Intelligent scheduling method and system for oil tankers based on multi-objective evaluation and meta-heuristic search

Abstract

This invention discloses an intelligent scheduling method and system for oil tankers based on multi-objective evaluation and metaheuristic search, relating to the field of oil waterway transportation scheduling technology. The method acquires daily basic data on oil tanker scheduling and a basic set of candidate vessels as a feasible search space; constructs a scheduling business objective evaluation function and a cost objective evaluation function, normalizes both, and then constructs a multi-objective fusion evaluation function, dynamically adjusting the objective weights according to the daily scheduling scenario; constructs a daily scheduling scheme encoding using transportation demand as the encoding position and vessel name as the gene value; iteratively optimizes within the feasible search space using a metaheuristic optimization algorithm; uses the multi-objective fusion evaluation value as the basis for accepting candidate schemes and updating schemes; and outputs the historical best scheme as the target daily scheduling scheme. This invention can generate daily oil tanker scheduling schemes that balance business acceptability, economy, and verifiability.

Description

Technical Field

[0001] This application relates to the field of oil waterway transportation scheduling and automatic ship dispatching technology, specifically to an intelligent scheduling method and system for oil tankers based on multi-objective evaluation and metaheuristic search. Background Technology

[0002] Daily vessel allocation is a crucial step in the scheduling of oil tanker waterway transportation. Its task is to allocate suitable vessels to the day's transportation needs after obtaining daily transportation demand, vessel dynamics, port restrictions, and liner information. The allocation process must simultaneously consider multiple objectives, including meeting demand, prioritizing urgent orders, tonnage matching, arrival time, liner priority, protecting scarce demand, and transportation costs.

[0003] Existing daily tanker scheduling methods mainly fall into two categories. One category is rule-based sequencing or human experience-based decision-making methods, which can effectively inherit the business experience of scheduling personnel. However, these methods typically lack a unified quantitative evaluation mechanism, making it difficult to compare the overall advantages and disadvantages of different schemes in terms of business objectives and costs on the same scale, and they lack a systematic cost evaluation process. The other category is cost optimization-based mathematical programming methods, which can effectively reduce transportation costs. However, if cost minimization is the sole objective, it may weaken business preferences such as prioritizing urgent orders, liner services, and protecting scarce demand, resulting in schemes that, while having lower costs, lack sufficient business acceptability.

[0004] The aforementioned vessel scheduling objectives and cost objectives have different dimensions, optimization directions, and levels of importance, and cannot be directly added together for comparison. Furthermore, the daily vessel scheduling scenario varies significantly from day to day, making fixed weights unsuitable for adapting to changing objective preferences under different scenarios. Existing multi-objective optimization methods, in the daily vessel scheduling scenario for oil tankers, fail to fully integrate the characteristics of vessel scheduling operations, such as demand fulfillment, tonnage matching, arrival time, liner priority, and protection against scarce demand. Moreover, the optimization results lack a direct correspondence with actual vessel names, making it difficult for scheduling personnel to understand and verify.

[0005] Therefore, there is an urgent need for a daily vessel scheduling method for oil tankers that can uniformly model, normalize, and dynamically integrate vessel scheduling business objectives and cost objectives, and embed multi-objective fusion evaluation into a metaheuristic optimization process, so as to generate a daily vessel scheduling plan that takes into account business acceptability, economy, and verifiability. Summary of the Invention

[0006] This application aims to provide a method and system for intelligent scheduling of oil tankers based on multi-objective evaluation and meta-heuristic search, in order to solve at least one technical problem existing in the prior art.

[0007] Firstly, the technical solution of this application provides an intelligent scheduling method for oil tankers based on multi-objective evaluation and meta-heuristic search, including: Obtain basic data on daily vessel scheduling for oil tankers and a basic candidate vessel set, and use the basic candidate vessel set as the feasible search space; Construct a boat scheduling business objective evaluation function, which is used to evaluate the performance of the daily boat scheduling plan on at least one boat scheduling business objective; Construct a cost target evaluation function, which is used to evaluate the performance of the daily boat dispatching scheme on at least one cost target; The evaluation function for the boat dispatching business objective and the evaluation function for the cost objective are normalized to obtain normalized evaluation values ​​for the boat dispatching business objective and the cost objective. Based on the normalized evaluation value of the boat dispatching business target and the normalized evaluation value of the cost target, a multi-objective fusion evaluation function is constructed, and the target weights in the multi-objective fusion evaluation function are dynamically adjusted according to the boat dispatching scenario of the day. Using transportation demand as the encoding location and ship name as the gene value, a daily ship scheduling scheme is encoded to generate an initial daily ship scheduling scheme. Starting from this initial scheme, iterative optimization is performed: in each iteration, a metaheuristic optimization algorithm is invoked to generate candidate daily ship scheduling schemes within the feasible search space. The multi-objective fusion evaluation value of the candidate schemes is calculated based on the multi-objective fusion evaluation function. Based on the multi-objective fusion evaluation value, it is determined whether to accept the candidate scheme, and the current daily ship scheduling scheme and the historical best daily ship scheduling scheme are updated. Iteration stops when a preset termination condition is met, and the historical best daily ship scheduling scheme is taken as the target daily ship scheduling scheme. Output the target day's boat scheduling plan and generate the corresponding boat scheduling business target score, cost target score, multi-target fusion score, target weight, cost change, reasons for non-matching, and plan description information.

[0008] Some solutions describe intelligent scheduling methods for oil tankers based on multi-objective evaluation and meta-heuristic search, wherein the scheduling business objectives include at least one of the following: demand satisfaction objective, priority for urgent orders objective, tonnage matching objective, arrival time objective, liner priority objective, and scarce demand protection objective. The cost targets include at least one of the following: owned vessel costs, social vessel costs, port operation costs, vessel occupancy costs, and mismatch penalty costs; The parameters for the daily vessel allocation scenario include at least one of the following: number of urgent orders, proportion of urgent orders, number of candidate vessels in short supply, proportion of social vessels, and cost pressure indicators. The cost pressure indicators are determined based on at least one of the following: average cost of social vessel candidates, proportion of owned vessels available, cost of non-matching penalty, budget cost threshold, and total daily demand.

[0009] Some solutions describe intelligent scheduling methods for oil tankers based on multi-objective evaluation and metaheuristic search. The step of acquiring daily scheduling data for oil tankers and a basic candidate vessel set, and using the basic candidate vessel set as a feasible search space, includes: Obtain the daily basic data of oil tanker scheduling, which includes daily dispatch data, daily oil transportation demand data, and port restriction information; Extract vessel status information from the daily vessel dispatch data; Extract the type and planned quantity of oil products from the daily oil transportation demand data; The set of available vessels for the day is determined based on the vessel status information; Based on at least one of the oil type, the planned volume, and the port restriction information, a basic feasibility assessment is made of the vessels and transportation demand in the available vessel set for the day. Ships that meet the basic feasibility judgment conditions are added to the candidate ship set corresponding to the transportation needs. The candidate ship sets of each transportation need constitute the basic candidate ship set, which serves as the feasible search space.

[0010] Some solutions describe intelligent scheduling methods for oil tankers based on multi-objective evaluation and metaheuristic search. The construction of a scheduling business objective evaluation function, used to evaluate the performance of a daily scheduling plan on at least one scheduling business objective, includes: Build requirements to meet goals The demand satisfaction target is used to characterize the degree of demand satisfaction of the daily ship scheduling scheme by the number or proportion of unmatched transportation demand. Build a priority target for urgent orders The urgent order priority target is used to evaluate the matching of urgent order transportation demand in the daily ship scheduling plan; Construct tonnage matching target The tonnage matching target is used to evaluate the deviation between the corresponding oil cargo carrying capacity of the ships in the daily ship dispatching plan and the planned transportation demand. Establish arrival time target The arrival time target is used to evaluate the degree of fit between the expected arrival time of ships at the target port of departure and the time requirements of transportation demand in the daily ship scheduling plan; Building Liner Priority Goals The liner priority target is used to evaluate the matching between liner vessels and liner demand in the daily vessel scheduling plan; Build scarce demand protection targets The scarce demand protection target is used to evaluate whether transportation demand with fewer than a preset threshold in the daily ship scheduling scheme is prioritized for matching. Based on at least one of the following objectives: demand fulfillment, priority for urgent orders, tonnage matching, arrival time, priority for liner services, and protection of scarce demand, construct the vessel scheduling business objective evaluation function. The evaluation function for the gantry operation target. The calculation expression is: ; in, Indicates the daily boat dispatch plan, The weights of each objective in the boat dispatching business are given, and the sum of the weights of each objective is 1.

[0011] Some solutions describe intelligent scheduling methods for oil tankers based on multi-objective evaluation and metaheuristic search. The construction of a cost-objective evaluation function, used to evaluate the performance of daily tanker scheduling schemes on at least one cost objective, includes: Calculate the cost of owning vessels when fulfilling transportation needs using owned vessels; Calculate the social vessel costs incurred when social vessels fulfill transportation demands; Calculate the mismatch penalty cost corresponding to the mismatched transportation demand; Based on the owned vessel cost, the social vessel cost, and the mismatch penalty cost, a cost objective evaluation function is constructed. The cost target evaluation function The calculation expression is: ; in, This is a collection of transportation needs for the day. For the assembly of ships, Indicates the first The daily oil transportation demand, Indicates the first Candidate ships, For the first The ship carried out the first Matching costs for transportation demand For the decision variable of arranging boats, when the first The transportation demand is determined by the first When the ship is in operation ,otherwise , For the first The penalty cost for mismatched transportation demand. For unmatched variables, when the first When the transport demand does not match the number of vessels ,otherwise .

[0012] Some solutions describe intelligent scheduling methods for oil tankers based on multi-objective evaluation and metaheuristic search. The normalization process for the scheduling business objective evaluation function and the cost objective evaluation function, to obtain normalized scheduling business objective evaluation values ​​and normalized cost objective evaluation values, includes: Obtain the business target value of each sub-target in the boat dispatching business target evaluation function, and the cost target value in the cost target evaluation function; Determine whether each objective is a minimization objective or a maximization objective; For the minimization objective, normalization is performed using the following expression:

[0013] in, For the first The goal is to schedule a boat trip in Japan. The normalized target value is below. For the first The goal is to schedule a boat trip in Japan. The target value below, and These are the first and second candidate solutions in the current candidate solution set. The minimum and maximum values ​​of each objective. To prevent constants with a denominator of zero; For a maximization objective, it is converted into a minimization objective and then normalized. We obtained the normalized evaluation values ​​for the shunting business objectives and the normalized evaluation values ​​for the cost objectives.

[0014] Some solutions describe intelligent scheduling methods for oil tankers based on multi-objective evaluation and metaheuristic search. The method involves constructing a multi-objective fusion evaluation function based on the normalized scheduling business objective evaluation value and the normalized cost objective evaluation value. The objective weights in the multi-objective fusion evaluation function are then dynamically adjusted according to the daily scheduling scenario. This includes: Based on the normalized evaluation value of the boat dispatching business target and normalized cost target evaluation value Construct a multi-objective fusion evaluation function Its calculation expression is: ; in, Weighting the integration of boat dispatching business objectives and cost objectives; Based on the parameters of the boat scheduling scenario for the day, adjust the internal weights of at least one sub-objective in the boat scheduling business objective evaluation function, and adjust the fusion weights between the boat scheduling business objective and the cost objective. Specifically, when the number or proportion of urgent orders exceeds a preset threshold, the internal weight corresponding to the urgent order priority target is increased; when the number of candidate vessels with scarce demand exceeds a preset threshold, the internal weight corresponding to the scarce demand protection target is increased; when the proportion of social vessels or cost pressure indicators exceed a preset threshold, the fusion weight is decreased. The value of is determined to increase the weight of cost targets in the fusion evaluation.

[0015] Some solutions describe an intelligent scheduling method for oil tankers based on multi-objective evaluation and metaheuristic search. The method involves constructing a daily scheduling scheme encoding using transportation demand as the encoding location and the ship name as the gene value, generating an initial daily scheduling scheme. Starting from this initial daily scheduling scheme, iterative optimization is performed: in each iteration, a metaheuristic optimization algorithm is invoked to generate candidate daily scheduling schemes within the feasible search space. The multi-objective fusion evaluation value of the candidate daily scheduling scheme is calculated based on the multi-objective fusion evaluation function. Based on the multi-objective fusion evaluation value, it is determined whether to accept the candidate daily scheduling scheme, and the current daily scheduling scheme and the historical best daily scheduling scheme are updated. When a preset termination condition is met, the iteration stops, and the historical best daily scheduling scheme is taken as the target daily scheduling scheme. This includes: Using each transport demand of the day as a coding position, and the ship name assigned to each transport demand as the gene value at the corresponding coding position, a daily ship scheduling scheme code is constructed; whereby a daily ship scheduling scheme is represented as: ,in, Indicates assignment to the first The name of the ship for the transportation demand, if the first If the transport demand is not matched with the vessels, then Not matched; In each iteration, the parent daily boat dispatching scheme is selected based on the multi-objective fusion evaluation value of each daily boat dispatching scheme; By exchanging the ship name genes at the transportation demand positions of the two parent generation daily ship dispatch schemes, a child generation daily ship dispatch scheme is formed. The ship name genes in the progeny daily ship dispatch scheme are mutated, and the feasibility of the mutated daily ship dispatch scheme is corrected to obtain candidate daily ship dispatch schemes. Calculate the difference in multi-objective fusion evaluation values ​​between the candidate daily ship dispatch scheme and the current daily ship dispatch scheme. , For candidate day scheduling plans; If the multi-objective fusion evaluation value of the candidate day's boat scheduling plan is better than that of the current day's boat scheduling plan, then the candidate day's boat scheduling plan is accepted as the current day's boat scheduling plan; if the cost target of the candidate day's boat scheduling plan is improved and the decrease in the boat scheduling business target does not exceed a preset threshold, then the candidate day's boat scheduling plan is accepted as the current day's boat scheduling plan; if the boat scheduling business target of the candidate day's boat scheduling plan is improved and the increase in the cost target does not exceed a preset threshold, then the candidate day's boat scheduling plan is accepted as the current day's boat scheduling plan. If the multi-objective fusion evaluation value of the current daily boat scheduling plan after acceptance is better than the multi-objective fusion evaluation value of the historical best daily boat scheduling plan, then the current daily boat scheduling plan will be updated to the historical best daily boat scheduling plan. Repeat the above iterative steps until the preset termination condition is met, and output the historical best daily boat scheduling scheme as the target daily boat scheduling scheme.

[0016] Secondly, the present application provides a computer-readable storage medium storing program information. After reading the program information, the computer executes the steps of the intelligent scheduling method for oil tankers based on multi-objective evaluation and meta-heuristic search as described in any one of the first aspects.

[0017] Thirdly, the present application provides an intelligent scheduling system for oil tankers based on multi-objective evaluation and meta-heuristic search, including a memory, a processor, and a computer program stored in the memory. The processor executes the computer program to implement the steps of the intelligent scheduling method for oil tankers based on multi-objective evaluation and meta-heuristic search as described in any of the first aspects.

[0018] The technical solution provided in this application has the following technical effects compared with the prior art: This application provides an intelligent scheduling method and system for oil tankers based on multi-objective evaluation and metaheuristic search. By constructing separate evaluation functions for scheduling business objectives and cost objectives, and then normalizing these functions to create a multi-objective fusion evaluation function, it can transform business objectives and cost objectives of different dimensions and directions into a unified scale for comprehensive evaluation, solving the problem of direct comparison of different objectives in existing technologies. Furthermore, this application dynamically adjusts the objective weights in the multi-objective fusion evaluation function according to the daily scheduling scenario, enabling the scheduling method to adapt to scheduling scenarios of different dates, demand structures, and vessel resource states. This avoids the failure of fixed weights in different scenarios and improves the adaptability of the scheduling scheme to different business pressures. This application constructs a daily scheduling scheme code using transportation demand as the encoding position and vessel name as the gene value, allowing the coding results in the metaheuristic optimization process to directly correspond to the real vessel names. This facilitates the review and on-site adjustment of optimization results by scheduling personnel, improving the interpretability and ease of execution of the scheme. In addition, this application performs iterative optimization within the feasible search space and uses multi-objective fusion evaluation values ​​as the basis for judging whether to accept candidate solutions and updating the current solution and the historical best solution during the candidate solution generation and evaluation process. This allows the candidate solution search process to be driven by changes in both the ship dispatching business objectives and cost objectives, thereby generating a daily oil tanker dispatching solution that takes into account both business acceptability and cost economy. Attached Figure Description

[0019] Figure 1 This is a flowchart of an embodiment of the intelligent scheduling method for oil tankers based on multi-objective evaluation and metaheuristic search described in this application; Figure 2 A flowchart illustrating the acquisition of basic daily vessel scheduling data and a basic candidate vessel set for oil tankers, provided in one embodiment of this application; Figure 3 A flowchart illustrating the construction of a boat dispatching business objective evaluation function provided in one embodiment of this application; Figure 4 A flowchart illustrating the construction of a cost target evaluation function provided in one embodiment of this application; Figure 5 A flowchart of multi-objective normalization processing provided in one embodiment of this application; Figure 6 A flowchart of multi-objective fusion evaluation and dynamic weight adjustment provided in one embodiment of this application; Figure 7 A metaheuristic optimization iterative flowchart based on multi-objective fusion evaluation values ​​is provided as an embodiment of this application; Figure 8 This is a schematic diagram of the structure of an intelligent scheduling system for oil tankers based on multi-objective evaluation and meta-heuristic search, provided as an embodiment of this application. Detailed Implementation

[0020] The specific embodiments of this application will be further described below with reference to the accompanying drawings.

[0021] It is readily understood that, based on the technical solution of this application, various structural and implementation methods can be interchanged by those skilled in the art without altering the essential spirit of this application. Therefore, the following detailed embodiments and accompanying drawings are merely illustrative examples of the technical solution of this application and should not be considered as the entirety of this application or as limitations or restrictions on the technical solution of the application.

[0022] This embodiment provides an intelligent scheduling method for oil tankers based on multi-objective evaluation and metaheuristic search, applied to a daily ship scheduling operating system, such as... Figure 1 As shown, it includes: S100: Obtain basic data on daily vessel scheduling for oil tankers and a basic candidate vessel set, and use the basic candidate vessel set as a feasible search space.

[0023] Specifically, this step acquires basic daily vessel scheduling data for oil tankers and standardizes this data, including unifying vessel names, oil product names, port names, region names, plan types, time fields, and status fields. Based on basic feasibility constraints, a basic feasibility assessment is performed on the vessels and transportation demands. Vessels meeting the basic feasibility assessment conditions are added to the candidate vessel set corresponding to the transportation demand. The candidate vessel sets for each transportation demand constitute the basic candidate vessel set, which serves as the feasible search space for subsequent multi-objective fusion evaluation and metaheuristic optimization.

[0024] S200: Construct a boat scheduling business objective evaluation function, which is used to evaluate the performance of the daily boat scheduling plan on at least one boat scheduling business objective.

[0025] Specifically, this step constructs a vessel scheduling business objective evaluation function based on the daily vessel scheduling business rules for oil tankers. This function expresses the non-cost objectives that scheduling personnel focus on in actual operations. In this embodiment, the system converts all objectives into a form where "the smaller the objective value, the better." The vessel scheduling business objective evaluation function is represented as a weighted sum of all sub-objectives, where at least one sub-objective is included in the evaluation function.

[0026] S300: Construct a cost target evaluation function, which is used to evaluate the performance of the daily boat dispatching scheme on at least one cost target.

[0027] Specifically, this step constructs a cost target evaluation function based on vessel type, route, fuel type, planned transport volume, port operation costs, vessel occupancy time, and mismatch penalty cost. When the matched vessel is a owned vessel, the matching cost includes the owned vessel's fixed costs, variable costs, port operation costs, and vessel occupancy costs. When the matched vessel is a privately owned vessel, the matching cost is determined based on the unit freight rate cost corresponding to the vessel type, route, fuel type, and planned transport volume. For mismatched transport demand, corresponding mismatch penalty costs are set.

[0028] S400: Normalize the boat dispatching business objective evaluation function and the cost objective evaluation function to obtain normalized boat dispatching business objective evaluation value and normalized cost objective evaluation value.

[0029] Specifically, in this step, since the boat scheduling business objective and the cost objective have different dimensions, the system first normalizes each objective, transforming them to a unified scale. For the minimization objective, the system uses extreme value normalization, mapping the objective value to a unified interval based on the minimum and maximum values ​​of each objective in the current candidate solution set. For the maximization objective, the system first converts it into a minimization objective before normalization. After normalization, the normalized evaluation values ​​for the boat scheduling business objective and the cost objective are obtained.

[0030] S500: Based on the normalized evaluation value of the boat dispatching business target and the normalized evaluation value of the cost target, a multi-objective fusion evaluation function is constructed, and the target weights in the multi-objective fusion evaluation function are dynamically adjusted according to the boat dispatching scenario of the day.

[0031] Specifically, this step constructs a multi-objective fusion evaluation function based on the normalized evaluation values ​​of the boat scheduling business objectives and the normalized evaluation values ​​of the cost objectives. This multi-objective fusion evaluation function is represented as a weighted sum of the evaluation values ​​of the boat scheduling business objectives and the cost objectives, where the fusion weights are used to adjust the relative importance of the boat scheduling business objectives and the cost objectives in the fusion evaluation. The system obtains the boat scheduling scenario parameters for the day and dynamically adjusts the objective weights in the multi-objective fusion evaluation function based on these parameters. Through dynamic weight adjustment, the system can automatically adjust optimization preferences according to different daily boat scheduling scenarios, avoiding the failure of fixed weights in different scenarios.

[0032] S600: Using transportation demand as the encoding location and ship name as the gene value, construct the encoding of the daily ship scheduling scheme to generate an initial daily ship scheduling scheme; starting from the initial daily ship scheduling scheme, perform iterative optimization: in each iteration, call the metaheuristic optimization algorithm to generate candidate daily ship scheduling schemes in the feasible search space, calculate the multi-objective fusion evaluation value of the candidate daily ship scheduling schemes according to the multi-objective fusion evaluation function, determine whether to accept the candidate daily ship scheduling scheme based on the multi-objective fusion evaluation value, and update the current daily ship scheduling scheme and the historical best daily ship scheduling scheme; when the preset termination condition is met, stop the iteration and take the historical best daily ship scheduling scheme as the target daily ship scheduling scheme.

[0033] Specifically, this step uses each transportation demand of the day as an encoding position, and the ship name assigned to each transportation demand as the gene value at the corresponding encoding position to construct the encoding of the daily ship scheduling plan. If a transportation demand does not match a ship, the position is set as a unmatched marker. The encoding result directly corresponds to the real ship name, which facilitates subsequent plan interpretation and manual review. Starting with the initial daily ship scheduling plan, a metaheuristic optimization algorithm is invoked for iterative optimization. In each iteration, the algorithm generates candidate daily ship scheduling plans within the feasible search space and calculates the multi-objective fusion evaluation value of the candidate plans based on the multi-objective fusion evaluation function. Based on this evaluation value, it is determined whether to accept the candidate plan and the current daily ship scheduling plan and the historical best daily ship scheduling plan are updated. Unlike plan selection based solely on cost increment, the candidate plan evaluation in this embodiment considers both changes in the ship scheduling business objective and changes in the cost objective, making the search process driven by both business and cost objectives. When a preset termination condition is met, the iteration stops, and the historical best daily ship scheduling plan is taken as the target daily ship scheduling plan.

[0034] S700: Output the target day's boat scheduling plan and generate the corresponding boat scheduling business target score, cost target score, multi-target fusion score, target weight, cost change, reason for non-match, and plan description information.

[0035] Specifically, this step outputs the vessel scheduling plan for the target day, including the vessel name matching each transportation demand, as well as the corresponding scheduling business objective score, cost objective score, multi-objective fusion score, objective weight, cost changes, reasons for mismatch, and plan explanation information. By outputting multi-objective evaluation explanation information, scheduling personnel can trace the basis for plan generation and understand the comprehensive trade-offs between business and cost objectives.

[0036] The intelligent tanker scheduling method based on multi-objective evaluation and metaheuristic search provided in this embodiment can simultaneously quantify both scheduling business preferences and transportation costs by constructing separate evaluation functions for scheduling business objectives and cost objectives, thus avoiding business or economic imbalances caused by single-objective evaluation. By normalizing the evaluation functions for scheduling business objectives and cost objectives, multiple objectives with different dimensions and directions can be transformed into a unified scale for comprehensive evaluation, solving the problem in existing technologies where business objectives and cost objectives are difficult to compare directly due to different dimensions. By dynamically adjusting the objective weights in the multi-objective fusion evaluation function according to the daily scheduling scenario, the scheduling method can adapt to changes in objective preferences under different daily scheduling scenarios, avoiding the failure of fixed weights in different scenarios. By constructing a daily scheduling scheme code with transportation demand as the encoding position and ship name as the gene value, the coding results in the metaheuristic optimization process directly correspond to the real ship names, facilitating the review and on-site adjustment of optimization results by scheduling personnel, and improving the interpretability and ease of execution of the scheme. By invoking a metaheuristic optimization algorithm, candidate daily vessel scheduling schemes are iteratively generated within the feasible search space. Multi-objective fusion evaluation values ​​are used as the basis for accepting and updating candidate schemes. This ensures the search process is driven by both changes in scheduling business objectives and cost objectives, generating daily oil tanker scheduling schemes that balance business acceptability and cost-effectiveness. In summary, the technical solution of this embodiment achieves an effective balance between business acceptability, economy, and verifiability in daily oil tanker scheduling schemes.

[0037] Preferably, in the above scheme, the vessel scheduling business objectives include at least one of the following: demand fulfillment objective, priority for urgent orders objective, tonnage matching objective, arrival time objective, liner priority objective, and scarce demand protection objective; the cost objectives include at least one of the following: cost of owned vessels, cost of social vessels, port operation costs, vessel occupancy costs, and non-matching penalty costs; the daily vessel scheduling scenario parameters include at least one of the following: number of urgent orders, proportion of urgent orders, number of candidate vessels with scarce demand, proportion of social vessels, and cost pressure index, wherein the cost pressure index is determined based on at least one of the following: average cost of candidate social vessels, available proportion of owned vessels, non-matching penalty costs, budgeted cost threshold, and total daily demand. Wherein: The demand satisfaction target is used to evaluate the number or proportion of transportation demand that is matched on a given day. The number or proportion of unmatched transportation demand is used to characterize the degree of demand satisfaction in the daily ship scheduling plan. The smaller the target value, the fewer the unmatched demand in the daily ship scheduling plan, and the higher the degree of demand satisfaction.

[0038] The urgent order priority target is used to evaluate whether urgent order transportation needs are prioritized for matching. The number of unmatched urgent orders or the degree of unmet urgent orders characterizes the urgent order fulfillment status of the daily ship scheduling plan. The smaller the target value, the higher the degree to which urgent order needs are prioritized for meeting.

[0039] The tonnage matching target is used to evaluate the degree of matching between the ship's corresponding oil cargo capacity and the planned transportation demand. The deviation between the ship's cargo capacity and the planned demand is used to characterize the accuracy of the tonnage matching. The smaller the target value, the closer the ship's cargo capacity is to the planned transportation demand.

[0040] The arrival time target is used to evaluate whether the estimated arrival time of a vessel at the target port of departure meets the scheduling requirements for the day. The deviation between the estimated arrival time of the vessel and the required time is used to characterize the degree of time fit. The smaller the target value, the better the selected vessel can meet the time requirements of transportation needs.

[0041] The liner priority target is used to evaluate whether liner vessels match the corresponding transportation demand in accordance with liner rules. The degree of mismatch between liner demand and liner service is used to characterize the liner service performance. The smaller the target value, the more likely the liner transportation demand can be matched with vessels that comply with the liner rules.

[0042] The scarcity demand protection target is used to evaluate whether the transportation demand of candidates with fewer vessels is given priority protection. The protection effect is characterized by the degree of mismatch between scarce demand and actual demand. The smaller the target value, the higher the degree to which the scarce transportation demand of candidate vessels is given priority protection.

[0043] The cost of owned vessels includes fixed costs and variable costs. The cost of outsourced vessels is determined based on the vessel type, route, fuel type, and planned transport volume of the outsourced vessel. Port operation costs and vessel occupancy costs are determined based on actual operations and vessel occupancy time. The cost of mismatch penalty is determined based on at least one of the planned transport demand, rush order markings, fuel type, and destination port.

[0044] The cost pressure index is determined based on at least one of the following: the average cost of social candidate vessels, the proportion of available self-owned vessels, the cost of mismatch penalties, the budgeted cost threshold, and the total daily demand. It is used to characterize the cost sensitivity under the current vessel allocation scenario. These scenario parameters collectively constitute the input basis for dynamic weight adjustment, enabling the system to automatically adjust its optimization preferences according to the actual vessel allocation scenario of the day.

[0045] Furthermore, such as Figure 2 As shown, S100 includes: S101: Obtain the basic data of the daily dispatch of oil tankers, which includes the daily dispatch data, the daily oil transportation demand data, and port restriction information.

[0046] Specifically, this step obtains daily vessel dispatch data, daily oil transportation demand data, basic vessel information, historical vessel scheduling data, port restriction information, liner information, and regional port information. Daily vessel dispatch data describes the dynamic status and basic carrying capacity of vessels on that day, including vessel name, vessel type, class, oil cargo tonnage, hold cleanliness, current anchorage, anchorage stay time, arrival time at the port of departure, and vessel status information. Daily oil transportation demand data describes the oil transportation tasks to be performed that day, including oil type, planned volume, loading port, destination port, consignee, urgent order identifier, and consolidation identifier. Basic vessel information describes the vessel's ownership (owned or outsourced), cargo capacity, consolidation capacity, and plan type. Historical vessel scheduling data determines whether a vessel is already booked for an existing plan. Port restriction information determines whether a vessel meets the tonnage, oil type, and berth restrictions corresponding to the loading or unloading port. Liner information determines whether liner vessels comply with fixed routes, oil type, and validity date requirements. Regional port information is used to determine the region to which the destination port belongs, thereby supporting the determination of plan type and region suitability.

[0047] S102: Extract vessel status information from the daily vessel dispatch data.

[0048] Specifically, this step extracts vessel status information from the daily dispatch data. This information includes vessel name, vessel type, class, oil cargo tonnage, hold cleanliness, current anchorage, anchorage stay time, arrival time at the port of departure, and vessel status identifier. The vessel status identifier indicates the vessel's current status, including whether it has a scheduled transport plan but has not yet completed loading, whether it is fully loaded and waiting to unload within the anchorage ring, whether it is empty and waiting to perform unscheduled tasks within the anchorage ring, and whether it is in a suspended or withdrawn state.

[0049] S103: Extract the oil type and planned quantity from the daily oil transportation demand data.

[0050] Specifically, this step extracts the oil type and planned quantity from the daily oil transportation demand data. The oil type is used to determine whether the vessel has the capacity to load the corresponding oil; the planned quantity is used for subsequent tonnage matching, that is, to determine whether the vessel's cargo capacity for the corresponding oil matches the planned quantity of transportation demand.

[0051] S104: Determine the set of available ships for the day based on the ship status information.

[0052] Specifically, this step determines whether a vessel is available for deployment that day based on its status identifier. If a vessel has a scheduled transport plan but has not yet completed loading, it will not be included in the available vessel pool for that day. If a vessel is fully loaded and waiting to unload within the anchorage, it will not be included in the available vessel pool for that day. If a vessel is empty and waiting to perform unscheduled tasks within the anchorage, it will not be included in the available vessel pool for that day. If a vessel is suspended, withdrawn, lacks effective cargo capacity, or does not meet the access requirements, it will not be included in the available vessel pool for that day. For vessels not at the target port of departure anchorage, the estimated arrival time at the target port of departure is calculated based on the return port conversion time corresponding to its current anchorage. Through the above determinations, the system generates the available vessel pool for that day.

[0053] S105: Based on at least one of the oil type, the planned volume, and the port restriction information, make a basic feasibility assessment of the vessels and transportation demand in the available vessel set for the day.

[0054] Specifically, this step, for each transportation demand, sequentially determines whether each vessel in the available vessel pool for the day meets the basic feasibility constraints. Fuel type is used to determine if the vessel is suitable for carrying the corresponding fuel; planned quantity is used to determine if the vessel's corresponding fuel carrying capacity meets the tonnage requirements of the transportation demand; port restriction information is used to determine if the vessel meets the tonnage restrictions, fuel restrictions, and berth restrictions corresponding to the loading or unloading port; liner information is used to determine if liner vessels comply with fixed routes, fuel types, and validity date requirements; historical vessel scheduling data is used to determine if the vessel is already occupied by an existing plan. This step can select one or more of the above criteria for feasibility assessment based on actual business needs.

[0055] S106: Add vessels that meet the basic feasibility judgment conditions to the candidate vessel set corresponding to the transportation needs. The candidate vessel sets of each transportation need constitute the basic candidate vessel set, and the basic candidate vessel set serves as the feasible search space.

[0056] Specifically, in this step, for vessels that meet the basic feasibility assessment criteria, the system adds them to the candidate vessel set for the corresponding transportation demand. For vessels that do not meet the criteria, the system records the reason for not entering the candidate set for later use in the scheme description information. After completing the feasibility assessment of all transportation demands, the candidate vessel sets for each transportation demand together constitute the basic candidate vessel set. The system uses this basic candidate vessel set as the feasible search space for subsequent multi-objective fusion evaluation and metaheuristic optimization.

[0057] In this embodiment, by extracting vessel status information from the daily vessel dispatch data to determine the set of available vessels for the day, vessels that are already occupied or in an unavailable state can be effectively excluded. By making a basic feasibility judgment based on at least one of the following: oil type, planned volume, port restriction information, liner information, and historical vessel dispatch data, infeasible matches can be filtered from multiple dimensions such as oil type suitability, tonnage matching, port restrictions, liner rules, and historical planned occupancy, and feasible vessels can be added to the candidate vessel set for the corresponding transportation needs. The candidate vessel sets for each transportation need constitute the basic candidate vessel set and serve as the feasible search space, ensuring that subsequent multi-objective fusion evaluation and metaheuristic optimization are only carried out within the feasible matching range, avoiding the generation of infeasible solutions, and improving the efficiency and effectiveness of the optimization search.

[0058] Furthermore, this solution also includes a step of standardizing the aforementioned data. This standardization includes unifying ship names, oil product names, port names, region names, plan types, time fields, and status fields. For example, if the same port has abbreviations, aliases, or different table formats, the system converts them into a unified port name; for oil product categories such as gasoline and diesel, the system unifies the oil product names from different data sources into standard oil product types; and for the ship status field, the system converts manually entered or historical status descriptions into calculable status identifiers.

[0059] After data standardization, the following objects are constructed: vessel objects, transport demand objects, port restriction objects, liner rule objects, and historical plan objects. Vessel objects store vessel name, vessel type, class, corresponding oil cargo capacity, hold cleanliness, current anchorage, estimated arrival time at the target port of departure, and vessel status. Transport demand objects store oil type, planned volume, loading port, destination port, rush order identifier, and consolidation identifier. Port restriction objects store restrictions on vessel tonnage, oil type, and berth conditions at loading or unloading ports. Liner rule objects store the loading port, destination port, oil type, and validity period for liner vessels. Historical plan objects store information on scheduled and incomplete voyages.

[0060] Preferably, such as Figure 3 As shown in S200, the construction of the boat scheduling business objective evaluation function, which is used to evaluate the performance of the daily boat scheduling plan on at least one boat scheduling business objective, includes: S201: Constructing a system that meets requirements and objectives The demand satisfaction target is used to characterize the degree of demand satisfaction of the daily ship scheduling scheme by the number or proportion of unmatched transportation demand.

[0061] In this step, the demand satisfaction target is defined as a target used to characterize the degree of unmet demand in the daily ship dispatching plan, which can be represented by the quantity or proportion of unmatched transportation demand. Let... This is a collection of transportation needs for the day. For the first The unmatched variable of transportation demand, when the first When the transport demand is not matched with the number of vessels, ,otherwise Then the requirement to satisfy the objective can be expressed as: ; in, This represents a daily boat scheduling plan. The smaller the target value, the fewer unmatched demands there are in the daily boat scheduling plan, and the higher the degree of demand satisfaction.

[0062] S202: Establish a priority target for urgent orders The urgent order priority target is used to evaluate the matching of urgent order transportation demand in the daily ship scheduling plan.

[0063] In this step, the priority target for urgent orders is defined as the number of unmatched urgent orders or the degree to which urgent orders are not fulfilled. Let... For the first The urgent order identifier for the transportation demand, when the first When the transportation demand is an urgent order, ,otherwise Then, the priority target for urgent orders can be expressed as: ; The smaller the target value, the higher the priority is given to meeting urgent orders.

[0064] S203: Constructing Tonnage Matching Targets The tonnage matching target is used to evaluate the deviation between the corresponding oil cargo capacity of the ships in the daily ship dispatching plan and the planned transportation demand.

[0065] In this step, the tonnage matching target is defined as the deviation between the vessel's corresponding oil cargo carrying capacity and the planned transportation demand. Let... For the first Planned volume of transportation demand For the first The ship targeted the first The transportation demand corresponds to the carrying capacity of oil products. For the decision variable of arranging boats, when the first The transportation demand is determined by the first When the ship is in operation, ,otherwise Then the tonnage matching target can be expressed as: ; in, To prevent a constant with a denominator of zero, the smaller this target value, the closer the ship's cargo capacity is to the planned transportation demand.

[0066] S204: Establishing Arrival Time Objectives The target arrival time is used to evaluate the degree of fit between the expected arrival time of ships at the target port of departure and the time requirements of transportation demand in the daily ship scheduling plan.

[0067] In this step, the arrival time target is defined as the deviation between the vessel's estimated arrival time at the target port of departure and the required vessel scheduling time for transportation demand. Let... For the first The ship is expected to arrive at the The time of departure from the port of shipment for each transport requirement For the first Given the time requirements corresponding to the transportation demand, the target arrival time can be expressed as: ; The smaller the target value, the better the selected vessel can meet the time requirements of transportation needs.

[0068] S205: Establishing Liner Priority Objectives The liner priority target is used to evaluate the matching between liner vessels and liner demand in the daily vessel scheduling plan.

[0069] In this step, the liner priority objective is defined as whether liner shipping demand is matched with vessels that comply with liner rules. Let... For the first Liner demand identifier for transport demand, when the first When the transportation demand falls under the category of liner shipping demand, ,otherwise ;set up For the first The ship and the first Liner matching identifier for transport demand, when the first The ship meets the requirements of the first When specifying the loading port, destination port, type of oil, and validity date requirements for a transportation request, ,otherwise The liner priority objective can then be expressed as: ; in, As a variable for ship scheduling decisions, the smaller this target value, the more likely the liner shipping demand can be matched with ships that comply with liner rules.

[0070] S206: Establishing targets for protecting scarce demand The scarcity demand protection target is used to evaluate whether transportation demands with fewer than a preset threshold in the daily ship scheduling scheme are prioritized for matching.

[0071] In this step, the scarcity demand protection objective is defined as whether transportation demands with fewer candidate vessels than a preset threshold are prioritized for matching. Let... For the first The scarcity indicator of transportation demand, when the first When the number of candidate vessels for a given transportation demand is less than a preset threshold, ,otherwise Then the target of protecting scarce demand can be expressed as: ; The smaller the target value, the higher the degree to which the scarce transportation needs of candidate vessels are prioritized for protection.

[0072] S207: Construct the vessel scheduling business objective evaluation function based on at least one of the following: demand satisfaction objective, rush order priority objective, tonnage matching objective, arrival time objective, liner priority objective, and scarce demand protection objective. .

[0073] The boat dispatching business objective evaluation function The calculation expression is: ; in, Indicates the daily boat dispatch plan, Each objective in the boat scheduling operation is assigned a weight, and the sum of these weights is 1. Through this step, this application transforms the boat scheduling preference, which was originally expressed by manual experience or rule-based sorting, into a computable objective function.

[0074] This embodiment quantifies the various business preferences that scheduling personnel focus on in actual operations into calculable objective functions by separately constructing objectives for demand fulfillment, priority of urgent orders, tonnage matching, arrival time, liner priority, and protection of scarce demand. The demand fulfillment objective evaluates the degree of demand fulfillment based on the number of unmatched demands; the urgent order priority objective evaluates the fulfillment of urgent orders based on the number of unmatched urgent orders; the tonnage matching objective evaluates the accuracy of tonnage matching based on the deviation between cargo capacity and planned volume; the arrival time objective evaluates the time suitability based on the deviation between the expected arrival time and the required time; the liner priority objective evaluates the liner execution based on the degree of liner demand mismatch; and the scarce demand protection objective evaluates the protection effect based on the degree of scarce demand mismatch. Furthermore, a unified evaluation function for scheduling business objectives is constructed through weighted summation, transforming scheduling preferences that originally relied on manual experience or rule-based ordering into calculable quantitative evaluation values, providing a complete quantitative foundation for subsequent multi-objective fusion evaluation.

[0075] Furthermore, such as Figure 4 As shown in S300, the construction of the cost target evaluation function, which is used to evaluate the performance of the daily boat dispatching scheme on at least one cost target, includes: S301: Calculate the cost of owning vessels when fulfilling transportation needs using owned vessels.

[0076] Specifically, when the matched vessel is a owned vessel, the system calculates the matching cost based on the owned vessel's fixed costs, variable costs, port operation costs, and vessel occupancy costs. The owned vessel's fixed costs refer to the vessel's inherent daily fixed expenses; its variable costs refer to the costs that change with each voyage during operation; port operation costs include loading port operation costs and unloading port operation costs; and vessel occupancy costs are related to the time the vessel occupies during its voyage. In one implementation, the owned vessel matching cost is expressed as: ; in, For the first Fixed daily cost of owned vessels For the first The daily cost of a single owned vessel For the first The ship carried out the first The voyage occupancy time for the transport demand, For loading port operating costs ， Costs for unloading port operations.

[0077] S302: Calculate the social vessel costs incurred when a social vessel fulfills a transportation demand.

[0078] Specifically, when the matched vessel is a private vessel, the system determines the unit freight cost based on the vessel type, route, fuel type, and planned transport volume. In one implementation, the private vessel matching cost is expressed as: ; in, For the first The social ship in the The corresponding shipping routes for each transportation demand and oil type The unit freight rate below For the first Planned volume of transportation demand.

[0079] The final matching cost is determined based on the vessel type. If the first If the vessel is owned, then ； If the first If the vessel is a social vessel, then .

[0080] S303: Calculate the mismatch penalty cost corresponding to the mismatched transportation demand.

[0081] For unmatched transportation demands, penalty costs are set based on the planned transportation demand volume, urgent order status, fuel type, and destination port. The penalty cost for unmatched orders, large planned volumes, or important destination ports can be set to a higher value to prevent the system from abandoning important transportation needs in order to reduce vessel usage costs.

[0082] S304: Construct a cost target evaluation function based on the owned vessel cost, the social vessel cost, and the mismatch penalty cost.

[0083] The cost target evaluation function The calculation expression is: ; in, This is a collection of transportation needs for the day. For the assembly of ships, Indicates the first The daily oil transportation demand, Indicates the first Candidate ships, For the first The ship carried out the first Matching costs for transportation demand For the decision variable of arranging boats, when the first The transportation demand is determined by the first When the ship is in operation ,otherwise , For the first The penalty cost for mismatched transportation demand. For unmatched variables, when the first When the transport demand does not match the number of vessels ,otherwise Through this step, the system can evaluate the cost differences of different vessel scheduling schemes from the perspective of the overall scheme, and incorporate the costs of owned vessels, social vessels, port operation costs, vessel occupancy costs, and mismatch penalty costs into the cost evaluation of the daily vessel scheduling scheme.

[0084] Because the objectives for scheduling and cost have different dimensions, the system first normalizes each objective. For example, the value of the demand fulfillment objective might be the quantity of unmatched demand, the value of the tonnage matching objective might be the tonnage deviation percentage, the value of the arrival time objective might be the number of hours, and the value of the cost objective might be the monetary value. If they are directly added together, a larger cost objective might suppress other business objectives, or a smaller, urgent order objective might be ineffective. Preferably, such as... Figure 5 As shown, the normalization process for the boat dispatching business objective evaluation function and the cost objective evaluation function in S400, to obtain the normalized boat dispatching business objective evaluation value and the normalized cost objective evaluation value, includes: S401: Obtain the business target value of each sub-target in the boat dispatching business target evaluation function, and the cost target value in the cost target evaluation function.

[0085] Specifically, the sub-target values ​​(including demand satisfaction target value, urgent order priority target value, tonnage matching target value, arrival time target value, liner priority target value, and scarce demand protection target value) in the vessel scheduling business target evaluation function of each daily vessel scheduling scheme in the current candidate scheme set, as well as the cost target value in the cost target evaluation function, are calculated as input data for normalization processing.

[0086] S402: Determine whether each objective is a minimization objective or a maximization objective; Specifically, the system determines whether each objective is a minimization or maximization objective. In this embodiment, the demand fulfillment objective (number of unmatched demands), the urgent order priority objective (number of unmatched urgent orders), the tonnage matching objective (deviation between cargo capacity and planned quantity), the arrival time objective (time deviation), and the cost objective (transportation cost) are all minimization objectives, meaning the smaller the objective value, the better. However, indicators such as demand fulfillment rate and urgent order fulfillment rate are maximization objectives, meaning the larger the objective value, the better. The system performs directional determination for each objective to facilitate subsequent unified normalization processing.

[0087] S403: For the minimization objective, normalization is performed using the following expression: ; in, For the first The goal is to schedule a boat trip in Japan. The normalized target value is below. For the first The goal is to schedule a boat trip in Japan. The target value below, and These are the first and second candidate solutions in the current candidate solution set. The minimum and maximum values ​​of each objective. To prevent constants with a denominator of zero.

[0088] In one implementation, the solution can be determined based on the set of candidate solutions generated during the current iteration. and Alternatively, the normalization upper and lower bounds can be determined jointly based on historical ship scheduling schemes, the initial scheme, and the current candidate schemes. This process maintains the relative stability of the target evaluation under different daily ship scheduling volumes and cost levels.

[0089] In this step, for the minimization objective, the system uses extreme value normalization to map the original objective value to the [0,1] interval based on the minimum and maximum values ​​of each objective in the current candidate solution set. After normalization, the values ​​of each minimization objective are compared on a uniform scale, and the smaller the objective value, the better the performance of that objective.

[0090] S404: For a maximization objective, convert it into a minimization objective and then normalize it.

[0091] In this step, for the maximization objective, it is first converted into a minimization objective before normalization. For example, the demand fulfillment rate can be converted into the number of unmatched demands, the rush order fulfillment rate into the number of unmatched rush orders, and the liner fulfillment rate into the number of unfulfilled liner orders. Through this conversion, all objectives can be uniformly processed in a way that the smaller the objective value, the better, so that all objectives have a consistent optimization direction in the subsequent integrated evaluation.

[0092] S405: Obtain the normalized evaluation values ​​of the ship dispatching business objectives and the normalized evaluation values ​​of the cost objectives.

[0093] In this step, after normalizing all sub-objectives, normalized evaluation values ​​for both the scheduling business objective and the cost objective are obtained for each daily scheduling plan. After direction unification and normalization, all objectives are compared according to the same optimization direction, resolving the problem of direct comparison between scheduling business objectives and cost objectives due to differences in dimensions and directions. When using a minimization evaluation method, a smaller multi-objective fusion evaluation value indicates a better overall performance of the daily scheduling plan.

[0094] This embodiment obtains the business target values ​​of each sub-target in the ship scheduling business target evaluation function and the cost target value in the cost target evaluation function, and determines whether each target is a minimization target or a maximization target, providing a basis for subsequent unified normalization processing. For minimization targets, extreme value normalization is used to map the original target value to a unified range. For maximization targets, they are first converted into minimization targets and then normalized. This ensures that targets with different dimensions and directions, such as demand satisfaction, priority for urgent orders, tonnage matching, arrival time, liner priority, protection of scarce demand, and transportation costs, are all converted to a unified scale and compared according to the same optimization direction. This solves the problem in the prior art that ship scheduling business targets and cost targets are difficult to directly compare and integrate due to different dimensions and directions, providing a unified quantitative basis for multi-target integrated evaluation.

[0095] In some preferred solutions, such as Figure 6 As shown in S500, the construction of a multi-objective fusion evaluation function based on the normalized boat scheduling business target evaluation value and the normalized cost target evaluation value, and the dynamic adjustment of the target weights in the multi-objective fusion evaluation function according to the boat scheduling scenario of the day, includes: S501: Evaluation value of boat dispatching business objectives based on normalized values and normalized cost target evaluation value Construct a multi-objective fusion evaluation function Its calculation expression is: ; in, The weighting of the integration between the objectives of the boat dispatching business and the cost objectives.

[0096] Specifically, the system constructs a multi-objective fusion evaluation function based on the normalized evaluation values ​​of the boat dispatching business objectives and the normalized evaluation values ​​of the cost objectives. Φ( X The smaller the value of ), the better the overall performance of the daily boat dispatching scheme. λ When the size is large, the system places more emphasis on the boat dispatching business objectives; when λ When the cost is relatively low, the system prioritizes cost objectives. Through this multi-objective fusion evaluation function, the system can unify boat dispatching business preferences and cost-effectiveness into a comparable evaluation value.

[0097] S502: Based on the boat scheduling scenario parameters for the day, adjust the internal weights of at least one sub-objective in the boat scheduling business objective evaluation function, and adjust the fusion weights between the boat scheduling business objective and the cost objective. Specifically, when the number or proportion of urgent orders exceeds a preset threshold, the internal weight corresponding to the urgent order priority target is increased; when the number of candidate vessels with scarce demand exceeds a preset threshold, the internal weight corresponding to the scarce demand protection target is increased; when the proportion of social vessels or cost pressure indicators exceed a preset threshold, the fusion weight is decreased. The value of is determined to increase the weight of cost targets in the fusion evaluation.

[0098] Specifically, the system dynamically adjusts the objective weights in the multi-objective fusion evaluation function based on the daily vessel scheduling scenario parameters. These parameters include at least one of the following: number of urgent orders, proportion of urgent orders, number of candidate vessels in short supply, proportion of social vessels, and cost pressure indicators. The system first sets the basic weights and fusion weights for each sub-objective of the vessel scheduling business objectives. λ The base value is then used to calculate the weight correction amount based on the scene parameters.

[0099] When the number or proportion of urgent orders exceeds a preset threshold, the system increases the internal weight corresponding to the urgent order priority target. This results in a greater penalty for unmatched urgent orders; when the number of candidate ships with scarce demand exceeds a preset threshold, the system increases the internal weight corresponding to the scarce demand protection target. This results in greater penalties for solutions that fail to meet scarce demand; when the proportion of social ships or cost pressure indicators exceed a preset threshold, the system reduces the fusion weight. λ The value of this parameter is chosen to give cost a greater weight in the multi-objective fusion evaluation, thereby increasing the importance of cost in the fusion evaluation. When liner shipping demand is high, the system can improve... This allows the degree of compliance with liner regulations to have a greater impact on the evaluation of the program. The weights of each sub-objective are normalized to a sum of 1.

[0100] In one implementation, basic weights for scheduling business objectives and cost objectives are first set. Then, weight adjustments are calculated based on the proportion of urgent orders, the proportion of scarce demand, the proportion of liner demand, the proportion of private vessels, and cost pressure indicators. The adjusted weights are then limited to preset upper and lower limits. Through dynamic weight adjustment, the system can avoid the failure of fixed weights in different daily scheduling scenarios, enabling the scheduling plan to automatically adjust and optimize preferences according to the daily demand structure, vessel resource status, and cost pressure.

[0101] S503: Output multi-objective fusion evaluation results.

[0102] Specifically, based on the constructed multi-objective fusion evaluation function and the dynamically adjusted objective weights, the multi-objective fusion evaluation value of the daily boat scheduling scheme is calculated. This value serves as the primary evaluation basis for candidate scheme generation, scheme adjustment, scheme acceptance, and historical best update in the subsequent metaheuristic optimization process. The system outputs the multi-objective fusion evaluation value of the target day's boat scheduling scheme, and simultaneously outputs the corresponding boat scheduling business objective score, cost objective score, and the weights of each objective.

[0103] This embodiment constructs a multi-objective fusion evaluation function, unifying the normalized evaluation values ​​of the ship scheduling business objectives and cost objectives into a single evaluation value through a weighted sum. This allows for comparison and weighing of ship scheduling preferences and cost-effectiveness on the same scale, solving the problem of difficulty in unifying the evaluation of multiple objectives in existing technologies. By adjusting the internal weights of sub-objectives in the ship scheduling business objective evaluation function and the fusion weight λ between the ship scheduling business objective and the cost objective based on the ship scheduling scenario parameters of the day, the system can adaptively adjust and optimize preferences according to the actual ship scheduling scenario of the day. When the number or proportion of urgent orders exceeds a preset threshold, the internal weight corresponding to the urgent order priority objective is increased; when the number of candidate ships with scarce demand exceeds a preset threshold, the internal weight corresponding to the scarce demand protection objective is increased; when the proportion of social ships or cost pressure indicators exceed a preset threshold, the value of the fusion weight λ is decreased to increase the proportion of the cost objective in the fusion evaluation. This allows the ship scheduling method to adapt to ship scheduling scenarios under different dates, different demand structures, and different cost pressures, avoiding the failure of fixed weights in different scenarios and improving the adaptability of the ship scheduling scheme to different business and cost pressures.

[0104] More preferably, such as Figure 7 As shown in S600, the encoding of the daily ship scheduling scheme is constructed using transportation demand as the encoding position and ship name as the gene value, generating an initial daily ship scheduling scheme; starting from the initial daily ship scheduling scheme, iterative optimization is performed: in each iteration, a metaheuristic optimization algorithm is called to generate candidate daily ship scheduling schemes within the feasible search space, the multi-objective fusion evaluation value of the candidate daily ship scheduling schemes is calculated according to the multi-objective fusion evaluation function, and it is determined whether to accept the candidate daily ship scheduling scheme based on the multi-objective fusion evaluation value, and the current daily ship scheduling scheme and the historical best daily ship scheduling scheme are updated; when a preset termination condition is met, the iteration stops, and the historical best daily ship scheduling scheme is taken as the target daily ship scheduling scheme, including: S601: Using each transport demand of the day as a coding position, and the ship name assigned to each transport demand as the gene value at the corresponding coding position, a daily ship scheduling scheme code is constructed; wherein, a daily ship scheduling scheme is represented as: ,in, Indicates assignment to the first The name of the ship for the transportation demand, if the first If the transport demand is not matched with the vessels, then This is a non-matching identifier.

[0105] Specifically, the daily transport demand sequence is used as the encoding position, and the ship name assigned to each transport demand is used as the gene value. Let the daily transport demand sequence be: ; When generating the initial daily boat scheduling plan, historical scheduling plans, random feasible plans, or heuristic plans based on multi-objective fusion evaluation values ​​can be used as a starting point. The specific method of generating the initial plan does not limit the scope of protection of this invention. The key to this invention lies in using multi-objective fusion evaluation values ​​to control the generation, adjustment, acceptance, and updating of candidate plans and historical best plans during the subsequent metaheuristic optimization process.

[0106] The advantage of using ship names as gene values ​​is that the encoding results can directly correspond to the actual ship deployment results. For the ship deployment personnel, each gene value output by the system is a real ship name, eliminating the need to convert abstract numbers into business objects, thereby improving the interpretability of the solution and the convenience of on-site execution.

[0107] S602: In each iteration, the parent daily boat dispatching scheme is selected based on the multi-objective fusion evaluation value of each daily boat dispatching scheme.

[0108] This embodiment uses a genetic algorithm as an example of a metaheuristic optimization algorithm. The system uses several feasible daily boat scheduling schemes as the initial population. Each individual adopts an encoding method where transportation demand is the encoding location and the boat name is the gene value. A multi-objective fusion evaluation function is used as the basis for evaluating individual fitness. During the selection phase, the system selects the parent daily boat scheduling scheme from the current population based on the multi-objective fusion evaluation value; the better the multi-objective fusion evaluation value, the higher the probability that the daily boat scheduling scheme will be selected as the parent scheme.

[0109] S603: Exchange the ship name genes at the transportation demand positions of two parent day-to-day ship dispatch schemes to form the offspring day-to-day ship dispatch scheme.

[0110] During the crossover phase, the system selects partial coding positions from the two parent generation daily boat arrangement schemes for boat name gene exchange, forming the offspring daily boat arrangement scheme. For example, for two parent generation schemes... X 1={ A , B , C , D}and X 2={ E , F , G , H After exchanging genes at positions 2 and 3, a progeny scheme is formed. X 1′={ A ,F , G , D}and X 2′={ E , B , C , H In cases where the same vessel is repeatedly used after crossover, or where a vessel does not meet the candidate set or assembly rules, the system will perform verification and adjustment based on the basic feasibility constraints.

[0111] S604: Mutate some of the ship name genes in the offspring daily ship dispatch scheme, and perform feasibility correction on the mutated daily ship dispatch scheme to obtain candidate daily ship dispatch schemes.

[0112] During the mutation phase, the system replaces, adjusts, or redistributes ship name genes at certain coding positions to create new candidate day ship scheduling schemes. For cases where duplicate usage, candidate set inconsistencies, or assembly rule violations occur after mutation, the system performs feasibility correction. For each candidate day ship scheduling scheme, the system calculates the ship scheduling business objective evaluation value, cost objective evaluation value, and multi-objective fusion evaluation value, and determines whether the candidate scheme should enter the next generation population based on the difference in the multi-objective fusion evaluation value.

[0113] S605: Calculate the difference in multi-objective fusion evaluation values ​​between the candidate daily boat dispatching scheme and the current daily boat dispatching scheme. , The candidate date for the boat scheduling plan.

[0114] Among them, the difference in multi-objective fusion evaluation value between the current daily boat scheduling scheme and the candidate daily boat scheduling scheme is used to determine whether the candidate scheme should enter the next generation of population; the system prioritizes retaining the candidate daily boat scheduling scheme with better multi-objective fusion evaluation value and that has passed feasibility correction.

[0115] Unlike selecting candidate solutions solely based on cost targets, this embodiment considers both changes in scheduling business objectives and cost targets during the selection, crossover, and mutation processes. For example, a candidate solution with lower cost might lead to unmet urgent orders or the occupation of scarce demand vessels, thus its multi-objective fusion evaluation value may not be optimal; conversely, a candidate solution with slightly higher cost might simultaneously meet the requirements of urgent orders, liner services, and tonnage matching, resulting in a potentially better multi-objective fusion evaluation value. In this way, the system can avoid business biases caused by a single objective during the genetic algorithm iteration process.

[0116] S606: If the multi-objective fusion evaluation value of the candidate day's boat scheduling scheme is better than the multi-objective fusion evaluation value of the current day's boat scheduling scheme, then the candidate day's boat scheduling scheme is accepted as the current day's boat scheduling scheme; if the cost target of the candidate day's boat scheduling scheme is improved and the decrease in the boat scheduling business target does not exceed a preset threshold, then the candidate day's boat scheduling scheme is accepted as the current day's boat scheduling scheme; if the boat scheduling business target of the candidate day's boat scheduling scheme is improved and the increase in the cost target does not exceed a preset threshold, then the candidate day's boat scheduling scheme is accepted as the current day's boat scheduling scheme.

[0117] Specifically, the system uses a multi-objective acceptance criterion to determine whether to accept a candidate day's boat scheduling plan. If the multi-objective fusion evaluation value of the candidate day's boat scheduling plan is better than the multi-objective fusion evaluation value of the current day's boat scheduling plan, i.e., ΔΦ<0, then the system accepts the candidate day's boat scheduling plan and updates it to the current day's boat scheduling plan.

[0118] If the cost target of the candidate day's scheduling plan improves, and the decrease in the scheduling business target does not exceed a preset threshold, the system will accept the candidate day's scheduling plan. This scenario applies when the candidate plan can significantly reduce costs while having minimal impact on rush orders, tonnage matching, liner priority, and protection of scarce demand.

[0119] If the scheduling business objectives of the candidate day's scheduling plan improve, and the increase in cost does not exceed a preset threshold, the system will also accept the candidate day's scheduling plan. This applies when the candidate plan can significantly improve the fulfillment of urgent orders, protection of scarce demand, or liner service performance, and the cost increase is within an acceptable range.

[0120] If a candidate day's boat scheduling plan does not meet the above acceptance criteria, the system can determine whether to accept the plan based on a preset acceptance probability. This probability-based acceptance mechanism is used to prevent the search process from prematurely getting stuck in a local optimum.

[0121] S607: If the multi-objective fusion evaluation value of the current daily boat scheduling scheme after acceptance is better than the multi-objective fusion evaluation value of the historical best daily boat scheduling scheme, then the current daily boat scheduling scheme will be updated to the historical best daily boat scheduling scheme.

[0122] Specifically, after each iteration completes the update of the current daily boat scheduling plan, the system compares the multi-objective fusion evaluation value of the accepted current daily boat scheduling plan with the multi-objective fusion evaluation value of the historical best daily boat scheduling plan. If the multi-objective fusion evaluation value of the current daily boat scheduling plan is better than that of the historical best daily boat scheduling plan, then the current daily boat scheduling plan is updated to the historical best daily boat scheduling plan. Through this step, the system continuously saves the optimal solution found during the iteration process.

[0123] S608: Repeat the above iterative steps until the preset termination condition is met, and output the historical best daily boat scheduling scheme as the target daily boat scheduling scheme.

[0124] Specifically, the system repeatedly executes iterative steps S602 to S607 until the maximum number of iterations is reached or the historically optimal daily vessel scheduling plan has not been improved after a preset number of consecutive rounds. At this point, the search terminates, and the currently saved historically optimal daily vessel scheduling plan is output as the target daily vessel scheduling plan. Through this process, the target daily vessel scheduling plan finally output by the system is the plan with the best multi-objective fusion evaluation value found throughout the entire iteration process, taking into account both the vessel scheduling business objectives and cost objectives. Specifically, the system outputs the target daily vessel scheduling plan, including the transport demand number, fuel type, planned transport volume, loading port, destination port, matching vessel name, vessel type, vessel class, gasoline cargo tonnage, diesel cargo tonnage, estimated arrival time at the target port of departure, and matching cost.

[0125] The system also outputs the following information: scheduling business target score, cost target score, multi-target fusion score, target weight, rush order fulfillment status, scarce demand protection status, liner matching status, cost changes, unmatched demand and reasons, and solution description information. The solution description information explains the balance between scheduling business objectives and cost objectives for the target day's scheduling plan compared to other candidate plans.

[0126] For example, for a given transportation demand, the system can output the reasons why it was assigned to a particular vessel, including that the vessel meets the fuel loading requirements, has a small tonnage matching deviation, has an early expected arrival time, will not occupy a scarce demand candidate vessel, has a low matching cost, or has a superior multi-objective fusion evaluation value. For unmatched transportation demands, the system can output the reasons for non-matching, including that there are no available candidate vessels, candidate vessels are occupied by higher priority demands, port restrictions are not met, liner rules are not met, or cost penalties were not arranged after weighing other objectives.

[0127] By outputting multi-objective evaluation information, the scheduling personnel can understand the basis of the selected plan. They can not only see the final ship name allocation result, but also the comprehensive trade-offs between demand satisfaction, priority for urgent orders, tonnage matching, arrival time, liner priority, protection of scarce demand, and cost control.

[0128] This application also provides a computer-readable storage medium storing program information. After reading the program information, the computer executes the steps of the intelligent scheduling method for oil tankers based on multi-objective evaluation and metaheuristic search described in the foregoing embodiments.

[0129] This application also provides an intelligent scheduling system for oil tankers based on multi-objective evaluation and meta-heuristic search, such as... Figure 8As shown, the electronic device includes at least one processor 81 and at least one memory 82. The memory 82 stores program information. After reading the program information, the processor 81 executes the intelligent tanker scheduling method based on multi-objective evaluation and metaheuristic search as described in any of the above method embodiments. The device may further include an input device 83 and an output device 84. The processor 81, memory 82, input device 83, and output device 84 can be communicatively connected. The memory 82, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules. The processor 81 executes various functional applications and data processing by running the non-volatile software programs, instructions, and modules stored in the memory 82, thereby implementing the intelligent tanker scheduling method based on multi-objective evaluation and metaheuristic search provided in any of the above embodiments. The memory 82 may include a program storage area and a data storage area. The program storage area may store the operating system and at least one application program required for a function; the data storage area may store data created based on the use of the intelligent tanker scheduling method based on multi-objective evaluation and metaheuristic search. Furthermore, memory 82 may include high-speed random access memory and non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, memory 82 may optionally include memory remotely located relative to processor 81, which can be connected via a network to an apparatus for performing the intelligent tanker scheduling method based on multi-objective evaluation and metaheuristic search. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof. Input device 83 may receive user clicks and generate signal inputs related to user settings and function controls of the intelligent tanker scheduling method based on multi-objective evaluation and metaheuristic search. Output device 84 may include a display device such as a screen. When one or more modules are stored in memory 82 and are run by one or more processors 81, the intelligent tanker scheduling method based on multi-objective evaluation and metaheuristic search in any of the above method embodiments is executed.

[0130] As needed, the above technical solutions can be combined to achieve the best technical effect.

[0131] The above are merely the principles and preferred embodiments of this application. It should be noted that, for those skilled in the art, several other modifications can be made based on the principles of this application, and these modifications should also be considered within the scope of protection of this application.

Claims

1. A smart scheduling method for oil tankers based on multi-objective evaluation and meta-heuristic search, characterized in that, include: Obtain basic data on daily vessel scheduling for oil tankers and a basic candidate vessel set, and use the basic candidate vessel set as the feasible search space; Construct a boat scheduling business objective evaluation function, which is used to evaluate the performance of the daily boat scheduling plan on at least one boat scheduling business objective; Construct a cost target evaluation function, which is used to evaluate the performance of the daily boat dispatching scheme on at least one cost target; The evaluation function for the boat dispatching business objective and the evaluation function for the cost objective are normalized to obtain normalized evaluation values ​​for the boat dispatching business objective and the cost objective. Based on the normalized evaluation value of the boat dispatching business target and the normalized evaluation value of the cost target, a multi-objective fusion evaluation function is constructed, and the target weights in the multi-objective fusion evaluation function are dynamically adjusted according to the boat dispatching scenario of the day. Using transportation demand as the encoding location and ship name as the gene value, a coding scheme for the daily ship dispatching plan is constructed to generate an initial daily ship dispatching plan; Starting with the initial daily boat scheduling scheme, iterative optimization is performed: In each iteration, a metaheuristic optimization algorithm is invoked to generate candidate daily boat scheduling schemes within the feasible search space. The multi-objective fusion evaluation value of the candidate daily boat scheduling scheme is calculated based on the multi-objective fusion evaluation function. Based on the multi-objective fusion evaluation value, it is determined whether to accept the candidate daily boat scheduling scheme, and the current daily boat scheduling scheme and the historical best daily boat scheduling scheme are updated. When a preset termination condition is met, the iteration stops, and the historical best daily boat scheduling scheme is taken as the target daily boat scheduling scheme. Output the target day's boat scheduling plan and generate the corresponding boat scheduling business target score, cost target score, multi-target fusion score, target weight, cost change, reasons for non-matching, and plan description information.

2. The intelligent scheduling method for oil tankers based on multi-objective evaluation and meta-heuristic search according to claim 1, characterized in that: The objectives of the vessel scheduling business include at least one of the following: demand fulfillment objective, priority for urgent orders objective, tonnage matching objective, arrival time objective, liner priority objective, and scarce demand protection objective. The cost targets include at least one of the following: owned vessel costs, social vessel costs, port operation costs, vessel occupancy costs, and mismatch penalty costs; The parameters for the daily vessel allocation scenario include at least one of the following: number of urgent orders, proportion of urgent orders, number of candidate vessels in short supply, proportion of social vessels, and cost pressure indicators. The cost pressure indicators are determined based on at least one of the following: average cost of social vessel candidates, proportion of owned vessels available, cost of non-matching penalty, budget cost threshold, and total daily demand.

3. The intelligent scheduling method for oil tankers based on multi-objective evaluation and meta-heuristic search according to claim 2, characterized in that, The process of acquiring basic daily vessel scheduling data for oil tankers and a basic candidate vessel set, and using the basic candidate vessel set as a feasible search space, includes: Obtain the daily basic data of oil tanker scheduling, which includes daily dispatch data, daily oil transportation demand data, and port restriction information; Extract vessel status information from the daily vessel dispatch data; Extract the type and planned quantity of oil products from the daily oil transportation demand data; The set of available vessels for the day is determined based on the vessel status information; Based on at least one of the oil type, the planned volume, and the port restriction information, a basic feasibility assessment is made of the vessels and transportation demand in the available vessel set for the day. Ships that meet the basic feasibility judgment conditions are added to the candidate ship set corresponding to the transportation needs. The candidate ship sets of each transportation need constitute the basic candidate ship set, which serves as the feasible search space.

4. The intelligent scheduling method for oil tankers based on multi-objective evaluation and meta-heuristic search according to claim 3, characterized in that, The construction of the boat scheduling business objective evaluation function, which is used to evaluate the performance of the daily boat scheduling plan on at least one boat scheduling business objective, includes: Build requirements to meet goals The demand satisfaction target is used to characterize the degree of demand satisfaction of the daily ship scheduling scheme by the number or proportion of unmatched transportation demand. Build a priority target for urgent orders The urgent order priority target is used to evaluate the matching of urgent order transportation demand in the daily ship scheduling plan; Construct tonnage matching target The tonnage matching target is used to evaluate the deviation between the corresponding oil cargo carrying capacity of the ships in the daily ship dispatching plan and the planned transportation demand. Establish arrival time target The arrival time target is used to evaluate the degree of fit between the expected arrival time of ships at the target port of departure and the time requirements of transportation demand in the daily ship scheduling plan; Building Liner Priority Goals The liner priority target is used to evaluate the matching between liner vessels and liner demand in the daily vessel scheduling plan; Build scarce demand protection targets The scarcity demand protection target is used to evaluate whether transportation demands with fewer than a preset threshold in the daily ship scheduling scheme are prioritized for matching. Based on at least one of the following objectives: demand fulfillment, priority for urgent orders, tonnage matching, arrival time, priority for liner services, and protection of scarce demand, construct the vessel scheduling business objective evaluation function. The evaluation function for the gantry operation target. The calculation expression is: ； in, Indicates the daily boat dispatch plan, The weights of each objective in the boat dispatching business are given, and the sum of the weights of each objective is 1.

5. The intelligent scheduling method for oil tankers based on multi-objective evaluation and meta-heuristic search according to claim 4, characterized in that, The construction of the cost target evaluation function, which is used to evaluate the performance of the daily boat dispatching scheme on at least one cost target, includes: Calculate the cost of owning vessels when fulfilling transportation needs using owned vessels; Calculate the social vessel costs incurred when social vessels fulfill transportation demands; Calculate the mismatch penalty cost corresponding to the mismatched transportation demand; Based on the owned vessel cost, the social vessel cost, and the mismatch penalty cost, a cost objective evaluation function is constructed. The cost target evaluation function The calculation expression is: ； in, This is a collection of transportation needs for the day. For the assembly of ships, Indicates the first The daily oil transportation demand, Indicates the first Candidate ships, For the first The ship carried out the first Matching costs for transportation demand For the decision variable of arranging boats, when the first The transportation demand is determined by the first When the ship is in operation ,otherwise , For the first The penalty cost for mismatched transportation demand. For unmatched variables, when the first When the transport demand does not match the number of vessels ,otherwise .

6. The intelligent scheduling method for oil tankers based on multi-objective evaluation and meta-heuristic search according to claim 5, characterized in that, The normalization process for the boat dispatching business objective evaluation function and the cost objective evaluation function to obtain normalized boat dispatching business objective evaluation values ​​and normalized cost objective evaluation values ​​includes: Obtain the business target value of each sub-target in the boat dispatching business target evaluation function, and the cost target value in the cost target evaluation function; Determine whether each objective is a minimization objective or a maximization objective; For the minimization objective, normalization is performed using the following expression: in, For the first The goal is to schedule a boat trip in Japan. The normalized target value is below. For the first The goal is to schedule a boat trip in Japan. The target value below, and These are the first and second candidate solutions in the current candidate solution set. The minimum and maximum values ​​of each objective. To prevent constants with a denominator of zero; For a maximization objective, it is converted into a minimization objective and then normalized. We obtained the normalized evaluation values ​​for the shunting business objectives and the normalized evaluation values ​​for the cost objectives.

7. The intelligent scheduling method for oil tankers based on multi-objective evaluation and meta-heuristic search according to claim 6, characterized in that, The step of constructing a multi-objective fusion evaluation function based on the normalized boat scheduling business target evaluation value and the normalized cost target evaluation value, and dynamically adjusting the target weights in the multi-objective fusion evaluation function according to the boat scheduling scenario of the day, includes: Based on the normalized evaluation value of the boat dispatching business target and normalized cost target evaluation value Construct a multi-objective fusion evaluation function Its calculation expression is: ； in, Weighting the integration of boat dispatching business objectives and cost objectives; Based on the parameters of the boat scheduling scenario for the day, adjust the internal weights of at least one sub-objective in the boat scheduling business objective evaluation function, and adjust the fusion weights between the boat scheduling business objective and the cost objective. Specifically, when the number or proportion of urgent orders exceeds a preset threshold, the internal weight corresponding to the urgent order priority target is increased; when the number of candidate vessels with scarce demand exceeds a preset threshold, the internal weight corresponding to the scarce demand protection target is increased; when the proportion of social vessels or cost pressure indicators exceed a preset threshold, the fusion weight is decreased. The value of is determined to increase the weight of cost targets in the fusion evaluation.

8. The intelligent scheduling method for oil tankers based on multi-objective evaluation and meta-heuristic search according to claim 3, characterized in that, The process involves constructing a daily ship scheduling scheme using transportation demand as the encoding location and ship name as the gene value, thereby generating an initial daily ship scheduling scheme. Starting from this initial daily ship scheduling scheme, iterative optimization is performed: in each iteration, a metaheuristic optimization algorithm is invoked to generate candidate daily ship scheduling schemes within the feasible search space. The multi-objective fusion evaluation value of the candidate daily ship scheduling schemes is calculated based on the multi-objective fusion evaluation function. Based on the multi-objective fusion evaluation value, it is determined whether to accept the candidate daily ship scheduling scheme, and the current daily ship scheduling scheme and the historical best daily ship scheduling scheme are updated. The iteration stops when a preset termination condition is met, and the historically optimal daily boat scheduling scheme is taken as the target daily boat scheduling scheme, including: Using each transport demand of the day as a coding position, and the ship name assigned to each transport demand as the gene value at the corresponding coding position, a daily ship scheduling scheme code is constructed; whereby a daily ship scheduling scheme is represented as: ,in, Indicates assignment to the first The name of the ship for the transportation demand, if the first If the transport demand is not matched with the vessels, then This is a non-matching identifier; In each iteration, the parent daily boat dispatching scheme is selected based on the multi-objective fusion evaluation value of each daily boat dispatching scheme; By exchanging the ship name genes at the transportation demand positions of the two parent generation daily ship dispatch schemes, a child generation daily ship dispatch scheme is formed. The ship name genes in the progeny daily ship dispatch scheme are mutated, and the feasibility of the mutated daily ship dispatch scheme is corrected to obtain candidate daily ship dispatch schemes. Calculate the difference in multi-objective fusion evaluation values ​​between the candidate daily ship dispatch scheme and the current daily ship dispatch scheme. , For candidate day scheduling plans; If the multi-objective fusion evaluation value of the candidate day's boat scheduling plan is better than that of the current day's boat scheduling plan, then the candidate day's boat scheduling plan is accepted as the current day's boat scheduling plan; if the cost target of the candidate day's boat scheduling plan is improved and the decrease in the boat scheduling business target does not exceed a preset threshold, then the candidate day's boat scheduling plan is accepted as the current day's boat scheduling plan; if the boat scheduling business target of the candidate day's boat scheduling plan is improved and the increase in the cost target does not exceed a preset threshold, then the candidate day's boat scheduling plan is accepted as the current day's boat scheduling plan. If the multi-objective fusion evaluation value of the current daily boat scheduling plan after acceptance is better than the multi-objective fusion evaluation value of the historical best daily boat scheduling plan, then the current daily boat scheduling plan will be updated to the historical best daily boat scheduling plan. Repeat the above iterative steps until the preset termination condition is met, and output the historical best daily boat scheduling scheme as the target daily boat scheduling scheme.

9. A computer-readable storage medium, characterized in that, The storage medium stores program information, and after the computer reads the program information, it executes the steps of the intelligent scheduling method for oil tankers based on multi-objective evaluation and meta-heuristic search as described in any one of claims 1-8.

10. An intelligent scheduling system for oil tankers based on multi-objective evaluation and meta-heuristic search, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the intelligent scheduling method for oil tankers based on multi-objective evaluation and meta-heuristic search as described in any one of claims 1-8.

Images