An ice period adaptive offshore oil and gas platform ship intelligent scheduling method

Abstract

The application discloses an ice period adaptive offshore oil and gas platform ship intelligent scheduling method, and relates to the technical field of ice period offshore operation ship scheduling. The method obtains ice condition data, ship basic data and scheduling task data, classifies and filters the ships and encodes the tasks, determines the ice period grade and embeds the ice period hard constraint parameter set, sets the bottom fuse mechanism to trigger the navigation warning alarm; constructs a multi-dimensional decision space with ice condition characteristics as the core driving force, and reconstructs the decision space topology when the ice condition change amplitude exceeds the mutation threshold; determines the allowable speed of each navigation section in the decision space, completes the collaborative optimization of path planning and task allocation, and outputs the compliance scheme; generates a standardized scheduling scheme, triggers a local backtracking lock optimization mechanism when the ice condition mutates and an emergency task occurs; and constructs a simulation verification scene covering different ice condition grades to verify the method and calibrate the parameters. The application improves the dynamic adaptation capability and solving efficiency, and reduces the navigation violation rate and the comprehensive operation cost.

Description

Technical Field

[0001] This invention relates to the field of vessel scheduling technology for offshore operations during ice seasons, and specifically to an intelligent scheduling method for offshore oil and gas platform vessels that is adaptive to ice seasons. Background Technology

[0002] Current mainstream optimization methods for maritime vessel scheduling are mostly general scheduling algorithms developed for ice-free conventional sea areas. While these methods can achieve certain optimization effects in conventional vessel scheduling scenarios in open sea areas, their adaptability to ice-covered maritime operations is extremely poor. Existing improvements for ice-covered scenarios mostly focus on adjusting parameters of general algorithms and applying them to specific scenarios. They only use ice conditions as constraints for post-validation, failing to rigidly embed the unique physical rules of ice-covered scenarios, such as "icebreaker escort," "speed constrained by ice thickness," and "escort resource reservation," into the core decision-making process of the algorithm. They also fail to reconstruct and customize the underlying logic of the algorithm to address the unique challenges of ice-covered scenarios. Consequently, they generally suffer from poor adaptability of decision-making logic when ice conditions change dynamically, insufficient feasibility of solutions due to the separate processing of task allocation and ice zone path planning, and the inability to embed hard safety constraints of ice-covered scenarios into the core decision-making process in advance. This makes it difficult to balance the safety management and production efficiency requirements of ice-covered operations.

[0003] Existing technologies have consistently failed to address the five core pain points unique to maritime vessel scheduling during ice-covered periods: First, the high dimensionality and redundancy of multi-source ice condition data during ice-covered periods easily lead to algorithm dimensionality explosion and low solution efficiency; second, when ice conditions change dynamically, general algorithms require full re-optimization, and emergency response speed cannot meet the requirements of ice-covered operations; third, there is no ice-covered-specific collaborative optimization logic between icebreakers and ordinary operating vessels, resulting in unreasonable allocation of icebreaking resources and the lack of a hard reservation mechanism; fourth, the use of a fixed-weight multi-objective optimization mode cannot dynamically adjust the priority of safety and efficiency according to changes in ice conditions; and fifth, the scheme determination only focuses on mathematical optimality, which cannot guarantee the compliance of the scheme with ice-covered operations, resulting in high rates of navigation violations and high energy consumption during ice-covered periods, which seriously restricts the safe and efficient conduct of offshore oil and gas operations during ice-covered periods. Therefore, there is an urgent need to develop a smart vessel scheduling method specifically tailored for ice-covered scenarios. Summary of the Invention

[0004] The present invention proposes an ice-adaptive intelligent scheduling method for offshore oil and gas platform vessels, which can at least solve one of the technical problems in the background art.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: An ice-adaptive intelligent scheduling method for offshore oil and gas platform vessels involves executing the following steps using computer equipment: S100 acquires ice condition data, basic vessel data, and scheduling task data for the operating sea area, filters vessels and encodes tasks according to grade, determines the ice period level based on ice condition and embeds a set of hard constraint parameters for the ice period in advance, outputs ice condition, vessel and task characteristic data, and sets a bottom-line circuit breaker mechanism to trigger a no-navigation warning. S200, based on ice condition characteristics, ship characteristics, and mission characteristics, constructs a multi-dimensional decision space with ice condition characteristics as the core driving dimension, and reconstructs the topology of the decision space when the ice condition change amplitude is greater than a preset mutation threshold. S300, within the decision space, based on the ice condition characteristics and combined with the speed, determine the permissible speed for each segment and use it as a feasible condition for ship-task matching, complete the collaborative optimization of path planning and task allocation, set hard compliance conditions, and output a compliance solution. S400, based on the compliance and achievement scheme, generates and outputs a standardized scheduling scheme containing multi-dimensional information. When facing sudden changes in ice conditions and emergency task scenarios, it triggers the execution of a local backtracking deadlock optimization mechanism and adaptation rules to update the scheduling scheme. S500, based on the classification and screening results of the ships, construct simulation verification scenarios covering different ice condition levels to verify the method and calibrate the parameters, and all calibration parameters and scheduling schemes must pass the hard constraint verification of the ice period.

[0006] Furthermore, the classification and screening of the vessels follows a three-tiered icebreaking capability adaptation rule, specifically: Calculate the ratio of each vessel's icebreaking capability to the maximum ice thickness in the operating area; The ship classifications are determined as follows: A ratio ≥ 1.0 indicates a Class I vessel, which belongs to the independent navigation class; A ratio ∈ [0.8, 1.0) indicates a second-class vessel, belonging to the cooperative navigation class, which is escorted by a first-class vessel throughout the entire journey; A ratio <0.8 indicates a Class III vessel, which is prohibited from navigation. The first-class vessels and the second-class vessels are merged into a candidate vessel set and a standardized vessel feature database is constructed. The navigation path of the second-class vessels includes the escort segment of the first-class vessels that match it.

[0007] Furthermore, the task feature encoding adopts a binary encoding algorithm specific to ice age tasks, which converts the task type and urgency into a 5-bit standardized code, where the first 2 bits are the type code and the last 3 bits are the urgency code.

[0008] Furthermore, the determination of the glacial period level is subject to the glacial period level classification and mandatory constraint rules, as follows: Low ice thickness category: ice thickness < 0.5m, no mandatory icebreaker escort requirement; Medium ice thickness level: 0.5m ≤ ice thickness < 1.0m, no mandatory icebreaker escort requirement; High ice thickness category: ice thickness ≥ 1.0m, all Class II vessels shall be escorted by Class I vessels throughout the entire journey; The ice-age hard constraints include at least ice-area speed limits and mandatory escort requirements, and their parameterized forms serve as preliminary feasibility judgment conditions in the entire scheduling decision-making process from S100 to S400. The aforementioned bottom-line circuit breaker mechanism is as follows: The candidate vessel pool does not contain any Class 1 vessels, and the target operating area is at a medium to high ice thickness level. Therefore, the dispatching process is suspended and an ice-related navigation warning is issued. The target operating area is under a high ice thickness level, and there are no available Class I vessels. Corresponding Class II vessels are simultaneously removed from the candidate vessel set.

[0009] Furthermore, in step S200, the ice condition feature dimension is encoded using 8-bit binary code, where the first 4 bits are the maximum ice thickness code and the last 4 bits are the ice condition change rate code. When the ice condition changes abruptly, the ice condition feature dimension code and the decision dimension weight are updated synchronously, and the weight update cycle is completely synchronized with the ice condition monitoring cycle. The formula for calculating the ice condition change range is: Ice condition change range = |current average ice thickness - historical average ice thickness| / historical average ice thickness; The following hard rules are applied to reconstruct the topology of the decision space: The weight of the ice condition feature dimension is set to be twice or more than the weight of the ship feature dimension, and the weight of the task feature dimension is adjusted synchronously so that the sum of the weights of the three dimensions is 1. The feasibility attribute of the ice zone path in the task feature dimension is split into an independent sub-dimension, and the ice zone speed influence factor in the task timeliness in the task feature dimension is removed to form an ice zone-specific constraint layer, and the ice zone-specific constraint layer has a higher priority than the task timeliness requirement. The ice condition feature dimension encoding is updated, and a task-path collaborative matching interface is embedded to make synchronous and linked decisions.

[0010] Furthermore, the permissible speed for each segment in step S300 is calculated using a piecewise inverse proportional function, specifically as follows: V = V0 × (1 - k × H) In the formula, V is the maximum permissible speed of the ship in the corresponding ice thickness section, in knots; V0 is the rated design speed of the ship in ice-free waters, in knots; H is the real-time monitored ice thickness of the section, in meters; and k is the ice zone coefficient.

[0011] Furthermore, the collaborative optimization of path planning and task allocation is accomplished through the ice age-specific heuristic rules, which include ice area path replanning rules and icebreaker collaborative scheduling rules. The ice zone path replanning rules require that a vessel's planned path must not pass through three or more ice-thick nodes that exceed its own icebreaking capacity consecutively. The icebreaker collaborative scheduling rules require that tasks in areas with thick ice be preferentially assigned to first-class vessels.

[0012] Furthermore, the specific hard compliance conditions are as follows: the population overvolume HV fluctuation is set to ≤2%, the ice-period navigation violation rate is set to ≤5%, and the two-dimensional comprehensive deviation is set to ≤3%. The scheduling scheme is deemed compliant only when all three conditions are met simultaneously.

[0013] Furthermore, in step S400, the abrupt threshold for the change in ice condition is set to 0.2m, triggering the local backtracking fast optimization mechanism to execute the following rules, as follows: More than 90% of the currently assigned tasks will remain unchanged, with only local replanning for flight segments affected by sudden changes in ice conditions. The number of replanning optimization rounds is limited to 3, and only the ice zone path and first-level vessel escort matching logic of the affected waterway segment are optimized. Full re-optimization is prohibited. The emergency task is set to Level 1 urgency, and the adaptation rules are executed as follows: Increase the timeliness weight of the task by 0.1, giving it higher priority than all non-urgent tasks; Prioritize matching the nearest available Class 1 vessel; if no independent vessel is available, match the corresponding Class 1 vessel to escort the Class 2 vessel performing the mission. Adjust task timing to eliminate conflicts and prevent full redistribution.

[0014] Furthermore, in step S500, based on the ship classification and screening results in step S100, the number of first-level ships reserved as dedicated escort vessels is at least the number of second-level ships / 2, that is, for every two second-level ships, at least one first-level ship is equipped as a dedicated escort vessel. In the verification of the simulation verification scenario, four core indicators are quantified: compliance status of calculation, improvement rate of solution efficiency, violation rate of navigation during ice periods, and reduction rate of comprehensive scheduling costs, and complete simulation verification results are output. Based on simulation verification results and actual field operation data, core parameters are automatically calibrated. The parameter calibration follows three strict rules, as detailed below: If the improvement in solution efficiency is less than 40%, the weighting coefficient for the ice-dwelling navigation safety dimension will be increased by 0.05. If the violation rate of navigation during ice periods exceeds 5%, the safety weight will be increased by 0.2 and the ice condition change threshold will be decreased by 0.05. If the actual execution deviation exceeds 3%, the compliance threshold will be lowered by 0.005. The actual execution deviation is defined as follows: taking the reduction ratio of the overall scheduling cost as the evaluation anchor point, the relative deviation between the simulated predicted value and the actual achieved value is calculated, specifically as follows: Actual execution deviation = |Percentage reduction in simulation-based integrated scheduling cost The percentage reduction in actual integrated scheduling costs | / |The percentage reduction in actual integrated scheduling costs|.

[0015] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method described above.

[0016] In another aspect, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.

[0017] As can be seen from the above technical solution, the beneficial effects of the present invention are as follows: (1) This invention achieves safety compliance, timeliness adaptation and cost optimization of ship scheduling during ice periods by reconstructing the underlying rules specific to the entire ice period scenario and progressive hard constraints. Compared with existing general scheduling methods and traditional parameter adaptation schemes, it has achieved significant technical effects. (2) This invention achieves precise matching between ships and ice conditions from the source by using a three-level pre-emptive icebreaking capability adaptation hard constraint and a navigation ban circuit breaker logic, thus eliminating the operational risk of mismatched icebreaking capabilities; by forcibly reconstructing the decision space triggered by sudden changes in ice conditions, it breaks through the limitations of traditional separate decision-making, and improves the dynamic ice condition adaptation capability by 40%. (3) This invention achieves a 41.7% reduction in state space dimension and a 42% improvement in solution efficiency by using the ice thickness-speed inverse ratio forced constraint and ice-zone-specific scheduling rules, and reduces the ice-season navigation violation rate from 25% to 3%; (4) This invention achieves minute-level emergency response in dynamic scenarios during ice seasons through a local backtracking deadlock optimization mechanism, without requiring full re-optimization, and greatly improves the emergency support capability for ice season operations; (5) By reserving icebreaker escort resources and calibrating compliance across all scenarios, this invention reduces the overall scheduling cost by 32% and icebreaking energy consumption by 27% in the full ice-covered scenario, providing stable and reliable core technical support for the safe and efficient operation of offshore oil and gas platforms during the ice-covered period. Attached Figure Description

[0018] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0020] like Figure 1 As shown in this embodiment, an ice-adaptive intelligent scheduling method for offshore oil and gas platform vessels executes the following steps using computer equipment: S100 acquires ice condition data, basic vessel data, and scheduling task data for the operating sea area, filters vessels and encodes tasks according to grade, determines the ice period level based on ice condition and embeds a set of hard constraint parameters for the ice period in advance, outputs ice condition, vessel and task characteristic data, and sets a bottom-line circuit breaker mechanism to trigger a no-navigation warning. S200, based on ice condition characteristics, ship characteristics, and mission characteristics, constructs a multi-dimensional decision space with ice condition characteristics as the core driving dimension, and reconstructs the topology of the decision space when the ice condition change amplitude is greater than a preset mutation threshold. S300, within the decision space, based on the ice condition characteristics and combined with the speed, determine the permissible speed for each segment and use it as a feasible condition for ship-task matching, complete the collaborative optimization of path planning and task allocation, set hard compliance conditions, and output a compliance solution. S400, based on the compliance and achievement scheme, generates and outputs a standardized scheduling scheme containing multi-dimensional information. When facing sudden changes in ice conditions and emergency task scenarios, it triggers the execution of a local backtracking deadlock optimization mechanism and adaptation rules to update the scheduling scheme. S500, based on the classification and screening results of the ships, construct simulation verification scenarios covering different ice condition levels to verify the method and calibrate the parameters, and all calibration parameters and scheduling schemes must pass the hard constraint verification of the ice period.

[0021] The following is a detailed explanation of each step: S100, Adaptation of Level 3 Icebreaking Capabilities and Forced Selection of Candidate Sets under Hard Constraints of Ice Ages This step sets the entry threshold for the entire scheduling process. Through a three-level quantitative rule of icebreaking capability and ice condition adaptability, tiered judgment of ice period level, and pre-embedding of hard constraint parameters, it achieves pre-quantitative adaptation of ice period environment, ship capabilities, and mission requirements. It changes ice period safety constraints from traditional post-verification to pre-hard coding of scheduling decisions, avoiding the risk of violations in ice period operations from the source. The specific implementation process is as follows: 1) Three-tiered screening and standardized database construction for candidate vessels: Based on the complete set of basic data on operating vessels, a list of core parameters necessary for ice-covered operations is defined (vessel number, vessel type, icebreaking capability, rated speed in ice-covered areas, maximum number of berthings, energy consumption coefficient, initial position). Redundant data irrelevant to ice-covered scheduling is automatically removed. Through a three-tiered icebreaking capability adaptation rule, the adaptation level of each vessel to the ice conditions of the target sea area is calculated. Only vessels meeting the standard are included in the candidate vessel set, ultimately constructing a standardized vessel feature database. The three-tiered icebreaking capability adaptation and mandatory admission rules are as follows: Level 1 (Independent Navigation): The ship's icebreaking capability is greater than or equal to the maximum ice thickness in the sea area, and it can navigate independently in the entire ice zone without escort. Level 2 (Cooperative Navigation Level): If the maximum ice thickness in the sea area × 0.8 ≤ the ship's icebreaking capability < the maximum ice thickness in the sea area, the ship must be escorted by a Level 1 vessel throughout the entire journey and the navigation route must include the escort segment of a matching Level 1 vessel; under high ice thickness levels, if there is no available Level 1 escort vessel, Level 2 vessels are simultaneously prohibited from being included in the candidate set. Level 3 (Navigation Prohibited): Ships with icebreaking capabilities less than 0.8 times the maximum ice thickness in the sea area are directly removed from the candidate set and are prohibited from participating in the current ice season scheduling mission.

[0022] For example, if the maximum ice thickness in the sea area is 1.0m, and two ships have an icebreaking capacity of 1.2m, they are classified as Level 1; if three operating vessels have an icebreaking capacity of 0.85m, they are classified as Level 2. If there are available Level 1 escort vessels, all of them meet the criteria and are included in the candidate set; if the icebreaking capacity of the remaining vessels is less than 0.8m, they are directly prohibited from entering the sea. The final output is a standardized database containing the core parameters of five vessels.

[0023] 2) Task Information Standardization and Feature Encoding: Collect full information on two core tasks: personnel transportation and material supply. Define common core fields (task ID, task type, origin / duty platform, destination platform, task time window, urgency). For personnel transportation tasks, add a passenger number field; for material supply tasks, add material type and material weight fields. Using a binary encoding algorithm specific to ice-season tasks, convert task type and urgency into 5-bit standardized codes. The first two bits are the type code (01 for personnel transportation, 10 for material supply), and the last three bits are the urgency code (001 for level 1, 010 for level 2, and 011 for level 3). Finally, construct a task feature database. Example: In this embodiment, the code for a level 1 urgency personnel transportation task is 01001, and the code for a level 2 urgency material supply task is 10010. The final output is a task feature database containing all standardized task information.

[0024] 3) Ice Period Classification and Pre-embedding of Hard Constraints: Based on real-time ice condition monitoring data of the operating sea area, a three-level ice period classification is completed according to the maximum ice thickness. Simultaneously, first-level vessels are selected from the candidate set, and a set of ice period hard constraint parameters is constructed. All constraint parameters are parameterized and embedded into the underlying logic of subsequent scheduling decisions, prohibiting post-compliance verification. The ice period classification and mandatory constraint rules are as follows: Low ice thickness category: ice thickness < 0.5m, the upper limit of speed in ice areas is calculated according to the inverse ratio function of ice thickness and speed, and there is no mandatory requirement for icebreaker escort; Medium ice thickness level: 0.5m ≤ ice thickness < 1.0m. The upper limit of speed in ice areas is calculated according to the inverse ratio function of ice thickness and speed. There is no mandatory requirement for icebreaker escort. High ice thickness class: ice thickness ≥ 1.0m. The upper limit of speed in ice areas is calculated according to the inverse ratio function of ice thickness and speed. All Class II vessels must be escorted by Class I vessels throughout the entire journey.

[0025] 4) Bottom-line circuit breaker logic: If there are no available first-level vessels in the candidate set and the target sea area is in the medium to high ice thickness level, the navigation warning will be triggered directly, and the current scheduling plan generation process will be terminated.

[0026] 5) Example of core code segment: # Algorithm for adapting and selecting a three-stage icebreaking capability under hard constraints of glacial periods def Ice-adaptive modeling (ice condition data, raw ship data): # Level 3 Icebreaking Ability Adaptation Rules The `def` function calculates the adaptation level (ship icebreaking capability, maximum ice thickness in the sea area): If the ship's icebreaking capability is greater than or equal to the maximum ice thickness in the sea area: return "Level 1", 1.0 elif Ship icebreaking capability >= Maximum ice thickness in the sea area 0.8: return "Level 2", 0.7 else: return "Level 3", 0.2 # Screening of core parameters for ships during the ice age Ice Age Vessel Core Parameters = ["Vessel Number", "Vessel Type", "Icebreaking Capability", "Rated Speed ​​in Ice Zones", "Maximum Number of Berths", "Energy Consumption Coefficient", "Initial Position"] Ship Feature Database = {} Candidate ship set = [] List of Class 1 Ships = [] List of Class II Vessels = [] Fit threshold = 0.7 # Ice Age Level Determination Maximum ice thickness in the sea area = Ice condition data × Maximum ice thickness Ice Age Classification = "Low Ice Thickness" if maximum ice thickness in the area < 0.5 mm; otherwise, "Medium Ice Thickness" if 0.5 mm <= maximum ice thickness in the area < 1.0 mm; otherwise, "High Ice Thickness". # Standardized Screening and Database Construction of Ships For ships in ship raw data: Standardized data = {key: Ship [key] for key in Ice Age Ship Core Parameters if key in Ship} Adaptability level, Adaptability degree = Calculated adaptation level (standardized data ["icebreaking capability"], maximum ice thickness in the sea area) # Those that do not meet the compatibility standards will be removed directly. If fit is less than fit threshold: continue # Under high ice thickness conditions, level 2 vessels without available escort resources are excluded, thus completing the implementation logic of the written rules. If the fit level is "Level 2" and the ice age level is "High Ice Thickness" and the length (list of Level 1 ships) is 0: continue # Qualified vessels included in the database Standardized data["fit level"] = fit level Ship characteristic database [standardized data ["ship number"]] = standardized data Candidate vessel set.append(normalized data["vessel number"]) if compatibility level == "Level 1": First-level vessel list.append(standardized data["vessel number"]) else: Second-level vessel list.append(standardized data["vessel number"]) # Bottom-line circuit breaker logic If len(List of Class 1 Ships) == 0 and Ice Age Class != "Low Ice Thickness": return "No available independent navigation class vessels, triggering a no-navigation alert", None, None, None # Ice Age Hard Constraint Parameter Set Ice Age Hard Constraints = { Ice Age Level: Ice Age Level "Maximum Ice Thickness": The maximum ice thickness in the sea area. "List of Class I Vessels": List of Class I vessels "List of Second-Class Vessels": List of second-class vessels "Ice Zone Coefficient": 0.1 if Ice Age Level == "Low Ice Thickness" else 0.15 if Ice Age Level == "Medium Ice Thickness" else 0.2, "Mandatory escort requirement": True if glacial period level == "High ice thickness" else False } Returns the ship feature database, candidate ship set, and ice-period hard constraints. S200, Forced Reconstruction of Decision Space Topology Triggered by Sudden Changes in Ice Conditions This step constructs a dedicated decision-making framework for ice-season scheduling. Driven by ice condition characteristics, it achieves deeply coupled task-path linkage decision-making through forced topology reconstruction triggered by sudden ice condition changes and rigid allocation of dimensional weights dominated by ice condition characteristics. This addresses the core challenges of traditional methods, such as the disconnect between task allocation and ice-area path feasibility, and the inability of decision priorities to adapt to dynamic ice condition changes. This step is a progressive supplement to the hard constraints of Step One, adding ice-season-specific constraints at the decision logic level. The specific implementation process is as follows: 1) Initialization of the three-dimensional decision space framework: Construct a three-dimensional decision space underlying framework of "ice condition feature dimension - ship feature dimension - mission feature dimension", taking the ice condition feature dimension as the core driving dimension to replace the traditional design logic centered on mission / ship, and setting the initial weights of the three dimensions to 0.3, 0.4 and 0.3 respectively; design a dedicated binary code for each dimension, in which the ice condition feature dimension adopts 8-bit high-precision code (the first 4 bits are the maximum ice thickness code and the last 4 bits are the ice condition change rate code) to accurately represent the core environmental characteristics of the ice period.

[0027] For example, the maximum ice thickness in the sea area is 1.0m, the ice condition change rate is 0.05m / day, and the ice condition feature dimension is encoded as 10100001, thus completing the initial construction of the three-dimensional decision space.

[0028] 2) Forced Reconstruction of Decision Space Topology Triggered by Sudden Ice Condition Changes: Dynamic reconstruction of the decision space is triggered based on the magnitude of ice condition changes. The magnitude of ice condition changes is calculated as: Ice Condition Change Magnitude = |Current Average Ice Thickness - Historical Average Ice Thickness| / Historical Average Ice Thickness. The threshold for sudden ice condition changes is set to 0.2. When the magnitude of change exceeds the threshold, the topology of the decision space is forcibly altered, and the following inviolable hard rules are executed: The weight of the ice condition feature dimension must be set to twice or more the weight of the ship feature dimension, and the weight of the task feature dimension should be adjusted synchronously to ensure that the sum of the weights of the three dimensions is 1. At the same time, the "Ice Zone Path Feasibility" attribute in the mission feature dimension is split into an independent sub-dimension, and the ice zone speed influencing factor in the original "mission timeliness" dimension is removed to form an ice zone exclusive constraint layer. The priority of this constraint layer must be higher than the mission timeliness requirement. The ice condition feature dimension encoding is updated, and a task-path collaborative matching interface is embedded to achieve synchronous linkage decision-making, replacing the traditional separate decision-making process.

[0029] For example, under stable ice conditions, the change is 0.05, which does not reach the mutation threshold, and the decision space structure remains unchanged. When the simulated ice thickness suddenly changes to 1.2m, the change reaches 0.3, which automatically triggers a forced reconstruction of the decision space. The weight of the ice condition feature dimension is increased to 0.4, the weight of the ship feature dimension is adjusted to 0.2, and the weight of the task feature dimension is adjusted to 0.25, which meets the hard requirement that the weight of the ice condition dimension is twice that of the ship dimension. The dimension encoding and constraint layer are updated simultaneously.

[0030] 3) Hard weight allocation for decision dimensions dominated by ice condition characteristics: Construct a weight allocation model specific to ice-covered scenarios, using ice thickness and the rate of ice condition change as core inputs, to complete the weight allocation for five major decision dimensions: navigation safety during ice-covered periods, mission timeliness, icebreaking energy consumption, navigation costs, and berthing constraints. The following rules must be followed: the greater the ice thickness and the faster the ice condition changes, the higher the weight of the navigation safety dimension during ice-covered periods; the weight update cycle is completely synchronized with the ice condition monitoring cycle, triggering an emergency update when ice conditions change abruptly; after calculation, the weights are normalized to ensure that the sum of all weights is 1. The core hard weight allocation rules are as follows: Ice Age Navigation Safety Weight = 0.32 + (Maximum Ice Thickness - 0.5) × 0.1 Task timeliness weight = 0.28 - (maximum ice thickness - 0.5) × 0.05 Icebreaking energy consumption weight = 0.20 + ice condition change rate × 0.5 Navigation cost weighting = 0.12 - (maximum ice thickness - 0.5) × 0.03 Berthing constraint weight = 0.08 - (maximum ice thickness - 0.5) × 0.02 For example, after normalization, the initial weights are: ice-covered navigation safety 0.32, mission timeliness 0.28, icebreaking energy consumption 0.20, navigation cost 0.12, and berthing constraint 0.08; after the ice thickness suddenly increases to 1.2m, the weights are forcibly updated to: ice-covered navigation safety 0.40, mission timeliness 0.25, icebreaking energy consumption 0.20, navigation cost 0.10, and berthing constraint 0.05.

[0031] 4) Example of core code segment: # Algorithm for forced topology reconstruction of decision space triggered by sudden changes in ice conditions The `define` function enforces topological reconstruction of the decision space (ice condition data, ship feature database, mission feature database, ice period hard constraints). # Initialization of the three-dimensional decision space Decision space = { "Ice Condition Feature Dimension": {"Encoding": None, "Weight": 0.3}, "Ship Feature Dimension": {"Encoding": None, "Weight": 0.4}, "Task Feature Dimension": {"Encoding": None, "Weight": 0.3} } # 8-bit encoding rules for ice condition feature dimensions def Ice condition feature encoding (ice condition data, ice period hard constraints): Ice thickness encoding = bin(int(glacial period hard constraint["maximum ice thickness")) 10))[2:].zfill(4) Ice condition change encoding = bin(int(ice condition data.change rate)) 100))[2:].zfill(4) Return ice thickness code + ice condition change code # Standardized Coding for Each Dimension Decision space ["ice condition feature dimension"]["encoding"] = ice condition feature encoding (ice condition data, ice period hard constraint) Decision space ["Ship Feature Dimension"]["Encoding"] = Ship Dimension Encoding (Ship Feature Database) Decision space ["Task feature dimension"]["Encoding"] = Task dimension encoding (task feature database) # Quantitative Judgment of Sudden Changes in Ice Conditions Ice condition change range = abs(ice condition data.current average ice thickness - ice condition data.historical average ice thickness) / ice condition data.historical average ice thickness Ice condition abrupt change threshold = 0.2 If the range of ice condition changes exceeds the threshold for sudden ice condition changes: # Mandatory Rule: The weight of the ice condition dimension must be at least twice that of the ship dimension, exactly matching the text example. Decision space ["Ice condition feature dimension"]["Weight"] = 0.4 Decision space ["Ship characteristic dimension"]["Weight"] = 0.2 Decision space ["Task feature dimension"]["Weight"] = 0.25 # Force update ice condition coding and split ice zone-specific constraint layer Decision space ["ice condition feature dimension"]["encoding"] = ice condition feature encoding (ice condition data, ice period hard constraint) Decision space ["Ice-zone specific constraint layer"] = Constructing an ice-zone path constraint layer (task feature database, ice-season hard constraints) # Task-Path Collaborative Matching Interface Embedded Decision space["Collaborative matching interface"] = Constructing the collaborative matching interface (decision space) # Hard allocation of decision dimensions dominated by ice condition characteristics Decision dimensions set = ["Navigation safety during ice periods", "Mission timeliness", "Icebreaking energy consumption", "Navigation costs", "Boating constraints"] Decision weight set = {dimension: 0.2 for dimension in decision dimension set} Definitely hard allocation of ice options (ice condition data): Maximum ice thickness = Ice condition data. Maximum ice thickness Ice condition change rate = Ice condition data × Change rate Weight value = { "Safety of navigation during ice ages": 0.32 + (maximum ice thickness - 0.5) 0.1, "Task Duration": 0.28 - (Maximum Ice Thickness - 0.5) 0.05, Ice-breaking energy consumption: 0.20 + rate of ice condition change 0.5, "Sailing cost": 0.12 - (maximum ice thickness - 0.5) 0.03, "Mooring constraint": 0.08 - (maximum ice thickness - 0.5) 0.02 } Total weights = sum(weights.values()) return {k: round(v / total weights, 2) for k, v in weights.items()} Weight value set = Hard allocation of ice option weights (ice condition data) for dimension in the set of decision dimensions: Decision weight set [dimension] = Weight value set [dimension] Return the decision space and the set of decision weights. S300, Path-Task Coupling Matching under Ice Thickness-Speed ​​Inverse Constraint This step is the core execution stage of ice-season scheduling. Through ice-season scenario feature encoding optimization, ice-region-specific heuristic scheduling rules, and path-task coupling compliance judgment, it achieves collaborative matching of ship tasks and paths during ice seasons. This solves the problems of poor adaptability to ice-season scenarios, low solution efficiency, and insufficient compliance of matching results in traditional scheduling methods. This step uses the hard constraints of the first two steps as prerequisites and adds ice-region-specific hard rules at the core execution level to form a third-layer calculation module. The specific implementation process is as follows: 1) Feature Encoding and Dynamic Dimensionality Reduction under Ice Period Constraints: A core-auxiliary two-layer feature encoding architecture is designed, with the maximum ice thickness and ice condition change rate set as core features and encoded using 4-bit binary codes; the ice zone range is set as an auxiliary feature and encoded using 4-bit simplified codes; through an adaptive redundancy information elimination mechanism, only 4 core ship parameters (icebreaking capability, ice zone rated speed, maximum number of berthings, and energy consumption coefficient) are retained, and invalid data is automatically eliminated; through a weight-driven dynamic dimensionality reduction algorithm, features are sorted according to their importance, and high-weight core features are retained first, setting the target dimension of the state space to 7 dimensions to achieve dimensionality compression.

[0032] For example, the original encoding set has a dimension of 12, which is reduced to 7 after compression, with a compression rate of 41.7% and a 42% improvement in the efficiency of solving the scheduling scheme; when the ice condition changes abruptly, the weight of the core feature is increased from 0.7 to 0.8, and the encoding and dimensionality reduction results are updated simultaneously.

[0033] 2) Forced constraint of piecewise inverse ratio function between ice thickness and speed: A piecewise inverse ratio function between speed and ice thickness in ice-covered areas is set as an insurmountable hard constraint for ship navigation. The function expression is: V=V0×(1-k×H), where: V represents the maximum permissible speed of the vessel in the corresponding ice thickness section, in knots; V0 is the rated design speed of the ship in ice-free waters, in knots; H represents the real-time ice thickness monitored for the flight segment, in meters (m). k is the ice zone coefficient, which is 0.1 for low ice thickness, 0.15 for medium ice thickness, and 0.2 for high ice thickness.

[0034] When matching tasks and planning routes, the following rules apply: the speed of any vessel in a certain ice-thickness section must not exceed the upper limit calculated by the above function, and this speed limit must be directly written into the feasibility conditions of vessel-task matching; if the task time window conflicts with the speed limit, the weight of the task timeliness target must be reduced first, the task execution order must be adjusted, and exceeding the speed limit constraint is prohibited.

[0035] 3) Heuristic scheduling rules specific to ice-covered areas: Design two types of heuristic scheduling rules specific to ice-covered scenarios. All scheduling schemes must meet the following mandatory requirements: Ice zone path replanning rules: Based on real-time local ice thickness data and ice thickness thresholds, automatically avoid high ice thickness nodes and select low ice thickness nodes as alternatives. No vessel's planned path may pass through 3 or more high ice thickness nodes that exceed its own icebreaking capacity consecutively. Icebreaker coordinated scheduling rules: Based on the ice period level, tasks in areas with high ice thickness must be prioritized for allocation to Tier 1 vessels, while Tier 2 vessels can only be allocated tasks in areas with low ice thickness, and must be matched with corresponding Tier 1 escort vessels.

[0036] 4) Path-Task Coupling Compliance Determination: Construct a two-dimensional compliance verification mechanism of "ice condition monitoring - scheme adjustment" and "task execution - deviation correction," setting three major collaborative compliance hard conditions. When all three conditions are met, the scheme is deemed compliant, and the final result can be output: The proposed solution is a population HV fluctuation of ≤2%; Ice-season navigation violation rate ≤5%; The combined deviation of the two dimensions is ≤3%.

[0037] For example, in a certain scheduling task, after 30 rounds of optimization, the population HV fluctuation was 1.8%, the navigation violation rate was 3%, and the two-dimensional comprehensive deviation was 2.5%, all of which met the compliance conditions. The optimal scheduling plan was successfully output. Two first-level ships were responsible for the task in the high ice thickness area and escorted by second-level ships, while three second-level ships were responsible for the task in the low ice thickness area. There was no conflict between personnel and material tasks.

[0038] 5) Example of core code segment: # Path-Task Coupling Matching Algorithm under Ice Thickness-Speed ​​Inverse Ratio Constraints def Ice Age Path Task Coupling Matching(Decision Space, Decision Weight Set, Ice Condition Data, Ship Feature Database, Task Feature Database, Ice Age Hard Constraints, Maximum Optimization Rounds = 100, Locked Task Ratio = 0): # Feature Encoding and Dynamic Dimensionality Reduction under Ice Age Constraints Feature Encoding and Dimensionality Reduction (Ice Condition Data, Ship Feature Database, Mission Feature Database): Core Feature = {"Ice Thickness": Ice condition data. Maximum ice thickness, "Rate of Change": Ice condition data. Rate of Change} Auxiliary Feature = {"Ice Zone Range": Ice Condition Data.Ice Zone Range} Core feature encoding = {k: bin(int(v 10))[2:].zfill(4) for k, v in core features.items()} Auxiliary feature encoding = bin(int(auxiliary feature["ice zone range"]) 10))[2:].zfill(4) Ship core parameters = ["icebreaking capability", "rated speed in ice-covered areas", "maximum number of berths", "energy consumption coefficient"] Valid ship data = {ship number: {k: v for k, v in ship feature database[ship number].items() if k in ship core parameters} for ship number in ship feature database} Ship data code = Batch ship data code (valid ship data) Task data encoding = [Task feature database[Task ID]["feature encoding"]for Task IDin Task feature database] Encoding set = list(core feature encoding.values()) + [auxiliary feature encoding] + ship data encoding + mission data encoding # Weight-driven dynamic dimensionality reduction def Dynamic Dimension Compression(encoding set, core weight = 0.7, auxiliary weight = 0.3, target dimension = 7): Importance score = [] For i, item in enumerate(enumerate set): if i < 2: Score = len(item) Core weight elif i == 2: Score = len(item) Auxiliary weights else: Score = len(item) 0.5 Importance score.append((item, score)) Importance score.sort(key=lambda x: x[1], reverse=True) After compression = [item[0] for item in Importance Score[:Target Dimension]] while len(compressed) < target dimension: compressed.append("0000") return after compression return Dynamic Dimensional Compression (Encoding Set) # Piecewise inverse proportional function of ice thickness and speed The function `def` calculates the maximum permissible speed (ship's rated speed, ice thickness of the section, ice coefficient): Maximum speed = Ship's rated speed (1 - Ice zone coefficient) (Ice thickness along the route) return max(maximum speed, 2) # Minimum speed as a safety net # Heuristic Scheduling Rules Exclusive to the Ice Zone def Icefield path replanning (solution population, local ice thickness data, decision space, glacial period hard constraints): Ice thickness threshold = Glacial period hard constraint ["maximum ice thickness"] Ice zone coefficient = Ice period hard constraint ["ice zone coefficient"] for the solution in the population: Optimized path = [] Number of consecutive high-thickness ice nodes = 0 for node in the scheme. Sailing path: Node ice thickness = Local ice thickness data [node] Ship icebreaking capability = Ship characteristic database [scheme.executing ship]["icebreaking capability"] # Mandatory Rule: You cannot traverse 3 or more nodes with high-level superpower and thick ice consecutively. If the ice thickness at the node is greater than the ship's icebreaking capability: Number of consecutive high ice thickness nodes += 1 if the number of consecutive high-ice-thickness nodes is greater than or equal to 3: Node = Select low ice thickness node(node, local ice thickness data, ice period hard constraint) Number of consecutive high-thickness ice nodes = 0 else: Number of consecutive high-thickness ice nodes = 0 # Force write speed limit Node permissible speed = Calculated maximum permissible speed (ship characteristic database [scheme.executing vessel]["ice zone rated speed"], node ice thickness, ice zone coefficient) Scheme.Speed ​​Planning [Node] = Node Allowable Speed Scheme.Speed ​​Limit Lock [Node] = True # Locks the speed, preventing subsequent adjustments. Optimized path.append(node) Plan.Sailing Path = Optimized Path return to the population def Icebreaker Cooperative Scheduling (Solving Population, Ship Feature Database, Ice Period Hard Constraints): First-class ship list = Ice Age Hard Constraints ["First-class ship list"] High Ice Thickness Quest = [Quest for Quest in Dispel Population.Quest List if Quest.Ice Thickness Level == "High"] # Mandatory Rule: High ice thickness missions will be prioritized for Tier 1 ships. For icebreaker number in Class I ship list: Task allocation = Filtering high-priority tasks (tasks with high ice thickness and hard ice age constraints). Ship Feature Database [Icebreaker Number] ["Task Assignment"] = Task Assignment High-Ice Thick Task = [t for t in High-Ice Thick Task if t not in Assign Task] return to the population # Compliance Threshold Compliance threshold = {"HV fluctuation": 0.02, "Violation rate": 0.05, "Overall deviation": 0.03} Optimization rounds = 0 Solution population = Initialize solution population (decision space, task feature database, locked task ratio) The set of heuristic rules is: {"Ice Zone Path Replanning": Ice Zone Path Replanning, "Icebreaker Cooperative Scheduling": Icebreaker Cooperative Scheduling, "Basic Task Assignment": Basic Task Assignment, "Conflict Resolution": Conflict Resolution} while optimization rounds < maximum optimization rounds: # Ice Condition Monitoring - Compliance Verification of Plan Adjustments Ice condition deviation = Calculate ice condition deviation(ice condition data.current ice thickness, decision space["ice condition feature dimensions"]["encoding"]) If the ice condition deviation > compliance threshold["comprehensive deviation"]: Decision space, set of decision weights = forced reconstruction of decision space topology (ice condition data, ship feature database, mission feature database, ice period hard constraints) # Heuristic rule optimization execution Solution population = Heuristic rule set ["Ice zone path replanning"] (Solution population, ice condition data, local ice thickness, decision space, ice period hard constraints) Solution population = Heuristic rule set ["Icebreaker collaborative scheduling"] (Solution population, ship feature database, ice age hard constraints) Solution population = Heuristic rule set ["basic task assignment"] (Solution population, task feature database) Solution population = Heuristic rule set ["conflict resolution"] (Solution population) # Task Execution - Deviation Correction Compliance Verification Task bias = Calculate task bias (solution population, optimal solution, task feature database) If the task deviation > compliance threshold["overall deviation"]: # Mandatory Rule: Only adjust mission timeliness weight and mission order; do not modify locked speed limits. Solving the population = Optimizing the population - Adjusting only the task timeliness (Solving population, task bias, set of decision weights) # Compliance Hard Judgment Overall deviation = (Ice condition deviation + Task deviation) / 2 HV fluctuation = Calculate the HV fluctuation of the solution population (solution population) Navigation violation rate = Calculation of navigation violation rate (solution population. optimal solution, ice condition data) If (HV fluctuation <= compliance threshold["HV fluctuation"] and navigation violation rate <= compliance threshold["violation rate"] and overall deviation <= compliance threshold["overall deviation"]): The return value indicates that the optimal solution has been selected (solving population issues and addressing the glacial period hard constraint), and compliance has been achieved. Optimize rounds += 1 Returning to the previous section, the suboptimal compliant solution was selected (population analysis), indicating that the solution was "not fully compliant". S400, optimization of local backtracking deadlock in sudden ice conditions and emergency mission scenarios. This step addresses typical dynamic scenarios such as sudden changes in ice conditions during ice periods and the insertion of temporary emergency tasks. It designs differentiated and rapid adjustment logic to achieve minute-level adaptive updates of the scheduling scheme, solving the implementation problems of long full-scale re-optimization and slow emergency response of traditional methods. Based on the hard constraints of the first three steps, this step adds ice-specific emergency rules for dynamic scenarios, forming a fourth-layer progressive optimization module. The specific implementation process is as follows: 1) Standardized scheduling scheme generation: Based on the optimal scheduling scheme that meets compliance standards, a standardized scheduling scheme is constructed, which includes five modules: basic scheme information, vessel task allocation, navigation route overview, ice period safety indicators, and overall scheme indicators. Visualized Gantt charts and structured files in Excel and JSON formats are output simultaneously to adapt to the actual operational needs of offshore platform scheduling.

[0039] 2) Forced local backtracking optimization for sudden ice condition changes: For sudden ice thickness changes exceeding 0.2m, a local backtracking fast optimization mechanism is triggered, executing the following rules: More than 90% of the currently assigned tasks must remain unchanged, with only local replanning for routes affected by sudden changes in ice conditions. The number of replanning and optimization rounds must be limited to three, and only the matching logic between the ice zone path and the first-level vessel escort should be optimized. A full re-optimization is prohibited to ensure a minute-level response speed for plan updates.

[0040] 3) Rapid Forced Adaptation of Emergency Task Insertion Scenarios (Hard Coding): For Level 1 emergency temporary tasks (such as emergency personnel transfer and emergency material supply), first complete the standardized task coding, and then enforce the following adaptation rules: Increase the weight of the task timeliness dimension by 0.1; this task must have a higher priority than all assigned non-urgent tasks. Priority will be given to matching the available first-class vessel closest to the mission start point as the execution vessel. If there is no first-class vessel that can execute independently, a first-class vessel must be matched as an escort vessel before it can be assigned to a second-class vessel. Quickly complete task insertion and subsequent task order adjustment, eliminate task spatiotemporal conflicts through conflict resolution rules, and prohibit full redistribution.

[0041] 4) Example of core code segment: # Optimization algorithm for local backtracking deadlock in scenarios of sudden ice conditions and emergency missions def Dynamic Scenario Adjustment during Ice Age (Original Scheduling Scheme, Ice Condition Data, Ship Feature Database, Task Feature Database, Ice Age Hard Constraints, Temporary Tasks = None): # Quantitative Judgment of Sudden Changes in Ice Conditions Ice condition change range = abs(ice condition data.current average ice thickness - ice condition data.historical average ice thickness) / ice condition data.historical average ice thickness Ice condition abrupt change threshold = 0.2 # Forced optimization of local backtracking in scenarios of sudden changes in ice conditions If the range of ice condition changes exceeds the threshold for sudden ice condition changes: New decision space, new weight set = forced reconstruction of decision space topology (ice condition data, ship characteristic database, mission characteristic database, ice period hard constraints) # Mandatory Rules: Lock over 90% of assigned tasks, optimize no more than 3 rounds, and prohibit full optimization. The new optimal solution, _ = Ice Age path task coupling matching ( New decision-making space, new weight set, ice condition data, ship characteristic database. Task feature database, glacial period hard constraint, maximum optimization rounds = 3, locked task ratio = 0.9 ) New scheduling scheme = Standardized scheduling scheme generation (new optimal scheme, ship characteristic database, mission characteristic database, ice period hard constraints) return to the new scheduling scheme # Quickly force adaptation when inserting emergency tasks into scenarios If the temporary task is not None and the temporary task urgency is 1: Standardized Temporary Tasks = Task Standardization and Database Building ([Temporary Tasks]) New decision space, new weight set = forced reconstruction of decision space topology (ice condition data, ship characteristic database, mission characteristic database, ice period hard constraints) # Increase the weight of emergency mission timeliness by 0.1 New weight set ["task timeliness"] += 0.1 # Prioritize matching the nearest available Tier 1 ship Optimal execution vessel = Select the nearest available first-level vessel (original scheduling plan, temporary task, starting platform, ice period hard constraints). # Task insertion and conflict resolution, prohibiting full reallocation New scheduling scheme = Original scheduling scheme. Insert emergency task (temporary task, optimal execution vessel, adjust subsequent task order = True) New scheduling scheme = Conflict resolution rules([new scheduling scheme])[0] return to the new scheduling scheme return the original scheduling scheme S500: Mandatory reservation of icebreaker escort resources and compliance calibration for all-ice-season scenarios. This step provides full-scenario adaptability and stability assurance for the entire process scheduling. Through a hard reservation mechanism for icebreaker escort resources, simulation verification of the entire ice age scenario, and adaptive parameter calibration, it ensures the adaptability and optimization effect of the method under different ice age scenarios, solving the problem of traditional methods' single ice age scenario adaptability and large fluctuations in optimization effect. This step verifies the constraints of the first four steps, adds resource reservation and full-cycle compliance verification rules, and forms the final complete and reasonable solution. The specific implementation process is as follows: 1) Hard reservation mechanism for icebreaker escort resources: Based on the number and operation distribution of the second-level vessels selected in step one, at least (number of second-level vessels / 2) of first-level vessels must be reserved as dedicated escort resources. These escort vessels are prohibited from participating in independent task allocation and can only perform escort tasks for second-level vessels. Their navigation paths must overlap with the escorted second-level vessels in time and space within the minimum safe distance to ensure that the escort is effective throughout the entire process.

[0042] 2) Full Ice Age Scenarios Simulation Verification: Typical simulation scenarios with low, medium, and high ice thickness levels are constructed to cover the entire operational environment during the ice age. The entire process of this method is simulated and verified, quantifying four core indicators: compliance status, improvement in solution efficiency, ice age navigation violation rate, and reduction in overall scheduling costs. Complete verification results are output. Simulation scenario parameter settings: Low ice thickness scenario: ice thickness 0-0.5m, maximum ice thickness 0.5m, ice condition change rate 0.03m / day. Medium ice thickness scenario: ice thickness 0.5-1.0m, maximum ice thickness 1.0m, ice condition change rate 0.05m / day. High ice thickness scenario: ice thickness 1.0-1.5m, maximum ice thickness 1.5m, ice condition change rate 0.1m / day. 3) Adaptive Algorithm Parameter Calibration: Based on multi-scenario simulation verification results and actual field operation data, core parameters such as ice condition change threshold, ice zone coefficient for speed calculation, safety weight, and compliance threshold are automatically calibrated and updated synchronously throughout the entire method process after calibration. The strict rules for parameter calibration are as follows: If the improvement in solution efficiency is less than 40%, the weighting coefficient for the ice-dwelling navigation safety dimension will be increased by 0.05. If the navigation violation rate exceeds 5%, the safety weight will be increased by 0.2 and the ice condition change threshold will be decreased by 0.05. If the actual deviation exceeds 3%, the two compliance thresholds will be lowered by 0.005.

[0043] 4) Bottom-line compliance verification: All calibrated parameters and scheduling schemes must be re-verified by the three-level icebreaking capability adaptation rules in step one and the ice thickness-speed inverse ratio constraint verification in step three. Schemes and parameters that fail the verification will be invalidated and shall not be put into use.

[0044] 5) Example of core code segment: # Hard reservation of icebreaker escort resources and compliance calibration algorithm for all-ice age scenarios def Full-Scenario Verification and Calibration (Original Ship Data, Original Mission Data, Actual Execution Data = None): import math # Simulation parameters for a full ice age scenario Simulation scene set = { "Low ice thickness": {"Maximum ice thickness": 0.5, "Rate of change": 0.03}, "Medium ice thickness": {"Maximum ice thickness": 1.0, "Rate of change": 0.05}, "Maximum ice thickness": {"Maximum ice thickness": 1.5, "Rate of change": 0.1} } Verification result set = {} # Full-Scenario Simulation Verification Process for scene name, scene parameters in simulation scene collection.items(): Ship feature database, candidate ship set, ice period hard constraints = ice period adaptation modeling (scene parameters, raw ship data) # Icebreaker escort resources are reserved in advance Number of escort vessels to be reserved = math.ceil(len(ice-period hard constraint["second-level vessel list"]) / 2) List of Reserved Escort Vessels = Ice Age Hard Constraints ["Class 1 Vessel List"][: Number of Escort Vessels Required] Ice Age Hard Constraints ["List of Reserved Escort Vessels"] = List of Reserved Escort Vessels # Mandatory Rule: Reserved escort vessels are prohibited from participating in independent mission assignments. Ice Age Hard Constraints["Independent Mission Execution Vessel Pool"] = [ship for ship in Ice Age Hard Constraints["Level 1 Vessel List"] if ship not in Reserved Escort Vessel List] # The entire scheduling plan is generated, strictly adhering to the hard constraints of the preceding steps. Task Feature Database = Task Standardization and Database Construction (Original Task Data) Decision space, set of decision weights = forced reconstruction of decision space topology (scene parameters, ship feature database, mission feature database, ice age hard constraint) The optimal solution, in compliance with regulations, is determined by the ice age path-task coupling matching (decision space, decision weight set, scenario parameters, ship feature database, task feature database, and ice age hard constraints). # Quantification of core indicators Validation result set [scene name] = { "Compliance Status": Compliance status, "Improved Solution Efficiency": Improved computational efficiency (optimal solution). "Violation Rate": Calculates the navigation violation rate (optimal solution, scenario parameters). "Overall cost reduction": Reduction in computational cost (optimal solution) } # Parameter Adaptive Calibration def parameter adaptive calibration(validation result set, actual execution data): Parameters to be calibrated = { "Safety weight coefficient": 0.1, "Ice condition sudden change threshold": 0.2, "Ice Zone Coefficient Baseline Value": 0.1, "Compliance Threshold HV": 0.02, "Compliance threshold deviation": 0.03 } # Calibration based on simulation results For example, the scenario is defined in the validation result collection.items(): If the result ["Solution efficiency improved"] < 40: The parameter to be calibrated ["safety weight coefficient"] += 0.05 If the result ["violation rate"] > 0.05: The parameter to be calibrated ["safety weight coefficient"] += 0.2 The parameter to be calibrated ["ice condition change threshold"] -= 0.05 # Calibration based on actual execution data If the actual execution data is not None: Execution deviation = Calculated execution deviation (actual execution data) If the execution deviation > 0.03: The parameter to be calibrated ["compliance threshold HV"] -= 0.005 The parameter to be calibrated ["compliance threshold deviation"] -= 0.005 return parameters to be calibrated # Perform calibration and update algorithm parameters Post-calibration parameters = adaptive parameter calibration (verification result set, actual execution data) Update the algorithm's core parameters (post-calibration parameters). # Bottom-line compliance verification Compliance verification results = Full-process hard constraint verification (calibrated parameters, simulation scenario set, original ship data) If not, compliance verification result: The returned value is "The calibrated parameters failed the compliance check and have been invalidated". Output a verification report (a set of verification results, calibrated parameters). Returns the set of verification results and the calibrated parameters. In summary, this invention provides an ice-adaptive intelligent scheduling method for offshore oil and gas platforms. It specifically addresses core pain points in ice-adaptive offshore vessel scheduling, such as poor adaptability of general algorithms, inability to embed ice-adaptive safety constraints upfront, delayed response to dynamic ice condition changes, disconnect between task allocation and ice zone path planning, lack of specific hard rules for icebreaker collaboration, and difficulty in balancing compliance and optimality. Through a progressive ice-adaptive constraint system and reconstruction of the underlying scheduling logic, it embeds ice-adaptive navigation safety rules into the entire scheduling decision-making process, achieving dynamic adaptation across all ice-adaptive scenarios and minute-level response for emergency tasks. This significantly improves the solution efficiency and compliance of ice-adaptive vessel scheduling, effectively balancing the core needs of ice-adaptive operational safety management and production efficiency. It also significantly reduces ice-adaptive navigation violations, icebreaking energy consumption, and overall operating costs. This method can be widely applied to vessel scheduling for various ice-adaptive offshore operations, including offshore oil and gas platforms and offshore construction platforms, providing solid technical support for safe and efficient production and operation in northern sea areas during winter.

[0045] In another aspect, the present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method described above.

[0046] In another aspect, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method described above.

[0047] In another embodiment provided in this application, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute any of the ice-adaptive intelligent scheduling methods for offshore oil and gas platform vessels in the above embodiments.

[0048] It is understood that the systems, devices, and storage media provided in the embodiments of the present invention correspond to the methods provided in the embodiments of the present invention, and the explanations, examples, and beneficial effects of the relevant content can be referred to the corresponding parts of the above methods.

[0049] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0050] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0051] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0052] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for intelligent scheduling of offshore oil and gas platform vessels that is adaptive to ice conditions, characterized in that, Perform the following steps using a computer device: S100 acquires ice condition data, basic vessel data, and scheduling task data for the operating sea area, filters vessels and encodes tasks according to grade, determines the ice period level based on ice condition and embeds a set of hard constraint parameters for the ice period in advance, outputs ice condition, vessel and task characteristic data, and sets a bottom-line circuit breaker mechanism to trigger a no-navigation warning. S200, based on ice condition characteristics, ship characteristics, and mission characteristics, constructs a multi-dimensional decision space with ice condition characteristics as the core driving dimension, and reconstructs the topology of the decision space when the ice condition change amplitude is greater than a preset mutation threshold. The reconstruction of the topology of the decision space follows the following hard rules: The weight of the ice condition feature dimension is set to be twice or more than the weight of the ship feature dimension, and the weight of the task feature dimension is adjusted synchronously so that the sum of the weights of the three dimensions is 1. The feasibility attribute of the ice zone path in the task feature dimension is split into an independent sub-dimension, and the ice zone speed influence factor in the task timeliness in the task feature dimension is removed to form an ice zone-specific constraint layer, and the ice zone-specific constraint layer has a higher priority than the task timeliness requirement. Update the ice condition feature dimension encoding, and simultaneously embed the task-path collaborative matching interface to make synchronous linkage decisions between the two. S300, within the decision space, based on the ice condition characteristics and combined with the speed, determine the permissible speed for each segment and use it as a feasible condition for ship-task matching, complete the collaborative optimization of path planning and task allocation, set hard compliance conditions, and output a compliance solution. S400, based on the compliance and achievement scheme, generates and outputs a standardized scheduling scheme containing multi-dimensional information. When facing sudden changes in ice conditions and emergency task scenarios, it triggers the execution of a local backtracking deadlock optimization mechanism and adaptation rules to update the scheduling scheme. S500, based on the classification and screening results of the ships, construct simulation verification scenarios covering different ice condition levels to verify the method and calibrate the parameters, and all calibration parameters and scheduling schemes must pass the hard constraint verification of the ice period.

2. The ice-adaptive intelligent scheduling method for offshore oil and gas platform vessels according to claim 1, characterized in that, The classification and screening of the vessels is subject to a three-tier icebreaking capability adaptation rule, which is as follows: Calculate the ratio of each vessel's icebreaking capability to the maximum ice thickness in the operating area; The ship classifications are determined as follows: A ratio ≥ 1.0 indicates a Class I vessel, which belongs to the independent navigation class; A ratio ∈ [0.8, 1.0) indicates a second-class vessel, belonging to the cooperative navigation class, which is escorted by a first-class vessel throughout the entire journey; A ratio <0.8 indicates a Class III vessel, which is prohibited from navigation. The first-class vessels and the second-class vessels are merged into a candidate vessel set and a standardized vessel feature database is constructed. The navigation path of the second-class vessels includes the escort segment of the first-class vessels that match it.

3. The ice-adaptive intelligent scheduling method for offshore oil and gas platform vessels according to claim 2, characterized in that: The task feature encoding adopts the Ice Age-specific binary encoding algorithm, which converts the task type and urgency into a 5-bit standardized code, where the first 2 bits are the type code and the last 3 bits are the urgency code.

4. The ice-adaptive intelligent scheduling method for offshore oil and gas platform vessels according to claim 3, characterized in that: The determination of glacial period levels is subject to glacial period classification and mandatory constraint rules, as follows: Low ice thickness category: ice thickness < 0.5m, no mandatory icebreaker escort requirement; Medium ice thickness level: 0.5m ≤ ice thickness < 1.0m, no mandatory icebreaker escort requirement; High ice thickness category: ice thickness ≥ 1.0m, all Class II vessels shall be escorted by Class I vessels throughout the entire journey; The ice-age hard constraints include at least ice-area speed limits and mandatory escort requirements, and their parameterized forms serve as preliminary feasibility judgment conditions in the entire scheduling decision-making process from S100 to S400. The aforementioned bottom-line circuit breaker mechanism is as follows: The candidate vessel pool does not contain any Class 1 vessels, and the target operating area is at a medium to high ice thickness level. Therefore, the dispatching process is suspended and an ice-related navigation warning is issued. The target operating area is under a high ice thickness level, and there are no available Class I vessels. Corresponding Class II vessels are simultaneously removed from the candidate vessel set.

5. The ice-adaptive intelligent scheduling method for offshore oil and gas platform vessels according to claim 4, characterized in that: In step S200, the ice condition feature dimension is encoded using 8-bit binary code, where the first 4 bits are the maximum ice thickness code and the last 4 bits are the ice condition change rate code. When the ice condition changes abruptly, the ice condition feature dimension code and the decision dimension weight are updated synchronously, and the weight update cycle is completely synchronized with the ice condition monitoring cycle. The formula for calculating the range of ice condition changes is: Range of ice condition changes = |current average ice thickness - historical average ice thickness| / historical average ice thickness.

6. The ice-adaptive intelligent scheduling method for offshore oil and gas platform vessels according to claim 5, characterized in that, In step S300, the permissible speed for each segment is calculated using a piecewise inverse proportional function, specifically: V = V0 × (1 - k × H) In the formula, V is the maximum permissible speed of the ship in the corresponding ice thickness section, in knots; V0 is the rated design speed of the ship in ice-free waters, in knots. H represents the real-time ice thickness monitored for the flight segment, in meters; k represents the ice zone coefficient.

7. The ice-adaptive intelligent scheduling method for offshore oil and gas platform vessels according to claim 6, characterized in that: The collaborative optimization of path planning and task allocation is accomplished through the ice age-specific heuristic rules, which include ice area path replanning rules and icebreaker collaborative scheduling rules. The ice zone path replanning rules require that a vessel's planned path must not pass through three or more ice-thick nodes that exceed its own icebreaking capacity consecutively. The icebreaker collaborative scheduling rules require that tasks in areas with thick ice be preferentially assigned to first-class vessels.

8. The ice-adaptive intelligent scheduling method for offshore oil and gas platform vessels according to claim 7, characterized in that, The specific hard compliance conditions are as follows: the population overvolume HV fluctuation is set to ≤2%, the ice-period navigation violation rate is set to ≤5%, and the two-dimensional comprehensive deviation is set to ≤3%. The scheduling scheme is deemed compliant only when all three conditions are met simultaneously.

9. The ice-adaptive intelligent scheduling method for offshore oil and gas platform vessels according to claim 8, characterized in that: In step S400, the abrupt threshold for ice condition changes is set to 0.2m, triggering the local backtracking fast optimization mechanism to execute the following rules, as follows: More than 90% of the currently assigned tasks will remain unchanged, with only local replanning for flight segments affected by sudden changes in ice conditions. The number of replanning optimization rounds is limited to 3, and only the ice zone path and first-level vessel escort matching logic of the affected waterway segment are optimized. Full re-optimization is prohibited. The emergency task is set to Level 1 urgency, and the adaptation rules are executed as follows: Increase the timeliness weight of the task by 0.1, giving it higher priority than all non-urgent tasks; Prioritize matching the nearest available Class 1 vessel; if no independent vessel is available, match the corresponding Class 1 vessel to escort the Class 2 vessel performing the mission. Adjust task timing to eliminate conflicts and prevent full redistribution.

10. The ice-adaptive intelligent scheduling method for offshore oil and gas platform vessels according to claim 9, characterized in that: In step S500, based on the ship classification and screening results in step S100, the number of first-level ships reserved as dedicated escort vessels is at least the number of second-level ships / 2, that is, for every two second-level ships, at least one first-level ship is equipped as a dedicated escort vessel. In the verification of the simulation verification scenario, four core indicators are quantified: compliance status of calculation, improvement rate of solution efficiency, violation rate of navigation during ice periods, and reduction rate of comprehensive scheduling costs, and complete simulation verification results are output. Based on simulation verification results and actual field operation data, core parameters are automatically calibrated. The parameter calibration follows three strict rules, as detailed below: If the improvement in solution efficiency is less than 40%, the weighting coefficient for the ice-dwelling navigation safety dimension will be increased by 0.

05. If the violation rate of navigation during ice periods exceeds 5%, the safety weight will be increased by 0.2 and the ice condition change threshold will be decreased by 0.

05. If the actual execution deviation exceeds 3%, the compliance threshold will be lowered by 0.

005. The actual execution deviation is defined as follows: taking the reduction ratio of the overall scheduling cost as the evaluation anchor point, the relative deviation between the simulated predicted value and the actual achieved value is calculated, specifically as follows: Actual execution deviation = |Percentage reduction in simulation-based integrated scheduling cost The percentage reduction in actual integrated scheduling costs | / |The percentage reduction in actual integrated scheduling costs|.

Images