Tanker scheduling method, device and chip

Abstract

Embodiments of the present invention provide a tanker scheduling method, apparatus, and chip. The tanker scheduling method includes: acquiring characteristic data and historical operational data; determining a crude oil type prohibition matrix corresponding to a target sea area based on the characteristic data; determining bulkhead residue data corresponding to each tanker based on the historical operational data; acquiring multi-port mixed loading task data, which includes multiple loading nodes and crude oil type loading sequences corresponding to the multiple loading nodes; determining a candidate tanker set based on the multi-port mixed loading task data and the multiple tankers; determining a compatibility risk coefficient for each tanker in the candidate tanker set; determining a risk response result for each tanker in the candidate tanker set based on the compatibility risk coefficient; and determining a scheduling instruction corresponding to the candidate tanker set based on the risk response result. The solution of the present invention significantly improves the tanker scheduling efficiency for continuous multi-port, multi-crude oil type loading tasks.

Description

Technical Field

[0001] This invention relates to the field of maritime management technology, and more specifically, to a tanker scheduling method, apparatus, and chip. Background Technology

[0002] In the current maritime crude oil transportation sector, particularly in multi-port mixed loading operations involving multi-floating production storage and offloading (FPSO) units, traditional methods heavily rely on the personal experience of dispatchers. This makes it difficult to systematically and quantitatively assess the potential chemical conflict risks, such as asphaltene precipitation and sedimentation, that may occur when a specific tanker, under historical loading conditions, performs a continuous loading operation containing multiple crude oil types. This often leads to a conservative, fixed-cycle tank washing strategy in the dispatching process. This can result in unnecessary washing costs and time delays, and may also lead to actual oil contamination incidents due to insufficient risk assessment. Achieving an effective balance between transportation safety and economic efficiency is difficult, resulting in inefficient tanker dispatching. Summary of the Invention

[0003] The purpose of this invention is to provide a tanker scheduling method, device, and chip that can solve the problem of insufficient efficiency in tanker scheduling, which makes it difficult to achieve an effective balance between transportation safety and economic efficiency during the transportation of crude oil at sea.

[0004] In view of this, an embodiment of the first aspect of the present invention provides a tanker scheduling method.

[0005] A second aspect of the present invention provides a tanker dispatching device.

[0006] An embodiment of the third aspect of the present invention provides a chip.

[0007] To achieve the above objectives, an embodiment of the first aspect of the present invention provides a tanker scheduling method, comprising: acquiring characteristic data of multiple crude oil varieties and historical operational data of multiple tankers within a target sea area; determining a crude oil type incompatibility matrix corresponding to the target sea area based on the characteristic data; determining bulkhead residue data corresponding to each tanker based on the historical operational data; acquiring multi-port mixed loading task data, the multi-port mixed loading task data including multiple loading nodes and crude oil type loading sequences corresponding to the multiple loading nodes; determining a candidate tanker set based on the multi-port mixed loading task data and the multiple tankers; for each tanker in the candidate tanker set, determining a compatibility risk coefficient based on the crude oil type incompatibility matrix, bulkhead residue data, and crude oil type loading sequence; determining a risk response result for each tanker in the candidate tanker set based on the compatibility risk coefficient; and determining a scheduling instruction corresponding to the candidate tanker set based on the risk response result.

[0008] This invention provides a tanker scheduling method. By constructing an oil type incompatibility matrix that quantifies the chemical compatibility between different crude oils and a bulkhead memory file that dynamically records the residual state of each tanker's bulkheads, upon receiving a mixed loading task containing multiple oil type sequences, it can perform a precise and quantitative dynamic compatibility risk assessment on candidate tankers. Based on the real-time assessed compatibility risk coefficient, a tiered response mechanism is activated, automatically generating a complete scheduling instruction that integrates target tanker assignment, customized tank cleaning instructions, navigation path, and risk warnings.

[0009] Understandably, dynamic quantitative assessment avoids a large number of unnecessary conservative tank cleaning operations, directly saving tank cleaning costs and berthing time. At the same time, its systematic risk assessment and conflict resolution mechanism can quickly match suitable tankers for executable tasks and generate optimal solutions while ensuring transportation safety. This reduces decision-making delays and ship waiting caused by risk uncertainty, and significantly improves the tanker scheduling efficiency for continuous loading tasks involving multiple ports and multiple oil types.

[0010] In some technical solutions, optionally, the risk response result for each tanker in the candidate tanker set is determined based on a compatibility risk coefficient, including: obtaining a first preset threshold and a second preset threshold, wherein the first preset threshold is less than the second preset threshold; when the compatibility risk coefficient is less than or equal to the first preset threshold, determining a first execution result for the tanker corresponding to the multi-port mixed loading task data; when the compatibility risk coefficient is greater than the first preset threshold and less than or equal to the second preset threshold, determining a second execution result for the tanker corresponding to the multi-port mixed loading task data; when the compatibility risk coefficient is greater than the second preset threshold, determining a third execution result for the tanker corresponding to the multi-port mixed loading task data; and determining a risk response result based on the first execution result, the second execution result, and the third execution result.

[0011] In this solution, by introducing two key quantitative benchmarks, the first preset threshold and the second preset threshold, the continuous risk coefficient is divided into three distinct level ranges: low, medium, and high. The first execution result, the second execution result, and the third execution result are generated accordingly. The risk response result is determined based on the multiple different execution results.

[0012] Understandably, by setting clear thresholds, any risk level can be instantly classified without repeated manual assessment. For low-risk situations (the first execution result), the system can immediately make a "no-wash straight-through" decision, eliminating unnecessary tank washing assessment and preparation time, and maximizing scheduling efficiency. This greatly improves the efficiency and accuracy of scheduling decisions.

[0013] In some technical solutions, optionally, determining the second execution result for the tanker corresponding to the multi-port mixed loading task data includes: determining multiple candidate tank cleaning parameters corresponding to the tanker based on bulkhead residue data and multiple preset tank cleaning depths; determining the comprehensive tank cleaning residue data corresponding to each candidate tank cleaning parameter based on bulkhead residue data and multiple preset tank cleaning depths; determining the post-tank cleaning risk coefficient corresponding to each candidate tank cleaning parameter based on the comprehensive tank cleaning residue data, the oil type incompatibility matrix, and the oil type loading sequence; obtaining the expected tank cleaning cost and preset weight coefficient corresponding to each candidate tank cleaning parameter; determining the comprehensive cost based on the post-tank cleaning risk coefficient, the expected tank cleaning cost, and the preset weight coefficient; determining the target tank cleaning parameter from the multiple candidate tank cleaning parameters based on the comprehensive cost; and determining the second execution result based on the target tank cleaning parameter.

[0014] In this scheme, the residual bulkhead state after each tank washing scheme is predicted by the tank washing depth and efficiency coefficient. The post-washing risk coefficient is then recalculated to assess its safety effect. Combined with the projected tank washing cost, which includes direct, time, and opportunity costs, and using preset weighting coefficients, a unified quantitative score is applied to all candidate schemes through a comprehensive cost function. The scheme with the lowest comprehensive cost value is selected as the target tank washing parameter, thereby determining the final operational instructions.

[0015] Understandably, this provides a precise, economical, and automated decision-making tool for medium-risk scenarios, fundamentally overcoming the drawbacks of relying on fixed-cycle tank cleaning or experience-based judgment, and significantly improving scheduling efficiency and economic benefits.

[0016] In some technical solutions, optionally, after determining the risk response result for each tanker in the candidate tanker set based on the compatibility risk coefficient, the tanker scheduling method further includes: determining whether there are available tankers in the candidate tanker set that meet preset safety conditions based on the risk response result; if there are no available tankers that meet the preset safety conditions, then performing task reconstruction processing on the multi-port mixed loading task data; wherein, the task reconstruction processing includes at least one of the following methods: re-screening and updating the candidate tanker set from multiple tankers based on the oil type taboo matrix and bulkhead residual data; adjusting the loading order of multiple oil types in the oil type loading sequence to determine the iterative oil type loading sequence; or splitting the multi-port mixed loading task data into multiple task subsets.

[0017] In this scheme, when the system determines that there are no available tankers that meet the preset safety conditions in the current candidate tanker set, it performs a task reconstruction process that includes multiple strategies, namely, re-screening tankers based on the chemical knowledge base, optimizing and adjusting the loading order of multiple oil types, or splitting the original task into multiple independently executable sub-tasks.

[0018] Understandably, this upgrades scheduling decisions from a one-off, static matching process to a dynamic optimization process that includes feedback, adjustments, and replanning. By re-screening, it can uncover potential vessels from the entire fleet that were initially overlooked but have better chemical compatibility, thereby improving the overall utilization of fleet resources.

[0019] In some technical solutions, optionally, an oil type incompatibility matrix corresponding to the target sea area is determined based on characteristic data, including: determining chemical compatibility data among multiple crude oil varieties based on characteristic data; determining the compatibility risk coefficient when mixing each pair of crude oil varieties based on chemical compatibility data; and determining the oil type incompatibility matrix based on all compatibility risk coefficients.

[0020] This scheme specifically defines the method for constructing the oil type incompatibility matrix. Its core steps include: extracting chemical compatibility data characterizing the interactions between oil products based on multi-source heterogeneous characteristic data (including static physicochemical indicators, historical loss records, laboratory mixing tests, and refinery feedback). Then, key indicators such as asphaltene precipitation tendency, precipitation formation rate, and the rate of change in American Petroleum Institute (API) specific gravity are selected. Through normalization and weighted summation using preset weighting coefficients, the quantitative compatibility risk coefficient for each pair of oil types is calculated. The risk coefficients of all oil type pairs are then organized into a symmetric matrix, namely the oil type incompatibility matrix.

[0021] Understandably, the oil type incompatibility matrix allows complex chemical compatibility issues to be abstracted into rapid matrix query operations, greatly improving the objectivity, consistency, and speed of risk assessment. It overcomes the efficiency bottleneck and subjective differences of manual judgment, thereby fundamentally improving the decision-making efficiency and reliability of the entire scheduling system in dealing with complex oil product mixing scenarios.

[0022] In some technical solutions, optionally, the bulkhead residue data corresponding to each tanker can be determined based on historical operating data, including: determining a list of historical voyages of the tanker based on historical operating data, wherein each voyage record in the historical voyage list includes the type of oil loaded, the tank cleaning depth, and the residue assessment value; and determining the bulkhead residue data of the tanker through a recursive relationship based on a preset tank cleaning efficiency coefficient, tank cleaning depth, and residue assessment value.

[0023] In this scheme, the bulkhead residual status is dynamically quantified through recursive relationships. The sequence of loaded oil type, tank cleaning depth and residual assessment value is extracted from historical operational data in units of voyage. Then, based on the tank cleaning efficiency coefficient corresponding to different preset tank cleaning depths, the historical residual assessment values ​​are calculated to decay sequentially. This simulates the superposition process of the cleaning effect of each tank cleaning and the new residual, thereby calculating the bulkhead residual data that reflects the current overall cleanliness of the cargo hold.

[0024] Understandably, transforming the historical loading information of tanker bulkheads from discrete, static records into a continuous, quantifiable, dynamic state model provides crucial state input for accurately assessing chemical compatibility risks. By mathematically recursively quantifying the combined residual impact of historical oil types after multiple tank cleaning operations, the risk assessment more accurately reflects the actual condition of the tanker, significantly improving the accuracy of risk warnings and avoiding potential contamination accidents due to insufficient estimation of historical bulkhead residues, or overly conservative cleaning decisions. This lays a reliable data foundation for subsequent cost-benefit analysis and optimized scheduling.

[0025] In some technical solutions, optionally, the risk response outcome for each tanker in the candidate tanker set is determined based on the compatibility risk coefficient, including: determining the adjacent conflict coefficient of adjacent oil types within the sequence based on the oil type loading sequence and the oil type taboo matrix; determining the historical residual conflict coefficient based on the residual oil type indicated by the bulkhead residual data, the first loaded oil type in the oil type loading sequence, and the oil type taboo matrix; and determining the compatibility risk coefficient based on the adjacent conflict coefficient and the historical residual conflict coefficient.

[0026] In this scheme, based on the pre-determined oil loading sequence for the mission, the inherent risks of all adjacent oil type pairs are comprehensively analyzed by querying the oil type taboo matrix to obtain the adjacent conflict coefficient, which characterizes the inherent continuous risk of the mission. Based on the residual oil type and its intensity indicated by the tanker's bulkhead residual data, and combined with the initial oil type loaded for the mission, the taboo matrix is ​​queried again to calculate the historical residual conflict coefficient, which reflects the conflict between the tanker's own historical state and the initial requirements of the mission. These two coefficients, derived from the mission plan and the vessel's state respectively, are combined using preset rules (such as taking the maximum value or weighted summation) to determine a comprehensive compatibility risk coefficient.

[0027] Understandably, the compatibility risk coefficient calculated by the model can more accurately reflect the true risk profile of a specific tanker when performing a specific mixed loading task. This provides a reliable and consistent quantitative benchmark for subsequent accurate risk classification, tank cleaning decisions, and conflict resolution, thereby fundamentally improving the scientific nature and safety of scheduling decisions and significantly reducing the risk of cargo damage accidents caused by incomplete risk assessment.

[0028] In some technical solutions, optionally, the dispatching instructions include target tanker identification, tank cleaning operation instructions, navigation path instructions, and risk warning information. The dispatching instructions corresponding to the candidate tanker set are determined based on the risk response results, including: determining the target tanker from the candidate tanker set based on the risk response results of each tanker; determining the tank cleaning operation instructions and risk warning information based on the risk response results of the target tanker; determining navigation path data based on multi-port mixed loading task data; and determining the dispatching instructions corresponding to the candidate tanker set based on the navigation path data.

[0029] In this scheme, based on the risk response results of each candidate tanker, the target tanker for the final task is determined through preset optimization rules (such as prioritizing direct execution and then selecting the tanker with the lowest overall cost). Then, specific and operable tank cleaning operation instructions are directly mapped from the risk response results of the target tanker, and corresponding risk warning information is generated based on its remaining risks. Simultaneously, based on the node sequence, loading volume, and time window in the multi-port mixed loading task data, combined with ship performance parameters, a detailed navigation path and time plan are calculated. The target tanker identification, tank cleaning operation instructions, navigation path instructions, and risk warning information are encapsulated into a complete scheduling instruction for execution.

[0030] Understandably, automated generation eliminates errors, omissions, and delays that might result from manual information integration and transmission, ensuring that decision-making intentions are fully and accurately conveyed to the execution end. The embedded risk warning information directly transforms the risk conclusions of the back-end assessment into preventative operational guidelines for the front end, enhancing the safety and controllability of the operational process. Ultimately, this significantly shortens the cycle from decision-making to execution for the entire multi-port mixed loading task, providing clear and unambiguous instructions and significantly improving the fleet's response speed, safety level, and management efficiency.

[0031] An embodiment of the second aspect of the present invention provides a tanker scheduling device, comprising: a data acquisition module for acquiring characteristic data of multiple crude oil varieties and historical operational data of multiple tankers within a target sea area; a matrix determination module for determining a crude oil type incompatibility matrix corresponding to the target sea area based on the characteristic data; a residue determination module for determining bulkhead residue data corresponding to each tanker based on the historical operational data; a task acquisition module for acquiring multi-port mixed loading task data, the multi-port mixed loading task data including multiple loading nodes and crude oil type loading sequences corresponding to the multiple loading nodes; a set determination module for determining a candidate tanker set based on the multi-port mixed loading task data and the multiple tankers; a risk determination module for determining a compatibility risk coefficient for each tanker in the candidate tanker set based on the crude oil type incompatibility matrix, bulkhead residue data, and crude oil type loading sequence; a result determination module for determining a risk response result for each tanker in the candidate tanker set based on the compatibility risk coefficient; and a tanker scheduling module for determining a scheduling instruction corresponding to the candidate tanker set based on the risk response result.

[0032] An embodiment of the third aspect of this application provides a chip including a processor and a communication interface, the communication interface and the processor being coupled together, the processor being used to run programs or instructions to implement the steps of the tanker scheduling method as described in the first aspect.

[0033] Additional aspects and advantages of the technical solutions of the present invention will become apparent in the following description or may be learned by practice of the invention. Attached Figure Description

[0034] Figure 1 One of the flowcharts of the tanker dispatching method according to this application is shown;

[0035] Figure 2 A second schematic flowchart of the tanker dispatching method according to this application is shown;

[0036] Figure 3 A third schematic flowchart of the tanker dispatching method according to this application is shown;

[0037] Figure 4 A fourth schematic flowchart of the tanker dispatching method according to this application is shown;

[0038] Figure 5 Fifth of the flowcharts illustrating the tanker dispatching method according to this application is shown;

[0039] Figure 6 Sixth schematic diagram of the tanker dispatching method according to this application is shown;

[0040] Figure 7 A flowchart of the tanker dispatching method according to this application is shown in diagram seven;

[0041] Figure 8 Eighth schematic diagram of the tanker dispatching method according to this application is shown;

[0042] Figure 9 A schematic block diagram of the structure of a tanker dispatching device according to this application is shown;

[0043] Figure 10 A system structure diagram corresponding to the tanker dispatching method according to this application is shown;

[0044] Figure 11 A schematic diagram of the workflow of the tanker dispatching method according to this application is shown.

[0045] Among them, 900: tanker dispatching device; 902: data acquisition module; 904: matrix determination module; 906: residue determination module; 908: task acquisition module; 910: set determination module; 912: risk determination module; 914: result determination module; 916: tanker dispatching module. Detailed Implementation

[0046] To better understand the above-described objectives, features, and advantages of the embodiments of the present invention, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0047] Many specific details are set forth in the following description in order to provide a full understanding of this application. However, embodiments of the invention may also be implemented in other ways different from those described herein. Therefore, the scope of protection of this application is not limited to the specific embodiments disclosed below.

[0048] The tanker scheduling method, apparatus, and chip provided in this application will be described in detail below with reference to specific embodiments and application scenarios.

[0049] This embodiment provides a tanker scheduling method, such as Figure 1 As shown, the tanker dispatching method includes:

[0050] Step S100: Obtain characteristic data of multiple crude oil varieties and historical operational data of multiple oil tankers within the target sea area;

[0051] Step S102: Determine the oil type taboo matrix corresponding to the target sea area based on the characteristic data;

[0052] Step S104: Determine the bulkhead residual data corresponding to each tanker based on historical operating data;

[0053] Step S106: Obtain multi-port mixed loading task data, which includes multiple loading nodes and oil type loading sequences corresponding to the multiple loading nodes;

[0054] Step S108: Determine the candidate tanker set based on multi-port mixed loading task data and multiple tankers;

[0055] Step S110: For each tanker in the candidate tanker set, determine the compatibility risk coefficient based on the oil type incompatibility matrix, bulkhead residual data, and oil type loading sequence;

[0056] Step S112: Determine the risk response outcome for each tanker in the candidate tanker set based on the compatibility risk coefficient;

[0057] Step S114: Determine the scheduling instructions corresponding to the candidate tanker set based on the risk response results.

[0058] This invention provides a tanker scheduling method. By constructing an oil type incompatibility matrix that quantifies the chemical compatibility between different crude oils and a bulkhead memory file that dynamically records the residual state of each tanker's bulkheads, upon receiving a mixed loading task containing multiple oil type sequences, it can perform a precise and quantitative dynamic compatibility risk assessment on candidate tankers. Based on the real-time assessed compatibility risk coefficient, a tiered response mechanism is activated, automatically generating a complete scheduling instruction that integrates target tanker assignment, customized tank cleaning instructions, navigation path, and risk warnings.

[0059] Understandably, dynamic quantitative assessment avoids a large number of unnecessary conservative tank cleaning operations, directly saving tank cleaning costs and berthing time. At the same time, its systematic risk assessment and conflict resolution mechanism can quickly match suitable tankers for executable tasks and generate optimal solutions while ensuring transportation safety. This reduces decision-making delays and ship waiting caused by risk uncertainty, and significantly improves the tanker scheduling efficiency for continuous loading tasks involving multiple ports and multiple oil types.

[0060] First, acquire characteristic data of multiple crude oil varieties in the target sea area. This characteristic data includes, but is not limited to, a series of key physicochemical indicators such as asphaltene content, gum content, sulfur content, acid value, wax content, and American Petroleum Institute (API) specific gravity of multiple crude oil varieties.

[0061] At the same time, historical operational data of multiple oil tankers were obtained. This historical operational data recorded in detail the specific type of oil loaded by each oil tanker on each past voyage, the depth of tank cleaning operations after unloading, and the estimated value of residual cargo oil in the tank walls through sensor detection or empirical formulas.

[0062] Based on the acquired crude oil characteristic data, a crude oil type incompatibility matrix corresponding to the target sea area is determined. Using historical cargo loss records, laboratory mixing experiment data, and refinery process feedback, the chemical compatibility relationship between any two crude oils is analyzed and quantified.

[0063] Specifically, based on measured data from multiple dimensions such as asphaltene precipitation tendency, precipitation formation rate, and API degree change rate, a compatibility risk coefficient between zero and one is assigned to each pair of oil types through normalization and weighted calculation using preset weighting coefficients. The closer the value is to one, the higher the risk of chemical conflicts such as precipitation and blockage when the two oils are mixed.

[0064] The risk coefficients of all oil pairs form a symmetric matrix, namely the oil taboo matrix.

[0065] The process of determining the crude oil incompatibility matrix involves transforming raw data into a risk quantification model. This involves drawing upon historical cargo loss case reports, laboratory datasets of paired crude oil blending compatibility tests, and process feedback data from downstream refineries regarding the stability of blended crude oil processing. Using this data, the chemical compatibility between any two crude oil varieties is quantified, determining the compatibility risk coefficient.

[0066] The compatibility risk coefficient ranges from zero to one; a higher value indicates poorer compatibility between the two oil types and a higher risk of mixing conflicts. Filling the compatibility risk coefficients of all oil type pairs into a symmetric matrix creates the oil type incompatibility matrix.

[0067] Simultaneously, based on the acquired historical operational data of the tankers, the bulkhead residue data corresponding to each tanker was determined. A recursive calculation model that considers the tank washing effect was used to dynamically quantify the current composite residue state of the tanker bulkheads.

[0068] The system iterates through the historical voyage list of oil tankers and, in conjunction with the preset cleaning efficiency coefficient corresponding to the tank cleaning depth after each voyage, calculates the cumulative attenuation of the residual assessment values ​​from each voyage.

[0069] For example, deep chemical cleaning can remove the vast majority of residues, while conventional seawater flushing can only remove a portion. The bulkhead residue data ultimately calculated through this recursive relationship is a quantitative value that characterizes the current level of residual risk to the bulkhead due to the influence of historically carried oil types.

[0070] When a new transportation task is available, the system retrieves multi-port mixed loading task data.

[0071] The multi-port mixed loading task data clearly specifies the multiple loading nodes that the task needs to visit in sequence, such as different floating production storage and offloading units, and the specific crude oil type planned to be loaded at each node, thus forming an orderly oil type loading sequence.

[0072] Subsequently, based on the data of multi-port mixed loading missions and the status of multiple tankers in the fleet, a preliminary screening was conducted to determine the candidate tanker set.

[0073] The basic criteria for selection include the tanker’s current availability, whether its total cargo capacity meets the mission’s total loading requirements, and whether its current location can reach the first loading node within the mission time window.

[0074] For each tanker in the candidate tanker set, the system initiates the core risk assessment process, which determines a comprehensive compatibility risk coefficient based on the oil type incompatibility matrix, the tanker's unique bulkhead residual data, and the oil type loading sequence given by the mission.

[0075] First, calculate the inherent sequence risk of the task, that is, query the compatibility risk coefficient between each pair of adjacent oil types in the oil type taboo matrix, and take the maximum value, which represents the direct mixing conflict that may occur when the planned oil products are loaded into the tank in sequence.

[0076] Second, calculate the risk introduced by historical residues on the carrier. That is, based on the main residual oil type indicated by the bulkhead residue data, look up the compatibility risk coefficient between it and the first oil type loaded in the mission from the oil type taboo matrix, and multiply the coefficient by the bulkhead residue data value representing the amount of residue to obtain the historical residue conflict coefficient.

[0077] The historical residual conflict coefficient is compared with the aforementioned maximum adjacent conflict coefficient, and the larger of the two values ​​is determined as the final compatibility risk coefficient for the tanker to perform this task.

[0078] Then, based on the calculated compatibility risk coefficient, the risk response outcome for each tanker in the candidate tanker set is determined.

[0079] By comparing the compatibility risk coefficient with preset safety and warning thresholds, risks are classified into different levels.

[0080] The system generates different execution strategies for different levels of risk. For example, if the risk is extremely low, tank cleaning can be performed directly without cleaning; if the risk is moderate, the most cost-effective tank cleaning solution should be recommended; and if the risk is too high, deep cleaning may be required or the tanker may be excluded from the evaluation. The risk response results include the final execution recommendations and requirements for that tanker.

[0081] Integrating all analytical conclusions, dispatch instructions corresponding to the candidate tanker set are determined based on the risk response results. The dispatch instructions are an executable operational document that clearly specifies the selected target tankers, stipulates the specific depth and requirements for tank cleaning operations, plans the navigation routes and timelines for their visits to various loading nodes, and includes risk warnings based on residual risk to guide on-site operators in taking appropriate monitoring measures.

[0082] In some embodiments, the oil type loading sequence in the multi-port mixed loading task data is preset and adjustable. The scheduling system can dynamically optimize the access order of multiple loading nodes based on the tanker fleet status and oil type taboo matrix, under the premise of meeting the total loading volume of the task, in order to explore a better task execution sequence that can reduce the overall compatibility risk coefficient, thereby proactively avoiding high-risk combinations in the task planning stage and improving the global optimization capability of the scheduling scheme.

[0083] In some embodiments, the tank cleaning efficiency coefficient used to determine bulkhead residue data is not a fixed value, but is determined by a sub-model through personalized calculation based on the specific tanker model, bulkhead material, historical tank cleaning effect records, and the type of cleaning agent and equipment capacity that the port where the tank cleaning operation is planned can provide. This makes the quantitative estimation of the residue state more consistent with the actual physical state and operating conditions of a single ship.

[0084] In some embodiments, optionally, in determining the risk response outcome based on the compatibility risk coefficient, the system considers not only the immediate risk of a single task but also the prediction of the long-term bulkhead condition evolution of the tanker. After simulating the execution of the current task, the system updates the predicted bulkhead residue data of the tanker in conjunction with the recommended tank cleaning scheme and assesses its compatibility impact on subsequent potential task queues, thereby incorporating considerations of the long-term maintenance value of the ship assets and the overall adaptability of the task pool into the scheduling decision.

[0085] In some embodiments, the final generated dispatch instructions may optionally include, in addition to basic elements such as the target tanker, tank cleaning instructions, and navigation route, personalized operational suggestions based on mission risk characteristics. These suggestions may include recommended delivery rates for specific oil type conversion nodes, tank temperature monitoring requirements for critical segments, or recommended online detection frequencies for high-risk mixed loading phases, extending macro-level dispatch decisions to micro-level operational guidance and forming a closed-loop operational safety assurance system.

[0086] In some embodiments, optionally, such as Figure 2 As shown, step S112: Determine the risk response outcome for each tanker in the candidate tanker set based on the compatibility risk coefficient, including:

[0087] Step S1120: Obtain a first preset threshold and a second preset threshold, wherein the first preset threshold is less than the second preset threshold;

[0088] Step S1122: When the compatibility risk coefficient is less than or equal to the first preset threshold, determine the first execution result of the tanker corresponding to the multi-port mixed loading task data;

[0089] Step S1124: When the compatibility risk coefficient is greater than the first preset threshold and less than or equal to the second preset threshold, determine the second execution result of the tanker corresponding to the multi-port mixed loading task data;

[0090] Step S1126: When the compatibility risk coefficient is greater than the second preset threshold, determine the third execution result of the tanker corresponding to the multi-port mixed loading task data;

[0091] Step S1128: Determine the risk response outcome based on the first execution result, the second execution result, and the third execution result.

[0092] In this embodiment, by introducing two key quantitative benchmarks, the first preset threshold and the second preset threshold, the continuous risk coefficient is divided into three distinct level ranges: low, medium, and high. The first execution result, the second execution result, and the third execution result are generated accordingly. The risk response result is determined based on the multiple different execution results.

[0093] Understandably, by setting clear thresholds, any risk factor can be instantly classified without repeated manual assessment. For low-risk situations (first execution result), the system can immediately make a "straight ahead without washing" decision, saving unnecessary tank washing assessment and preparation time, and maximizing scheduling efficiency. For medium-risk situations (second execution result), the system can automatically trigger and guide the next stage of refined cost-benefit analysis, avoiding decision-making stagnation or overly conservative tank washing due to ambiguous risk assessment, ensuring that the optimal solution for safety and cost is found within a controllable timeframe. For high-risk situations (third execution result), the system can immediately take measures such as exclusion, escalation, or trigger task refactoring, preventing high-risk tasks from entering subsequent lengthy assessment stages, effectively shortening the overall scheduling cycle, and avoiding vessel scheduling conflicts caused by slow decision-making. This hierarchical triggering mechanism enables the scheduling system to operate as efficiently as a "decision tree," quickly matching the optimal processing path for different risk levels, fundamentally solving the problems of decision delays, inconsistent standards, and low efficiency caused by traditional reliance on manual experience judgment, and greatly improving the efficiency and accuracy of scheduling decisions.

[0094] Obtain the first preset threshold and the second preset threshold from the preset configuration parameters.

[0095] The first and second preset thresholds are key decision boundaries jointly set based on long-term operational experience, industry safety standards, and the acceptable probability of accidents for different risk levels.

[0096] The first preset threshold is usually set to a low value, representing the boundary between "safe" and "needs attention"; the second preset threshold is set to a high value, representing the boundary between "needs mitigation measures" and "high-risk prohibition".

[0097] The first preset threshold must be less than the second preset threshold, so as to clearly divide the three logical intervals of low, medium and high on the number line.

[0098] For example, the first preset threshold can be set to 0.3, and the second preset threshold can be set to 0.6.

[0099] These thresholds are not fixed and can be dynamically adjusted by the operator in the system management interface according to the risk preferences of different routes, the value of the fuel carried, or specific safety activity requirements.

[0100] After obtaining the thresholds, the previously calculated compatibility risk coefficients for specific tankers are compared sequentially with these two thresholds. This comparison follows a rigorous logical judgment process, aiming to assign a unique and clear risk status label to each tanker.

[0101] When the compatibility risk coefficient is determined to be less than or equal to the first preset threshold, the tanker is identified as the first execution result for the current multi-port mixed loading task data. This indicates that, based on comprehensive assessment, the chemical conflict risk of the tanker performing this task is extremely low, placing it within a completely acceptable "safe zone." The internal instruction meaning of the first execution result is "Direct execution permitted, without any additional tank cleaning operations."

[0102] When the system determines that the compatibility risk coefficient is greater than the first preset threshold and simultaneously less than or equal to the second preset threshold, it determines the tanker as the second execution result corresponding to the current multi-port mixed loading task data. This indicates that the risk has exceeded the safe range and entered the "alert zone," indicating a clear but not urgent chemical compatibility hazard.

[0103] The second execution result itself is a signal that triggers a more complex analysis process, meaning "the tanker has a moderate risk and the tank cleaning scheme cost-benefit analysis subsystem needs to be activated for further evaluation." It does not directly output the final operation command, but rather indicates that the decision-making process needs to enter the next refined analysis stage, which includes multi-scheme simulation, cost calculation, and optimization comparison.

[0104] When the system determines that the compatibility risk coefficient is greater than the second preset threshold, it determines that the tanker corresponds to the third execution result of the current multi-port mixed loading task data.

[0105] This situation indicates that the risk has entered a "high-risk zone," and the possibility of chemical conflict is very high.

[0106] The initial instruction corresponding to the third execution result within the system usually means "the risk is too high, it is not recommended to assign it directly or it is difficult to solve by conventional tank washing alone".

[0107] The third execution result will lead to two subsequent paths: one is that the system will directly downgrade or exclude the tanker from the current task's feasible candidate list; the other is that it will trigger a special handling mechanism for such high risks, such as forcibly evaluating the effectiveness of deep tank cleaning, or directly transferring to a higher-level task conflict resolution process.

[0108] For example, if the first preset threshold is set to 0.3 and the second preset threshold is 0.6, tanker A has a compatibility risk coefficient of 0.25. Since it is less than or equal to 0.3, it is assigned the first execution result, meaning it can directly perform the task without tank cleaning. Tanker B has a coefficient of 0.45, falling between 0.3 and 0.6, and is assigned the second execution result. The system will calculate the costs of no cleaning, regular tank cleaning, and deep cleaning for it. Tanker C has a coefficient of 0.75, higher than 0.6, and is assigned the third execution result. The system may first consider forcing it to undergo deep cleaning and then reassessing. If the requirements are still not met, it may suggest replacing it with another tanker.

[0109] Based on the first, second, or third execution results determined for each candidate tanker, the final risk response results for the tankers are summarized and determined.

[0110] This "Risk Response Outcome" is a structured output object that integrates the aforementioned classification tags and more specific decision-making details under each tag. For tankers marked as the first execution outcome, the risk response outcome is "direct execution"; for tankers marked as the second execution outcome, the risk response outcome will include the "optimal tank cleaning plan and execution instructions" recommended after cost-benefit analysis; for tankers marked as the third execution outcome, the risk response outcome may be "reject assignment," "deep tank cleaning and reassessment required," or "initiate mission reconfiguration recommendations." This finalized risk response outcome will serve as the direct basis for generating specific dispatch instructions.

[0111] In some embodiments, optionally, the values ​​of the first preset threshold and the second preset threshold are determined dynamically or in real-time based on the total value of the oil types involved in the multi-port mixed loading task data, the urgency of the task, or the current environmental regulatory level of the target sea area, through a preset mapping rule or calculation model. This allows the risk classification standard to adapt to the importance of different tasks and external regulatory requirements, achieving scenario-based adaptation of risk tolerance, and further improving the flexibility of the scheduling strategy while ensuring a safety baseline.

[0112] In some embodiments, optionally, after determining the third execution result, the system further refines the result into different handling instruction levels based on the specific extent to which the compatibility risk coefficient exceeds the second preset threshold. For example, when the exceedance is within a small range, the handling instruction is "a deep tank cleaning of the specified standard must be performed, and a reassessment must be conducted after the operation"; when the exceedance is extremely large, the handling instruction is directly "assignment is prohibited, and it is recommended to remove it from the candidate set." This refined classification based on the degree of exceedance makes the handling of high-risk situations more precise and targeted, avoiding the potential waste of capacity or misjudgment of risks that may result from a "one-size-fits-all" approach.

[0113] In some embodiments, the logic for determining the first, second, and third execution results may optionally be temporarily overridden or weighted under specific conditions. These specific conditions include, but are not limited to, an extreme shortage of available tankers in the fleet, or a mission with the highest priority.

[0114] In this situation, the system can, based on a pre-set emergency decision-making rule, allow tankers originally classified as having a third-order execution outcome to be downgraded to undergo tank cleaning assessment according to the second-order execution outcome process, while simultaneously attaching the highest-level risk warning to the final instruction. This provides a pre-set decision-making pathway for the dispatch system to balance the conflicting demands of "safety" and "mission completion" in extreme circumstances.

[0115] In some embodiments, optionally, such as Figure 3 As shown, the second execution result for determining the tanker's corresponding multi-port mixed loading task data includes:

[0116] Step S11240: Determine multiple candidate tank cleaning parameters corresponding to the tanker based on the bulkhead residual data and multiple preset tank cleaning depths;

[0117] Step S11242: Determine the comprehensive tank cleaning residue data corresponding to each candidate tank cleaning parameter based on the bulkhead residue data and multiple preset tank cleaning depths;

[0118] Step S11244: Determine the post-washing risk coefficient corresponding to each candidate washing parameter based on the comprehensive data of tank washing residue, the oil type incompatibility matrix, and the oil type loading sequence;

[0119] Step S11246: Obtain the estimated tank cleaning cost and preset weight coefficient corresponding to each candidate tank cleaning parameter;

[0120] Step S11248: Determine the comprehensive value based on the post-washing risk coefficient, the estimated washing cost, and the preset weighting coefficient;

[0121] Step S11250: Determine the target tank washing parameter from multiple candidate tank washing parameters based on the comprehensive cost value;

[0122] Step S11252: Determine the second execution result based on the target tank washing parameters.

[0123] In this embodiment, the residual bulkhead state after each tank washing scheme is predicted by the tank washing depth and efficiency coefficient. The post-washing risk coefficient is then recalculated to assess its safety effect. Combined with the projected tank washing cost, which includes direct, time, and opportunity costs, and using preset weighting coefficients, a unified quantitative score is applied to all candidate schemes through a comprehensive cost function. The scheme with the lowest comprehensive cost value is selected as the target tank washing parameter, thereby determining the final operational command.

[0124] Understandably, this provides a precise, economical, and automated decision-making tool for medium-risk scenarios, fundamentally overcoming the drawbacks of relying on fixed-cycle tank cleaning or experience-based judgment, and significantly improving scheduling efficiency and economic benefits.

[0125] Specifically, by conducting parallel simulations and quantitative comparisons of various possible paths such as "no-wash", "regular tank cleaning", and "deep tank cleaning", the solution with the lowest total cost that meets the safety threshold can be accurately identified. This avoids unnecessary costs and time delays caused by excessive tank cleaning, and also prevents potential risks caused by insufficient tank cleaning.

[0126] This data-driven optimization decision-making replaces the difficult trade-offs between cost, safety, and time made manually. It can automatically generate the optimal solution in a short time, greatly shortening the assessment and decision-making cycle for medium-risk tankers and ensuring the rapid response capability of fleet scheduling.

[0127] Based on the known bulkhead residue data of the tanker and multiple preset tank cleaning depths defined in the system, multiple candidate tank cleaning parameters corresponding to the tanker are determined.

[0128] The preset tank cleaning depth usually includes at least three levels: "no washing", "regular seawater rinsing" and "deep chemical cleaning", with each level corresponding to a specific operation standard code.

[0129] Candidate wash parameters are a set of alternative schemes consisting of these wash depth level codes, such as {0,1,2}, representing all potential operational paths that can be evaluated.

[0130] After determining the candidate tank washing parameters, the system needs to predict the tanker's state after implementing each tank washing scheme.

[0131] Therefore, based on the bulkhead residue data and multiple preset tank cleaning depths, the comprehensive tank cleaning residue data corresponding to each candidate tank cleaning parameter was determined.

[0132] The system pre-stores the cleaning efficiency coefficient corresponding to each cleaning depth. The cleaning efficiency coefficient represents the average removal rate of historical residues on the bulkhead by the graded cleaning operation.

[0133] For each candidate tank cleaning parameter, the system multiplies the current tanker bulkhead residue data by a coefficient to calculate the predicted residue state of the tanker bulkhead after the tank cleaning depth is performed, i.e., the comprehensive tank cleaning residue data.

[0134] Subsequently, the system needs to assess whether the actual risk of the tanker performing the mission has been reduced to an acceptable level after each tank washing option is adopted.

[0135] Therefore, based on the comprehensive data of tank cleaning residues, the oil type incompatibilities matrix, and the oil type loading sequence, the post-tank cleaning risk coefficient corresponding to each candidate tank cleaning parameter is determined.

[0136] For each candidate tank washing parameter and its predicted comprehensive tank washing residue data, the system replaces the original bulkhead residue data with the comprehensive tank washing residue data and re-introduces it into the complete calculation process of the compatibility risk coefficient.

[0137] Specifically, the maximum value of adjacent oil type conflicts is first calculated based on the oil type loading sequence and the oil type taboo matrix. Then, the updated historical residual conflict coefficient is calculated based on the comprehensive data of tank cleaning residues and the oil type taboo matrix. Finally, the maximum value of the two is taken to obtain the post-tank cleaning risk coefficient under this tank cleaning scheme. The post-tank cleaning risk coefficient quantifies the chemical conflict risk that the tanker still faces when performing mixed loading tasks after implementing the tank cleaning scheme.

[0138] However, it is not enough to only consider risk reduction; economic costs must also be taken into account.

[0139] Therefore, the system retrieves the estimated tank cleaning cost for each candidate tank cleaning parameter from the integrated operating cost database. This estimated tank cleaning cost is a comprehensive cost estimate that considers multiple dimensions: direct material and treatment costs, including the cost of fresh water and chemical cleaning agents, and the cost of treating the resulting oily wastewater; time costs, i.e., the berthing time caused by the tank cleaning operation occupying the berth, calculated according to the daily charter rate of the tanker; and potential opportunity costs, i.e., the estimated loss of revenue from other adjacent tasks that may be missed due to the tank cleaning delay.

[0140] At the same time, preset weighting coefficients are invoked to balance the relative importance of economy and safety in subsequent comprehensive evaluation.

[0141] Having obtained the post-washing risk coefficient and estimated washing cost for each option, the system needs a unified metric for comparison.

[0142] Accordingly, the comprehensive cost is determined based on the post-washing risk coefficient, the estimated washing cost, and the preset weighting coefficient. For each candidate washing parameter, its comprehensive cost is calculated using a linear weighted summation function. The formula shows that both high costs and high residual risks increase the cost, while the preset weighting coefficient determines which is more costly. After calculation, each washing scheme corresponds to a quantified comprehensive cost.

[0143] Compare the comprehensive cost of all candidate tank washing parameters and select the option with the smallest value. The selected option, that is, the tank washing intervention level that can achieve the minimum total cost while meeting risk control requirements under model evaluation, is determined as the target tank washing parameter.

[0144] For example, for a certain oil tanker, the comprehensive cost of the three options—no cleaning, routine cleaning, and deep cleaning—is calculated to be 1.5, 0.8, and 1.2, respectively. Among them, the routine cleaning option has the lowest cost of 0.8, so the target cleaning parameter is determined to be the code "1" representing routine cleaning.

[0145] The second execution result is a structured decision output. It explicitly includes the final recommended tank washing depth level, i.e., the specific operational instructions represented by the target tank washing parameters. For example, it generates a second execution result of "After performing routine seawater flushing, this mixed loading task is permitted." This result, along with the intermediate values ​​of the calculations on which it is based, will be passed to the subsequent dispatch instruction generation module as part of the tanker's final risk response result.

[0146] In some embodiments, the tank cleaning efficiency coefficient is optionally not a globally uniform set of preset constants, but rather dynamically calculated or matched using a regression model based on the specific tanker model, the type of bulkhead lining material, and the actual average cleaning effect records of similar tank cleaning operations statistically derived from the tanker's historical data. This allows the prediction of the residual state after tank cleaning to better reflect the actual physical characteristics and historical performance of a single vessel, improving the prediction accuracy of "comprehensive tank cleaning residue data," thereby making subsequent risk and cost assessments more reliable.

[0147] In some embodiments, the calculation of the estimated tank cleaning cost may also incorporate real-time or forecasted market variables. These variables include current water and electricity rates and service rates at the specific port where tank cleaning is scheduled, market price fluctuations of cleaning agents, and estimates of the impact of fuel prices on additional voyage fuel consumption due to tank cleaning delays. By incorporating dynamic market factors into the cost model, the comparison of the comprehensive cost function can reflect a more realistic current economic situation, avoiding decision-making biases caused by using outdated price data.

[0148] In some embodiments, optionally, during the process of determining the target tank washing parameters based on the comprehensive cost value, the system not only compares the absolute magnitude of the cost values ​​but also jointly verifies the operation time corresponding to each candidate tank washing scheme and the time window constraint of the task. If the total tank washing and operation time required by the scheme with the lowest cost value would result in failure to meet the latest completion deadline of the task, then the scheme would be considered infeasible and excluded. The system will then select the scheme with the lowest comprehensive cost value from all feasible schemes that meet the time window constraint as the target tank washing parameters. This ensures that the finally recommended scheme is not only optimal in terms of cost and risk but also practically feasible in terms of scheduling timing.

[0149] In some embodiments, optionally, such as Figure 4 As shown, after determining the risk response result for each tanker in the candidate tanker set based on the compatibility risk coefficient in step S112, the tanker scheduling method further includes:

[0150] Step S1130: Determine whether there are any available tankers in the candidate tanker set that meet the preset safety conditions based on the risk response results;

[0151] Step S1132: If there are no available tankers that meet the preset safety conditions, perform task reconstruction processing on the multi-port mixed loading task data;

[0152] The task reconstruction process includes at least one of the following methods: re-selecting and updating the candidate tanker set from multiple tankers based on the oil type taboo matrix and bulkhead residual data; adjusting the loading order of multiple oil types in the oil type loading sequence to determine the iterative oil type loading sequence; or splitting the multi-port mixed loading task data into multiple task subsets.

[0153] In this embodiment, when the system determines that there are no available tankers that meet the preset safety conditions in the current candidate tanker set, it performs a task reconstruction process that includes multiple strategies, namely, re-screening tankers based on the chemical knowledge base, optimizing and adjusting the loading order of multiple oil types, or splitting the original task into multiple independently executable sub-tasks.

[0154] Understandably, this upgrades scheduling decisions from a one-off, static matching process to a dynamic optimization process that includes feedback, adjustments, and replanning. By re-screening, it can uncover potential vessels from the entire fleet that were initially overlooked but have better chemical compatibility, thereby improving the overall utilization of fleet resources.

[0155] By adjusting the loading sequence, it can avoid high-risk combinations from the source of mission design without changing physical constraints, providing a final guarantee for extreme conflicts and ensuring that critical missions can continue to proceed at an acceptable cost (such as using two ships). This multi-layered system-level reconfiguration capability enables the scheduling system to flexibly respond to various complex scenarios, significantly reducing the risk of order cancellations or major delays due to chemical compatibility issues, and ensuring the flexibility and reliability of operational plans.

[0156] Based on the risk response results determined for each tanker, it is necessary to conduct an overall review of the candidate tanker set to determine whether there are any usable tankers that meet the preset safety conditions.

[0157] The preset safety condition is defined as follows: the estimated final execution risk (the original compatibility risk coefficient for the first execution result, and the post-tank washing risk coefficient corresponding to the recommended tank washing scheme for the second and third execution results) of the operation instructions corresponding to the final risk response result of the tanker must be less than or equal to the first preset threshold, i.e., the safety threshold.

[0158] Simultaneously, the tanker must not be marked as "rejected" or in other unavailable states. The system iterates through the entire set of candidate tankers, checking whether the final risk response outcome for each tanker simultaneously meets both conditions.

[0159] If, after traversal, the system determines that there is at least one available tanker that meets the above-mentioned preset safety conditions, it means that there is a suitable vessel in the current candidate set that can safely and directly (or after recommended tank cleaning) perform the task. The process will then terminate, and a scheduling instruction will be generated directly based on the tanker's result.

[0160] If it is determined that there is no available tanker that meets the preset safety conditions, it means that under the current mission definition, tanker fleet status, and chemical knowledge base, it is impossible to find a tanker that can complete the mission safely through individual optimization.

[0161] At this point, task reconstruction processing is performed on the multi-port mixed loading task data itself. Task reconstruction processing is not an adjustment of individual ships, but rather a replanning of task constraints or definitions. It includes at least one of the following systematic solution paths:

[0162] The first reconstruction method is to re-select and update the candidate tanker set from multiple tankers based on the oil type taboo matrix and the bulkhead residue data of all tankers in the fleet (including tankers that were not initially in the candidate set).

[0163] By accessing the complete tanker fleet list and utilizing the oil type contraindication matrix and detailed bulkhead residue data for each tanker, the theoretical compatibility risk coefficient between each tanker and the oil type loading sequence for this mission can be quickly calculated.

[0164] The system will prioritize tankers with good compatibility between the historical bulkhead residue and the first type of oil loaded for the mission, or tankers with very low bulkhead residue data, even if they may not be in the initial candidate set due to their slightly remote location.

[0165] The second reconstruction method involves adjusting the loading order of multiple oil types in the oil type loading sequence to determine the iterative oil type loading sequence. This method assumes that the access order of multiple loading nodes in the task has a certain degree of flexibility.

[0166] Based on the oil type taboo matrix, a rearrangement and combination analysis is performed on a given oil type loading sequence. The goal is to find a new oil type order with lower chemical compatibility risk. For example, the system attempts to calculate all possible node access permutations (or, in practice, uses a heuristic search algorithm). For each new sequence, the maximum value of its adjacent oil type conflicts is calculated. The system then selects the sequence that minimizes this maximum value as the new iterative oil type loading sequence.

[0167] Then, this new sequence will replace the original sequence, and the compatibility risk coefficient of the current candidate tanker set (or the newly selected set) will be recalculated and evaluated. This is equivalent to avoiding high-risk combinations of adjacent oil types by adjusting the loading order without changing the total loading volume and node list.

[0168] The third reconstruction approach is to split the multi-port mixed loading task data into multiple task subsets. This approach decomposes the original task when the above two methods are not feasible or still cannot meet the requirements. The core of the splitting logic is to find a high-risk "conflict point" in the oil type loading sequence based on the oil type taboo matrix, and use this point as the boundary to divide the task in two. Subsequently, each sub-task needs to independently perform fleet screening, risk assessment, and scheduling, possibly performed by two different tankers, thereby completely avoiding the mixing of incompatible oil types in the same cargo hold.

[0169] After the task refactoring is completed, the main scheduling decision process will be re-entered based on the refactoring results (updated tanker set, new oil type sequence, or split sub-tasks) to conduct a new round of risk assessment and response determination, forming a decision loop that includes feedback and optimization, until a solution that meets the preset safety conditions is found, or the final conclusion that the task is not feasible is given after exhausting all possible refactoring.

[0170] In some embodiments, the process of re-screening and updating the candidate tanker set from multiple tankers may not involve blindly traversing all tankers, but rather using a pre-built tanker-oil type affinity index for efficient targeted retrieval. The affinity index is based on the main residual oil type lineage reflected in the historical bulkhead residue data of each tanker, as well as an oil type taboo matrix, and pre-calculates the "basic conflict potential" between each tanker and various common oil types. When re-screening is required, the system prioritizes recalling the subset of tankers with the lowest average basic conflict potential with all oil types in the task's "oil type loading sequence" for in-depth evaluation, thereby introducing chemical compatibility heuristics into the global screening and significantly improving the efficiency and accuracy of finding suitable tankers.

[0171] In some embodiments, the specific splitting points for dividing multi-port mixed loading task data into multiple task subsets are optionally determined by a multi-objective optimization model. The multi-objective optimization model not only considers that the chemical conflict between the two oil types split at the splitting point should be as high as possible to maximize the splitting benefits, but also balances the resulting sub-tasks in terms of vessel load factor utilization, voyage economy, and overall completion time. The system evaluates multiple potential splitting schemes and selects the scheme that achieves the best trade-off between resolving key chemical conflicts and maintaining the operational efficiency of the sub-tasks, avoiding uneconomical or difficult-to-execute sub-tasks due to brute-force splitting.

[0172] In some embodiments, optionally, such as Figure 5 As shown, step S102: Determine the oil type taboo matrix corresponding to the target sea area based on the characteristic data, including:

[0173] Step S1020: Determine the chemical compatibility data among multiple crude oil varieties based on the characteristic data;

[0174] Step S1022: Determine the compatibility risk coefficient when mixing each two crude oil varieties based on chemical compatibility data;

[0175] Step S1024: Determine the oil type incompatibility matrix based on all compatibility risk coefficients.

[0176] In this embodiment, the method for constructing the oil type incompatibility matrix is ​​specifically defined. Its core steps include: extracting chemical compatibility data characterizing the interactions between oil products based on multi-source heterogeneous characteristic data (including static physicochemical indicators, historical cargo loss records, laboratory mixing tests, and refinery feedback). Then, key indicators such as asphaltene precipitation tendency, precipitation formation rate, and API gravity change rate are selected. Through normalization and weighted summation with preset weighting coefficients, the quantitative compatibility risk coefficient for each pair of oil types is calculated. The risk coefficients of all oil type pairs are then organized into a symmetric matrix, i.e., the oil type incompatibility matrix.

[0177] Understandably, the oil type incompatibility matrix allows complex chemical compatibility issues to be abstracted into rapid matrix query operations, greatly improving the objectivity, consistency, and speed of risk assessment. It overcomes the efficiency bottleneck and subjective differences of manual judgment, thereby fundamentally improving the decision-making efficiency and reliability of the entire scheduling system in dealing with complex oil product mixing scenarios.

[0178] First, based on the characteristic data of multiple crude oil varieties obtained in the target sea area, the chemical compatibility data among these crude oil varieties is determined. The chemical compatibility data not only includes the independent static physicochemical indicators of each oil type, such as asphaltene content, gum content, sulfur content, total acid value, API gravity index, wax content, etc., but more importantly, it includes dynamic correlation data that reflects the interactions between oil types.

[0179] These related data mainly come from three dimensions: actual cargo loss case reports recorded in historical operations, which clearly record problems such as precipitation and blockage caused by specific oil type mixing; a dataset of paired oil type mixing compatibility test conducted in a laboratory environment, which tests and simulates mixing at different ratios and temperatures, and records the onset time and rate of phenomena such as asphaltene precipitation and sedimentation; and process feedback data from downstream refineries, reflecting the impact of different mixed crude oil feedstocks on unit stability and product quality.

[0180] These multi-source heterogeneous data are cleaned, aligned, and correlated to extract quantitative indicators that can directly characterize the compatibility between any two oil products, forming the basis of chemical compatibility data.

[0181] After obtaining the above chemical compatibility data, the system enters the quantitative calculation stage, that is, to determine the compatibility risk coefficient when mixing each two crude oil varieties based on the chemical compatibility data.

[0182] For each pair of oil types to be evaluated, several key measured values ​​are extracted from the chemical compatibility data, including the measured values ​​of asphaltenes precipitation tendency, precipitation formation rate, and API gravity index change rate of the blended oil. To eliminate dimensions and make different indicators comparable, the system calls the maximum values ​​of all oil type pairs in the historical database for each indicator, and normalizes these measured values ​​to obtain a series of ratios between zero and one.

[0183] Pre-defined weighting coefficients are introduced. These weighting coefficients are pre-set by operations experts based on the degree of attention paid to different risk types in actual business operations. For example, a higher weight can be assigned to the risk of asphalt precipitation, a medium weight to the risk of sedimentation, and a lower weight to the risk of density characteristic changes. The sum of all weighting coefficients is one.

[0184] The compatibility risk coefficient is calculated using a weighted summation method. This involves multiplying the normalized asphaltene precipitation tendency ratio by its corresponding weighting coefficient, adding the normalized precipitation formation rate ratio by its weighting coefficient, and then adding the normalized API change rate ratio by its weighting coefficient. The sum of these three values ​​yields a scalar value between zero and one. A value closer to zero indicates excellent chemical compatibility between the two oil products and a very low risk of mixing; a value closer to one indicates extremely poor compatibility and a very high risk of stability problems after mixing.

[0185] Based on the compatibility risk coefficients calculated among all oil type pairs, the final oil type taboo matrix is ​​determined. The oil type taboo matrix is ​​a square matrix with the number of rows and columns equal to the total number of crude oil types considered within the target sea area.

[0186] The element in the i-th row and j-th column of the matrix is ​​the compatibility risk coefficient of the two oil products.

[0187] Since the risk of mixing any two oil products, A and B, is equal to the risk of mixing B and A, the matrix is ​​a symmetric matrix. The elements on the main diagonal represent the mixing risk of the same type of oil itself, which is theoretically zero.

[0188] This complete matrix is ​​stored in a structured manner to form an oil type incompatibilities matrix knowledge base. The oil type incompatibilities matrix serves as the foundation for rapid queries in all subsequent risk assessment steps. The system can instantly retrieve the preset risk coefficients between any two oil types using the oil type number, thereby supporting efficient real-time scheduling decisions.

[0189] In some embodiments, optionally, such as Figure 6 As shown, step S104: Determine the bulkhead residual data corresponding to each tanker based on historical operating data, including:

[0190] Step S1040: Determine the historical voyage list of the tanker based on historical operating data. Each voyage record in the historical voyage list includes the type of oil loaded, tank washing depth, and residual assessment value.

[0191] Step S1042: Based on the preset tank cleaning efficiency coefficient, tank cleaning depth and residual assessment value, determine the tanker bulkhead residual data through recursion.

[0192] In this embodiment, the bulkhead residual status is dynamically quantified through a recursive relationship. The sequence of loaded oil type, tank cleaning depth and residual assessment value per voyage is extracted from historical operational data. Then, based on the tank cleaning efficiency coefficients corresponding to different preset tank cleaning depths, the historical residual assessment values ​​are calculated to decay sequentially. This simulates the superposition process of the cleaning effect of each tank cleaning and the new residual, thereby calculating the bulkhead residual data that reflects the current overall cleanliness of the cargo hold.

[0193] Understandably, transforming the historical loading information of tanker bulkheads from discrete, static records into a continuous, quantifiable, dynamic state model provides crucial state input for accurately assessing chemical compatibility risks. By mathematically recursively quantifying the combined residual impact of historical oil types after multiple tank cleaning operations, the risk assessment more accurately reflects the actual condition of the tanker, significantly improving the accuracy of risk warnings and avoiding potential contamination accidents due to insufficient estimation of historical bulkhead residues, or overly conservative cleaning decisions. This lays a reliable data foundation for subsequent cost-benefit analysis and optimized scheduling.

[0194] Based on the obtained historical operating data of oil tankers, a list of historical voyages of oil tankers is determined.

[0195] Historical operational data is a comprehensive data set, mainly derived from the tanker's logbook, loading and unloading operation records, and tank cleaning certificates and inspection reports issued by official or third parties.

[0196] Structured information directly related to changes in cargo hold status is extracted from these historical operational data, and a record is created for each historical voyage in the list.

[0197] Each voyage record must include three key data fields: the type of crude oil loaded, i.e., the type or name of the crude oil that is mainly carried in the cargo holds during the voyage; the depth of tank cleaning, i.e., the level of tank cleaning performed before loading the crude oil of this voyage, such as ordinary tank cleaning, crude oil tank cleaning, special cleaning, etc., each level corresponding to different cleaning standards and work intensity; and the residual assessment value, which is a comprehensive assessment of the nature and amount of residues of the crude oil from the previous voyage in the tank walls, pipelines, etc., obtained by laboratory analysis (such as gas chromatography) through in-tank sampling or by on-site assessment by the surveyor after the end of this voyage or before loading of the next voyage. It is usually expressed as a level or quantitative score.

[0198] After constructing a complete list of historical voyages, the system enters the core calculation phase, which involves determining the tanker's bulkhead residue data through a recursive relationship based on preset tank cleaning efficiency coefficients, tank cleaning depths, and residual assessment values. The tank cleaning efficiency coefficient is an empirical parameter table derived in advance through statistical analysis of a large amount of historical data, defining the average percentage efficiency for removing residual oil types from the previous voyage at different tank cleaning depth levels.

[0199] For example, the efficiency coefficient of a special wash may be as high as 98%, while that of a regular tank wash may only be 80%. The recursive relationship starts from the known or assumed initial cleanliness state of the tanker, usually with an empty or known residual data vector of bulkhead residues, which records the residual intensity of each oil component.

[0200] The system iterates through the historical voyage list in chronological order. For each voyage in the list, the calculation logic is as follows: The system first queries the preset tank washing efficiency coefficient table based on the tank washing depth recorded in the voyage record to obtain the corresponding efficiency value.

[0201] Subsequently, by applying a recursive relationship, using the bulkhead residue data at the end of the previous voyage, the tank washing efficiency coefficient used in this voyage, and the information on the type of oil actually loaded in this voyage, the new bulkhead residue data after the loading operation of this voyage is calculated.

[0202] Specifically, the recursive process simulates two stages: physical cleaning and recontamination. The residual data from the previous voyage is multiplied by a decay factor determined by the cleaning efficiency coefficient, representing the removal effect of the original residue after cleaning. Simultaneously, the type of oil loaded in the current voyage contributes a "new residue" value due to its adhesion characteristics to the bulkhead and potential interactions with old residues. This contribution value is partially calibrated and corrected by the residue assessment values ​​recorded in the voyage log. Through this recursive calculation, the bulkhead residual data is updated for each voyage, ensuring it always reflects the theoretical residual state of the cargo hold after the latest operation.

[0203] Once the system has processed the last voyage record in the historical voyage list, the resulting bulkhead residue data is a quantitative description of the cargo hold cleanliness status inferred from all historical operations at the current moment.

[0204] In some embodiments, optionally, such as Figure 7 As shown, step S112: Determine the risk response outcome for each tanker in the candidate tanker set based on the compatibility risk coefficient, including:

[0205] Step S1120: Determine the adjacency conflict coefficient of adjacent oil types within the sequence based on the oil type loading sequence and the oil type taboo matrix;

[0206] Step S1122: Determine the historical residual conflict coefficient based on the residual oil type indicated by the bulkhead residual data, the first loaded oil type in the oil type loading sequence, and the oil type taboo matrix;

[0207] Step S1124: Determine the compatibility risk coefficient based on the adjacent conflict coefficient and the historical residual conflict coefficient.

[0208] In this embodiment, based on the predetermined oil type loading sequence for the mission, the inherent risks of all adjacent oil type pairs are comprehensively analyzed by querying the oil type taboo matrix to obtain the adjacent conflict coefficient, which characterizes the inherent continuous risk of the mission. Based on the residual oil type and its intensity indicated by the tanker's bulkhead residual data, and combined with the first oil type loaded for the mission, the taboo matrix is ​​queried again to calculate the historical residual conflict coefficient, which reflects the conflict between the tanker's own historical state and the initial requirements of the mission. These two coefficients, which are derived from the mission plan and the carrier state respectively, are combined according to preset rules (such as taking the maximum value or weighted summation) to determine a comprehensive compatibility risk coefficient.

[0209] Understandably, the compatibility risk coefficient calculated by the model can more accurately reflect the true risk profile of a specific tanker when performing a specific mixed loading task. This provides a reliable and consistent quantitative benchmark for subsequent accurate risk classification, tank cleaning decisions, and conflict resolution, thereby fundamentally improving the scientific nature and safety of scheduling decisions and significantly reducing the risk of cargo damage accidents caused by incomplete risk assessment.

[0210] Based on the oil loading sequence provided by the mission party and the pre-constructed oil taboo matrix, the adjacent conflict coefficients of adjacent oil types within the sequence are determined.

[0211] The oil loading sequence is an ordered list that clearly specifies the types of crude oil to be loaded sequentially at multiple ports and their order.

[0212] Process each pair of adjacent oil types in the oil loading sequence in sequence, that is, the oil type loaded at the previous port and the oil type loaded at the next port.

[0213] For each pair of adjacent oil types, the system uses the numbers of these two oil types as an index to directly query the corresponding cell in the oil type incompatibility matrix and obtain its preset compatibility risk coefficient value. The compatibility risk coefficient value directly reflects the inherent risk when these two oil types come into contact in the cargo hold and may mix.

[0214] Next, it is necessary to synthesize these inherent risk values ​​for all adjacent oil type pairs in the entire loading sequence. A typical synthesis method is to calculate the arithmetic mean of the risk values ​​of all adjacent oil type pairs, which represents the sum of internal continuity risks faced by the entire loading sequence during sequential execution. This synthesized value is the adjacency conflict coefficient. The higher the adjacency conflict coefficient, the more drastic the oil type switching required by the task sequence itself, and the greater the inherent compatibility challenge.

[0215] Secondly, based on the residual oil types indicated by the bulkhead residual data, the first loaded oil type in the oil type loading sequence, and the oil type taboo matrix, the historical residual conflict coefficient is determined.

[0216] Bulkhead residue data is a quantitative vector that not only indicates which historical oil types remained in the cargo hold, but more importantly, it provides the residual intensity or concentration ratio of each residual oil type.

[0217] The first type of crude oil loaded in the oil loading sequence is the first type of crude oil to be loaded into the empty tank during the mission.

[0218] For each type of residual oil recorded in the bulkhead residual data, the system queries the oil type incompatibility matrix using its number and the number of the first loaded oil type to obtain the basic compatibility risk coefficient between the two.

[0219] Subsequently, this basic risk factor will be multiplied or weighted by the specific residual intensity of the residual oil type, because the higher the residual intensity, the greater its potential impact on adverse reactions with the first-fill oil type.

[0220] The risk values ​​of all residual oil types calculated in this way and corrected for residual strength are summed to obtain a comprehensive evaluation value, namely the historical residual conflict coefficient.

[0221] Based on the calculated adjacent conflict coefficient and historical residual conflict coefficient, the final compatibility risk coefficient for performing this task on this tanker is determined.

[0222] Assign a preset weight to each of these two types of conflict coefficients. The weight setting can be based on business experience. For example, it may be considered that conflicts between adjacent oil types within a task sequence are the main source of risk, and therefore a higher weight is given to adjacent conflict coefficients; at the same time, it may be considered that tank washing can partially eliminate the impact of historical residual conflicts, and therefore a lower weight is given to historical residual conflict coefficients.

[0223] The compatibility risk coefficient is obtained by multiplying the adjacent conflict coefficient by its weight, adding the historical residual conflict coefficient multiplied by its weight, and then summing the results. This final coefficient is a single, comparable value that comprehensively reflects the sum of dual risks faced by a specific tanker when performing a specific mixed loading task, from both the internal risks within the task sequence and the tanker's own condition. It provides a direct quantitative basis for subsequent risk classification, tanker sequencing, and response decisions.

[0224] In some embodiments, the dispatching instructions may optionally include a target tanker identifier, tank cleaning operation instructions, navigation path instructions, and risk warning information, such as... Figure 8 As shown, step S114: Determine the scheduling instructions corresponding to the candidate tanker set based on the risk response results, including:

[0225] Step S1140: Determine the target tanker from the candidate tanker set based on the risk response results of each tanker;

[0226] Step S1142: Determine the tank cleaning operation instructions and risk warning information based on the risk response results of the target oil tanker;

[0227] Step S1144: Determine the navigation route data based on the multi-port mixed loading mission data;

[0228] Step S1146: Determine the scheduling instructions corresponding to the candidate tanker set based on the navigation path data.

[0229] In this embodiment, based on the risk response results of each candidate tanker, the target tanker for the final task is determined through preset optimization rules (such as prioritizing direct execution and then selecting the tanker with the lowest overall cost). Then, specific and operable tank cleaning operation instructions are directly mapped from the risk response results of the target tanker, and corresponding risk warning information is generated based on its remaining risks. Simultaneously, based on the node sequence, loading volume, and time window in the multi-port mixed loading task data, combined with ship performance parameters, a detailed navigation path and time plan are calculated. The target tanker identifier, tank cleaning operation instructions, navigation path instructions, and risk warning information are encapsulated into a complete scheduling instruction and issued for execution.

[0230] Understandably, automated generation eliminates errors, omissions, and delays that might result from manual information integration and transmission, ensuring that decision-making intentions are fully and accurately conveyed to the execution end. The embedded risk warning information directly transforms the risk conclusions of the back-end assessment into preventative operational guidelines for the front end, enhancing the safety and controllability of the operational process. Ultimately, this significantly shortens the cycle from decision-making to execution for the entire multi-port mixed loading task, providing clear and unambiguous instructions and significantly improving the fleet's response speed, safety level, and management efficiency.

[0231] The system determines the target tanker for the final mission from the candidate tanker set based on the risk response results of each tanker. It examines the risk response results of each tanker in the candidate set, clarifying whether the tanker is available and, if so, the additional conditions required.

[0232] Tankers that can be "directly executed" in terms of risk response are usually given priority, that is, those tankers that do not require additional tank cleaning operations and whose compatibility risk coefficient already meets safety requirements, in order to pursue the highest efficiency.

[0233] If there are multiple such tankers, they can be sorted and selected based on auxiliary indicators such as the estimated time of arrival at the first loading node or the historical mission execution efficiency.

[0234] If no suitable tanker is available for direct deployment, the system selects the tanker with the lowest overall cost and the best balance between total execution cost and risk from among those tankers whose risk response outcomes include a "recommended tank cleaning plan." This selection logic ensures that the final assignment is a comprehensive optimization of economy and timeliness under the premise of safety and feasibility.

[0235] After identifying the target tanker, the system needs to generate specific operational instructions. Therefore, the tank cleaning operation instructions and risk warning information are determined based on the risk response results of the target tanker. The generation of the tank cleaning operation instructions follows a direct mapping relationship: if the tanker's risk response result is "execute directly," then the tank cleaning operation instruction will be "no tank cleaning required"; if the risk response result includes a "recommended tank cleaning plan," such as "perform deep chemical cleaning," then the tank cleaning operation instruction will specify the tank cleaning depth level in detail, along with the corresponding standard operating requirements, such as the type of cleaning agent, flushing pressure, and duration.

[0236] At the same time, risk warning information will be generated based on the tanker's final remaining risk level after the tank cleaning plan is revised.

[0237] If the remaining risk coefficient is zero or extremely low, the risk warning message may be "None"; if there is an acceptable residual risk, the system will generate a structured text that clearly points out the risk points, such as "Note the potential compatibility risk between the first-loaded oil type A and the historical residual oil type B in the bulkhead", and suggest corresponding monitoring measures, such as "It is recommended to increase the frequency of visual inspections and sample testing inside the tank during the initial loading stage".

[0238] The navigation path data is determined based on the multi-port mixed loading mission data, which includes a list of loading nodes that must be visited in sequence, the planned loading volume of each node, and the mission time window.

[0239] The determination of navigation path data begins with connecting these loading nodes in a prescribed order to form a logical access path.

[0240] By combining the target tanker’s current position, service speed, and known standard sailing distances between ports, the estimated sailing time from the tanker’s current position to the first loading node, and the sailing time between subsequent loading nodes, is calculated.

[0241] Based on the planned workload and standard loading / unloading efficiency at each node, the estimated berthing time at each node is estimated. Finally, the system integrates the access sequence, the estimated travel distance and time for each segment, and the estimated arrival and departure times for each node to form a complete flight path and timetable, i.e., flight path data.

[0242] The unique identifier of the target tanker, the specific tank cleaning operation instructions, the detailed planned navigation route data, and the corresponding risk warning information are assembled according to the predetermined message or document format to generate a complete final dispatch instruction corresponding to the decision result of the candidate tanker set.

[0243] The dispatch instructions are automatically sent to the assigned target tanker via the network and simultaneously sent to the port dispatch centers of each loading node involved in the task, thereby driving the entire multi-port mixed loading task into the actual execution stage.

[0244] In some embodiments, the navigation path instructions in the scheduling commands are optionally not static, fixed routes, but rather associated with a set of dynamic path selection rules based on preset conditions. The system integrates real-time or predicted maritime traffic density information, regional weather warnings, and temporary waterway control notices. When generating navigation path data, the system selects an optimal sub-path from multiple pre-stored alternative navigation sub-paths for each port-to-port voyage based on the currently integrated dynamic information, and combines these sub-paths into the final recommended navigation path instructions. This allows the final scheduling instructions to adapt to changes in the external environment, improving the safety and economy of navigation.

[0245] In some embodiments, the output of the dispatch instruction may optionally include a machine-readable instruction core data packet and one or more human-machine interface views adapted to different recipients. The system renders and presents the same instruction core data packet differently based on the role of the instruction recipient. For example, the view sent to the target tanker focuses on detailed tank cleaning procedures and voyage plans; the view sent to the port dispatch center highlights the tanker's estimated arrival and departure times, the types of services required, and risk management requirements. This ensures that information from the same decision-making source is delivered to different execution stages in the most efficient and easily understandable form.

[0246] In one specific embodiment, optionally, the system structure block diagram of the present invention is as follows: Figure 10As shown, the system comprises a data layer, a computation layer, an input layer, and an output layer. The data layer consists of a bulkhead memory database (storing the tanker's historical voyages and current bulkhead remnant status) and an oil type taboo database (generating an oil type taboo matrix). The computation layer receives multi-port mixed loading tasks from the input layer and performs core analysis and decision-making through its internal risk calculation module (including comprehensive risk coefficient calculation), tank cleaning decision module, and task allocation module (including candidate tanker screening and conflict resolution). The core of the output layer is the scheduling instruction generation module, which, based on the results of the computation layer, ultimately generates complete scheduling instructions including tanker assignment, oil retrieval route, tank cleaning instructions, and risk warnings, and issues them for execution through the task output interface.

[0247] The method flowchart of the present invention is as follows: Figure 11 As shown, it includes:

[0248] Step S200: Construct an oil type taboo matrix;

[0249] Step S202: Construct a tanker bulkhead memory archive;

[0250] Step S204: Receive multi-port mixed loading task;

[0251] Step S206: Screening of candidate tankers;

[0252] Step S208: Determine if the candidate tanker set is empty;

[0253] If the judgment result of step S208 is yes, execute step S228: the task is not feasible, adjust the parameters, and return to step S206.

[0254] If the judgment result of step S208 is negative, proceed to step S210: calculate the chemical compatibility risk coefficient;

[0255] Step S212: Determine whether the risk coefficient is less than or equal to the safety threshold;

[0256] When the judgment result of step S212 is yes, step S224 is executed: directly assign the tanker without washing, and then proceed to step S226: output the dispatch instruction;

[0257] If the judgment result of step S212 is negative, proceed to step S214: determine whether the risk coefficient is less than or equal to the warning threshold.

[0258] If the judgment result of step S214 is yes, proceed to step S216: perform a cost-benefit analysis of the tank washing scheme;

[0259] Step S218: After the tank is washed, is the risk less than or equal to the safety threshold?

[0260] If the judgment result of step S218 is yes, proceed to step S224: directly assign the tanker, no washing required;

[0261] If the result of step S218 is negative, proceed to step S220: conflict resolution.

[0262] If the result of step S214 is negative, proceed directly to step S220: conflict resolution.

[0263] After step S220: conflict resolution, proceed to step S222: whether a feasible solution has been found;

[0264] If the judgment result of step S222 is negative, proceed to step S228: the task is not feasible, adjust the parameters, and return to step S206.

[0265] If the judgment result of step S222 is yes, then step S224 is executed;

[0266] Step S226: Output scheduling instructions.

[0267] Specifically, step one: construct an oil type contraindication matrix:

[0268] First, chemical property data of common crude oil varieties in the target sea area are collected, including but not limited to: asphaltene content, gum content, sulfur content, acid value, API gravity, and wax content. Through historical cargo loss data, laboratory blending test data, and refining process feedback data, chemical compatibility relationships between different crude oil types are established, forming a crude oil incompatibility matrix M.

[0269] Oil Type Taboo Matrix Let n be an n×n symmetric matrix, where n is the number of oil product types, and the matrix elements are... The compatibility risk coefficient when oil type i and oil type j are mixed is defined as follows:

[0270] ;

[0271] in, The measured value represents the tendency of asphaltene precipitation after mixing oil types i and j. This represents the historical maximum value of the tendency for asphaltene precipitation after mixing oil types i and j. This represents the measured value of the precipitation rate. This represents the historical maximum rate of precipitation formation. This represents the measured value of the change rate of API gravity of the mixed oil. This represents the historical maximum value of the API gravity change rate of the blended oil. These are weighting coefficients, set according to actual operational needs (e.g., α=0.5, β=0.3, γ=0.2).

[0272] The value range is [0,1]. The closer the value is to 1, the worse the compatibility (i.e., the higher the risk of conflict). The closer the value is to 0, the more compatible it is.

[0273] The oil type incompatibilities matrix is ​​stored in the oil type incompatibilities database and supports dynamic updates (when new oil types appear or new test data is generated).

[0274] Step Two: Constructing a Tanker Bulkhead Memory Archive:

[0275] A "bulwark memory file" is established for each registered shuttle tanker to record its historical operational data. The file data structure includes the following fields: Tanker unique identifier. List of historical voyages Each voyage Includes: Loading oil type Tank washing depth after unloading (Divided into three levels: 0 - Unwashed tank, 1 - Routine seawater flushing, 2 - Deep chemical cleaning); Post-unloading bulkhead residue assessment value (Estimated by sensors or empirical formulas, with values ​​ranging from [0,1], where 0 indicates no residue and 1 indicates severe residue).

[0276] Current bulkhead remnant status Calculated using the recursive formula:

[0277] ;

[0278] in, This is the tank cleaning efficiency coefficient, which is related to the tank cleaning depth: =0 (no cleaning effect if the tank is not washed) =0.6 (60% of residue is removed by regular rinsing), =0.95 (deep cleaning removes 95% of residue), k is the sequence number of the current voyage. This represents the initial residual assessment value obtained after loading oil during the k-th voyage (i.e., the previous voyage). Let be the efficiency coefficient of the tank washing operation performed in preparation for the next voyage after the k-th voyage, and t be the summation index, representing the sequence number of the historical voyage. The initial residual assessment value obtained after loading oil during the t-th historical voyage. Let be the efficiency coefficient of the tank washing operation performed after the j-th voyage.

[0279] The records are stored in the bulkhead memory database and are automatically updated after each voyage.

[0280] Step 3: Receive multi-port mixed loading tasks:

[0281] The scheduling system receives new multi-port mixed loading tasks through the task input interface. Task information includes: task number. The list of nodes for the Floating Production Storage and Offloading (FPSO) units that need to be berthed sequentially is P={P1,P2,…,Pm}, where each node… Includes oil types to be filled and loading capacity Task time window [ ]; Optional tanker list (can be specified or automatically filtered by the system).

[0282] Step 4: Screening of candidate tankers:

[0283] Based on the mission requirements, tankers that meet the basic physical conditions are selected from the fleet as a candidate set. The screening criteria include: the tanker is currently available (not undergoing maintenance or performing other tasks);

[0284] Tanker capacity:

[0285] ;

[0286] in, The total tank capacity of the candidate tankers. Let be the planned loading quantity of a specific oil type that needs to be loaded at port j in this multi-port mixed loading operation, and m represent the total number of ports that need to be berthed sequentially in this multi-port mixed loading operation.

[0287] The tanker is currently positioned to reach the first FPSO node within the time window.

[0288] If the candidate set is empty, the feedback task is not feasible and the task parameters need to be adjusted.

[0289] Step 5: Calculate the chemical compatibility risk factor:

[0290] For each oil tanker in the candidate set Based on the current residual state in its bulkhead memory archive and the sequence of oil types to be loaded Calculate the comprehensive risk coefficient for mixed loading at multiple ports. .

[0291] Sub-step 5.1: Calculate the conflict risk between adjacent oil types:

[0292] For each pair of adjacent oil types in the task Read the compatibility risk coefficient from the oil type contraindication matrix .

[0293] Sub-step 5.2: Calculate the risk of conflict between historical residues and the first-fill oil type:

[0294] First-time oil type Compared with the oil type currently remaining on the tanker (i.e., the oil type from the previous voyage) The conflict risk coefficient (if it exists) is However, residual amount needs to be considered. The impact. The historical residual conflict coefficient is defined as:

[0295] ;

[0296] If the tanker has no history (new ship), then .

[0297] Sub-step 5.3: Calculate the overall risk coefficient:

[0298] Taking into account both conflicts between adjacent oil types and historical conflicts, the maximum value is taken as the comprehensive risk coefficient for the tanker's mission:

[0299] ;

[0300] in, This represents the compatibility risk coefficient.

[0301] Step Six: Risk Threshold Determination

[0302] Set a safety threshold (For example =0.3) and warning threshold (For example =0.6).

[0303] like If the chemical compatibility is good, the tanker can be directly assigned to perform the task without additional tank cleaning.

[0304] like This indicates a moderate risk, and the tank cleaning plan needs to be evaluated.

[0305] like This indicates a high risk, requiring deep tank cleaning or replacement of the tanker.

[0306] Step 7: Cost-benefit analysis of the tank washing program:

[0307] For candidate tankers requiring tank cleaning, the system evaluates the costs and residual risks of three cleaning options (no cleaning, conventional cleaning, and deep cleaning) and recommends the optimal option.

[0308] Sub-step 7.1: Calculate the risk of residue after tank cleaning

[0309] For the depth of the tank cleaning The residual state after tank cleaning for:

[0310] ;

[0311] in, For the depth of the tank washing The tank washing efficiency coefficient under the given conditions.

[0312] Following tank cleaning, the risk of conflict between the tanker and the first type of oil loaded is updated as follows:

[0313] ;

[0314] Comprehensive risk factor after tank cleaning Still calculate according to step five, but with Alternative .

[0315] Sub-step 7.2: Calculate the tank cleaning cost:

[0316] Tank cleaning costs This includes: Direct costs: freshwater consumption, chemical cleaning agent consumption, and oily wastewater treatment costs; Time costs: extended berthing time due to tank cleaning operations, calculated on a daily rental basis; Opportunity costs: potential missed revenue from subsequent tasks due to tank cleaning delays.

[0317] Cost data is pre-stored in the system and can be dynamically adjusted according to port and tanker type.

[0318] Sub-step 7.3: Select the optimal solution:

[0319] For the plan Define its comprehensive cost function:

[0320] ;

[0321] in, , Weighting coefficients (e.g.) =0.7, This is used to balance costs and risks.

[0322] Choose to The minimum cost option is recommended. If the no-wash option (d=0) has the lowest cost and... ≤ If the cost is low, the output will be "No need to wash"; otherwise, if a regular tank cleaning is the least expensive option, the output will be "No need to wash". ≤ If the output is "Standard tank cleaning", then "Deep tank cleaning" will be recommended; otherwise, "Deep tank cleaning" will be recommended.

[0323] Step 8: Conflict Resolution and Task Reassignment

[0324] If for all tankers in the current candidate tanker set, even after deep tank cleaning... This indicates that the multi-port mixed loading operation involves chemical conflicts that cannot be resolved through tank washing. In this case, the system activates the conflict resolution mechanism:

[0325] Option A: Replace the tanker. If the presence of other tankers in the fleet (e.g., tankers with different historical loading records) can reduce the risk, then repeat steps four through seven.

[0326] Option B: Adjust the oil extraction sequence. If the oil type sequence can be interchanged (e.g., swapping the oil extraction sequences of two FPSOs), the system generates a new oil type sequence. Recalculate the risk.

[0327] Option C: Split the task. The original task is split into two sub-tasks, which are executed by different tankers to avoid mixing incompatible oil types.

[0328] Option D: Force tank cleaning and accept remaining risks. If none of the above options are feasible and the task must be carried out, select the tanker with the lowest risk and issue a "deep tank cleaning + risk warning" instruction to remind operators to strengthen monitoring.

[0329] Step 9: Output scheduling instructions:

[0330] The system generates complete scheduling instructions based on the final decision, including:

[0331] Tanker assignment: Specify the tanker number to perform the task;

[0332] Oil extraction route: order of visits to each FPSO and estimated arrival and departure times;

[0333] Tank cleaning instructions: Clearly define the tank cleaning depth (0 - no cleaning required, 1 - regular tank cleaning, 2 - deep tank cleaning) and operational requirements;

[0334] Risk warning: If the remaining risk is still higher than It outputs early warning information and suggested monitoring measures.

[0335] The instructions were sent to the corresponding tankers and port operations units via the network.

[0336] like Figure 9As shown in the illustration, this application embodiment also provides a tanker scheduling device 900, comprising: a data acquisition module 902, used to acquire characteristic data of multiple crude oil varieties and historical operating data of multiple tankers within a target sea area; a matrix determination module 904, used to determine a crude oil type incompatibility matrix corresponding to the target sea area based on the characteristic data; a residue determination module 906, used to determine bulkhead residue data corresponding to each tanker based on historical operating data; a task acquisition module 908, used to acquire multi-port mixed loading task data, the multi-port mixed loading task data including multiple loading nodes and crude oil type loading sequences corresponding to the multiple loading nodes; a set determination module 910, used to determine a candidate tanker set based on the multi-port mixed loading task data and multiple tankers; a risk determination module 912, used to determine a compatibility risk coefficient for each tanker in the candidate tanker set based on the crude oil type incompatibility matrix, bulkhead residue data, and crude oil type loading sequence; a result determination module 914, used to determine the risk response result for each tanker in the candidate tanker set based on the compatibility risk coefficient; and a tanker scheduling module 916, used to determine a scheduling instruction corresponding to the candidate tanker set based on the risk response result.

[0337] This application also provides a chip, which includes a processor and a communication interface. The communication interface and the processor are coupled. The processor is used to run programs or instructions to implement the various processes of the above-described tanker scheduling method embodiments and achieve the same technical effects. To avoid repetition, it will not be described again here. Furthermore, the chip improves the data processing speed of the tanker scheduling method in this application.

[0338] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0339] In this invention, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance; the term "multiple" refers to two or more unless otherwise explicitly defined. The terms "install," "connect," "link," and "fix" should be interpreted broadly. For example, "connect" can be a fixed connection, a detachable connection, or an integral connection; "link" can be a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0340] In the description of this invention, it should be understood that the terms "upper," "lower," "left," "right," "front," "rear," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or unit referred to must have a specific orientation or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0341] In the description of this specification, the terms "one embodiment," "some embodiments," "specific embodiment," etc., refer to specific features, structures, materials, or characteristics described in connection with an embodiment or example that are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0342] The above are merely preferred embodiments of the present invention and are not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for scheduling oil tankers, characterized in that, include: Acquire characteristic data of multiple crude oil varieties and historical operational data of multiple oil tankers within the target sea area; Based on the aforementioned characteristic data, a prohibited oil type matrix corresponding to the target sea area is determined; Based on the historical operational data, determine the bulkhead residue data corresponding to each of the tankers; Acquire multi-port mixed loading task data, which includes multiple loading nodes and oil type loading sequences corresponding to the multiple loading nodes; A candidate tanker set is determined based on the multi-port mixed loading task data and the multiple tankers; For each tanker in the candidate tanker set, a compatibility risk coefficient is determined based on the oil type incompatibility matrix, the bulkhead residual data, and the oil type loading sequence. The risk response outcome for each tanker in the candidate tanker set is determined based on the compatibility risk coefficient. Based on the risk response results, determine the scheduling instructions corresponding to the candidate tanker set; The step of determining the oil type taboo matrix corresponding to the target sea area based on the characteristic data includes: Based on the aforementioned characteristic data, chemical compatibility data among multiple crude oil varieties were determined; The compatibility risk coefficient for mixing each pair of the crude oil varieties is determined based on the chemical compatibility data. The oil type incompatibilities matrix is ​​determined based on all the aforementioned compatibility risk coefficients; The step of determining the bulkhead residue data corresponding to each of the tankers based on the historical operational data includes: The historical voyage list of the tanker is determined based on the historical operating data. Each voyage record in the historical voyage list includes the type of oil loaded, the depth of tank cleaning, and the residual assessment value. Based on the preset tank cleaning efficiency coefficient, the tank cleaning depth, and the residual assessment value, the residual data of the tanker's bulkhead is determined through a recursive relationship.

2. The tanker dispatching method according to claim 1, characterized in that, The step of determining the risk response result for each tanker in the candidate tanker set based on the compatibility risk coefficient includes: Obtain a first preset threshold and a second preset threshold, wherein the first preset threshold is less than the second preset threshold; When the compatibility risk coefficient is less than or equal to the first preset threshold, the first execution result of the tanker corresponding to the multi-port mixed loading task data is determined. When the compatibility risk coefficient is greater than the first preset threshold and less than or equal to the second preset threshold, the second execution result of the tanker corresponding to the multi-port mixed loading task data is determined. When the compatibility risk coefficient is greater than the second preset threshold, the third execution result of the tanker corresponding to the multi-port mixed loading task data is determined. The risk response result is determined based on the first execution result, the second execution result, and the third execution result.

3. The tanker dispatching method according to claim 2, characterized in that, The second execution result for determining that the tanker corresponds to the multi-port mixed loading task data includes: Based on the bulkhead residue data and multiple preset tank cleaning depths, multiple candidate tank cleaning parameters corresponding to the tanker are determined; Based on the bulkhead residue data and multiple preset tank cleaning depths, determine the comprehensive tank cleaning residue data corresponding to each candidate tank cleaning parameter; Based on the comprehensive data of tank cleaning residue, the oil type taboo matrix, and the oil type loading sequence, determine the post-tank cleaning risk coefficient corresponding to each of the candidate tank cleaning parameters; Obtain the estimated tank cleaning cost and preset weighting coefficient corresponding to each of the candidate tank cleaning parameters; The comprehensive cost is determined based on the post-washing risk coefficient, the estimated washing cost, and the preset weighting coefficient. The target tank cleaning parameter is determined from among the candidate tank cleaning parameters based on the comprehensive cost value. The second execution result is determined based on the target tank washing parameters.

4. The tanker dispatching method according to claim 3, characterized in that, After determining the risk response outcome for each tanker in the candidate tanker set based on the compatibility risk coefficient, the tanker scheduling method further includes: Based on the risk response results, determine whether there are any available tankers in the candidate tanker set that meet the preset safety conditions; If no available tanker meets the preset safety conditions, then the multi-port mixed loading task data will be reconstructed. The task refactoring process includes at least one of the following methods: Based on the oil type incompatibilities matrix and the bulkhead residue data, the candidate tanker set is re-selected and updated from the multiple tankers; Adjust the loading order of multiple oil types in the oil type loading sequence to determine the iterative oil type loading sequence; Alternatively, the multi-port mixed loading task data can be split into multiple task subsets.

5. The tanker dispatching method according to claim 1, characterized in that, The step of determining the risk response result for each tanker in the candidate tanker set based on the compatibility risk coefficient includes: The adjacent conflict coefficients of adjacent oil types within the sequence are determined based on the oil type loading sequence and the oil type taboo matrix; The historical residual conflict coefficient is determined based on the residual oil type indicated by the bulkhead residual data, the first oil type loaded in the oil type loading sequence, and the oil type taboo matrix. The compatibility risk coefficient is determined based on the adjacent conflict coefficient and the historical residual conflict coefficient.

6. The tanker dispatching method according to any one of claims 1 to 5, characterized in that, The dispatch instructions include target tanker identifiers, tank cleaning operation instructions, navigation path instructions, and risk warning information. The step of determining the dispatch instructions corresponding to the candidate tanker set based on the risk response results includes: From the candidate tanker set, the target tanker is determined based on the risk response results of each tanker; The tank cleaning operation instructions and the risk warning information are determined based on the risk response results of the target oil tanker; Determine navigation route data based on the multi-port mixed loading mission data; The scheduling instructions corresponding to the candidate tanker set are determined based on the navigation path data.

7. A tanker dispatching device, characterized in that, include: The data acquisition module is used to acquire characteristic data of multiple crude oil varieties and historical operational data of multiple oil tankers within the target sea area; The matrix determination module is used to determine the oil type taboo matrix corresponding to the target sea area based on the characteristic data; The residue determination module is used to determine the bulkhead residue data corresponding to each of the tankers based on the historical operating data; The task acquisition module is used to acquire multi-port mixed loading task data, which includes multiple loading nodes and oil type loading sequences corresponding to the multiple loading nodes. The set determination module is used to determine a candidate tanker set based on the multi-port mixed loading task data and the multiple tankers; The risk determination module is used to determine the compatibility risk coefficient for each tanker in the candidate tanker set based on the oil type incompatibility matrix, the bulkhead residual data and the oil type loading sequence. The result determination module is used to determine the risk response result for each tanker in the candidate tanker set based on the compatibility risk coefficient. The tanker scheduling module is used to determine the scheduling instructions corresponding to the candidate tanker set based on the risk response results; The matrix determination module is specifically used to determine the chemical compatibility data among multiple crude oil varieties based on the characteristic data. The compatibility risk coefficient for mixing each pair of the crude oil varieties is determined based on the chemical compatibility data. The oil type incompatibilities matrix is ​​determined based on all the aforementioned compatibility risk coefficients; The residue determination module is specifically used to determine the historical voyage list of the tanker based on the historical operating data. Each voyage record in the historical voyage list includes the type of oil loaded, the tank cleaning depth, and the residue assessment value. Based on the preset tank cleaning efficiency coefficient, the tank cleaning depth, and the residue assessment value, the bulkhead residue data of the tanker is determined through a recursive relationship.

8. A chip, characterized in that, The chip includes a processor and a communication interface, the communication interface being coupled to the processor, the processor being used to run programs or instructions to implement the steps of the tanker scheduling method as described in any one of claims 1 to 6.

Images