading schemes, the load optimization equipment enters the iterative optimization stage to determine the optimal load layout result.

[0181] In addition to mandatory constraints, loading constraints also include priority constraints. Priority constraints are soft control rules that are subordinate to mandatory constraints. They are derived from charterer special transport agreements, port operation specifications, and recommended configuration clauses for fire protection systems. These include specific requirements such as parking high-value Type 1 vehicles near fire lanes and parking vehicles to be unloaded at intermediate ports near loading and unloading docks. Violating priority constraints will not result in the loading scheme being directly eliminated, but will only deduct corresponding points during the fitness calculation.

[0182] The output value of the multi-objective optimization function is the fitness, which is used to quantify the overall performance of the loading scheme. The higher the value, the better the overall performance of the scheme. The fitness is calculated by weighting the three sub-objectives of cargo capacity utilization, navigation stability and loading and unloading efficiency in the multi-objective optimization function according to the configured weight coefficients.

[0183] The loading optimization equipment uses fitness as the evaluation criterion and iteratively optimizes multiple initial loading schemes using an improved adaptive genetic algorithm. The improved adaptive genetic algorithm is an optimization algorithm designed to address the shortcomings of traditional genetic algorithms, which are prone to premature convergence or slow convergence due to fixed crossover and mutation probabilities. It is adapted to the complex scenario of multi-constraint and multi-objective loading of roll-on / roll-off ships.

[0184] In each iteration, the load optimization equipment first eliminates load schemes that do not meet the mandatory constraints. The mandatory constraints include battery load symmetry constraints, deck load strength constraints, fire compartment isolation constraints, and stability reserve constraints. Schemes that do not meet any of the mandatory constraints are not included in subsequent calculations.

[0185] The loading optimization equipment modifies the fitness of the retained loading schemes based on priority constraints. The modification range is preset according to the importance of the priority constraints. Priority constraints with higher importance correspond to larger modification ranges. If there are cases in the scheme where high-value Type I vehicles are not close to the fire lane, the corresponding points will be deducted from the fitness according to the corresponding deduction rules. If there are cases in the scheme where vehicles to be unloaded at intermediate ports are close to the loading and unloading operation port, the corresponding points will be added to the fitness according to the corresponding bonus rules.

[0186] The load allocation optimization device adaptively adjusts the crossover probability based on the fitness of each load allocation scheme. The crossover probability is defined as the probability that chromosome segment exchange and recombination occurs between two parent load allocation schemes in the genetic algorithm.

[0187] The preset fitness threshold is the statistical average fitness of all retained loading schemes in the current generation, used to distinguish between high-fitness schemes and low-fitness schemes.

[0188] The crossover probability of a high-fitness solution is lower than that of a low-fitness solution. This allows for the preservation of the advantages of the high-fitness solution while exploring potential high-quality solutions for the low-fitness solution through high-probability crossover, thus balancing the efficiency of gene preservation and new solution exploration.

[0189] The load allocation optimization equipment adaptively adjusts the mutation probability according to the iteration process. The mutation probability is defined as the probability that the parking position of a certain vehicle in a single load allocation scheme will randomly change.

[0190] The iteration process is determined by the ratio of the current iteration count to the total preset iteration count. In the early stages of iteration (when the ratio is less than one-third), a higher mutation probability is set to expand the search range and prevent the algorithm from getting trapped in local optima prematurely. In the later stages of iteration (when the ratio is greater than two-thirds), a lower mutation probability is set to reduce invalid perturbations and accelerate convergence.

[0191] The load optimization equipment incorporates a dynamic correction operator in its fitness calculation. This dynamic correction operator is a module used to predict dynamic changes in operating conditions throughout the voyage. It includes at least one of a fuel consumption center of gravity offset correction factor and a mid-voyage port loading / unloading load change correction factor.

[0192] The process of determining the fuel consumption center of gravity offset correction factor is to calculate the center of gravity offset caused by fuel consumption during the voyage based on the total fuel consumption of the voyage, the location of the fuel tanks, and the order of fuel consumption. The center of gravity offset is then quantified and added to the fitness calculation logic. If the loading scheme does not reserve the corresponding center of gravity margin, the corresponding fitness score is deducted.

[0193] The process of determining the load change correction factor at intermediate ports is to calculate the load distribution and center of gravity change caused by intermediate port loading and unloading operations based on the unloading list, loading list, and loading and unloading sequence in the voyage call port sequence. After quantifying the center of gravity change, it is added to the fitness calculation logic. If the loading scheme does not reserve the corresponding load margin, the corresponding fitness score is deducted.

[0194] When preset termination conditions are met, the loading optimization equipment stops iterating and outputs the loading scheme with the highest fitness as the loading layout result. Preset termination conditions include no improvement in fitness after a preset number of consecutive iterations or reaching a preset maximum number of iterations. The preset number of consecutive iterations is a pre-set threshold for the number of iterations, corresponding to the typical convergence period for Ro-Ro vessel loading, used to prevent the algorithm from iterating ineffectively after reaching its optimal state. The preset maximum number of iterations is a pre-set maximum threshold for the number of iterations, used to ensure that the loading calculation is completed within the port's required response time limit, avoiding wasted computing power.

[0195] In some embodiments, the sources of priority constraints can optionally be extended to battery health status data, insurance institution risk control clauses, and port special operation specifications. When the battery health of a Type 1 vehicle is lower than a preset health threshold, a priority constraint of reserving a safety distance of no less than one meter around the low-health vehicle is added. When a voyage is insured with high-value transportation insurance, a priority constraint of requiring the parking area of ​​high-value vehicles to cover twice the coverage of fire sprinklers is added. This covers special scenarios such as the transportation of old new energy vehicles and high-value customized transportation, adapting to differentiated operational needs without touching the bottom line of mandatory constraints.

[0196] In some embodiments, the preset termination condition can optionally be dynamically adjusted based on the real-time requirements of the voyage. When an emergency port operation instruction is received, the number of consecutive preset generations is adjusted to ten, and the upper limit of the preset number of iterations is adjusted to one hundred, prioritizing the timeliness of the loading response. For regular voyages, the upper limit of twenty consecutive generations without improvement and two hundred iterations is maintained to ensure the optimality of the solution. At the same time, an early termination rule is added: when the margin of all mandatory constraints of the currently output loading solution exceeds 20%, and the fitness fluctuation of five consecutive generations is less than 0.5%, the iteration can be terminated early to reduce unnecessary computing power consumption and improve loading efficiency.

[0197] In some embodiments, optionally, such as Figure 6 As shown, the optimized loading method for roll-on / roll-off (Ro-Ro) ships also includes:

[0198] Step S200: During the navigation of the target vessel, acquire real-time navigation data at time intervals not exceeding a preset period;

[0199] Step S202: Determine the dynamic optimization triggering characteristics based on real-time navigation data and load layout results;

[0200] Step S204: When the dynamic optimization triggering feature meets the preset triggering condition, the loading layout result is locally iteratively corrected based on real-time navigation data and loading constraint features to obtain the iterative loading layout result;

[0201] Step S206: Update the loading layout results based on the iterative loading layout results;

[0202] The preset triggering conditions include at least one of the following: the target vessel's center of gravity offset exceeds the preset offset threshold, the real-time initial stability height is lower than the preset stability threshold, and the wind and wave level corresponding to the real-time sea state is higher than the predicted wind and wave level by no less than the preset level difference.

[0203] Real-time navigation data includes at least one of the following: real-time draft, real-time center of gravity, fuel consumption, ballast water status, and real-time sea state.

[0204] The dynamic optimization triggering features are defined based on battery load distribution data and load constraint features.

[0205] In this embodiment, the load optimization equipment collects real-time navigation data at a preset cycle. Based on the battery load distribution data and load constraint characteristics in the output load layout results, it determines the dynamic optimization trigger characteristics. When the trigger characteristics meet the conditions such as the center of gravity shift exceeding the preset threshold, the initial stability height being lower than the stability threshold, or the sea state rising suddenly exceeding the preset level difference, it performs local iterative correction. After correction, the load results are updated and synchronized to the ship's bridge.

[0206] Understandably, after the ship sets sail, it continuously collects real-time data at a preset cycle. When a preset threshold is triggered, it automatically performs dynamic correction and issues an updated plan. It can offset the risk of stability degradation caused by fuel consumption and deteriorating sea conditions without global recalculation. Small-scale adjustments can avoid operational interruptions, provide real-time and accurate load support for the bridge, solve the pain point of the disconnect between dynamic changes in operating conditions during navigation and the initial load, and balance navigation safety and operational efficiency.

[0207] The roll-on / roll-off (Ro-Ro) vessel loading optimization method provided by this invention does not terminate the execution process after outputting the loading layout results, but instead switches to a full-voyage dynamic monitoring mode to cope with the time-varying characteristics of the ship's own state and environmental parameters during the voyage.

[0208] The load optimization equipment acquires real-time navigation data at intervals not exceeding a preset period. The preset period is a data acquisition interval dynamically set by the load optimization equipment's built-in scheduling module based on the sea state level initially set for the voyage. It is automatically shortened in severe sea states to increase monitoring frequency and appropriately extended in calm sections to reduce system power consumption.

[0209] Real-time navigation data includes at least one of the following: real-time draft, real-time center of gravity, fuel consumption, ballast water status, and real-time sea state.

[0210] Real-time draft is a comprehensive depth parameter obtained by collecting the submersion depth of the ship at various locations in the water from draft sensors installed at the bow, stern, and foreshore. This parameter is used to calculate the current displacement of the ship.

[0211] The real-time center of gravity is the three-dimensional coordinate of the ship's center of gravity, calculated by the inertial navigation system in conjunction with the load monitoring module based on real-time data of fuel consumption, ballast water allocation, and minor cargo displacement.

[0212] Fuel consumption is the total fuel consumption accumulated since departure from port, as recorded by the flow meters at the outlets of each fuel tank, and is used to estimate the center of gravity shift caused by fuel reduction.

[0213] Ballast water status refers to the water volume, density, and pump operation status of each compartment as monitored by level sensors in each ballast water tank, which is used to adjust the ship's buoyancy and overall longitudinal strength in real time.

[0214] Real-time sea conditions are environmental parameters such as wind speed, wind direction, significant wave height, and surge period at the current location of the ship, collected by shipborne meteorological instruments and wave radar.

[0215] The load optimization equipment determines the dynamic optimization trigger features based on real-time navigation data and load layout results. The dynamic optimization trigger features are comprehensive indicators that quantify the degree of deviation between the current actual state of the ship and the preset benchmark of the load layout results. Their definition is entirely based on battery load distribution data and load constraint features.

[0216] The loading optimization equipment first performs a vector subtraction between the real-time center of gravity coordinates and the initial center of gravity coordinates recorded in the loading layout results to obtain the center of gravity offset vector. The lateral component of the center of gravity offset vector is the lateral offset of the center of gravity, which is used to characterize the ship's heeling tendency.

[0217] Secondly, based on the change model of fuel consumption and ballast water status, the real-time load distribution of each deck area is calculated. The real-time load distribution is then compared with the battery load distribution data recorded in the loading layout results in a grid, and the areas of abnormal local load accumulation caused by minor cargo displacement or structural deformation are identified.

[0218] Finally, the significant wave height in the real-time sea state is compared with the wave height threshold of the predicted wind and wave levels corresponding to the loading layout results to quantify the degree of sea state deterioration. The aforementioned lateral offset of the center of gravity, local load deviation, and sea state deterioration coefficient together constitute the dynamic optimization triggering characteristics.

[0219] When the dynamic optimization triggering feature meets the preset triggering conditions, the load optimization device performs local iterative correction on the load layout result to obtain the iterative load layout result.

[0220] The preset trigger conditions include at least one of the following: the target vessel's center of gravity offset exceeds a preset offset threshold, which is the maximum allowable offset ratio set according to the upper limit of the weight difference between port and starboard sides recorded in the loading constraint characteristics, and is used to define the degree of danger of the center of gravity offset; the real-time initial stability height is lower than a preset stability threshold, which is the minimum allowable stability margin set according to the lower limit of the initial stability height recorded in the loading constraint characteristics, and is used to reserve a safety margin for navigation; the wind and wave level corresponding to the real-time sea state increases by no less than a preset level difference relative to the forecast wind and wave level, which is the maximum allowable jump in wind and wave level set according to the error range of the marine environment forecast, and is used to define the degree of danger of sea state deterioration.

[0221] Local iterative correction involves limited adjustments to specific deck or grid areas most affected by dynamic factors, without altering the overall layout, aiming to restore safety margins with minimal intervention. During the correction process, an improved adaptive genetic algorithm is used to fine-tune the parking positions of a small number of vehicles. Adjustments include adjusting the preload of vehicle lashing rigging or replacing the parking positions of a few vehicles within the available parking spaces of their respective loading zones to balance port and starboard weight or reduce local load peaks. All adjustments strictly adhere to all mandatory constraints recorded in the loading constraint characteristics, prohibiting adjustments across fire zones and exceeding deck load strength constraints.

[0222] The loading optimization equipment updates the loading layout results based on the iterative loading layout results. The update process involves writing the new vehicle parking positions, lashing parameters, and stability calculation data recorded in the iterative loading layout results into the storage module of the loading optimization equipment. At the same time, the updated data is transmitted to the ship's bridge terminal through the ship-shore communication link to provide the captain with real-time decision-making basis for ship handling.

[0223] In some embodiments, the preset period can be dynamically adjusted in conjunction with the voyage segment attributes and sensor health status. The preset period corresponding to the transoceanic long-distance voyage segment is longer than that for the near-shore voyage segment. When a certain type of sensor fails and cannot collect effective data, the collection interval of the other related sensors is shortened to form data complementarity, avoiding dynamic monitoring blind spots caused by single-point sensor failure, and covering scenarios of sensor failure and different monitoring needs of different voyage segments.

[0224] In some embodiments, the scope of real-time navigation data collection can be optionally extended to include temperature monitoring data around the first type of vehicle in the cargo hold, real-time stress monitoring data of lashing components, navigation control data of special navigation areas, and temporary port scheduling instructions. Temperature monitoring data is used to identify abnormal heating trends of the battery pack of the first type of vehicle, stress monitoring data of lashing components is used to quantify the risk of vehicle displacement, and special navigation area data and port instructions are used to respond to sudden control requirements, enriching the decision-making basis for dynamic optimization.

[0225] In some embodiments, optionally, the process of updating the loading layout results based on the iterative loading layout results can also be linked with the shore-based monitoring center and the maritime regulatory platform. The updated layout data is synchronously sent to the shore-based operation and maintenance terminal and the maritime reporting system. At the same time, a loading change report is automatically generated, specifying the reason for the change, the adjusted parking spaces, and the stability change parameters. If the deviation between the adjusted layout and the initial layout exceeds a preset deviation threshold, an inspection task is automatically issued to the crew terminal, requiring on-site verification of the vehicle lashing status and displacement of the changed parking spaces. The updated layout can also be input into the shipborne digital twin model to conduct real-time stability simulation verification to ensure that the corrected layout is compliant under all operating conditions.

[0226] In some embodiments, the dynamic optimization process can optionally be adapted to special navigation scenarios. When the target vessel enters restricted waters, polar seas, or pirate warning zones, the sensitivity of preset triggering conditions is automatically increased, and the preset cycle is shortened to increase the monitoring frequency. When the vessel encounters extreme weather such as typhoons or cold waves, the adjustment range of local iterative correction is automatically expanded to reserve a larger stability margin. When the vessel is carrying out loading and unloading operations at a port, the dynamic optimization process is paused to avoid disrupting the operation order. Monitoring is restarted after the operation is completed, covering the needs of extreme environments and special operation scenarios, and improving the robustness of the solution.

[0227] In some embodiments, optionally, such as Figure 7 As shown, the optimized loading method for roll-on / roll-off (Ro-Ro) ships also includes:

[0228] Step S300: In response to the berthing instruction at the intermediate port, obtain the current vessel status characteristics, the unloading list and loading list corresponding to the intermediate port;

[0229] Step S302: Based on the current vessel status characteristics and unloading list, and in conjunction with the preset isolation distance, determine the stability prediction characteristics after unloading;

[0230] Step S304: Based on the stability prediction characteristics, the loading list and the stowage constraint characteristics, the vehicle parking positions in the remaining cargo hold space are re-iterated to determine the stowage layout results at intermediate ports.

[0231] Step S306: Generate loading and unloading operation instructions corresponding to the intermediate ports based on the intermediate port loading layout results, and synchronize the loading and unloading operation instructions to the ship's bridge terminal, deck work team terminal and port dispatch center terminal through the ship-shore communication link.

[0232] In this embodiment, the loading optimization equipment responds to the berthing instruction at the intermediate port and retrieves the current ship status characteristics within a preset time window before the target ship arrives at the port. The preset time window takes into account both data timeliness and calculation reserve period. After deducting the corresponding load in combination with the unloading list, the stability parameters after unloading are predicted strictly according to the loading constraint characteristics. The loading list and the remaining compliant cargo hold space are matched, and the optimal loading layout at the intermediate port is iteratively solved under the mandatory constraint framework. Loading and unloading operation instructions containing loading and unloading sequence, operation path and lashing requirements are generated and synchronized to the ship bridge terminal, deck operation team terminal and port dispatch center terminal through the ship-shore communication link.

[0233] Understandably, the integrated loading and unloading optimization scheme at intermediate ports is regenerated based on real-time ship status and synchronized to each execution terminal through ship-shore communication links. This allows for the early completion of loading and unloading solutions, reducing port operation waiting time, optimizing operation routes, improving loading and unloading efficiency, adapting to the needs of multi-segment continuous transportation scenarios, and reducing port demurrage costs and navigation safety hazards.

[0234] The intermediate port berthing instruction is issued by the intermediate port dispatch center to the loading optimization equipment through the ship-to-shore communication link. The instruction includes the estimated berthing time, the designated berth number, the maximum allowed operation time, and the operation priority requirements. It is the trigger signal to start the intermediate port loading adjustment process.

[0235] The preset duration is an advance response window determined by the load optimization equipment based on the target vessel's current real-time speed, remaining voyage distance, and the shortest preparation time for port operations at intermediate ports. The advance response window must meet two conditions simultaneously: first, it must be earlier than the expected berthing time to ensure the completion of load calculation; second, it must be later than the most recent dynamic optimization update time to ensure the timeliness of the input data.

[0236] The current ship status characteristics are a set of all operating parameters of the target ship collected by the loading optimization equipment at the current moment, including real-time draft, real-time center of gravity, fuel consumption, ballast water status, currently effective loading layout results, distribution of remaining available parking spaces in the cargo hold, and the lashing status of loaded vehicles. All of the above parameters come from the real-time monitoring data storage module during the navigation phase and are the basic input for predicting the ship status after unloading.

[0237] The unloading list is provided simultaneously by the intermediate port dispatch center, recording the unique identifier, current parking location, vehicle type, and battery pack parameters of all transport vehicles to be unloaded during this berth. The loading list is also provided simultaneously by the intermediate port dispatch center, recording the unique identifier, vehicle type, battery pack parameters, and scheduled loading zone requirements of all transport vehicles to be loaded onto the ship during this berth. Both lists must correspond one-to-one with the port operation plan to ensure that the loading adjustment matches the actual operation needs.

[0238] The loading optimization equipment determines the stability prediction characteristics after unloading based on the current ship status characteristics and unloading list, combined with the preset isolation distance requirements.

[0239] The determination process involves first deducting the weight contribution, battery load contribution, and center of gravity influence of all vehicles to be unloaded from the unloading list based on the current ship's current state characteristics, to obtain preliminary state parameters after unloading. Then, combined with the predicted fuel consumption during the voyage to the intermediate port and the ballast water adjustment plan, the estimated center of gravity coordinates, estimated initial stability height, and estimated load distribution in each deck area after unloading are calculated. These parameters together constitute the stability prediction characteristics. During the calculation process, the deck load strength constraint, battery load symmetry constraint, and stability reserve constraint in the loading constraint characteristics must be strictly referred to to ensure that the prediction results meet the statutory safety requirements.

[0240] The remaining cargo hold space is the set of parking spaces in the cargo hold that can be used by newly loaded vehicles after unloading. The determination process is as follows: remove all parking spaces corresponding to the unloading list from the currently effective loading layout results, and then, based on the parking area allowed for each type of transport vehicle in the loading constraint characteristics, remove parking spaces that do not meet the vehicle zoning requirements in the loading list. At the same time, exclude parking spaces occupied by fixed facilities, blind spots for lashing, and reserved fire lanes, and finally obtain the set of remaining available parking spaces that meet the safety requirements.

[0241] The loading optimization equipment iteratively solves for the parking positions of vehicles in the remaining cargo hold space based on stability prediction characteristics, the loading list, and loading constraints, thus determining the loading layout at intermediate ports. The solution process uses a multi-objective optimization function with pre-configured weight coefficients as the evaluation benchmark, assigning each transport vehicle in the loading list to a compliant parking space within the remaining cargo hold. The process strictly adheres to all mandatory constraints in the loading constraints, including preset isolation distance requirements, fire compartment isolation constraints, battery load symmetry constraints, and deck load strength constraints. Simultaneously, a dynamic correction operator for the navigation phase is introduced to predict fuel consumption and sea state impacts in subsequent segments. An improved adaptive genetic algorithm is used for iterative optimization, ultimately outputting the loading layout at intermediate ports that satisfies all constraints and has the highest overall fitness.

[0242] The loading optimization equipment generates loading and unloading operation instructions corresponding to the intermediate ports based on the loading layout results of the intermediate ports. The instructions include the unloading sequence of vehicles to be unloaded, the deck driving path, and the temporary storage area; the loading sequence of vehicles to be loaded, the target parking space, the lashing operation requirements, and the operation conflict avoidance rules. The generation process must ensure that the loading and unloading paths do not intersect, the operation line is the shortest, and high-value vehicles and high-energy-density type 1 vehicles are given priority to maximize port operation efficiency.

[0243] The loading optimization equipment synchronizes loading and unloading operation instructions to the ship's bridge terminal, deck work team terminal, and port dispatch center terminal via the ship-to-shore communication link.

[0244] The ship's bridge terminal is used to display the overall operation plan, post-unloading stability parameters, and emergency response plan to the captain. The deck operation team terminal is used to push specific vehicle movement paths and lashing operation instructions to on-site personnel. The port dispatch center terminal is used to provide port dispatchers with operation progress tracking and resource allocation basis. The data of the three terminals are consistent in real time to avoid operation conflicts and information asymmetry.

[0245] In some embodiments, optionally, such as Figure 8 As shown, in step S112: after iteratively optimizing multiple initial loading schemes according to the multi-objective optimization function and determining at least one loading layout result, the ro-ro ship loading optimization method further includes:

[0246] Step S1140: Determine the three-dimensional visualization data corresponding to the loading layout results using a digital twin model;

[0247] Step S1142: Perform static compliance verification on the load layout results based on the 3D visualization data to obtain the static compliance verification results;

[0248] Step S1144: If the static compliance verification result indicates compliance, perform full-condition dynamic simulation verification of the load distribution layout result.

[0249] Step S1146: When the load distribution layout result passes the full-condition dynamic simulation verification, determine the stability calculation result and fire compartment verification data based on the load distribution layout result;

[0250] The 3D visualization data includes vehicle parking locations, deck load distribution cloud maps, ship center of gravity coordinates, fire protection zone divisions, and loading and unloading route planning.

[0251] The full-condition dynamic simulation verification includes at least the extreme sea state verification. The verification parameters for the extreme sea state verification include a roll angle not less than a preset roll angle threshold and a wind and wave level not less than a preset wind and wave level threshold.

[0252] In this embodiment, after obtaining the load layout result through iterative optimization, the load optimization equipment imports it into the digital twin model, performs a digital twin three-dimensional visualization of the optimal solution, and sequentially passes static compliance verification and full-condition dynamic simulation verification. After all verifications are passed, the stability calculation results and fire compartment verification data are output.

[0253] Understandably, 3D visualization solves the problem that traditional 2D loading drawings cannot identify hidden compliance risks, and dynamic simulation covers extreme conditions throughout the voyage, thus eliminating safety hazards caused by layout defects from a mechanism perspective.

[0254] After determining the loading layout through iterative optimization, a digital twin model is used to pre-verify the compliance and adaptability of the loading layout results to avoid potential safety hazards during subsequent navigation or operations.

[0255] The digital twin model is a virtual simulation system built into the loading optimization equipment, which models the target ship one-to-one. Its modeling data source is the basic inherent data of the target ship, including deck structure drawings verified by the classification society, fire compartment planning data, lashing arrangement diagrams, ship hydrostatic parameters, and stability calculation reports. It can accurately reproduce the physical properties, navigation characteristics, and fire protection system coverage of the target ship, providing a high-fidelity simulation environment for virtual verification.

[0256] The load allocation optimization equipment imports the load allocation layout results into a digital twin model to generate three-dimensional visualization data corresponding to the load allocation layout results.

[0257] The 3D visualization data is a collection of all elements that can be interactively viewed in a virtual scene. Among them, the vehicle parking position is the specific location information of each transport vehicle in the cargo hold in the loading layout result. In the digital twin model, it is mapped to the coordinates of the physical parking space on the corresponding deck layer, which corresponds one-to-one with the parking space in the physical cargo hold.

[0258] The deck load distribution cloud map renders the battery load and vehicle load in each grid of the cargo hold as gradient layers of different colors according to their numerical values, intuitively showing the load concentration and distribution uniformity in each deck area.

[0259] The ship's center of gravity coordinates are three-dimensional spatial coordinates calculated in a digital twin model based on the weight and center of gravity position of all vehicles in the loading layout, combined with the pre-set fuel and ballast water reserves for navigation. They are the core parameters for stability verification.

[0260] Fire compartments are defined in the digital twin model as the first loading compartment, the boundary of the second loading compartment, the location of fire-resistant partitions, and the coverage of fire sprinklers and explosion-proof ventilation facilities, which visually represent the physical isolation status of the two types of loading compartments.

[0261] The loading and unloading route planning is a digital twin model that plans a non-intersecting driving route from the deck to the dock for each vehicle based on the parking position of the vehicle in the loading layout. The route automatically avoids fire lanes, lashing blind spots and temporary work areas to ensure the orderly operation of loading and unloading.

[0262] The load optimization equipment performs static compliance verification on the load layout results based on 3D visualization data, and obtains the static compliance verification results. The verification basis for static compliance verification is all the mandatory constraints listed in the load constraint characteristics. The verification process involves comparing each element in the 3D visualization data to check whether the vehicle parking position is within the allowable parking range of the corresponding loading zone, whether the load in each area shown in the deck load distribution cloud map does not exceed the upper limit of the deck load, whether the weight difference of the battery load on the port and starboard sides does not exceed the upper limit of the weight difference on the port and starboard sides, whether the parking interval between the first type of vehicles and the second type of vehicles is not less than the preset isolation distance, whether the center of gravity height corresponding to the ship's center of gravity coordinates meets the center of gravity height constraint requirements, and whether the calculated initial stability height and complete stability criterion number are not lower than the lower limit of the initial stability height and the lower limit of the complete stability criterion number in the stability reserve constraints, respectively.

[0263] The static compliance verification result is the output of the verification process. Only when all verification items meet the mandatory constraints can the static compliance verification result be considered compliant, and the load layout result can enter the subsequent verification stage. If any verification fails, the load optimization equipment will automatically backtrack and adjust the load layout result until all static compliance requirements are met.

[0264] Under the condition that the static compliance verification results indicate compliance, the load optimization equipment performs dynamic simulation verification of the load layout results under all working conditions.

[0265] The full-condition dynamic simulation verification covers all typical operations and extreme environmental scenarios throughout the entire ship navigation cycle, including at least normal sea state cruise conditions, continuous fuel consumption conditions, intermediate port loading and unloading operations, and extreme sea state conditions.

[0266] The extreme sea state calibration verifies the safety of the loading layout under extreme marine conditions. Its calibration parameters include a roll angle not less than a preset roll angle threshold and a wave level not less than a preset wave level threshold. The preset roll angle threshold is the maximum permissible roll angle set according to the target vessel's stability calculation report, typically ranging from 15 to 30 degrees, corresponding to the vessel's roll tolerance limit under extreme sea states. The preset wave level threshold is the highest seaworthy wave level set according to the International Maritime Organization's (IMO) Regulations for Navigation Safety, typically Beaufort scale 8 to 10, corresponding to the upper limit of sea states where the vessel can safely navigate.

[0267] During the verification process, the digital twin model loads verification parameters for extreme sea state conditions, simulating the ship's navigation attitude in extreme environments where the roll angle is not less than the preset roll angle threshold and the wind and wave level is not less than the preset wind and wave level threshold. It monitors the vehicle lashing status, deck load changes, and stability parameter fluctuations in real time to determine whether there are risks such as vehicle displacement, lashing failure, load exceeding limits, or stability falling below the safety threshold.

[0268] In addition to extreme sea state conditions, the full-condition dynamic simulation verification also includes scenarios such as center of gravity shift after fuel depletion and load changes after unloading at intermediate ports, to ensure that the loading layout results meet safety requirements in all navigation stages.

[0269] When the load distribution layout results are verified by dynamic simulation under full operating conditions, the load distribution optimization equipment determines the stability calculation results and fire compartment verification data based on the load distribution layout results.

[0270] The stability calculation results are a formal stability accounting report output by the digital twin model that conforms to the verification specifications of the classification society and can be used as legal materials for maritime filing.

[0271] Fire compartment verification data is a special verification report on the fire isolation requirements in the loading layout results. It includes whether the parking area of ​​the first type of vehicle is completely within the first loading compartment, whether the parking interval between the first type of vehicle and the second type of vehicle is not less than the preset isolation distance, whether the fire lane and the work lane are unobstructed, and whether the coverage of fire protection facilities meets the requirements. It is the core certificate of fire protection compliance.

[0272] The stability calculation results and fire protection zone verification data will be uploaded to the maritime supervision platform simultaneously to complete the compliance reporting process for voyage loading.

[0273] In one specific embodiment, optionally, the present invention constructs a full-process dynamic loading optimization system based on a shipping big data platform, a ship static model, an improved adaptive genetic algorithm, digital twin simulation verification, and ship-shore 5G / satellite communication technology. The overall architecture is divided into five layers: shore-side big data computing layer, ship-side on-site execution layer, data interaction and transmission layer, simulation verification layer, and operation instruction output layer.

[0274] Step 1: Multi-source heterogeneous data acquisition and standardization processing:

[0275] The data required for load optimization is divided into five categories, totaling 32 core parameters. After collection, the data undergoes standardized cleaning, deduplication, completion, and unit conversion.

[0276] (1) Basic inherent data of the ship, including: ship main dimensions (overall length, beam, depth, design draft), lightship weight and center of gravity coordinates, parameters of each deck (total deck area, maximum uniformly distributed load intensity (unit: t / m²), maximum concentrated load, boundary limits), fire compartment division data (zone number, size, location of enclosed partition, maximum number of vehicles, explosion-proof rating), ramp parameters (number, location, weight limit, width limit, height limit, maximum passage gradient, single vehicle passage time), ballast tank parameters, ship stability limit thresholds (minimum stability height, maximum roll angle, maximum pitch angle, upper limit of center of gravity height).

[0277] (2) Full data of transport vehicles, including: vehicle identification number (VIN), vehicle type (pure electric / plug-in hybrid / fuel / hazardous chemical), total vehicle weight, center of gravity height, external dimensions, battery pack parameters (weight, installation position, dimensions, load strength per unit area), unloading port code, loading and unloading priority, and special restrictions (overwidth / overheight / overweight / cannot be inverted).

[0278] (3) Route and navigation conditions data, including: port of departure, port of destination, sequence of intermediate ports of call, sea condition forecast along the route (roll extreme value, wind and wave level), fuel consumption plan (hourly fuel consumption, total fuel consumption, corresponding center of gravity offset), ballast water adjustment plan, and segment stability specifications (distinguishing between inland waterway segment and ocean segment).

[0279] (4) Port operation constraints data, including: time limits for each berth at each port of call, available time periods for operation, mandatory requirements for unloading sequence, number of simultaneous operations on ramps, collection and distribution capacity, and operation restrictions in severe weather.

[0280] (5) Compliance data, including: International Maritime Organization (IMO) International Maritime Stability Code (IS Code), China Classification Society Ro-Ro Ship Fire Prevention Code, New Energy Vehicle Transport Load Limit Standard, and Differentiated Compliance Requirements for Inland Waterways and Ocean Waterways.

[0281] Data standardization rules: weight units are uniformly converted to tons (t), size units are uniformly converted to meters (m), and the center of gravity coordinates are based on the ship's baseline, with the midship as the longitudinal origin and the bow as the positive direction; missing data is automatically filled in using the historical average of the same model, and abnormal data is automatically removed and an alert is triggered.

[0282] Step Two: Construction of a Dedicated Load Constraint Library for New Energy Vehicles

[0283] This step constructs a loading constraint library specifically for the transportation characteristics of new energy vehicles, divided into three levels: mandatory constraints, priority constraints, and prohibition constraints. Schemes that do not meet the mandatory constraints are directly judged as invalid schemes.

[0284] Mandatory constraints (if any one of them is not satisfied, the solution is invalid):

[0285] (1) Battery load symmetry constraint: The total weight deviation of vehicles on the port and starboard sides of the battery vehicle centralized parking deck shall not exceed 3%, and the left and right load deviation of a single battery centralized area shall not exceed 2%;

[0286] (2) Deck load strength constraint: The actual uniformly distributed load strength of the battery vehicle centralized parking area shall not exceed 85% of the maximum design load strength of the deck, with a 15% safety margin reserved;

[0287] (3) Center of gravity height constraints: For vehicles with a center of gravity height exceeding 1.2 m, the number of such vehicles parked shall not exceed 40% of the total number of parking spaces on the deck; the overall center of gravity height of the ship shall not exceed 90% of the standard limit.

[0288] (4) Fire compartment isolation constraints: All pure electric vehicles must be assigned to independent fire compartments, with a horizontal net distance of no less than 3 m from fuel vehicles. Parking across fire compartments is strictly prohibited, and vehicles of different explosion-proof levels are strictly prohibited from being parked together in the same compartment.

[0289] (5) Stability reserve constraints: The initial stability height (GM) calculated by the loading scheme shall not be less than 0.8 m, and the integrity stability criterion number K shall not be less than 1.0, which meets the requirements of the IMO integrity stability specification;

[0290] (6) Binding and fitting constraints: Over-height and over-weight battery vehicles must be parked in parking spaces with a binding point density that meets the requirements, and are prohibited from being parked in parking spaces without binding conditions.

[0291] Priority constraints (satisfying these constraints will improve the score of the proposed solution):

[0292] (1) Vehicles unloading at intermediate ports should be parked in the area around the corresponding ramps to shorten the loading and unloading route;

[0293] (2) Vehicles unloading at the same port should be parked together as much as possible to avoid scattered operations;

[0294] (3) Heavy vehicles and battery vehicles should be parked on the lower deck first, which helps to lower the overall center of gravity of the ship.

[0295] Prohibited constraints (absolutely prohibited):

[0296] (1) Battery-powered vehicles and fuel-powered vehicles are prohibited from being parked together in the same fire compartment;

[0297] (2) It is prohibited for vehicles unloading first to be blocked by vehicles unloading later, forming "dead parking spaces" that cannot be driven out.

[0298] Step 3: Multi-objective optimization function and adaptive weight model:

[0299] This step constructs a three-dimensional multi-objective optimization model, employing a weighted comprehensive evaluation method. The overall objective function is:

[0300] F = ω1×F1+ω2×F2+ω3×F3;

[0301] Where F1 is the cabin capacity utilization target, F2 is the navigation stability target, F3 is the loading and unloading efficiency target, F is the overall objective function, and each sub-objective function is normalized to the interval [0,1], with weights satisfying ω1+ω2+ω3=1.

[0302] Three sub-objective functions (all normalized):

[0303] (1) Cabin capacity utilization target F1: measured by the ratio of the actual number of vehicles to the theoretical maximum number of vehicles, guiding the algorithm to maximize the number of vehicles while satisfying the constraints;

[0304] (2) Navigation stability target F2: The normalized value of the difference between the initial stability height GM and the minimum value of the standard is used to guide the algorithm to maximize the stability reserve;

[0305] (3) Loading and unloading efficiency target F3: The algorithm is guided to minimize the loading and unloading path and eliminate vehicle blockage by measuring the inverse normalized value of the total loading and unloading travel path.

[0306] Weight adaptive adjustment rule:

[0307] (1) Severe sea conditions in the open ocean: ω1=0.2, ω2=0.7, ω3=0.1, priority is given to ensuring navigation stability. Severe sea conditions in the open ocean include wind and wave levels ≥6 or roll angle forecast ≥8°;

[0308] (2) Port expedited operations (berth duration ≤ 6 hours): ω1=0.2, ω2=0.2, ω3=0.6, prioritizing loading and unloading efficiency;

[0309] (3) Conventional near-ocean / inland waterway routes: ω1=0.4, ω2=0.3, ω3=0.3, balancing the three major objectives;

[0310] (4) Dedicated transportation for new energy vehicles: ω1=0.3, ω2=0.5, ω3=0.2, prioritizing the safety and stability of the load.

[0311] Step 4: Initial Scheme Population Generation and Initial Selection:

[0312] Based on standardized data and a constraint library, 100 initial loading schemes were generated using a combination of random generation and experience-guided methods. Each scheme included the parking deck, parking space number, port and starboard positions, and lashing requirements for each vehicle. Subsequently, an initial screening was performed using the constraint library to eliminate all invalid schemes that did not meet the mandatory constraints, retaining no fewer than 30 valid schemes for the iterative optimization phase.

[0313] Step 5: Iterative optimization using an improved adaptive genetic algorithm:

[0314] This step employs an improved adaptive genetic algorithm, which, compared to the standard genetic algorithm, adds a constraint penalty mechanism, adaptive crossover and mutation probability adjustment, and a navigation dynamic correction operator:

[0315] Fitness calculation: The overall objective function F value is used as the fitness function value; solutions that slightly violate priority constraints have their fitness scores reduced proportionally; solutions that violate mandatory constraints are directly eliminated.

[0316] Adaptive crossover probability: The crossover probability of high-fitness schemes is set to 0.5-0.7, and the crossover probability of low-fitness schemes is set to 0.8-0.9, in order to accelerate the spread of high-quality genes;

[0317] Adaptive mutation probability: 0.1 to 0.2 in the early stage of iteration (to expand the search range), and 0.01 to 0.05 in the later stage of iteration (to protect high-quality solutions).

[0318] Dynamic correction operator: In the fitness calculation, the fuel consumption center of gravity offset correction factor, the intermediate port loading and unloading load change correction factor and the sea state roll effect correction factor are preset to realize the dynamic adaptation of the scheme in advance.

[0319] Termination conditions: If the fitness value does not improve for 20 consecutive generations, or the number of iterations reaches 200, output the optimal solution with the highest fitness and 2 alternative solutions.

[0320] Step Six: Digital Twin 3D Visualization and Static Compliance Verification:

[0321] After the optimal solution is output, it is simultaneously imported into the ship's digital twin model to generate a 3D visualized loading plan, displaying the parking position of each vehicle, deck load distribution cloud map, ship center of gravity coordinates, fire compartment division, and loading and unloading route planning. Then, static compliance verification is performed (deck load strength, stability height, center of gravity height, fire compartment, and constraint satisfaction). If all criteria are met, the dynamic simulation phase begins; if any criterion is not met, the system automatically reverts to step five for re-optimization.

[0322] Step 7: Full-condition dynamic simulation verification:

[0323] This step simulates the dynamic operating environment of a ship throughout its entire voyage and under all operating conditions, and verifies the loading scheme under the following five types of operating conditions:

[0324] (1) Normal navigation conditions: Verification of floating state, stability, roll and pitch at rated speed and under normal wind and waves;

[0325] (2) Extreme sea state conditions: Stability verification under extreme conditions of wind and wave level 10, roll angle 15°, and pitch angle 5° to ensure that the ship does not capsize and the stability does not fail;

[0326] (3) Fuel consumption conditions: The center of gravity shift is checked when the fuel consumption reaches the lowest value throughout the entire process to ensure that the stability meets the standard throughout the entire process;

[0327] (4) Loading and unloading conditions at intermediate ports: load distribution, center of gravity change and stability verification of the remaining voyage after unloading at each intermediate port;

[0328] (5) Turning and braking conditions: load stability verification under full rudder turning and emergency braking conditions to ensure that the vehicle does not shift abnormally.

[0329] Solutions that fail verification are automatically marked as problematic. After adjusting the constraint parameters, the algorithm is iterated again until all operating conditions are verified.

[0330] Step 8: Real-time dynamic correction during navigation:

[0331] After the vessel departs, the system enters dynamic correction mode, refreshing real-time data (including draft, center of gravity, fuel consumption, ballast water status, real-time sea state, and changes in loading and unloading plans at intermediate ports) every 30 minutes. The system will automatically initiate local dynamic optimization when any of the following conditions occur:

[0332] The ship's center of gravity shifted by more than 0.1 m;

[0333] Stability is reduced to below 95% of the standard limit;

[0334] The real-time sea state and wave level is 2 or more levels higher than the forecast value;

[0335] The loading / unloading schedule or unloading sequence at intermediate ports has changed;

[0336] The alarm indicating loose vehicle fasteners caused a change in local load distribution.

[0337] The dynamic optimization calculation time is no more than 30 seconds, and the correction plan is automatically sent to the bridge and deck crew terminals without manual intervention.

[0338] Step Nine: Real-time Loading and Unloading Optimization at Intermediate Ports:

[0339] Twelve hours before a vessel arrives at an intermediate port, the system automatically imports the unloading list, loading list, and berth operation duration of the intermediate port. Based on the current real-time vessel status, the system re-optimizes the loading and generates an integrated optimization plan that includes "unloading sequence - loading position - load adjustment - ballast water coordination". This ensures that the unloading path is unobstructed, the loading plan meets stability requirements, and the stability meets the standards immediately after departure.

[0340] Step 10: Issuance and Execution of Ship-Shore Integration Commands:

[0341] The final payload allocation plan was simultaneously transmitted to the three execution terminals via 5G / satellite communication links:

[0342] (1) Ship bridge terminal: Displays key safety parameters such as the ship's center of gravity, stability height, and load distribution for navigation decision-making reference;

[0343] (2) Deck work team terminal: Displays the parking location of each vehicle, boarding order, lashing requirements and fire compartment guidance;

[0344] (3) Port dispatch center terminal: Displays unloading sequence, ramp operation plan and collection and distribution coordination requirements.

[0345] The system automatically generates a loading scheme compliance report, stability calculation sheet, and fire protection zone verification report, which can be directly used for ship inspection and maritime filing.

[0346] In one specific embodiment, optionally, the complete execution steps include:

[0347] Step 1: Data Collection and Standardization

[0348] The system accesses all ship parameters, details of 3,420 vehicles, route sea condition forecasts, and port operation constraint data (a total of 32 parameters), and completes standardized cleaning; it automatically completes the missing center of gravity data of 12 vehicles (using the historical average of the same vehicle type), and removes 3 abnormal sea condition data, achieving 100% data integrity.

[0349] Step 2: Loading the constraint library and setting weights:

[0350] The system incorporates 6 mandatory constraints, 3 priority constraints, and 2 prohibition constraints for new energy vehicles. Due to the forecast of severe sea conditions in the strait through which the route passes, the adaptive weights are automatically set as follows: cabin capacity weight ω1=0.2, stability weight ω2=0.7, and efficiency weight ω3=0.1, prioritizing navigation safety.

[0351] Step 3: Initial scheme generation and iterative optimization:

[0352] 100 initial schemes are generated. After constraint screening, 30 effective schemes are retained for iteration. After 200 iterations (118 seconds of computation time), the optimal loading scheme is output, with a comprehensive fitness score of 98.7 out of 100.

[0353] Step 4: Core parameters of the optimal solution:

[0354] (1) Fire compartment configuration: All 1,980 pure electric vehicles are assigned to two dedicated explosion-proof and fire-proof compartments and are parked in the lower level of decks 1 and 2, with 990 vehicles parked on each side. The load deviation on the left and right sides is 2.07%, which meets the mandatory constraint of ≤3%. Fuel vehicles and plug-in hybrid vehicles are parked on deck 3 and are separated from electric vehicles in a closed compartment. The measured horizontal distance is 4.2 m, which meets the specification requirement of ≥3 m.

[0355] (2) Load control: The actual uniformly distributed load intensity in the battery vehicle area of ​​decks 1# and 2# is 82.3% of the design maximum value, which meets the safety margin requirement of ≤85% and there is no local overload;

[0356] (3) Stability parameters: The overall center of gravity height of the ship is 6.82 m, the initial metacentric height GM=1.12 m (the minimum value in the standard is 0.8 m), and the complete stability criterion number K=1.32, which meets the requirements of the IMO standard;

[0357] (4) Loading and unloading adaptation: 360 vehicles unloaded at Port A are all parked within 50 m of the stern ramp; 480 vehicles unloaded at Port B are all parked in the area around the bow ramp; there are no vehicles blocking the way, and the loading and unloading route is shortened by about 40% compared to the random loading scheme.

[0358] Step 5: Full-condition simulation verification:

[0359] The simulation covered the entire navigation process, including extreme conditions such as a Force 7 gale in the strait, a roll of 12°, fuel consumption throughout the journey, and continuous loading and unloading at intermediate ports. Verification results: Under all conditions, the initial stable height was no less than 0.9 m, with no center of gravity exceeding the limit, no load imbalance, and no stability failure; all verifications were passed.

[0360] Step 6: Dynamic corrections during navigation:

[0361] While navigating a strait, the measured wind and waves reached Force 7, with a roll angle of 11°. The system automatically refreshed the data and performed dynamic fine-tuning every 30 minutes, making three consecutive adjustments involving the positions of 18 vehicles. Throughout the adjustments, the stability height remained above 1.0 m, and no safety incidents occurred. Twelve hours before arriving at Ningbo Port, the system automatically generated an optimized intermediate port plan, with actual port operations lasting 4.2 hours, approximately 32% shorter than the baseline reference plan.

[0362] Step 7: Comparison of Implementation Results:

[0363] (1) The stability reserve (initial stability height GM) is increased by about 28% compared with the traditional loading scheme of the same type of ship, which meets the IMO specifications;

[0364] (2) The cabin capacity utilization rate was 92.4%, which is about 13.6% higher than the historical benchmark level of the same batch;

[0365] (3) The average loading and unloading time at intermediate ports is reduced by about 34%, which corresponds to a reduction in port waiting costs per voyage;

[0366] (4) The fire compartment configuration meets 100% of the inspection requirements of the China Classification Society;

[0367] (5) No manual intervention is required for the loading scheme throughout the process, and the workload related to loading on the driver's cab is significantly reduced compared to the traditional method.

[0368] For example, the implementation object is a 6200-car-capacity ocean-going car carrier with a length of 199.9 m, a beam of 38 m, a depth of 14.5 m, and a design draft of 8.2 m.

[0369] The vehicle main deck has three decks, with a maximum uniformly distributed load strength of approximately 3.2 t / m² per deck (total deck area is approximately 3900 m², corresponding to a maximum total load of approximately 12500 t); it is divided into eight fire compartments, two of which are dedicated explosion-proof fire compartments for pure electric vehicles; it has one ramp at the bow and one at the stern; and the design initial stability height GM is not less than 0.8 m.

[0370] The route includes a combined transport route connecting ports C, A, B, and D, with a total sailing time of approximately 12 days. The route from port C to port A is an inland waterway segment, while the route from port A to port D is an ocean segment. The route passes through Strait A, with forecasted wind and wave levels of 5 to 7 and extreme roll values ​​of 8° to 12°.

[0371] Transport vehicles: A total of 3,420 vehicles were loaded, including 1,980 pure electric vehicles, 560 plug-in hybrid vehicles, and 880 traditional fuel vehicles. Intermediate port loading and unloading: 360 vehicles (pure electric) were unloaded at Port A, 480 vehicles (pure electric and fuel vehicles) were unloaded at Port B, and the remaining vehicles were unloaded at the Port of Singapore.

[0372] Operational constraints: The duration of operation at Port A berth shall not exceed 6 hours, and at Port B it shall not exceed 8 hours. The entire process must comply with IMO stability and fire protection regulations.

[0373] like Figure 9As shown in the figure, this application embodiment also provides a roll-on / roll-off ship loading optimization device 900, including: a data acquisition module 902, used to acquire basic inherent data of the target ship and vehicle data of multiple transport vehicles to be loaded into the cargo hold of the target ship, the transport vehicles including a first type of vehicle and a second type of vehicle, the first type of vehicle being a new energy vehicle equipped with a battery pack, and the second type of vehicle being a fuel vehicle, the basic inherent data including deck load strength data of each deck and fire protection zone planning data, the vehicle data including battery pack parameters of the first type of vehicle; a battery load module 904, used to determine the battery load distribution data corresponding to each deck area in the cargo hold according to the battery pack parameters and deck load strength data; a zone planning module 906, used to determine the first loading zone corresponding to the first type of vehicle, the second loading zone corresponding to the second type of vehicle, and the preset isolation distance between the first type of vehicle and the second type of vehicle according to the fire protection zone planning data; and a loading constraint module 908, used for... Based on battery load distribution data, the first loading zone, the second loading zone, and the preset isolation distance, the loading constraint characteristics are determined. The loading constraint characteristics include mandatory constraints that must be met, including at least battery load symmetry constraints, deck load strength constraints, and fire compartment isolation constraints. The scheme planning module 910 is used to determine multiple initial loading schemes corresponding to the target vessel based on the loading constraint characteristics and basic inherent data. Each initial loading scheme includes loading position data of multiple transport vehicles. The function construction module 912 is used to obtain a multi-objective optimization function. The multi-objective optimization function takes cargo capacity utilization, navigation stability, and loading and unloading efficiency as sub-objectives, and assigns weight coefficients to each sub-objective. The loading optimization module 914 is used to iteratively optimize multiple initial loading schemes based on the multi-objective optimization function under the premise of meeting the mandatory constraints, and determine at least one loading layout result. The loading layout result is used to indicate the loading layout of multiple transport vehicles in the cargo hold of the target vessel.

[0374] like Figure 10 As shown, this application embodiment also provides an electronic device 1000, including a processor 1110, a memory 1109, and a program or instructions stored in the memory 1109 and executable on the processor 1110. When the program or instructions are executed by the processor 1110, they implement the various processes of the above-described roll-on / roll-off ship loading optimization method and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0375] Optionally, the processor 1110 is used to acquire basic inherent data of the target vessel and vehicle data of multiple transport vehicles to be loaded into the cargo hold of the target vessel.

[0376] Optionally, the processor 1110 is also used to determine the battery load distribution data corresponding to each deck area in the cargo hold based on the battery pack parameters and deck load intensity data.

[0377] Optionally, the processor 1110 is also used to determine, based on the fire compartment planning data, the first loading compartment corresponding to the first type of vehicle, the second loading compartment corresponding to the second type of vehicle, and the preset isolation distance between the first type of vehicle and the second type of vehicle.

[0378] Optionally, the processor 1110 is also used to determine the load constraint characteristics based on the battery load distribution data, the first loading zone, the second loading zone, and the preset isolation distance;

[0379] Optionally, the processor 1110 is also configured to determine multiple initial loading schemes corresponding to the target vessel based on loading constraint characteristics and basic inherent data;

[0380] Optionally, the processor 1110 is also used to obtain a multi-objective optimization function, which takes cabin capacity utilization, navigation stability and loading and unloading efficiency as sub-objectives, and assigns weight coefficients to each sub-objective.

[0381] Optionally, the processor 1110 is also configured to iteratively optimize multiple initial loading schemes according to a multi-objective optimization function under the premise of satisfying mandatory constraints, and determine at least one loading layout result, the loading layout result being used to indicate the loading layout of multiple transport vehicles in the cargo hold of the target vessel.

[0382] The memory 1109 can be used to store software programs and various data. The memory 1109 may primarily include a first storage area for storing programs or instructions and a second storage area for storing data. The first storage area may store the operating system, application programs or instructions required for at least one function (such as sound playback, image playback, etc.). Furthermore, the memory 1109 may include volatile memory or non-volatile memory, or both. The non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus RAM (DRRAM). The memory 1109 in this embodiment includes, but is not limited to, these and any other suitable types of memory.

[0383] In this invention, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance; the term "multiple" refers to two or more unless otherwise explicitly defined. The terms "install," "connect," "link," and "fix" should be interpreted broadly. For example, "connect" can be a fixed connection, a detachable connection, or an integral connection; "link" can be a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0384] In the description of this invention, it should be understood that the terms "upper," "lower," "left," "right," "front," "rear," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or unit referred to must have a specific orientation or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0385] In the description of this specification, the terms "one embodiment," "some embodiments," "specific embodiment," etc., refer to specific features, structures, materials, or characteristics described in connection with an embodiment or example that are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0386] The above are merely preferred embodiments of the present invention and are not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for optimizing the loading of roll-on / roll-off (Ro-Ro) vessels, characterized in that, Performed by load optimization equipment, including: Acquire basic inherent data of the target vessel and vehicle data of multiple transport vehicles to be loaded into the cargo hold of the target vessel. The transport vehicles include a first type of vehicle and a second type of vehicle. The first type of vehicle is a new energy vehicle equipped with a battery pack, and the second type of vehicle is a fuel vehicle. The basic inherent data includes deck load strength data and fire compartment planning data of each deck. The vehicle data includes battery pack parameters of the first type of vehicle. The battery load distribution data for each deck area in the cargo hold is determined based on the battery pack parameters and the deck load strength data. Based on the fire protection zone planning data, determine the first loading zone corresponding to the first type of vehicle, the second loading zone corresponding to the second type of vehicle, and the preset isolation distance between the first type of vehicle and the second type of vehicle; The loading constraint features are determined based on the battery load distribution data, the first loading zone, the second loading zone, and the preset isolation distance. The loading constraint features include mandatory constraints that must be met. The mandatory constraints include at least battery load symmetry constraints, deck load strength constraints, and fire compartment isolation constraints. Based on the loading constraint characteristics and the basic inherent data, multiple initial loading schemes corresponding to the target vessel are determined, and each initial loading scheme includes loading position data of multiple transport vehicles. Obtain a multi-objective optimization function, wherein the multi-objective optimization function has cabin capacity utilization, navigation stability and loading and unloading efficiency as sub-objectives, and assigns weight coefficients to each of the sub-objectives respectively; Under the premise of satisfying the mandatory constraints, the multiple initial loading schemes are iteratively optimized according to the multi-objective optimization function to determine at least one loading layout result, which is used to indicate the loading layout of the multiple transport vehicles in the cargo hold of the target vessel.

2. The method for optimizing the loading of roll-on / roll-off ships according to claim 1, characterized in that, The step of determining the load constraint characteristics based on the battery load distribution data, the first loading zone, the second loading zone, and the preset isolation distance includes: The upper limit of the deck load is determined based on the battery load distribution data, and the upper limit of the deck load does not exceed a preset proportion of the deck load strength data of the corresponding deck. The upper limit of the weight difference between the port and starboard sides in the battery load concentration area is determined based on the battery load distribution data. The center of gravity height constraint is determined based on the installation location of the battery pack and the overall center of gravity height of the vehicle in the vehicle data. The center of gravity height constraint includes the upper limit of the proportion of transport vehicles whose center of gravity height exceeds a preset height threshold on their respective decks, and the upper limit of the overall center of gravity height of the target ship. The minimum distance between the first type of vehicle and the second type of vehicle is determined based on the preset isolation distance, and the first type of vehicle is restricted to parking in an independent fire compartment; The stability reserve constraint is determined based on the stability limit threshold of the target vessel. The stability reserve constraint includes the lower limit of the initial stability height and the lower limit of the complete stability criterion number. The permitted parking area for each type of transport vehicle is determined based on the first loading zone and the second loading zone; The loading constraint characteristics are determined based on the upper limit of the deck load, the upper limit of the weight difference between the port and starboard sides, the center of gravity height constraint, the minimum interval value, and the area range.

3. The method for optimizing the loading of roll-on / roll-off ships according to claim 1, characterized in that, The step of determining multiple initial loading schemes corresponding to the target vessel based on the loading constraint characteristics and the basic inherent data includes: The available parking spaces within the cargo hold are determined based on the aforementioned inherent data. Based on the area range corresponding to each of the transport vehicles in the loading constraint features, a candidate parking space that conforms to the corresponding loading zone is assigned to each transport vehicle; Within the range of the candidate parking spaces, multiple sets of correspondences between the transport vehicles and parking spaces are randomly generated to obtain multiple initial loading schemes; Based on the mandatory constraints, a preliminary screening is performed on multiple initial loading schemes, and initial loading schemes that do not meet the mandatory constraints are eliminated.

4. The method for optimizing the loading of roll-on / roll-off ships according to claim 1, characterized in that, After obtaining the multi-objective optimization function, the roll-on / roll-off ship loading optimization method further includes: Obtain the route and sea state data and port call duration data of the target vessel; The forecast wind and wave level is determined based on the sea state data of the aforementioned route. The urgency of the operation is determined based on the port call duration data. When the predicted wind and wave level is not less than the preset sea state level threshold, the weight coefficient corresponding to the navigation stability is increased, and the weight coefficients corresponding to the other sub-targets are decreased accordingly. When the port call operation time data is not greater than the preset time threshold, the weight coefficient corresponding to the loading and unloading efficiency is increased, and the weight coefficients corresponding to the other sub-targets are decreased accordingly. The multi-objective optimization function is updated based on the adjusted weighting coefficients; The sum of the weight coefficients corresponding to each of the sub-objectives is a constant.

5. The method for optimizing the loading of roll-on / roll-off ships according to claim 1, characterized in that, The loading constraint features also include priority constraints. Under the premise of satisfying the mandatory constraints, the iterative optimization of multiple initial loading schemes according to the multi-objective optimization function to determine at least one loading layout result includes: Using the function value of the multi-objective optimization function as the fitness, the improved adaptive genetic algorithm iteratively optimizes multiple initial loading schemes. In each iteration, loading schemes that do not meet the mandatory constraints are eliminated, and the fitness of the retained loading schemes is modified according to the priority constraints. The crossover probability is adaptively adjusted based on the fitness of each loading scheme, so that the crossover probability corresponding to the loading scheme with fitness higher than the preset fitness threshold is less than the crossover probability corresponding to the loading scheme with fitness lower than the preset fitness threshold. The mutation probability is adaptively adjusted according to the iteration process, so that the mutation probability in the early stage of the iteration is greater than the mutation probability in the later stage of the iteration. A dynamic correction operator is introduced into the fitness calculation, which includes at least one of a fuel consumption center of gravity offset correction factor and a mid-port loading and unloading load change correction factor. When a preset termination condition is met, the loading scheme with the highest fitness is output as the loading layout result. The preset termination condition includes no improvement in fitness for a consecutive preset number of generations or the number of iterations reaching a preset upper limit.

6. The method for optimizing the loading of roll-on / roll-off ships according to claim 1, characterized in that, The method for optimizing the loading of roll-on / roll-off ships also includes: During the navigation of the target vessel, real-time navigation data is acquired at time intervals not exceeding a preset period. The real-time navigation data includes at least one of the following: real-time draft, real-time center of gravity, fuel consumption, ballast water status, and real-time sea state. Dynamic optimization triggering features are determined based on the real-time navigation data and the load layout results. The dynamic optimization triggering features are defined based on the battery load distribution data and the load constraint features. When the dynamic optimization triggering feature meets the preset triggering conditions, the loading layout result is locally iteratively corrected according to the real-time navigation data and the loading constraint feature to obtain the iterative loading layout result; Update the loading layout results based on the iterative loading layout results; The preset triggering conditions include at least one of the following: the center of gravity offset of the target vessel exceeds a preset offset threshold, the real-time initial stability height is lower than a preset stability threshold, and the wind and wave level corresponding to the real-time sea state is higher than the predicted wind and wave level by no less than a preset difference.

7. The method for optimizing the loading of roll-on / roll-off ships according to claim 6, characterized in that, The method for optimizing the loading of roll-on / roll-off ships also includes: In response to a berthing instruction at a transit port, obtain the current vessel status characteristics, the unloading list and loading list corresponding to the transit port; Based on the current ship status characteristics and the unloading list, and in conjunction with the preset isolation distance, the stability prediction characteristics after unloading are determined; Based on the stability prediction characteristics, the loading list and the stowage constraint characteristics, the vehicle parking positions in the remaining cargo space in the cargo hold are re-iterated and solved to determine the stowage layout results at intermediate ports. Based on the intermediate port loading layout results, loading and unloading operation instructions corresponding to the intermediate port are generated, and the loading and unloading operation instructions are synchronized to the ship's bridge terminal, deck work team terminal and port dispatch center terminal through the ship-shore communication link.

8. The method for optimizing the loading of roll-on / roll-off (Ro-Ro) vessels according to any one of claims 1 to 7, characterized in that, After iteratively optimizing multiple initial loading schemes according to the multi-objective optimization function to determine at least one loading layout result, the roll-on / roll-off ship loading optimization method further includes: The digital twin model is used to determine the three-dimensional visualization data corresponding to the loading layout results. The three-dimensional visualization data includes vehicle parking locations, deck load distribution cloud maps, ship center of gravity coordinates, fire protection zone division, and loading and unloading route planning. Based on the three-dimensional visualization data, a static compliance check is performed on the loading layout result to obtain the static compliance check result. Under the condition that the static compliance verification result indicates compliance, the loading layout result is subjected to full-condition dynamic simulation verification. The full-condition dynamic simulation verification includes at least extreme sea state verification. The verification parameters of the extreme sea state verification include a roll angle not less than a preset roll angle threshold and a wind and wave level not less than a preset wind and wave level threshold. When the load distribution layout result passes the full-condition dynamic simulation verification, the stability calculation result and fire compartment verification data are determined based on the load distribution layout result.

9. A roll-on / roll-off ship loading optimization device, characterized in that, include: The data acquisition module is used to acquire basic inherent data of the target ship and vehicle data of multiple transport vehicles to be loaded into the cargo hold of the target ship. The transport vehicles include a first type of vehicle and a second type of vehicle. The first type of vehicle is a new energy vehicle equipped with a battery pack, and the second type of vehicle is a fuel vehicle. The basic inherent data includes deck load strength data and fire protection zone planning data of each deck. The vehicle data includes battery pack parameters of the first type of vehicle. The battery load module is used to determine the battery load distribution data corresponding to each deck area in the cargo hold based on the battery pack parameters and the deck load intensity data. The zoning planning module is used to determine, based on the fire protection zoning planning data, the first loading zone corresponding to the first type of vehicle, the second loading zone corresponding to the second type of vehicle, and the preset isolation distance between the first type of vehicle and the second type of vehicle; The loading constraint module is used to determine loading constraint features based on the battery load distribution data, the first loading zone, the second loading zone, and the preset isolation distance. The loading constraint features include mandatory constraints that must be met. The mandatory constraints include at least battery load symmetry constraints, deck load strength constraints, and fire compartment isolation constraints. The scheme planning module is used to determine multiple initial loading schemes corresponding to the target vessel based on the loading constraint characteristics and the basic inherent data. Each initial loading scheme includes loading position data of multiple transport vehicles. The function construction module is used to obtain a multi-objective optimization function, wherein the multi-objective optimization function takes cabin capacity utilization, navigation stability and loading and unloading efficiency as sub-objectives, and assigns weight coefficients to each of the sub-objectives respectively; The loading optimization module is used to iteratively optimize multiple initial loading schemes according to the multi-objective optimization function under the premise of satisfying the mandatory constraints, and determine at least one loading layout result. The loading layout result is used to indicate the loading layout of the multiple transport vehicles in the cargo hold of the target vessel.

10. An electronic device, characterized in that, It includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the roll-on / roll-off ship loading optimization method as described in any one of claims 1 to 8.

Images