A method for optimizing marine oil spill emergency recovery

Abstract

The present invention discloses a method for optimizing marine oil spill emergency recovery, belonging to the field of oil recovery technology. The method comprises the following steps: obtaining marine oil spill accident risk factors; constructing a three-objective function consisting of maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact; solving the three-objective function using a multi-objective genetic-particle swarm hybrid algorithm, combining a non-dominated sorting genetic algorithm with a particle swarm optimization algorithm, to obtain an optimal solution set for different configurations of different numbers of oil skimmers under different weights for maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact; and employing a fuzzy superior-inferior solution distance method to find the optimal number of oil skimmers under certain weights from the optimal solution set for different configurations of different numbers of oil skimmers, thereby obtaining a marine oil spill emergency recovery solution. This method can reduce environmental and economic losses, thereby achieving efficient and low-cost optimization of light crude oil recovery strategies.

Description

Technical Field

[0001] The present invention relates to the field of oil recovery, and in particular to an optimization method for emergency recovery of offshore oil spills. Background Art

[0002] Oil spills at sea not only pose a serious threat to the marine ecosystem, but may also have a profound impact on economic activities such as fisheries and tourism. The spread rate of oil spills, changes in environmental factors, and the weathering process of crude oil all make timely and effective emergency responses particularly complicated. In this case, crude oil recovery operations have become an important means to mitigate the impact of oil spills and restore marine ecology. Current marine oil spill emergency response technologies have significant defects. Existing methods mostly focus on a single goal, making it difficult to coordinately optimize multiple conflicting goals, and generally ignore the nonlinear effects of dynamic environmental parameters such as wind speed and temperature on oil film diffusion and skimmer efficiency; traditional optimization algorithms are prone to falling into local optimality or insufficient solution set diversity, and are unable to balance global search and local development capabilities; most models lack dynamic modeling of the oil weathering process, resulting in the inability to adapt to the attenuation of oil film thickness in real time during recovery; in addition, existing studies mostly use static equipment efficiency coefficients, without considering the coupling effect of environment and equipment performance. Summary of the Invention

[0003] In order to solve the problems of insufficient multi-objective optimization capability, poor adaptability to dynamic environments, easy local optimization and insufficient solution diversity in existing marine oil spill emergency response methods, a marine oil spill emergency recovery optimization method is proposed, which includes the following steps:

[0004] Obtain risk factors for marine oil spills;

[0005] The risk parameters of marine oil spill accidents are preprocessed by Monte Carlo simulation method to obtain Monte Carlo sample set;

[0006] Based on the pre-processed marine oil spill risk factors, a three-objective function is constructed, including maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact.

[0007] The objective function for maximizing oil recovery is expressed as follows:

[0008]

[0009] in: is the volume of recovered oil, For operation time, The upper limit of the operating time for oil recovery using skimmers is for The number of oil skimmers, For the Oil skimmers in arrive The amount of oil recovered during the phase;

[0010] The objective function of minimizing the recovery cost is expressed as follows:

[0011]

[0012] in: is the total cost, It is a type of oil skimmer. is the upper limit of the types of skimmers, It is The operating cost of a skimmer is related to the volume of the spilled oil. It is The transportation cost of the skimmers from the warehouse to the spill site;

[0013] The objective function expression of minimizing environmental impact is as follows:

[0014]

[0015] in, is the base rate, is the volume, For geographical location, Indicates a special management area. The pollutant characteristics, : indicates environmental impact;

[0016] Step 4: Construct the constraints of the three-objective function considering the petroleum weathering process, including: the equality constraint of volume decay and the resource constraint of recovery equipment;

[0017] The equality constraint for the volume decay is:

[0018]

[0019]

[0020] in: for The oil film thickness of the stage, for The amount of oil remaining after oil recovery and natural weathering; is the initial volume of the oil spill, A is the oil spill area, for arrive The amount of oil lost during this period through oil recovery and natural weathering processes, for to The volume of oil lost during the evaporation and weathering process during this period, for to The volume of oil lost during the oil dispersion process during the period;

[0021] in:

[0022] Oil recovery The expression is as follows:

[0023]

[0024] in: For the Oil skimmers in arrive The amount of oil recovered in each stage, , is the empirical coefficient of the oil skimmer, is the oil film thickness;

[0025] The volume of oil lost during oil evaporation and weathering The equation is as follows:

[0026]

[0027]

[0028] in: and is the evaporation equation parameter of a specific oil, for The evaporation rate of the stage, is the temperature, For time, Express Taking the natural logarithm, For the to Volume of evaporated oil in each stage (m 3 ), for The amount of oil remaining after oil recovery and natural weathering;

[0029] The volume of oil lost during oil dispersion The equation is as follows:

[0030]

[0031]

[0032]

[0033]

[0034] in: for is the water content, and the water content is used to simulate the oil emulsification phenomenon. is the curve fitting constant that varies with wind speed, is the molar viscosity constant, represents the exponential function, is the wind speed, For time; is the dispersion rate, is the wind speed, is the dynamic viscosity of the oil, for The oil film thickness of the stage, is the oil-water interfacial tension, for The dynamic viscosity of the stage oil, for The dynamic viscosity of the stage oil, for The evaporation rate of the stage, for The moisture content of the stage, is the molar viscosity constant, represents the exponential function, For the to Volume of dispersed oil in the stage, for The dispersion rate of the stage, for The amount of oil remaining after oil recovery and natural weathering;

[0035] The recycling equipment resource constraints are as follows:

[0036]

[0037]

[0038] in: For the The number of oil skimmers, The maximum limit for the number of skimmers is Represents an integer;

[0039] Combined with the Monte Carlo sample set obtained after preprocessing, the three objective functions are transformed into a robust form containing expected value and variance, balancing the stability and volatility of the objective function to obtain the processed objective function;

[0040] A multi-objective genetic-particle swarm optimization algorithm is combined with a non-dominated sorting genetic algorithm and a particle swarm optimization algorithm to solve the three objective functions. The optimal solution set for different types of skimmers with different configuration combinations of different numbers under different weights of maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact is obtained.

[0041] The importance of the three objective functions is compared pairwise using the analytic hierarchy process to form a judgment matrix, and the relative weights of the importance of the three objective functions are calculated. The fuzzy superiority and inferiority solution distance method is then used to find the optimal number configuration combination of skimmers under the determined relative weights from the optimal solution set of different number configuration combinations of different types of skimmers, which is the marine oil spill emergency recovery plan.

[0042] Furthermore: the risk factors of marine oil spill accidents include: oil spill parameters, marine environmental factors and oil spill response conditions;

[0043] The oil spill parameters include oil spill type, oil spill area, oil film thickness, oil viscosity and oil weathering degree;

[0044] Said marine environmental factors include wind speed, sea water temperature and interfacial tension;

[0045] The oil spill response conditions include skimmer parameters, emergency response team arrival time, recovery efficiency and response cost.

[0046] Furthermore, the Monte Carlo sample set is obtained after the combined preprocessing, and the three objective functions are converted into a robust form including expected value and variance, balancing the stability and volatility of the objective function. The process of obtaining the processed objective function is as follows:

[0047] N independent parameter combinations are randomly sampled from the defined probability distribution to generate a sample set containing wind speed, temperature, skimmer efficiency coefficient, and recovery cost. These samples are used as inputs for the optimization of the three objective functions. The three objective functions are robustified in the following way:

[0048]

[0049]

[0050]

[0051] in: Refers to the expected value of the recovery amount, which represents the average performance of the recovery amount under the parameter distribution. Refers to the standard deviation of the recovery amount, which indicates the volatility of the recovery amount. Refers to the expected value of the recovery cost, which represents the average performance of the recovery cost under the parameter distribution. Refers to the standard deviation of recovery costs, indicating the volatility of recovery costs. It refers to the expected value of environmental impact, which indicates the average performance of environmental impact under the parameter distribution. Refers to the standard deviation of environmental impact, indicating the volatility of environmental impact. and is a weight parameter used to balance the influence of expected value and variance, where As the decision variable, it represents the number of configuration combinations of skimmers.

[0052] Furthermore, the process of solving the processed objective function by combining the non-dominated sorting genetic algorithm with the particle swarm optimization algorithm to form a multi-objective genetic-particle swarm hybrid algorithm to obtain the optimal solution set for different types of skimmers with different number configuration combinations under different weights of maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact is as follows:

[0053] Initialize the position and velocity of the particle swarm;

[0054] For each particle, the speed and position are updated according to the particle group speed and position update formula;

[0055] At the beginning of each generation, the inertia weight is dynamically updated to balance the global search and local search capabilities;

[0056] The updated positions of particles in the particle swarm are repaired to ensure that the configuration combinations of different types of skimmers with different quantities meet the constraints of the three objective functions including maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact.

[0057] By simulating the binary crossover operator, the individuals in the updated population are crossovered to generate new offspring individuals;

[0058] The polynomial mutation operator is used to perform mutation operations on the offspring individuals after the crossover operation to increase the diversity of solutions;

[0059] Evaluate the new population after particle swarm update and genetic operation, and calculate the objective function value and constraint value of the new population;

[0060] Compare the objective function value of the new population with the individual optimal objective function value. If the new solution is better, update the individual optimal position and the corresponding objective function value.

[0061] Combine the feasible solutions in the new population with the solutions in the archive, recalculate the non-dominated sort and congestion, and update the solution set in the archive. The archive stores the current optimal non-dominated solution set.

[0062] After iterative optimization with the maximum number of iterations, the non-dominated solution set and the corresponding objective function in the archive are returned, and the optimal solution set of different types of skimmers with different numbers of configuration combinations under different weights for maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact is obtained.

[0063] Furthermore, the process of using the fuzzy superior-inferior solution distance method to find the optimal number configuration combination of oil skimmers under a certain weight from the optimal solution set of different number configuration combinations of different types of oil skimmers, that is, the marine oil spill emergency recovery plan, is as follows:

[0064] Perform triangular fuzzification on each element in the decision matrix of the optimal solution set;

[0065] Standardize the fuzzified decision matrix;

[0066] Weighting the standardized decision matrix to obtain a weighted decision matrix;

[0067] Determine the fuzzy positive ideal solution and the fuzzy negative ideal solution for the weighted decision matrix;

[0068] Calculate the Euclidean distance of each skimmer configuration combination scheme to the positive ideal solution and the negative ideal solution respectively;

[0069] Based on the Euclidean distance between the positive ideal solution and the negative ideal solution, the relative proximity of each skimmer quantity configuration combination scheme is calculated, and the skimmer quantity configuration combination schemes are ranked according to the relative proximity. The scheme with the largest relative proximity is the optimal skimmer quantity configuration combination.

[0070] The process of using the hierarchical analysis method to compare the importance of the three objective functions in pairs to form a judgment matrix and calculate the relative weights of the importance of the three objective functions is as follows:

[0071] The decision-making problem is decomposed into three levels: the objective level, the criterion level, and the solution level. The decision-making problem is: in an offshore oil spill emergency recovery scenario, based on the three conflicting objectives of maximizing oil recovery volume, minimizing recovery costs, and minimizing environmental impact, the optimal skimmer quantity configuration combination with the best overall performance is selected from the Pareto optimal solution set.

[0072] Compare each factor at the criterion level or solution level, determine the relative importance, and construct a judgment matrix;

[0073] Normalize the judgment matrix and calculate the weight vector of each factor;

[0074] The consistency index and random consistency ratio are calculated to determine whether the matrix is ​​consistent, and then the relative weights of the importance of the three objective functions are obtained.

[0075] The proposed method for optimizing marine oil spill emergency recovery can effectively support emergency response to marine oil spills under limited resources, helping decision-makers quickly develop efficient recovery plans and reduce environmental and economic losses, thereby achieving efficient and low-cost optimization of light crude oil recovery strategies.

[0076] Secondly, it achieves a balance between recycling efficiency and cost, optimizing resource allocation. By combining the non-dominated sorting of NSGA-II (non-dominated sorting genetic algorithm) and the global search capabilities of PSO (particle swarm optimization), it generates a high-quality Pareto optimal solution set, providing a more comprehensive and cost-effective solution for practical applications. The Pareto solution set generated by the hybrid algorithm is more evenly distributed and closer to the ideal solution than the separate NSGA-II and PSO algorithms.

[0077] In addition, by dynamically simulating the weathering process of oil, the complex situations in actual oil spill accidents can be reflected more accurately, providing more practical decision-making support for emergency response. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0079] Figure 1 is a flow chart of the method of the present invention;

[0080] Figure 2 A framework diagram for identifying risks of oil spills;

[0081] Figure 3 This is the algorithm flow chart of the NSGA-II-PSO hybrid algorithm (multi-objective genetic-particle swarm hybrid algorithm);

[0082] Figure 4 is the Pare front graph of the NSGA-II -PSO hybrid algorithm;

[0083] Figure 5 It is the algorithm performance evaluation index diagram;

[0084] Figure 6 This is the optimal solution graph obtained by the NSGA-II-PSO hybrid algorithm using fuzzy TOPSIS (top-to-bottom solution distance method) on the Pareto front;

[0085] Figure 7 is the convergence curve of the NSGA-II-PSO hybrid algorithm on the objective of maximizing oil recovery;

[0086] Figure 8 is the convergence curve of the NSGA-II-PSO hybrid algorithm on the objective of minimizing the recovery cost;

[0087] Figure 9 is the convergence curve of the NSGA-II-PSO hybrid algorithm on the objective of minimizing environmental impact;

[0088] Figure 10 This is a heat map of the effect of wind speed on oil recovery;

[0089] Figure 11 This is a thermal diagram of the effect of temperature on oil recovery;

[0090] Figure 12 Sobol sensitivity analysis diagram for the combined effects of wind speed and wind speed oil recovery. DETAILED DESCRIPTION

[0091] It should be noted that, unless there is any conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0092] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. The following description of at least one exemplary embodiment is actually only illustrative and is in no way intended to limit the present invention and its application or use. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0093] Figure 1 is a flow chart of the method of the present invention;

[0094] A method for optimizing marine oil spill emergency recovery comprises the following steps:

[0095] S1: Obtain risk factors for marine oil spill accidents;

[0096] S2: Preprocessing the risk parameters of marine oil spill accidents by Monte Carlo simulation method to obtain Monte Carlo sample set;

[0097] S3: Based on the pre-processed marine oil spill risk factors, a three-objective function is constructed, including maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact;

[0098] S4: Construct the constraints of the three-objective function considering the petroleum weathering process, including: the equality constraint of volume decay and the resource constraint of recovery equipment;

[0099] S5: Combining the Monte Carlo sample set obtained after preprocessing, the three objective functions are transformed into a robust form including expected value and variance, balancing the stability and volatility of the objective function to obtain the processed objective function;

[0100] S6: A multi-objective NSGA-II-PSO hybrid algorithm is constructed by combining the NSGA-II algorithm with the PSO algorithm to solve the three-objective function, obtaining the optimal solution set for different types of skimmer configuration combinations with different weights for maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact. This algorithm combines the advantages of NSGA-II's global Pareto solution generation with PSO's dynamic inertia weight local optimization to address the imbalance between convergence and diversity of a single algorithm.

[0101] S7: The importance of the three objective functions is compared pairwise using the analytic hierarchy process to form a judgment matrix, and the relative weights of the importance of the three objective functions are calculated. The fuzzy TOPSIS (top-to-bottom solution distance method) method is used to find the optimal marine oil spill emergency recovery plan with a certain relative weight from the optimal solution set of different types of skimmers with different number configuration combinations, replacing the traditional linear weighting method.

[0102] The method also includes comparing the hybrid algorithm with NSGA-II and MOPSO algorithms and analyzing the sensitivity of parameters to output the optimal combination of skimmer quantity configuration;

[0103] In this case, crude oil recovery operations become an important means to mitigate the impact of oil spills and restore marine ecology. Figure 2 A systematic oil spill risk identification framework is presented, which categorizes influencing factors into three main categories: oil spill parameters, marine environment, and oil spill response conditions. These factors together determine the spread pattern, impact range, and effectiveness of response measures.

[0104] The risk factors of marine oil spill accidents include: oil spill parameters, marine environmental factors and oil spill response conditions;

[0105] Oil spill parameters include spill type, spill area, oil film thickness, oil viscosity, and degree of oil weathering. Different types of crude oil have varying physical and chemical properties, such as density, viscosity, and volatility, which directly influence the spread and weathering of the spill. The spill area and oil film thickness determine the reach and recovery difficulty of the spill, while the oil's viscosity and degree of weathering affect the recovery efficiency of skimmers. These factors profoundly influence the oil's behavior in the marine environment and its potential harm to the ecosystem.

[0106] Marine environmental factors, including wind speed, seawater temperature, and interfacial tension, significantly influence the spread, migration, and weathering of oil spills. Wind speed can affect the spread of oil spills and increase the area covered by the spill. Seawater temperature and interfacial tension, on the other hand, influence the evaporation and dissolution of oil. High water temperatures accelerate oil volatilization, reduce its viscosity, and affect recovery efficiency.

[0107] Oil spill response criteria include skimmer specifications, emergency response team arrival time, recovery efficiency, and response costs. Different types of skimmers are suitable for different spill conditions and environments. Choosing the right skimmer for oil recovery can improve the efficiency and effectiveness of oil spill response. These typically include disc skimmers, weir skimmers, and belt skimmers. Arrival time refers to the time it takes for the emergency response team to arrive at the incident site. Prompt arrival can minimize the spread of the spill and improve recovery efficiency. Recovery efficiency is affected by various factors, such as oil film thickness, sea conditions, and equipment performance. Response costs include equipment costs, transportation costs, and labor costs, which directly impact the economic viability of the emergency response.

[0108] The uncertainty risk factors of marine oil spill accidents are preprocessed, and the parameter uncertainty is quantified by Monte Carlo simulation.

[0109] First, we select the parameters that have a significant impact on oil spill recovery, including wind speed ( ),temperature( ), skimmer efficiency ( ) and recovery costs ( ). Based on historical data or actual observations, define the probability distribution of each parameter. For example, wind speed ( ) obeys a normal distribution with a mean of 10 m / s and a standard deviation of 2 m / s. Temperature ( ) obeys a uniform distribution and ranges from 275.15K to 281.15 K. A large number of samples are randomly drawn from the defined probability distribution to generate multiple sets of parameter combinations, and the random sample sets are used as the output of data preprocessing.

[0110] These parameter combinations will be used in the subsequent three-objective function optimization solution process as input data for optimization calculations of NSGA-II-PSO hybrid algorithms.

[0111] The three objective functions of maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact include an objective function of maximizing oil recovery and an objective function of minimizing recovery costs;

[0112] The objective function for maximizing oil recovery is expressed as follows:

[0113]

[0114] in: is the volume of recovered oil, For operation time, The upper limit of the operating time for oil recovery using skimmers is for The number of oil skimmers, For the Oil skimmers in arrive The amount of oil recovered at each moment;

[0115] The objective function of minimizing the recovery cost is expressed as follows:

[0116]

[0117] in: is the total cost, It is a type of oil skimmer. is the upper limit of the types of skimmers, It is The operating cost of a skimmer is related to the volume of oil spilled. It is The transportation cost of dispatching a skimmer from the warehouse to the spill site.

[0118] The objective function expression of minimizing environmental impact is as follows:

[0119]

[0120] in, is the base rate ($1.00), is the volume (m 3 ), is the geographical location (8 = near the coast; 5 = nearshore waters; 1 = offshore), Indicates a special regulatory area (2 = yes; 1 = no), is the pollutant characteristic (8 = heavy oil; 4 = medium oil; 1 = light oil); EI stands for environmental impact;

[0121] The expression of the nonlinear equation for oil recovery by the skimmer is as follows:

[0122]

[0123] in: For the Oil skimmers in arrive The amount of oil recovered in each stage, , is the empirical coefficient of the oil skimmer, is the oil film thickness;

[0124]

[0125]

[0126] in: for The oil film thickness of the stage, for The amount of oil remaining after oil recovery and natural weathering; is the initial volume of the oil spill, A is the oil spill area, for arrive The amount of oil lost during this period through oil recovery and natural weathering processes, for to The volume of oil lost during the evaporation and weathering process during this period, for to The volume of oil lost during the oil dispersion process during the period;

[0127] The equations constraining volume loss due to petroleum weathering include the equations for evaporation and dispersion during weathering;

[0128] The volume of oil lost during oil evaporation and weathering The evaporation equation is as follows:

[0129]

[0130]

[0131] in: and is the evaporation equation parameter of a specific oil, for Evaporation rate of the stage (%), is the temperature (K), is time (h), Express Taking the natural logarithm, For the to Volume of evaporated oil in each stage (m 3 ), for The amount of oil remaining after oil recovery and natural weathering (m 3 ).

[0132] The emulsification phenomenon is simulated using the embedding rate of water in the oil film:

[0133]

[0134] in, is the moisture content, is the curve fitting constant that varies with wind speed (2×10 -6 ), is the molar viscosity constant (0.7, applicable to crude oil and heavy fuel oil), represents the exponential function, is the wind speed (m / s), is time (h);

[0135] The volume of oil lost during oil dispersion The equation is as follows:

[0136]

[0137]

[0138]

[0139]

[0140] in: for The water content of the stage is used to simulate the oil emulsification phenomenon. is the curve fitting constant that varies with wind speed, is the molar viscosity constant, represents the exponential function, is the wind speed, For time; is the dispersion rate, is the wind speed, is the dynamic viscosity of the oil, for The oil film thickness of the stage, is the oil-water interfacial tension, for The dynamic viscosity of the stage oil, for The dynamic viscosity of the stage oil, for The evaporation rate of the stage, for The moisture content of the stage, is the molar viscosity constant, represents the exponential function, For the to Volume of dispersed oil in the stage, for The dispersion rate of the stage, for The amount of oil remaining after oil recovery and natural weathering;

[0141] The recycling equipment resource constraints are as follows:

[0142]

[0143]

[0144] in: For the The number of oil skimmers, The maximum limit for the number of skimmers is Represents an integer.

[0145] The Monte Carlo sample set is obtained after the combined preprocessing, and the three objective functions are converted into a robust form containing expected value and variance, balancing the stability and volatility of the objective function. The process of obtaining the processed objective function is as follows:

[0146] From the sample set obtained by Monte Carlo simulation, N sets of independent parameter combinations are randomly selected from the defined probability distribution to generate a sample set containing wind speed, temperature, skimmer efficiency coefficient, and recovery cost as the input for the three-objective function optimization. The three-objective function is robustified by the following method:

[0147]

[0148]

[0149]

[0150] in: Refers to the expected value of the recovery amount, which represents the average performance of the recovery amount under the parameter distribution. Refers to the standard deviation of the recovery amount, which indicates the volatility of the recovery amount. Refers to the expected value of the recovery cost, which represents the average performance of the recovery cost under the parameter distribution. Refers to the standard deviation of recovery costs, indicating the volatility of recovery costs. It refers to the expected value of environmental impact, which indicates the average performance of environmental impact under the parameter distribution. Refers to the standard deviation of environmental impact, indicating the volatility of environmental impact. and is a weight parameter used to balance the influence of expected value and variance.

[0151] A multi-objective genetic-particle swarm optimization algorithm is combined with a non-dominated sorting genetic algorithm and a particle swarm optimization algorithm to solve the three-objective function. The optimal solution set for different types of skimmers with different weights of maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact is obtained as follows:

[0152] Initialize the position and velocity of the particle swarm;

[0153] For each particle, the speed and position are updated according to the particle group speed and position update formula;

[0154] At the beginning of each generation, the inertia weight is dynamically updated to balance the global search and local search capabilities;

[0155] The updated positions of particles in the particle swarm are repaired to ensure that the configuration combinations of different types of skimmers with different quantities meet the constraints of the three objective functions including maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact.

[0156] By simulating the binary crossover operator, the individuals in the updated population are crossovered to generate new offspring individuals;

[0157] The polynomial mutation operator is used to perform mutation operations on the offspring individuals after the crossover operation to increase the diversity of solutions;

[0158] Evaluate the new population after particle swarm update and genetic operation, and calculate the objective function value and constraint value of the new population;

[0159] Compare the objective function value of the new population with the individual optimal objective function value. If the new solution is better, update the individual optimal position and the corresponding objective function value.

[0160] Combine the feasible solutions in the new population with the solutions in the archive, recalculate the non-dominated sort and congestion, and update the solution set in the archive. The archive stores the current optimal non-dominated solution set.

[0161] After iterative optimization for the maximum number of iterations, the non-dominated solution set and the corresponding objective function in the archive are returned, and the optimal solution set for different types of skimmers with different number configuration combinations under different weights of maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact is obtained.

[0162] The process of using the fuzzy TOPSIS method to find the optimal marine oil spill emergency recovery plan under a certain weight from the optimal solution set is as follows:

[0163] Perform triangular fuzzification on each element in the decision matrix of the optimal solution set;

[0164] Standardize the fuzzified decision matrix;

[0165] Weighting the standardized decision matrix to obtain a weighted decision matrix;

[0166] Determine the fuzzy positive ideal solution and the fuzzy negative ideal solution for the weighted decision matrix;

[0167] Calculate the Euclidean distance of each skimmer configuration combination scheme to the positive ideal solution and the negative ideal solution respectively;

[0168] The relative closeness of the number configuration combination scheme of each skimmer is calculated based on the Euclidean distance between the positive ideal solution and the negative ideal solution. Sort all the solutions: the number of oil skimmer configuration combinations, The largest solution is the optimal skimmer quantity configuration combination.

[0169] The process of using the hierarchical analysis method to compare the importance of the three objective functions in pairs to form a judgment matrix and calculate the relative weights of the importance of the three objective functions is as follows:

[0170] The decision-making problem is decomposed into three levels: the objective level, the criterion level, and the solution level. The decision-making problem is: in an offshore oil spill emergency recovery scenario, based on the three conflicting objectives of maximizing oil recovery volume, minimizing recovery costs, and minimizing environmental impact, the optimal skimmer quantity configuration combination with the best overall performance is selected from the Pareto optimal solution set.

[0171] Compare each factor at the criterion level or solution level, determine the relative importance, and construct a judgment matrix;

[0172] Normalize the judgment matrix and calculate the weight vector of each factor;

[0173] The consistency index and random consistency ratio are calculated to determine whether the matrix is ​​consistent, and then the relative weights of the importance of the three objective functions are obtained.

[0174] Example 1: Assume that a light crude oil spill occurs in a certain sea area.

[0175] S1: Obtaining Risk Factors for Marine Oil Spills. The initial spill volume is 4,000 cubic meters, and the oil slick is 40 mm thick. During the "golden 24-hour" window for emergency response, the oil slick reaches a quasi-steady state due to the combined effects of wind shear and turbulent diffusion, with the area expansion rate dropping to less than 5% of the initial level. To improve computational efficiency, the spill area is assumed to remain relatively stable. To address this incident, the emergency response team plans to use three different types of skimmers for oil recovery. Assuming that due to vehicle weight restrictions, ten skimmers of each type are required, the maximum capacity of the oil recovery vessel is 20 skimmers. Due to the suddenness of the incident and limited transportation, additional skimmers could not be supplied within 48 hours. Therefore, the key objective at this stage is to determine the optimal combination of the three skimmers within 24 hours to maximize oil recovery and minimize recovery costs. Table 1 lists the empirical coefficients and transportation costs of the three skimmers.

[0176] Table 1 Oil skimmer coefficients

[0177]

[0178] The weathering parameters of Arabian light crude oil involved in the case are shown in Table 2:

[0179] Table 2 Weathering parameters of Arabian light crude oil

[0180]

[0181] S2: Monte Carlo simulation is used to randomly sample parameters such as wind speed and temperature to quantify the impact of parameter uncertainty on the recycling strategy.

[0182] Random sampling of parameters involved in the weathering process is shown in the following table:

[0183] Table 3 Parameter probability distribution definition

[0184]

[0185] S3: Based on the pre-processed marine oil spill risk factors, a three-objective function is constructed, which includes maximizing oil recovery, minimizing recovery costs, and minimizing environmental impacts. It is expressed as:

[0186]

[0187]

[0188]

[0189] Step 4: Construct the constraints of the three-objective function considering the petroleum weathering process, including: the equality constraint of volume decay and the resource constraint of recovery equipment;

[0190] The equality constraint of volume decay and the resource constraint of recycling equipment are:

[0191]

[0192]

[0193]

[0194]

[0195]

[0196]

[0197]

[0198]

[0199]

[0200]

[0201]

[0202]

[0203] in: To maximize oil recovery; To minimize recycling costs; To minimize environmental impact; is the decision variable of the objective function, i.e., the number of different types of skimmers, For the type of oil skimmer, For the Oil skimmers in Oil recovery rate at each stage; For the The operating cost of a skimmer, For the Shipping costs for oil skimmers; is the initial volume of the oil spill (m 3 ); for The amount of oil remaining after oil recovery and natural weathering (m 3 ); 、 、 They are arrive h Volume of oil recovered, evaporated and dispersed during the period (m3 ); for Oil film thickness at each stage (mm), is the oil spill area (m 2 ); for Evaporation rate of the stage (%), and is the parameter of the oil evaporation equation, is the temperature (K), is time (h), for Moisture content of the stage (%), for Oil viscosity (cP) at each stage, for Dispersion rate of the stage (%), is the wind speed (m / s), is the interfacial tension (dyne / m).

[0204] Step 5: Combine the Monte Carlo sample set obtained after preprocessing and transform the three objective functions into a robust form containing expected value and variance, balance the stability and volatility of the objective function, and obtain the processed objective function;

[0205] The random samples generated by Monte Carlo simulation in data preprocessing are coupled with three-objective optimization to construct a robust decision framework. Monte Carlo simulation is used to generate random samples of parameters, which are then used to construct the robust decision framework. The sample generation step can be considered data preprocessing. Robust decision optimization is an optimization method that modifies the three-objective function based on the parameter combinations generated by Monte Carlo simulation to find the optimal solution under uncertainty. This enhances the robustness of the objective function, ensuring a stable output solution set under different conditions. This step is part of the optimization process. The robust decision framework is the objective function below, which accounts for the randomness and volatility of the parameters.

[0206] Each parameter in the Monte Carlo simulation has its own probability distribution, and the joint distribution of these parameters describes the probability of their simultaneous occurrence. From this joint distribution, N groups of independent samples are randomly drawn. Each group of samples (s = 1, 2, ..., N) represents a set of possible parameter combinations, such as It may be a set of parameters with a wind speed of 10 m / s and a temperature of 278 K. Another set of parameter combinations may be wind speed 12 m / s and temperature 280 K. Substituting different combinations into the following robust multi-objective problem can ensure that the optimization results remain stable under various uncertainties. The sensitivity analysis of wind speed and temperature was carried out.

[0207] For each set of parameters , making the multi-objective problem robust:

[0208]

[0209]

[0210]

[0211] in: Refers to the expected value of the recovery amount, which represents the average performance of the recovery amount under the parameter distribution. Refers to the standard deviation of the recovery amount, which indicates the volatility of the recovery amount. Refers to the expected value of the recovery cost, which represents the average performance of the recovery cost under the parameter distribution. Refers to the standard deviation of recovery costs, indicating the volatility of recovery costs. It refers to the expected value of environmental impact, which indicates the average performance of environmental impact under the parameter distribution. Refers to the standard deviation of environmental impact, indicating the volatility of environmental impact. and is a weight parameter used to balance the influence of expected value and variance.

[0212] Each set of parameters in the Monte Carlo sample set (s=1,2,…,N) and the position vector of the particle swarm Combination, calculate each group ( , ) corresponding to the recycling amount ( , ) and cost ( , );

[0213] For each particle , based on N groups ( ( , ), ( , ), ( , ))Calculate the expected value and standard deviation to generate the robust three-objective value ( , );

[0214] The Sobol method is used to quantify the contribution of parameters to the target and perform sensitivity analysis:

[0215]

[0216] in: Representation parameters The first-order sensitivity index of . ∈[0,1], the larger the value, the more significant the impact of the parameter on the objective function. Indicates that the parameter When the value of the objective function RV is determined, Indicates that the parameter When the value of RV is determined, the variance of the average value of the objective function RV is calculated. represents the variance of the recovered amount;

[0217] S6: A multi-objective genetic-particle swarm optimization algorithm is combined with a non-dominated sorting genetic algorithm and a particle swarm optimization algorithm to solve the processed objective function, and the optimal solution set for different types of skimmers with different number configuration combinations under different weights of maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact is obtained;

[0218] Combining the advantages of NSGA-II's global Pareto solution generation and PSO's dynamic inertia weight local optimization, it solves the imbalance between the convergence and diversity of a single algorithm. Figure 3 This is the algorithm flow chart of the NSGA-II-PSO hybrid algorithm (multi-objective genetic-particle swarm hybrid algorithm). The specific steps of the combined hybrid algorithm are as follows:

[0219] S61: Initialize the position of the particle swarm and speed ;

[0220]

[0221]

[0222] in: Indicates the The position vector of a particle, Respectively represent Particles in Position component in 2 dimensions; Indicates the The velocity vector of a particle, Respectively represent Particles in The velocity components in the dimensions, is the total number of particles in the swarm.

[0223] S62: Iterative Optimization:

[0224] For each particle, the speed and position update formula of the PSO particle swarm algorithm is updated as follows:

[0225]

[0226]

[0227] in: It is a particle exist The first iteration The speed of dimension, is the current iteration number, It is a particle exist The first iteration The speed of dimension, is the inertia parameter, is the individual learning coefficient, is the global learning coefficient, and It is a random number uniformly distributed in the range of [0,1], which is used to introduce randomness and help the algorithm escape from the local optimum. is the individual extreme value, is the global extreme value, It is a particle exist The first iteration Dimensional location. is the current position of the particle, For particles Updated location, For particles Updated speed.

[0228] At the beginning of each generation, the inertia weight is dynamically updated , to balance global search and local search capabilities.

[0229]

[0230] in: It is The inertia weight of the iteration, The maximum value of the inertia weight, is the minimum value of the inertia weight, is the current iteration number, is the maximum number of iterations.

[0231] Solution repair: Repair the updated position to ensure that the solution meets the constraints.

[0232] Genetic Operations:

[0233] Crossover operation: The individuals in the population are crossover-operated by simulating the binary SBX crossover operator to generate new offspring individuals.

[0234] Mutation operation: The polynomial mutation operator is used to perform mutation operations on the offspring individuals after crossover to increase the diversity of solutions.

[0235] New population evaluation: Evaluate the new population after particle swarm update and genetic operation, and calculate its objective function value and constraint value.

[0236] Individual optimal update: Compare the objective function value of the new population with the individual optimal objective function value. If the new solution is better, update the individual optimal position and the corresponding objective function value.

[0237] Archive update: merge the feasible solutions in the new population with the solutions in the archive, recalculate the non-dominated sort and congestion, and update the solution set in the archive. The archive stores the current optimal non-dominated solution set.

[0238] Return results

[0239] After the maximum number of iterations of iterative optimization, the set of non-dominated solutions and the corresponding objective function in the archive are returned. These solutions represent the Pareto optimal trade-off between the amount of recovery and the cost.

[0240] Figure 4 is the Pare front graph of the NSGA-II -PSO hybrid algorithm;

[0241] Figure 5 This paper compares the NSGA-II-PSO hybrid algorithm with other multi-objective intelligent optimization algorithms, and quantifies the differences in convergence and distribution of each algorithm using Hypervolume and Spacing values. The Hypervolume metric refers to the volume of the target space enclosed by the non-dominated solution set obtained by the algorithm and the reference point. The larger the HV value, the better the overall performance of the algorithm. Its calculation formula is:

[0242]

[0243] in Represents a volume calculation function, which is used to calculate the volume of a certain area in the target space. For all solutions v i Take the union of the regions defined in the target space, represents the number of solutions in the solution set S.

[0244] The smaller the Spacing value, the more evenly the solution set is distributed. The calculation formula is:

[0245]

[0246] Where N represents the number of all solutions, d i represents the distance from solution i to its nearest neighbor solution, It represents the average distance of all solutions to their nearest neighbor solutions.

[0247] In summary, NSGA-II-PSO demonstrates stronger comprehensive optimization performance in the Hypervolume indicator by integrating the non-dominated sorting mechanism of NSGA-II with the global search capability of PSO. At the same time, it maintains a high distribution quality in the Spacing indicator, effectively balancing convergence, diversity, and uniformity, and providing a higher quality and more comprehensive Pareto solution set for offshore oil spill recovery decision-making, verifying its superiority in complex multi-objective optimization problems.

[0248] S7: Use the hierarchical analysis method to compare the importance of the three objective functions in pairs to form a judgment matrix, calculate the relative weights of the importance of the three objective functions, and use the fuzzy superiority and inferiority solution distance method to find the optimal number configuration combination of skimmers under the determined relative weights from the optimal solution set of different number configuration combinations of different types of skimmers, which is the marine oil spill emergency recovery plan.

[0249] Triangular fuzzy weights and proximity calculation are used to quickly screen the optimal solution from the Pareto solution set, replacing the traditional linear weighted method.

[0250] S71: Fuzzify each element in the decision matrix of the optimal solution set;

[0251]

[0252] Among them, each element is a triangular fuzzy number, indicating the The plan in The fuzzy evaluation value of the indicator,

[0253] represents the fuzzy evaluation value of the first option on the first indicator,

[0254] represents the fuzzy evaluation value of the first option on the mth indicator,

[0255] represents the fuzzy evaluation value of the nth solution on the first indicator,

[0256] It represents the fuzzy evaluation value of the nth scheme on the mth indicator;

[0257] is the minimum possible value of the fuzzy evaluation value, is the most likely value of the fuzzy evaluation value, is the maximum possible value of the fuzzy evaluation value;

[0258] S72: Standardizing the fuzzified decision matrix;

[0259] For the benefit-type index (the larger the better), in the invention, it refers to the oil recovery amount:

[0260]

[0261] Cost-type indicators (the smaller the value, the better), in the invention, refer to the recovery cost:

[0262]

[0263] in, It is the fuzzy evaluation value after standardization; Indicates the The maximum value of the index, Indicates the The minimum value of the indicator, is the minimum possible value of the fuzzy evaluation value, is the most likely value of the fuzzy evaluation value, is the maximum possible value of the fuzzy evaluation value;

[0264] S73: Weighting the standardized decision matrix to obtain a weighted decision matrix ;

[0265]

[0266] in, is the weighted fuzzy evaluation value; Indicates the weight assigned to each indicator, satisfying , Indicates the number of rows in the decision matrix, representing the number of decision options; It represents the number of columns in the decision matrix and the number of evaluation indicators.

[0267] S74: Determine a fuzzy positive ideal solution (FPIS) and a fuzzy negative ideal solution (FNIS) for the weighted decision matrix;

[0268] Fuzzy Positive Ideal Solution (FPIS):

[0269]

[0270] For benefit attributes:

[0271]

[0272] For cost attributes:

[0273]

[0274] Fuzzy Negative Ideal Solution (FNIS):

[0275]

[0276] For benefit attributes:

[0277]

[0278] For cost attributes:

[0279]

[0280] in, represents the ideal optimal solution vector obtained by the weighted decision matrix, It is its weight; represents the ideal worst solution vector obtained by the weighted decision matrix, It's its weight, is the number of components of the solution vector. Represents fuzzy positive ideal solution No. A quantity, Represents fuzzy positive ideal solution No. A portion.

[0281] S75: Calculate the Euclidean distances of different configuration combinations of different numbers of different types of skimmers to the positive ideal solution and the negative ideal solution respectively:

[0282]

[0283]

[0284] in, Indicates the The Euclidean distance from the solution to the fuzzy positive ideal solution, Indicates the The Euclidean distance from each solution to the fuzzy negative ideal solution; is the weighted fuzzy evaluation value, Represents fuzzy positive ideal solution No. A quantity, Represents fuzzy positive ideal solution No. A portion.

[0285] S76: Calculate the relative closeness of different solutions based on the Euclidean distance between the positive ideal solution and the negative ideal solution of each solution :

[0286]

[0287] in Indicates the The relative closeness of the solutions, , the larger the value, the closer the solution is to the ideal solution.

[0288] S77: Calculate the relative closeness of each solution , according to the relative proximity Sort all configurations of different types of skimmers with different quantities. The largest solution is the optimal solution.

[0289] The process of using the hierarchical analysis method to compare the importance of the three objective functions in pairs to form a judgment matrix and calculate the relative weights of the importance of the three objective functions is as follows:

[0290] Step 1: Build a hierarchical model

[0291] Decompose the decision-making problem into three levels: goal level, criterion level and solution level.

[0292] Target layer: Select the optimal solution on the Pareto front.

[0293] Criteria layer: three goals (maximizing oil recovery, minimizing recovery costs, and minimizing environmental pollution).

[0294] Solution layer: candidate solutions on the Pareto front.

[0295] Step 2: Construct a judgment matrix;

[0296] By comparing the factors of the criterion layer or the solution layer in pairs, we can determine their relative importance to the target layer (or the previous layer). The comparison results are expressed using the 1-9 scale proposed by Satie, as shown in the following table. Based on the values ​​of the scale table, the judgment matrix A=(a ij ) 3*3 .

[0297] Table 4 Judgment matrix scale values ​​and meanings

[0298]

[0299] Step 3: Calculate the weight vector

[0300] By normalizing the judgment matrix, the weight vector of each factor is calculated. In the judgment matrix A, its element is a ij , the normalized matrix R The calculation formula is:

[0301]

[0302] Where rij is the element in the i-th row and j-th column of the normalized matrix R, is the sum of the jth column of the judgment matrix A, and n is the order of the judgment matrix. After normalization, calculate the weight vector W, which is the average value of each row of the normalized matrix R. The calculation formula is:

[0303]

[0304] And verify, the weight vector W The sum of all elements of is 1:

[0305]

[0306] Step 4: Consistency Check

[0307] By calculating the consistency index (CI) and the random consistency ratio (CR), we can check whether the judgment matrix is ​​consistent and ensure that there is no logical contradiction. The formula of the consistency index CI is:

[0308]

[0309] is the maximum eigenvalue of the judgment matrix A;

[0310] The formula for the consistency ratio is:

[0311]

[0312] Among them, RI is the average random consistency index, corresponding to n=1 to 9, and the RI values ​​are shown in the following table:

[0313] Table 5 Average random consistency index (RI) table

[0314]

[0315] When CR < 0.1, the consistency of the matrix is ​​considered acceptable.

[0316] S8: Dynamic Strategy Output and Verification

[0317] Output the optimal equipment combination, compare the hybrid algorithm with NSGA-II and MOPSO algorithms, and perform parameter sensitivity analysis.

[0318] Figure 6 The optimal solution obtained by the NSGA-II-PSO hybrid algorithm using fuzzy TOPSIS (top-to-bottom solution distance method) on the Pareto front is shown in the figure. In marine oil spills, rapid oil recovery is a core environmental protection goal. The spread of oil directly threatens marine ecosystems, fishery resources, and coastal communities, and its environmental and social costs far exceed the economic investment in emergency response operations. Therefore, maximizing oil recovery (G1) is considered to be strongly more important than minimizing recovery costs (G2), slightly more important than minimizing environmental impacts (G3), and significantly more important than minimizing recovery costs (G2). A judgment matrix is ​​constructed, and the weights of the three objective functions are calculated using the AHP method:

[0319]

[0320] The maximum eigenvalue Substituting the RI value into the consistency test formula yields CR = 0.026 < 1, indicating satisfactory matrix consistency. Therefore, the weights of the three objective functions are set to maximize oil recovery (0.643), total cost (0.074), and minimize environmental impact (0.283), and then substituted into the fuzzy TOPSIS method. This reflects the emergency management principle of "pollution control takes precedence over short-term costs," resulting in the optimal solution tending to choose a solution with a higher recovery rate and lower environmental impact, even if the cost is relatively high. The optimal solution is as follows: Figure 6 As shown, the distribution of skimmers at this time is SK1=0, SK2=6, SK3=9. The oil recovery volume at this time is 2777.1m3, the total cost is 28632421.2 yuan, and the environmental impact is 47558862.9 yuan.

[0321] Figure 7 is the convergence curve of the NSGA-II-PSO hybrid algorithm on the objective of maximizing oil recovery;

[0322] Figure 8 is the convergence curve of the NSGA-II-PSO hybrid algorithm on the objective of minimizing the recovery cost;

[0323] Figure 9 is the convergence curve of the NSGA-II-PSO hybrid algorithm on the objective of minimizing environmental impact;

[0324] The results demonstrate that the NSGA-II-PSO algorithm effectively balances recovery efficiency and cost during the optimization process. During the first 20 iterations, the objective function experienced significant fluctuations, indicating that the algorithm was able to quickly find optimal solutions early on. This is due to the PSO algorithm's global search capabilities, which enabled rapid exploration of the solution space in this early stage. After 20 iterations, the decline slowed, indicating that the algorithm entered the local search phase. The NSGA-II algorithm's local refinement capabilities played a significant role in this phase, further optimizing the distribution and quality of the solution set through non-dominated sorting and congestion calculation.

[0325] Figure 10 It is a three-dimensional surface diagram of the effect of wind speed on oil recovery;

[0326] Figure 11 The three-dimensional surface diagram of the effect of temperature on oil recovery;

[0327] Figure 12 is the sobol index for the effects of wind speed and temperature on oil recovery;

[0328] The results show that increasing wind speed and temperature accelerate oil weathering, increasing the amount of oil evaporated and dispersed, leading to a decrease in recovered oil. Rising temperature accelerates oil evaporation but has a relatively small impact on recovered oil. When the temperature rises from 274 K to 281 K, recovered oil decreases by approximately 11.8%. When the wind speed increases from 5 m / s to 15 m / s, the volume of oil lost through dispersion increases, resulting in a decrease in recovered oil by approximately 5.4%. In the first-order Sobol index, wind speed (U) is 0.2334, and temperature (T) is 0.7711, indicating that temperature has a greater impact on recovered oil, consistent with the trend shown in the heat map.

[0329] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A marine oil spill emergency recovery optimization method, characterized by: The following steps are involved: Step 1: Obtain risk factors for marine oil spill accidents; Step 2: Preprocess the marine oil spill risk parameters using the Monte Carlo simulation method to obtain a Monte Carlo sample set; Step 3: Based on the pre-processed marine oil spill risk factors, a three-objective function is constructed, which includes maximizing oil recovery, minimizing recovery costs, and minimizing environmental impacts. The objective function for maximizing oil recovery is expressed as follows: in: is the volume of recovered oil, is the time period number, The upper limit of the time period, for The number of oil skimmers, For the Oil skimmers in The amount of oil recovered during the period; The objective function to minimize the recovery cost is expressed as follows: in: is the total cost, It is a type of oil skimmer. is the upper limit of the types of skimmers, It is The operating cost of a skimmer is related to the volume of oil spilled. It is The transportation cost of the skimmers from the warehouse to the spill site; The objective function expression of minimizing environmental impact is as follows: in, is the base rate, is the volume, For geographical location, Indicates a special management area. The pollutant characteristics, : indicates environmental impact; Step 4: Construct the constraints of the three-objective function considering the petroleum weathering process, including: the equality constraint of volume decay and the resource constraint of recovery equipment; The equality constraint for the volume decay is: in: for The oil film thickness during the period, for The amount of oil remaining after oil recovery and natural weathering during the period; is the initial volume of the oil spill, A is the oil spill area, for The amount of oil lost during this period through oil recovery and natural weathering processes, for The volume of oil lost during the evaporation and weathering process during this period, for The volume of oil lost during the oil dispersion process during the period; Of which: Oil recovery The expression is as follows: in: For the Oil skimmers in The amount of oil recovered during the period, , is the empirical coefficient of the oil skimmer, is the oil film thickness; The volume of oil lost during evaporation and weathering The equation is as follows: in: and is the evaporation equation parameter of a specific oil, for The evaporation rate during the period, is the temperature, For time, Express Taking the natural logarithm, For the The volume of evaporated oil during the period, for The amount of oil remaining after oil recovery and natural weathering during the period; The volume of oil lost during oil dispersion The equation is as follows: in: for The water content of the time period is used to simulate the oil emulsification phenomenon. is the curve fitting constant that varies with wind speed, is the molar viscosity constant, represents the exponential function, is the wind speed, For time; is the dispersion rate, is the wind speed, is the dynamic viscosity of the oil, for The oil film thickness during the period, is the oil-water interfacial tension, for The dynamic viscosity of the oil during the period, for The dynamic viscosity of the oil during the period, for The evaporation rate during the period, for The moisture content of the period, is the molar viscosity constant, represents the exponential function, for The volume of oil dispersed during the period, for The dispersion rate of the time period, for The amount of oil remaining after oil recovery and natural weathering during the period; The recycling equipment resource constraints are as follows: in: For the The number of oil skimmers, The maximum limit for the number of skimmers is Represents an integer; Step 5: Combine the Monte Carlo sample set obtained after preprocessing and transform the three objective functions into a robust form containing expected value and variance, balance the stability and volatility of the objective function, and obtain the processed objective function; Step 6: A multi-objective genetic-particle swarm optimization algorithm is constructed by combining a non-dominated sorting genetic algorithm with a particle swarm optimization algorithm to solve the processed objective function, and the optimal solution set for different types of skimmers with different configuration combinations under different weights of maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact is obtained; Step 7: Use the hierarchical analysis method to compare the importance of the three objective functions in pairs to form a judgment matrix, calculate the relative weights of the importance of the three objective functions, and use the fuzzy superiority and inferiority solution distance method to find the optimal number configuration combination of skimmers under the determined relative weights from the optimal solution set of different types of skimmers with different number configuration combinations, which is the marine oil spill emergency recovery plan.

2. The method for optimizing marine oil spill emergency recovery according to claim 1, wherein: The risk factors of marine oil spill accidents include: oil spill parameters, marine environmental factors and oil spill response conditions; The oil spill parameters include oil spill type, oil spill area, oil film thickness, oil viscosity and oil weathering degree; Said marine environmental factors include wind speed, sea water temperature and interfacial tension; The oil spill response conditions include skimmer parameters, emergency response team arrival time, recovery efficiency and response cost.

3. The method for optimizing marine oil spill emergency recovery according to claim 1, wherein: The Monte Carlo sample set is obtained after the combined preprocessing, and the three objective functions are converted into a robust form containing expected value and variance, balancing the stability and volatility of the objective function. The process of obtaining the processed objective function is as follows: N independent parameter combinations are randomly sampled from the defined probability distribution to generate a sample set containing wind speed, temperature, skimmer efficiency coefficient, and recovery cost. These samples are used as inputs for the optimization of the three-objective function, which is then robustified in the following way: in: Refers to the expected value of the recovery amount, which indicates the average performance of the recovery amount under the parameter distribution. Refers to the standard deviation of the recovery amount, which indicates the volatility of the recovery amount. Refers to the expected value of the recovery cost, which indicates the average performance of the recovery cost under the parameter distribution. Refers to the standard deviation of the recovery cost, which indicates the volatility of the recovery cost. Refers to the expected value of environmental impact, which indicates the average performance of environmental impact under parameter distribution. Refers to the standard deviation of environmental impact, indicating the volatility of environmental impact. and is a weight parameter used to balance the influence of expected value and variance, where As the decision variable, it represents the number of configuration combinations of skimmers.

4. The method for optimizing marine oil spill emergency recovery according to claim 1, wherein: The process of solving the processed objective function by combining the non-dominated sorting genetic algorithm with the particle swarm optimization algorithm to form a multi-objective genetic-particle swarm hybrid algorithm to obtain the optimal solution set for different types of skimmers with different configuration combinations of different quantities under different weights of maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact is as follows: Initialize the position and velocity of the particle swarm; For each particle, the speed and position are updated according to the particle group speed and position update formula; At the beginning of each generation, the inertia weight is dynamically updated to balance the global search and local search capabilities; The updated positions of particles in the particle swarm are repaired to ensure that the configuration combinations of different types of skimmers with different quantities meet the constraints of the three objective functions including maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact. By simulating the binary crossover operator, the individuals in the updated population are crossovered to generate new offspring individuals; The polynomial mutation operator is used to perform mutation operations on the offspring individuals after the crossover operation to increase the diversity of solutions; Evaluate the new population after particle swarm update and genetic operation, and calculate the objective function value and constraint value of the new population; Compare the objective function value of the new population with the individual optimal objective function value. If the new solution is better, update the individual optimal position and the corresponding objective function value. Combine the feasible solutions in the new population with the solutions in the archive, recalculate the non-dominated sort and congestion, and update the solution set in the archive. The archive stores the current optimal non-dominated solution set. After iterative optimization with the maximum number of iterations, the non-dominated solution set and the corresponding objective function in the archive are returned, and the optimal solution set of different types of skimmers with different numbers of configuration combinations under different weights for maximizing oil recovery, minimizing recovery costs, and minimizing environmental impact is obtained.

5. The method for optimizing marine oil spill emergency recovery according to claim 1, wherein: The process of using the fuzzy superior-inferior solution distance method to find the optimal number configuration combination of oil skimmers with a certain relative weight from the optimal solution set of different number configuration combinations of different types of oil skimmers, that is, the marine oil spill emergency recovery plan, is as follows: Perform triangular fuzzification on each element in the decision matrix of the optimal solution set; Standardize the fuzzified decision matrix; Weighting the standardized decision matrix to obtain a weighted decision matrix; Determine the fuzzy positive ideal solution and the fuzzy negative ideal solution for the weighted decision matrix; Calculate the Euclidean distance of each skimmer configuration combination scheme to the positive ideal solution and the negative ideal solution respectively; Based on the Euclidean distance between the positive ideal solution and the negative ideal solution, the relative proximity of each skimmer quantity configuration combination scheme is calculated, and the skimmer quantity configuration combination schemes are ranked according to the relative proximity. The scheme with the largest relative proximity is the optimal skimmer quantity configuration combination.

6. The method for optimizing marine oil spill emergency recovery according to claim 1, characterized in that: The process of using the hierarchical analysis method to compare the importance of the three objective functions in pairs to form a judgment matrix and calculate the relative weights of the importance of the three objective functions is as follows: The decision-making problem is decomposed into three levels: the objective level, the criterion level, and the solution level. The decision-making problem is: in an offshore oil spill emergency recovery scenario, based on the three conflicting objectives of maximizing oil recovery volume, minimizing recovery costs, and minimizing environmental impact, the optimal skimmer quantity configuration combination with the best overall performance is selected from the Pareto optimal solution set. Compare each factor at the criterion level or solution level, determine the relative importance, and construct a judgment matrix; Normalize the judgment matrix and calculate the weight vector of each factor; The consistency index and random consistency ratio are calculated to determine whether the matrix is ​​consistent, and then the relative weights of the importance of the three objective functions are obtained.

Images