A design method for mooring system of floating platform

Abstract

The present invention discloses a floating platform mooring system design method, including: establishing an initial mooring model through OrcaFlex; defining optimization variables, objective functions and constraints, setting each variable to change within a reasonable range; establishing an initial sample set through an adaptive Latin hypercube sampling method, and then calling OrcaFlex for time domain analysis; constructing an initial proxy model using the optimization variable values ​​and the objective function calculation results; calculating the expected improvement, determining the sampling points, iteratively updating, and determining and outputting a solution set that meets the requirements. The present invention uses an improved algorithm to optimize the constructed proxy model, thereby improving the optimization efficiency and reducing the computing power requirements for the machine; realizing the intelligent optimization of the mooring system design scheme, and obtaining the optimal design scheme within the range by setting the initial parameter range, so that the mooring design is simple and fast, and the economic benefits of the mooring system are greatly improved under the premise of meeting the requirements of the specification.

Description

Technical Field

[0001] The present invention relates to the technical field of offshore photovoltaic power generation, and in particular to a design method for a floating platform mooring system. Background Art

[0002] While the technology for floating photovoltaic systems is maturing, the design of mooring systems presents increasing challenges. Mooring systems are crucial for ensuring the safe movement of the platform. Variations in the layout and length of mooring lines can lead to uneven load distribution. Therefore, the mooring and anchoring installation tolerances for floating photovoltaic arrays are much tighter, further complicating the installation process. The mooring system must adapt to changing water levels, and its mechanisms and engineering design must be able to cope with year-round fluctuations.

[0003] In the harsh shallow water environment, the optimal design of the floating photovoltaic mooring system is crucial. For catenary mooring systems, the size and layout of the anchor chain and buoys are important factors in determining the system configuration and strength. The optimal design of floating photovoltaic mooring systems is fundamental to ensuring system safety and economic efficiency.

[0004] Currently, mooring system design is divided into two approaches: manual iteration and algorithmic optimization. Manual iteration requires a certain level of designer experience, is time-consuming, and rarely results in an optimal solution. Algorithmic optimization, on the other hand, mostly employs methods such as genetic algorithms and neural networks. These methods require high computing power and long optimization cycles. For example, genetic algorithms, in order to obtain an optimal mooring solution, require a long iteration cycle and population size, which is time-consuming, expensive, and results in low optimization efficiency. Summary of the Invention

[0005] In order to overcome the defects of the prior art, the technical problem to be solved by the present invention is: how to improve the algorithm efficiency of the floating platform mooring system without requiring high computing power.

[0006] To solve the above technical problems, the present invention provides a method for designing a floating platform mooring system, comprising the following steps:

[0007] S1: Establish the initial mooring model through OrcaFlex;

[0008] S2: Define optimization variables, objective functions, and constraints, and set each variable to vary within a reasonable range;

[0009] S3: The initial sample set is established through the adaptive Latin hypercube sampling method, and then OrcaFlex is called for time domain analysis;

[0010] S4: Construct an initial proxy model using the optimized variable values ​​and the objective function calculation results;

[0011] S5: Calculate the expected improvement, determine the sampling points, iteratively update, determine and output the solution set that meets the requirements.

[0012] On the basis of the above technical solution, the present invention can also be improved as follows.

[0013] Preferably, in step S3, the adaptive Latin hypercube sampling method enhances the diversity of sampling by improving the minimum distance between sample points in multidimensional space; maintains the uniformity of samples in each dimension and optimizes the minimum distance value by iteratively and dynamically adjusting the closest sample point pairs in the initial Latin hypercube sampling sample set; and repeats the process through an adaptive fine-tuning mechanism until the minimum distance between sample points is no longer significantly improved or a predetermined number of iterations is reached.

[0014] Preferably, step S5 further comprises:

[0015] S5.1: Define the parameters of the genetic algorithm, perform optimization on the genetic algorithm, find the point corresponding to the maximum expected improvement, and calculate the expected improvement;

[0016] S5.2: Determine the sampling points used in each iteration;

[0017] S5.3: Calculate the true value corresponding to the sampling point through time domain analysis and assign it to the proxy model to iteratively update it;

[0018] S5.4: Determine whether the result satisfies the constraint conditions (or reaches the maximum number of iterations). If so, the iteration ends and proceeds to step S5.5. If not, return to step S5.1 and continue calculation and iteration.

[0019] S5.5: Output a set of solutions that meet the requirements.

[0020] Preferably, step S1 further includes: setting sea environment parameters, and calculating resistance load, wherein the resistance load includes force and moment.

[0021] Preferably, the optimization variables include but are not limited to: mooring line length, diameter, buoyancy, and position.

[0022] Preferably, platform motion, mooring line tension safety factor and construction cost are used as the objective functions for optimization.

[0023] Preferably, the total weight of the mooring lines is used instead of the cost calculation. When the number is fixed, the lighter the single mooring line, the lower the cost.

[0024] Compared with the prior art, the present invention can produce the following beneficial technical effects:

[0025] 1. This invention uses a multi-objective parallel optimization algorithm to optimize the design and analysis of floating photovoltaic mooring systems in response to the environmental conditions of irregular wind and wave areas. This method is highly efficient and does not require high computing power. By combining the algorithm with professional mooring analysis software, it optimizes and designs a mooring solution that meets regulatory requirements and is highly economical.

[0026] 2. The adaptive Latin hypercube sampling method in this paper combines spatial layout optimization with the mandatory requirements of traditional uniform sampling, significantly improving the overall quality of multidimensional parameter space sampling without increasing the number of additional samples. This method provides a new and effective tool for experimental design in complex system modeling and simulation, and is particularly suitable for applications with strict requirements for high-quality sampled data.

[0027] 3. This invention solves the mooring system design problem for floating platforms. It uses an improved algorithm to optimize the constructed proxy model, improving optimization efficiency and reducing the computing power requirements of the machine. The entire analysis process realizes intelligent optimization of the mooring system design scheme. By setting the initial parameter range, the optimal design scheme within the range is obtained, reducing the design experience requirements for users, making mooring design simple and quick, and greatly improving the economic benefits of the mooring system while meeting regulatory requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] 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 work.

[0029] Figure 1 A flowchart of the steps of the floating platform mooring system design method of the present invention;

[0030] Figure 2 (a) is a sampling diagram using the Latin hypercube sampling method;

[0031] Figure 2 (b) is a sampling diagram using the adaptive Latin hypercube sampling method of the present invention;

[0032] Figure 3 Schematic diagram of the specific process of step S5;

[0033] Figure 4 A top view of the floating photovoltaic (2×2) mooring system;

[0034] Figure 5 Side view of the floating photovoltaic (2×2) mooring system. DETAILED DESCRIPTION

[0035] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of 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. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0036] Please refer to Figure 1 As shown in FIG, it is a flowchart of the steps of the floating platform mooring system design method of the present invention; the design method includes the following steps:

[0037] S1: Establish an initial mooring model using OrcaFlex. OrcaFlex is a professional dynamic analysis software widely used in the field of marine engineering. Developed by the British company Orcina, it is mainly used to analyze the mechanical properties of marine structures and simulate installation processes.

[0038] Wherein, step S1 further includes:

[0039] Calculate the resistance load, which includes force and moment. Specifically, set the marine environment parameters, input coefficients such as wind and flow forces, and use the method provided by the International Petroleum Company Shipping Association to calculate the resistance load generated by ocean currents and air passing through the floating platform in the horizontal direction. See formula ① as follows:

[0040]

[0041] Where: f x 、f y and f z is the force in the X, Y and Z directions; m x 、m y and m z is the moment in the X, Y and Z directions; C surge 、C sway 、C heave 、C roll 、C pitch and C yaw are the surge, sway, heave, roll, pitch and yaw coefficients relative to the current or wind direction of the low-frequency ship heading; ρ is the seawater density or air density; v is the relative speed of the current or wind passing through the floating platform.

[0042] S2: Define the optimization variable x, objective function f(x), and constraints (according to the API specification), and set each variable to change within a reasonable range;

[0043] 1) Optimization variables x include but are not limited to: mooring line length, diameter, buoyancy, position, etc.; see formula ② as follows:

[0044]

[0045] Where n is the number of objective functions and k is the number of optimization variables.

[0046] 2) When optimizing the mooring system, it is necessary not only to meet the requirements of the specification but also to consider its cost. Therefore, the platform motion, mooring line tension safety factor SF and cost q are combined. obj etc. as the objective function for optimization, minimizing the safety factor and cost under the constraint of meeting the API (American Petroleum Institute) requirements. Preferably, the mooring line tension safety factors according to API RP 2SK quasi-static and dynamic analysis are 2.0 and 1.67, respectively. The mooring line tension safety factor ensures that each mooring line has sufficient strength to withstand the load effects imposed by extreme environmental conditions. See formula ③ as follows:

[0047]

[0048] Where: T d is the mooring line tension limit, T is the maximum mooring tension; please see formula ④ as follows:

[0049] T d =cd 2 (44-80d) ④

[0050] Where: c is the constant related to the steel chain grade, d is the nominal diameter.

[0051] The total weight of the mooring lines is used instead of the cost. Under the condition of a fixed number of mooring lines, the lighter the single mooring line, the lower the cost. Please see formula ⑤ as follows:

[0052]

[0053] Where: m i L is the wet weight per unit length of the i-th section material; i is the length of the i-th section of material.

[0054] S3: The initial sample set ω is established by the adaptive Latin hypercube sampling method, and then OrcaFlex is called for time domain analysis;

[0055] In the field of statistical experimental design, Latin hypercube sampling (LHS) is widely used due to its uniform sampling property in a single dimension; however, this method does not significantly optimize the minimum distance between sample points in the multidimensional parameter space, sometimes resulting in reduced discreteness and efficiency of parameter space exploration.

[0056] This paper proposes an adaptive Latin hypercube sampling method to improve the minimum distance d of sample points in multidimensional space. min , thereby enhancing the diversity of sampling. The minimum distance refers to the minimum value of the Euclidean distance between any two sample points in the sample point set. The Euclidean distance measures the true distance between two points in multidimensional space; see formula ⑥ as follows:

[0057]

[0058] Where: X ik is the value of the i-th sample point in the k-th dimension; m is the sample dimension; the sample set is an n×m matrix, representing n m-dimensional points. Please see formula 7 as follows when written in matrix form:

[0059]

[0060] Therefore, the adaptive Latin Hypercube sampling method of the present invention dynamically adjusts the closest sample point pairs in the initial LHS sample set in an iterative manner, maintaining the uniformity of the samples in all dimensions while exploring new sample layouts to maximize the Euclidean distance between sample point pairs. In each iteration, the distance between sample point pairs is calculated and the minimum distance value is optimized. This is repeated through an adaptive fine-tuning mechanism until the minimum distance between sample points no longer increases significantly or the predetermined number of iterations is reached. See formula ⑧ below:

[0061]

[0062] Where: is the value of the i-th sample point in the l-th dimension after the t-th iteration; i, j are the numbers of two randomly selected sample points; l is the number of a randomly selected dimension; n is the number of samples.

[0063] Specifically,

[0064] 1) Randomly select two sample points i and j, and a dimension l;

[0065] 2) Exchange the values ​​of sample points i and j on dimension l to obtain new sample points and Where t represents the number of iterations;

[0066] 3) If the minimum distance of the sample point set after the exchange becomes larger, then accept the exchange, otherwise keep it as it is.

[0067] The adaptive Latin hypercube sampling method in this invention combines the optimization of spatial layout with the mandatory requirements of traditional uniform sampling, significantly improving the overall quality of multidimensional parameter space sampling without increasing the number of additional samples. This method provides a new and effective tool for experimental design in complex system modeling and simulation, and is particularly suitable for application scenarios with strict requirements for high-quality sampling data.

[0068] Please refer to Figure 2 As shown, Figure 2 (a) in the figure is a sampling diagram using the Latin hypercube sampling method. Figure 2 (b) is a sampling diagram using the adaptive Latin hypercube sampling method of the present invention; in the two-dimensional space [0,1), the Latin hypercube sampling method and the adaptive Latin hypercube sampling method are used to generate ten sample points each to compare the sampling effects of the two methods; Figure 2 As shown in (a) and (b), the sample points generated by the adaptive Latin hypercube sampling method are more dispersed in space than the sample points generated by the Latin hypercube sampling method, which shows that the adaptive Latin hypercube sampling method in the present invention can improve the uniformity of sampling, so that the sample points better cover the entire space.

[0069] S4: Construct an initial proxy model using the optimized variable values ​​and the objective function calculation results Please see formula 9 as follows:

[0070]

[0071] Where: μ is the mean of the Gaussian process; ε(x) is the error term. The error terms of any two points are correlated. See formula ⑩ as follows:

[0072]

[0073] Where: n is the number of optimization variables; θ k and p k is the undetermined coefficient.

[0074] S5: Calculate the expected improvement, determine the sampling point, iteratively update, determine and output the solution set that meets the requirements; for details, please refer to Figure 3 As shown, it is a specific flow chart of step S5;

[0075] S5.1: Define the parameters of the genetic algorithm, optimize the genetic algorithm, find the point corresponding to the maximum expected improvement value, and calculate the expected improvement (EI);

[0076] S5.2: Determine the sampling points used in each iteration;

[0077] S5.3: Calculate the true value corresponding to the sampling point through time domain analysis and assign it to the proxy model to iteratively update it;

[0078] S5.4: Determine whether the result satisfies the constraint conditions (or reaches the maximum number of iterations). If so, the iteration ends and proceeds to step S5.5. If not, return to step S5.1 and continue calculation and iteration.

[0079] S5.5: Output a set of solutions that meet the requirements.

[0080] In this method, new sampling points are selected during iterations using a genetic algorithm that maximizes the expected criterion and are added to the sampling set. By establishing an initial population and defining a fitness function as the inverse of the expected improvement model, the genetic algorithm selects sampling points with potential for improvement in each generation. These points are then integrated into the surrogate model, enabling an efficient search for the global non-dominated frontier in high-dimensional design spaces. This integrated approach is particularly suitable for complex multi-objective optimization problems.

[0081] The fitness function expression is shown in the formula and as follows:

[0082]

[0083]

[0084] Where: EI(x) is the expected improvement function; EIM(x) is the expected improvement matrix; Φ(·) and φ(·) are the cumulative distribution function and probability density function of the standard normal distribution, respectively; r i is the reference point for setting; f i j is the element in the jth row and ith column of the approximate Pareto front; and s i (x) are the predicted value and standard deviation of the ith objective function surrogate model, i = 1, 2, …, m, j = 1, 2, …, k.

[0085] Therefore, establishing initial sample points in this invention does not require high computing power; a proxy model can be constructed using a relatively small number of initial calculation examples. Parallel processing is employed for iteration, determining multiple sampling points for iteration, fully utilizing the computing power of existing computers and thereby improving optimization efficiency. Based on the industry-recognized OrcaFlex professional mooring analysis software, the design method of this invention only requires setting an initial parameter range to derive the optimal mooring design within that range. This allows floating platforms to obtain a highly economical and adaptable solution for various complex sea areas.

[0086] In summary, this invention, used for mooring floating platforms, solves the challenges of mooring system design under complex sea conditions by iteratively optimizing the mooring system. This reduces mooring design costs and improves mooring optimization efficiency, achieving economic maximization while meeting relevant regulatory requirements. This invention integrates the traditional EGO algorithm with a genetic algorithm. The sampling process utilizes an improved sampling method that combines adaptive sampling with Latin hypercube sampling. Finally, the algorithm uses the specialized mooring analysis software OrcaFlex for time-domain analysis and optimization of the results, ultimately yielding a feasible mooring solution.

[0087] Next, the present invention will be further described in detail with reference to an embodiment; Figure 4 and Figure 5 As shown, Figure 4 This is a top view of the floating photovoltaic (2×2) mooring system. Figure 5 It is a side view of a floating photovoltaic (2×2) mooring system; the mooring system is connected to a floating platform 1 and includes: a mooring line polyester cable 2, a connection point 3 between the mooring line and the platform, a buoy 4, an upper steel chain 5, a lower heavy chain 6 and an anchor point 7.

[0088] In the initial mooring scheme, the steel chain used was Grade 3 with a total length of 94m; the diameter of the upper steel chain 5 was 0.038m, and the length of each chain was 64m; the diameter of the lower heavy chain 6 was 0.08m, and the length of each chain was 30m (94m-64m); the mass of a single buoy 4 in the mooring line was 80kg and the volume was 0.5m 3 , with a height of 0.7m and a position at (0,0,12) from connection point 3. The polyester cable 2 connecting the platform has a diameter of 0.1m, and a single length of 7m and 9.899495m. The operating environment is a 50-year return period with a water depth of 12.68m. Time-domain analysis shows that the mooring system experiences the greatest stress when the environmental direction is 45°. Therefore, a 45° wind, wave, and current direction is selected for analysis. The initial model safety factor is calculated to be 4.27, which is redundant compared to the API specification requirement of 1.67 and is not economically viable.

[0089] The above model is used as the initial model for optimization. The variables are set as the steel chain length, diameter, polyester diameter, buoy mass, volume and position. The parameter variation range is shown in Table A. The objective function is set as the mooring line safety factor and wet weight, and the API specification is used as the constraint condition.

[0090] Table A Parameter Variation Range

[0091] variable Minimum Maximum Upper steel chain length (m) 60 84 Upper steel chain diameter (m) 0.02 0.04 Lower heavy chain diameter (m) 0.04 0.1 Polyester cable diameter (m) 0.04 0.1 <![CDATA[Buoy volume (m 3 )]]> 0.4 0.8 Buoy mass (kg) 50 120 Buoy position (m) 8 15

[0092] Through iterative optimization, the final optimization results are shown in Table B. The safety factor is reduced to 1.91, which not only meets the API specification requirement of 1.67, but also significantly reduces the cost of the anchor chain. The wet weight is 61% of the original, and the wet weight of a single mooring line is 2.51te.

[0093] Table B Comparison of mooring options

[0094] variable Before optimization After optimization Upper steel chain length (m) 64 68 Lower heavy chain length (m) 30 26 Upper steel chain diameter (m) 0.038 0.028 Lower heavy chain diameter (m) 0.08 0.055 Polyester cable diameter (m) 0.1 0.045 <![CDATA[Buoy volume (m 3 )]]> 0.5 0.522 Buoy mass (kg) 80 118 Buoy position (m) (0,0,12) (0,0,14) Safety factor 4.27 1.91 Wet weight of single anchor chain (te) 4.14 2.51

[0095] The method according to the present invention is used to optimize the design of a floating photovoltaic mooring system, thereby reducing the material cost of the mooring system and improving the design efficiency and economy of the mooring system while meeting regulatory requirements.

[0096] The floating platform mooring system design method of the present invention solves the mooring system design problem of a floating platform. By using an improved algorithm to optimize the constructed proxy model, the optimization efficiency is improved and the computing power requirement of the machine is reduced. The entire analysis process realizes the intelligent optimization of the mooring system design scheme. By setting the initial parameter range, the optimal design scheme within the range is obtained, which reduces the design experience requirement for the user, makes the mooring design simple and quick, and greatly improves the economic benefit of the mooring system while meeting the requirements of the specifications.

[0097] 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 method for designing a floating platform mooring system, characterized in that: The steps include: S1: Establish initial mooring model through OrcaFlex; S2: Define optimization variables, objective functions, and constraints, and set each variable to vary within a reasonable range; S3: The initial sample set is established by adaptive Latin hypercube sampling method, and then OrcaFlex is called for time domain analysis; The adaptive Latin hypercube sampling method aims to improve the minimum distance dmin of sample points in multidimensional space, thereby enhancing the diversity of sampling. The minimum distance refers to the minimum value of the Euclidean distance between any two sample points in the sample point set. The Euclidean distance measures the true distance between two points in multidimensional space, as shown in formula ⑥: Where: X ik is the value of the i-th sample point in the k-th dimension; m is the sample dimension; the sample set is an n×m matrix, representing n m-dimensional points, written in matrix form, as shown in formula ⑦: The adaptive Latin hypercube sampling method dynamically adjusts the closest sample point pairs in the initial LHS sample set iteratively, while maintaining the uniformity of the samples in each dimension, and exploring new sample layouts to maximize the Euclidean distance between sample point pairs; in each iteration, the distance between sample point pairs is calculated and the minimum distance value is optimized, and this is repeated through an adaptive fine-tuning mechanism until the minimum distance between sample points is no longer significantly improved or reaches the predetermined number of iterations, as shown in Formula ⑧: Where: is the value of the i-th sample point in the l-th dimension after the t-th iteration; i, j are the numbers of two randomly selected sample points; l is the number of a randomly selected dimension; n is the number of samples; S4: construct an initial proxy model using the optimization variable values ​​and the objective function calculation results; S5: Calculate the expected improvement, determine the sampling point, iteratively update, determine and output the solution set that meets the requirements. Step S5 further includes: S5.1: Define the parameters of the genetic algorithm, optimize the genetic algorithm, find the point corresponding to the maximum expected improvement value, and calculate the expected improvement; S5.2: Determine the sampling points used in each iteration; S5.3: Calculate the true value corresponding to the sampling point through time domain analysis and assign it to the proxy model to iteratively update it; S5.4: Determine whether the result meets the constraint condition. If so, the iteration ends and enters step S5.5; if not, return to step S5.1 and continue calculation and iteration; S5.5: Output a solution set that meets the requirements.

2. The floating platform mooring system design method according to claim 1, characterized in that: Step S1 further includes: setting sea environment parameters and calculating resistance load, wherein the resistance load includes force and moment.

3. The floating platform mooring system design method according to claim 1, characterized in that: The optimization variables include but are not limited to: mooring line length, diameter, buoyancy and position of buoys.

4. The method for designing a floating platform mooring system according to claim 1, characterized in that: The platform motion, mooring line tension safety factor and construction cost are taken as the objective functions of optimization.

5. The floating platform mooring system design method according to claim 4, characterized in that: The total weight of the mooring lines is used instead of cost calculation. When the number is fixed, the lighter the single mooring line, the lower the cost.

Images