FPSO deck parallel cable layout method

Abstract

The present application relates to cable layout technical field, specifically to FPSO deck parallel cable layout method, using improved dream optimization algorithm for FPSO deck parallel cable layout, including the following steps: S1, initialization basic parameter, set up the related parameter of improved dream optimization algorithm, construct layout space;S2, the population size of hybrid multi-strategy initialization dream optimization algorithm is used, solution is generated, and path is obtained;S3, enter the exploration stage, update solution, judge whether the path exists interference, if there is interference, then it is corrected by breadth-first search, and path update is completed;S4, enter the development stage, determine the global optimal solution, update solution again, realize path optimization;S5, based on the global optimal solution, the horizontal and vertical coordinates of cable path point are determined, and the final path result is output.The present application guarantees the feasibility and quality of path, can completely avoid equipment interference, and can improve the convenience of path laying through simplification and smoothing.

Description

Technical Field

[0001] This invention relates to the field of cable layout technology, and more specifically to a method for laying parallel cables on an FPSO deck. Background Technology

[0002] Floating production storage and offloading (FPSO) units house natural gas liquefaction plants on ships, integrating the production, storage, and unloading of liquefied natural gas. They can be moored directly above gas fields for operations and are important equipment for offshore natural gas field development. The deck area of ​​an FPSO is much smaller than that of an onshore natural gas liquefaction plant, and the need to arrange numerous production, processing, and storage equipment within the limited deck space results in very crowded conditions.

[0003] As a crucial carrier for connecting various devices and transmitting power and signals, cables require meticulous planning within limited space to avoid interference with other equipment while simultaneously meeting the power and signal transmission needs of those devices. In practice, multiple cables are often designed in parallel. Furthermore, cables for equipment in different functional areas, such as separation modules, compression modules, and liquefaction modules, need to be arranged systematically to ensure safe cable laying while facilitating maintenance and management. Traditional ship cable layout technology involves designing cable channels on two-dimensional drawings. Existing automated cable layout methods for oil and gas production platforms primarily rely on single-pipe layouts, which are prone to problems such as unreasonable spatial arrangement and interference with other equipment. Modifying drawings after changes to parallel pipeline designs is cumbersome and inefficient, making it difficult to meet the complex cable layout design requirements of FPSO decks.

[0004] Dream Optimization Algorithm (DOA), as an emerging intelligent optimization algorithm, simulates the cognitive mechanism of human dreams, transforming the "memory-forgetting-sharing" characteristics into optimization strategies. Multi-strategy collaborative design effectively balances exploration and development, making it particularly suitable for high-dimensional, nonlinear, and multi-constraint practical problems. However, there is currently little research in the field of cable layout. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide a method for parallel cable layout on FPSO deck.

[0006] This invention is achieved through the following technical solution: a parallel cable layout method for FPSO decks, using an improved dream optimization algorithm for FPSO deck parallel cable layout, comprising the following steps:

[0007] S1. Initialize basic parameters, set parameters related to the improved dream optimization algorithm, and construct the layout space;

[0008] S2. Initialize the population size of the dream optimization algorithm using a hybrid multi-strategy approach, generate solutions, and obtain paths;

[0009] S3. Enter the exploration phase, update the solution, determine if there is any interference in the path, and if there is interference, correct it through breadth-first search to complete the path update;

[0010] S4. Enter the development phase, determine the globally optimal solution, update the solution again, and achieve path optimization;

[0011] S5. Based on the global optimal solution, determine the horizontal and vertical coordinates of the cable path points and output the final path result.

[0012] S1 includes the following sub-steps:

[0013] S1-1, Set the maximum number of iterations, population size, and number of path points for the improved dream optimization algorithm;

[0014] S1-2. Construct the layout space. Based on the FPSO deck area and the location of each piece of equipment on the FPSO deck, set the map size, cable start point, cable end point, and obstacle parameters of the layout space. The obstacles are models of equipment that cables cannot pass through in the layout space, and then generate a grid map.

[0015] S2 includes the following sub-steps:

[0016] S2-1. Store the path data as a 1×4Dim matrix, where the data in the first 2Dim dimensions correspond to the path points of the first cable, and the data in the last 2Dim dimensions correspond to the path points of its parallel cables.

[0017] S2-2. A hybrid multi-strategy approach is adopted to generate solutions for the population individuals in the dream optimization algorithm using three strategies in proportion;

[0018] S2-3. When generating solutions, first generate solutions in the first 2Dim dimensions, and then generate solutions in the second 2Dim dimensions by filtering non-obstacle points in the neighborhood of the solutions in the first 2Dim dimensions through an outer diffusion method.

[0019] S2-2 includes the following sub-steps:

[0020] S2-2-1. Using 30% of the population as an A* heuristic, move point by point from the starting point of the cable towards the ending point to generate a solution:

[0021] S2-2-2, The solution is obtained by linear interpolation of 40% of the population between the cable start point and the cable end point;

[0022] S2-2-3. After the remaining 30% of the population individuals generate a solution using linear interpolation, random perturbation is added to the solution obtained by linear interpolation.

[0023] The calculation formula in S2-2-1 is as follows:

[0024] next = current + dir × step;

[0025] dir=goalpost-current;

[0026] step=max(gridheight,gridwidth) / (Dim+1);

[0027] In the formula, next is the next path point, current is the current node, goalpos is the endpoint of the cable, gridheight is the number of rows in the grid map, gridwidth is the number of columns in the grid map, Dim is the solution dimension, dir is the direction vector, and step is the movement distance.

[0028] The formulas for obtaining the solution based on linear interpolation in S2-2-2 and S2-2-3 are as follows:

[0029] next = startpos + (goalpos-startpos);

[0030] =(j-1) / (Dim+1), j=2:Dim+1;

[0031] In the formula, startpos is the cable starting point. Here, j is the interpolation control variable, and j is the interpolation dimension.

[0032] After adding a random perturbation to S2-2-3, the solution is calculated using the following formula:

[0033] next=current+rand·min(gridheight,gridwidth) / 5;

[0034] In the formula, rand is a random number that takes the value [0, 1].

[0035] S3 includes the following sub-steps:

[0036] S3-1. The solution for the first 2Dim dimensions is updated using memory strategy, forgetting and supplementation strategy. The solution for the last 2Dim dimensions is generated by adding random perturbation to the solution for the first 2Dim dimensions.

[0037] S3-2. Update the updated solution using a repair strategy, where the repair strategy includes: determining whether the path of the parallel cable interferes with the device. If interference occurs, obtain the collision points between the interfering line segments, and use the breadth-first search strategy to expand and search layer by layer outward from the collision points until the path formed by the updated collision points and the two endpoints of the original interfering line segment does not interfere.

[0038] The S4 includes the following sub-steps:

[0039] S4-1. Calculate the parallel fitness of the population individuals to obtain the global optimal solution;

[0040] S4-2. Update the solution of the first 2Dim dimensions using the forgetting and replenishment strategy. The solution of the last 2Dim dimensions is obtained by adding random perturbations based on the solution of the first 2Dim dimensions. If the solution of a certain dimension exceeds the boundary, the solution of that dimension is randomly generated, and the solution is updated using the simplification strategy and the smoothing strategy.

[0041] The formula for calculating the parallel fitness in the S4-1 is:

[0042] Fit = ;

[0043] fit = w1·pathlength + w2·(collision + segment) + w3(smoothness + sharpturn);

[0044] Among them, Fit is the parallel cable fitness, fit is the single cable fitness, parallelism is the parallelism, pathlength is the cable length, collision is the path point collision penalty, segment is the line segment interference penalty, smoothness is the 90° smoothness penalty, sharpturn is the 60° smoothness penalty, and w1, w2, w3, w4 are the weights corresponding to the targets respectively.

[0045] The simplification strategy in the S4-2 is:

[0046] Connect a virtual straight line between the start and end points of the curve to be processed, calculate the distances between all intermediate points and the straight line, find the maximum distance value dmax, and compare dmax with the thinning threshold threshold: if dmax < threshold, all intermediate points on this curve are discarded; if dmax ≥ threshold, then take this point as the boundary, divide the curve into two parts, and repeat the above process for these two parts of the curve until all points are processed.

[0047] The smoothing strategy in the S4-2 is:

[0048] By using a cubic β-spline smoothing curve, other cables in a parallel cable will shift towards the corresponding point of another cable path during smoothing. The smoothing calculation formula is as follows:

[0049] newpos = (1-a) 3 p0+3·(1-a) 2 ·a·p1+3·(1-a)·a·p2+a 3 p3;

[0050] Where newpos is the smoothed path point, a is the control parameter, a=0.5, and p0, p1, p2, and p3 are four consecutive points on the original cable path.

[0051] Compared with the prior art, the beneficial effects of the present invention are:

[0052] This application proposes a parallel cable layout method for FPSO decks based on an improved dream optimization algorithm. It generates initial paths through a hybrid multi-strategy approach, iterates through exploration and development phases, and balances multiple objective requirements using a dynamic weight fitness function. This application ensures path feasibility and quality, completely avoiding equipment interference, improving path laying convenience through simplification and smoothing, and enhancing the spacing stability of parallel cables for easier maintenance. It also improves algorithm adaptability and efficiency, avoids local optima problems in traditional algorithms, reduces unnecessary computation, and is applicable to parallel pipeline layouts in various equipment-dense scenarios.

[0053] This application breaks away from the traditional single-path generation mode, integrating A* heuristic, linear interpolation, and linear interpolation + random perturbation strategies in a 3:4:3 ratio. At the same time, it generates parallel cable paths through outer diffusion, ensuring that the initial solution has directionality, feasibility, and diversity, laying a high-quality starting point for subsequent optimization.

[0054] In the exploration phase of this application, the path is updated through grouped memory and forgetting supplementation strategies, and the interference problem is solved by synchronously calling the repair strategy. In the development phase, the focus is on path optimization. First, the global optimal solution is screened through a parallel fitness function, and then the path is upgraded from feasible to high-quality through simplification and smoothing strategies.

[0055] This application transforms the four core requirements of path length, obstacle avoidance, smoothness, and parallelism into a single-objective weighted function. By dynamically adjusting the weights in the exploration / development phase, it guides the algorithm to converge toward the optimal solution that is "short path, collision-free, with few detours, and highly parallel". Attached Figure Description

[0056] Figure 1 This is a flowchart of the application process;

[0057] Figure 2This is a schematic diagram of the parallel cable layout result using a conventional DOA in environment 1 in Example 2;

[0058] Figure 3 This is a schematic diagram of the parallel cable layout result in environment 1 using the method described in this invention in Example 2;

[0059] Figure 4 This is a schematic diagram of the parallel cable layout result using a conventional DOA in environment 2 in Example 2;

[0060] Figure 5 This is a schematic diagram of the parallel cable layout result in environment 2 using the method described in this invention in Example 2. Detailed Implementation

[0061] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0062] Example 1

[0063] The Dream Optimization Algorithm (DOA) is a metaheuristic algorithm inspired by human dreams. By simulating the cognitive mechanism of human dreams, it integrates basic memory strategies, forgetting and replenishment strategies that balance exploration and development capabilities, and dream sharing strategies that enhance the ability to escape local optima. The optimization process is divided into exploration and development phases, thereby obtaining satisfactory optimization results.

[0064] Combining the characteristics of human dreams with optimization algorithm knowledge, the DOA algorithm is mainly divided into three stages: initialization stage, exploration stage, and development stage.

[0065] During the initialization phase, DOA first generates a random solution within the search space as the initial solution, thus initiating the optimization process. The formula for obtaining the initial population is as follows:

[0066] X i =X l +rand×(X u -X l ), i = 1, 2, ..., N;

[0067] Where N represents the number of individuals, i.e., the population size; X i It is the i-th individual in the population; X l and X uThese represent the lower and upper bounds of the search space, respectively; rand is a random number between 0 and 1.

[0068] During the exploration phase, the iteration count ranges from 0 to T. d First, the population is divided into 5 groups based on memory capacity. Individuals in each group are updated as follows: Each iteration is considered a dream behavior. By continuously performing this behavior, we obtain the optimal solution and optimal value. Before each dream session, all individuals in each group are shown to the entire population (i.e., the best individuals from the previous few iterations). Since individuals randomly forget some information (i.e., information in certain dimensions) while dreaming, only the positions of the forgotten dimensions are updated. Groups with different memory capacities imply different numbers of forgotten dimensions, represented by parameters k1, k2, k3, k4, and k5. Therefore, the position of each individual is first reset to the position of the best individual from the previous few iterations in that group, and then kq dimensions are randomly selected, and the positions of these dimensions are updated, where q = 1, 2, 3, 4, and 5 represent the group number. Updates are performed sequentially from the 1st individual to the Nth individual. The specific update method and formula are as follows:

[0069] The memory strategy, serving as the basic memory strategy for individuals in group q, involves remembering the position information of the best individual in the group before dreaming, and then resetting that position information to the position of that best individual during dreaming. The calculation formula is as follows:

[0070] ;

[0071] in, This represents the solution for the i-th individual at iteration i+1; This represents the best individual in group q after iteration t.

[0072] The forgetting and replenishment strategy combines global and local search capabilities. This strategy follows memory strategies, allowing individuals to forget and self-organize location information within the forgetting dimension. The update formula is as follows:

[0073] ;

[0074] in, This represents the solution for the i-th individual at the j-th dimension in iteration t+1. This represents the position of the best individual in the entire population in the j-th dimension during iteration t; and Let represent the lower and upper bounds of the search space in the j-th dimension, respectively.

[0075] The dream-sharing strategy in DOA enhances the ability to escape local optima. This strategy operates in parallel with forgetting and replenishment strategies, follows the memory strategy, and allows individuals to randomly acquire positional information from other individuals in the forgetting dimension. The update formula is as follows:

[0076] ;

[0077] m is a natural number randomly selected from [1, N] and used for updating each dimension.

[0078] During the development phase, the iteration count starts from T. d To T max No further grouping is performed. Before each dream session, the best dream (i.e., the best individual) of the entire population from the previous few iterations is shown to the population. Then, the position of the forgetting dimension for each individual is updated. Since all individuals in the population have the same number of forgetting dimensions, denoted as k, k is the position of the forgetting dimension in the Dim dimension. r The forgetting dimensions are randomly selected, and the positions of these dimensions are updated according to the forgetting and replenishment strategies of the exploration phase.

[0079] Based on this, refer to Figure 1 This application discloses a method for laying out parallel cables on an FPSO deck, which uses an improved dream optimization algorithm to lay out parallel cables on an FPSO deck, including the following steps:

[0080] S1. Initialize basic parameters, set parameters related to the improved dream optimization algorithm, and construct the layout space.

[0081] S1 includes the following sub-steps:

[0082] S1-1. Set the maximum number of iterations, population size, number of path points (solution dimension Dim), and other relevant parameters for the improved Dream Optimization Algorithm (DOA).

[0083] S1-2. Construct the layout space. Based on the FPSO deck area and the location of each piece of equipment on the FPSO deck, set the map size, cable start point, cable end point, and obstacle parameters of the layout space. The obstacles are models of equipment that cables cannot pass through in the layout space, and then generate a grid map.

[0084] This application solves the problem of continuous space being difficult to process by algorithms by transforming the actual deck space into passable grids and obstacle grids, and quantifying the cable path with grid coordinates, wherein the equipment corresponds to the obstacle grid.

[0085] S2. Initialize the population size of the dream optimization algorithm using a hybrid multi-strategy approach, generate solutions, and obtain paths. The population is the set of individuals; a complete path of a single parallel cable corresponds to one individual in the population.

[0086] S2 includes the following sub-steps:

[0087] S2-1. Store the path data as a 1×4Dim matrix, where the data in the first 2Dim dimensions correspond to the path points of the first cable, and the data in the last 2Dim dimensions correspond to the path points of its parallel cables.

[0088] This application uses a 1×4Dim matrix to structurally store two cable paths, where the first 2Dim represents the coordinates of the first cable and the last 2Dim represents the coordinates of the parallel cable. During iteration, the logic of updating the parallel cable path based on the first cable path is always maintained, avoiding the failure of parallelism caused by the optimization of the two cables separately, and ensuring that the layout meets the engineering requirements of orderly parallel cables between FPSO deck equipment.

[0089] S2-2. A hybrid multi-strategy approach is adopted to generate solutions for the population in the Dream Optimization Algorithm using three strategies in proportion. To achieve a high-quality initial solution for the parallel cable and ensure a certain degree of parallelism from the outset, the traditional single-path initialization strategy of the standard DOA is extended to generate two paths simultaneously, ensuring that they start from the same starting point, reach adjacent endpoints, and maintain a certain degree of parallelism at intermediate points. Specifically, three different initialization methods are used and applied randomly in proportion to increase population diversity.

[0090] Furthermore, S2-2 includes the following sub-steps:

[0091] S2-2-1. Using a strategy that employs the A* heuristic, 30% of the population moves point by point from the starting point of the cable towards the ending point, generating a solution. The calculation formula is as follows:

[0092] next = current + dir × step;

[0093] dir=goalpost-current;

[0094] step=max(gridheight,gridwidth) / (Dim+1);

[0095] In the formula, next is the next path point, current is the current node, goalpost is the endpoint of the cable, gridheight is the number of rows in the grid map, gridwidth is the number of columns in the grid map, Dim is the solution dimension, dir is the direction vector, and step is the movement distance.

[0096] S2-2-2, The solution is obtained by linear interpolation of 40% of the population between the cable start point and the cable end point;

[0097] S2-2-3. After the remaining 30% of the population individuals generate solutions using linear interpolation, random perturbations are added to the solutions obtained by linear interpolation to increase path diversity.

[0098] The formula for calculating the solution using linear interpolation is as follows:

[0099] next = startpos + (goalpos-startpos);

[0100] =(j-1) / (Dim+1), j=2:Dim+1;

[0101] In the formula, startpos is the cable starting point. Here, j is the interpolation control variable, and j is the interpolation dimension.

[0102] The formula for calculating the solution after adding random perturbation in S2-2-3 is as follows:

[0103] next=current+rand·min(gridheight,gridwidth) / 5;

[0104] In the formula, rand is a random number that takes the value [0, 1].

[0105] Therefore, this application generates feasible routes through deterministic and uncertain strategies, ensuring that the initial solution has a certain degree of feasibility and improving the speed of iterative convergence.

[0106] S2-3. When generating solutions, first generate solutions in the first 2Dim dimensions, and then generate solutions in the second 2Dim dimensions by filtering non-obstacle points in the neighborhood of the solutions in the first 2Dim dimensions through an outer diffusion method.

[0107] S3. Enter the exploration phase, update the solution, determine if there is any interference in the path, and if there is interference, correct the collision points through breadth-first search to ensure that the path is free of interference and complete the path update.

[0108] S3 includes the following sub-steps:

[0109] S3-1. The solution for the first 2Dim dimensions is updated using memory, forgetting and supplementation strategies. The solution for the last 2Dim dimensions is generated by adding random perturbations to the solution for the first 2Dim dimensions.

[0110] S3-2. To ensure the feasibility of the solution, a repair strategy is adopted to update the updated solution. The repair strategy includes: determining whether the path of the parallel cable interferes with the device; if interference occurs, obtaining the collision point between the interfering segments, and using a breadth-first search strategy to expand the search layer by layer from the collision point until the path formed by the updated collision point and the two ends of the original interfering segment does not interfere. During the iterative update of DOA, the search using the memory, forgetting, supplementation, and dream-sharing strategies may lead to the parallel cable solution interfering with the device. Relying solely on the penalty function in the fitness function will result in low iteration convergence speed and poor solution quality. Therefore, this application calls the repair strategy after the solution is updated during the iterative process of the exploration phase.

[0111] S4. Enter the development phase, determine the globally optimal solution, update the solution again, and achieve path optimization;

[0112] S4 includes the following sub-steps:

[0113] S4-1. Calculate the parallel fitness of individuals in the population to obtain the global optimal solution;

[0114] When laying parallel cables, to ensure low economic cost, the bending angle should primarily be obtuse, and the two cables should run parallel as much as possible. Therefore, the fitness function for parallel cables includes length, bending penalty, collision penalty, and smoothness penalty. Simultaneously, a linear weighted approach is used to simplify the multi-objective calculation into a single-objective function.

[0115] The formula for calculating parallel fitness in S4-1 is as follows:

[0116] Fit= ;

[0117] fit=w1·pathlength+w2·(collision+segment)+w3(smoothness+sharpturn);

[0118] Where Fit represents the fitness of parallel cables, fit represents the fitness of a single cable, parallelism represents the degree of parallelism, pathlength represents the cable length, collision represents the path point collision penalty, segment represents the segment interference penalty, smoothness represents the 90° smoothness penalty, sharpturn represents the 60° smoothness penalty, and w1, w2, w3, and w4 represent the weights of the corresponding targets. Different settings exist at different stages. In the exploration phase, the weights change with the number of iterations, as detailed below:

[0119] w1 = 1 - 0.3 × (i / 0.7 × T);

[0120] w2 = 500 + 500×(i / 0.7×T);

[0121] w3 = 2 + 3×(i / 0.7×T);

[0122] w4 = i / 0.7×T;

[0123] In the development stage, the values of the four are fixed values, which are 0.6, 1000, 5, and 3 respectively.

[0124] S4-2. Update the solutions of the first 2Dim dimensions using the forgetting and supplement strategy. The solutions of the last 2Dim dimensions are obtained by adding random perturbations to the solutions of the first 2Dim dimensions. If the solution of a certain dimension exceeds the boundary, the solution of that dimension is randomly generated. The reduction strategy and smoothing strategy are used to update the solutions.

[0125] Too many nodes will cause more bends and redundant parts in the parallel cable path. Therefore, the Douglas-Peucker algorithm for solving line thinning is introduced for path simplification. Specifically, the operation process of the reduction strategy is to virtually connect a straight line between the start and end points of the curve to be processed, calculate the distances from all intermediate points to the straight line, and find the maximum distance value dmax. Compare dmax with the thinning threshold threshold: if dmax < threshold, all intermediate points on this curve are discarded; if dmax ≥ threshold, then take this point as the boundary, divide the curve into two parts, and repeat the above process for these two parts of the curve (that is, the two divided parts of the curve are used as new curves for the reduction strategy operation) until all points are processed.

[0126] The thinning threshold threshold is set according to the acceptable degree of deviation of the parallel pipeline. In a grid environment with a grid granularity of 1, it is usually set to 1.

[0127] Acute bends are not conducive to the use and laying of cables. The solutions obtained by the algorithm should have high smoothness. When iteratively solving in the early exploration stage, feasible solutions are mainly considered. Therefore, in the algorithm development stage, smoothing processing is added after updating the solutions. Further, the smoothing strategy includes: smoothing the curve through cubic β-spline. Other cables of the parallel cable will shift to the corresponding points of another cable path during smoothing, enhancing parallelism. The smoothing calculation formula is as follows:

[0128] newpos = (1 - a) 3 p0 + 3·(1 - a) 2 ·a·p1 + 3·(1 - a)·a·p2 + a 3 ·p3;

[0129] Among them, newpos is the path point after smoothing, a is the control parameter, a = 0.5, and p0, p1, p2, and p3 are four consecutive points on the original cable path point respectively.

[0130] This application avoids the algorithm getting stuck in a single local optimum by incorporating multiple different path schemes based on the diversity of solutions.

[0131] During the iteration process, the optimization is not targeted at a single path, but rather all individuals (i.e., paths) in the entire population are operated on simultaneously, such as in the exploration phase and the development phase. By selecting high-quality individuals and eliminating low-quality individuals, the population is gradually guided to converge toward the globally optimal path.

[0132] S5. Based on the global optimal solution, determine the x and y coordinates of the cable path points and output the final path result. In this step, the first 2Dim dimensions in the global optimal solution correspond to the x and y coordinates of the first cable path point, where the 1st to Dim dimensions are the x coordinates, the Dim+1 to 2Dim dimensions are the y coordinates, and so on. The last 2Dim dimensions correspond to the x and y coordinates of the parallel cable path points of the first cable.

[0133] Example 2

[0134] Reference Figures 2-5 In the diagram, red and blue lines represent two parallel pipeline paths, with red grids indicating the starting points and blue and dark green grids indicating the ending points. To demonstrate the effectiveness of the improved algorithm, simulations were conducted in grid maps with different obstacles and different starting and ending points (referred to as Environment 1 and Environment 2, respectively). The algorithm used in this invention was compiled using MATLAB in a Windows 11 environment. First, a simulation model space for obstacles was established. The device models in the layout space can be considered as obstacles that cannot be passed through by cables. The grid method was used to divide the space into equal parts, with a grid granularity of 1. The size of the entire layout space was 20×20. The population size popSize was 40, the maximum number of iterations maxIter was set to 100, the exploration phase ratio was 0.7, the development phase ratio was 0.3, and the number of path points was 20. Compared with the traditional DOA algorithm, the experimental results are shown in Table 1. It can be seen that the algorithm proposed in this invention showed good performance in cable layout in different environments.

[0135] Table 1 Comparison between traditional DOA algorithm and the algorithm proposed in this application

[0136]

[0137] The above description is merely an optional embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural transformations made using the content of the present invention under the concept of the present invention, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. A method for parallel cable layout on an FPSO deck, characterized in that, The improved dream optimization algorithm is used for parallel cable layout on FPSO decks, including the following steps: S1. Initialize basic parameters, set parameters related to the improved dream optimization algorithm, and construct the layout space; S2. Initialize the population size of the dream optimization algorithm using a hybrid multi-strategy approach, generate solutions, and obtain paths; S2 includes the following sub-steps: S2-1. Store the path data as a 1×4Dim matrix, where the data in the first 2Dim dimensions correspond to the path points of the first cable, and the data in the last 2Dim dimensions correspond to the path points of its parallel cables. S2-2, Using a hybrid multi-strategy approach, the individuals in the dream optimization algorithm are proportionally assigned to three strategies to generate solutions; S2-2 includes the following sub-steps: S2-2-1. Using 30% of the population as an A* heuristic, move point by point from the starting point of the cable towards the ending point to generate a solution: S2-2-2, The solution is obtained by linear interpolation of 40% of the population between the cable start point and the cable end point; S2-2-3. After the remaining 30% of the population individuals are used to generate a solution using linear interpolation, random perturbation is added to the solution obtained by linear interpolation. S2-3. When generating solutions, first generate solutions in the first 2Dim dimensions, and then generate solutions in the second 2Dim dimensions by filtering non-obstacle points in the neighborhood of the solutions in the first 2Dim dimensions through an outer diffusion method. S3. Enter the exploration phase, update the solution, and determine if there is any interference in the path. If interference exists, correct it using breadth-first search to complete the path update. S3 includes the following sub-steps: S3-1. The solution for the first 2Dim dimensions is updated using memory strategy, forgetting and supplementation strategy. The solution for the last 2Dim dimensions is generated by adding random perturbation to the solution for the first 2Dim dimensions. S3-2. The updated solution is updated using a repair strategy. The repair strategy includes: determining whether the path of the parallel cable interferes with the equipment; if interference occurs, obtaining the collision point between the interference segments; and using a breadth-first search strategy to expand the search layer by layer from the collision point until the updated collision point and the path formed by the two ends of the original interference segment do not interfere. S4. Enter the development phase, determine the globally optimal solution, update the solution again, and achieve path optimization; S4 includes the following sub-steps: S4-1. Calculate the parallel fitness of individuals in the population to obtain the global optimal solution; The formula for calculating parallel fitness in S4-1 is as follows: Fit= ; fit=w1·pathlength+w2·(collision+segment)+w3(smoothness+sharpturn); Where Fit is the fitness of parallel cables, fit is the fitness of a single cable, parallelism is the degree of parallelism, pathlength is the cable length, collision is the path point collision penalty, segment is the segment interference penalty, smoothness is the 90° smoothness penalty, sharpturn is the 60° smoothness penalty, and w1, w2, w3, and w4 are the weights of the corresponding targets. S4-2. Update the solution of the first 2Dim dimensions using the forgetting and supplementation strategy. The solution of the last 2Dim dimensions is obtained by adding random perturbations to the solution of the first 2Dim dimensions. If the solution of a certain dimension exceeds the boundary, the solution of that dimension is randomly generated. The solution is updated using the simplification strategy and the smoothing strategy. Introducing Douglas into the simplification strategy Puk algorithm simplifies paths; The smoothing strategy is to smooth the curve through a cubic β-spline. When smoothing other cables of the parallel cables, they will shift to the corresponding points of another cable path. S5. Based on the global optimal solution, determine the horizontal and vertical coordinates of the cable path points and output the final path result.

2. The FPSO deck parallel cable layout method according to claim 1, characterized in that, The S1 includes the following sub-steps: S1-1. Set the maximum number of iterations, population size, and number of path points of the improved dream optimization algorithm. S1-2. Construct the layout space. According to the FPSO deck area and the positions of each device on the FPSO deck, set the map size, cable start point, cable end point, and obstacle parameters of the layout space. The obstacle is a device model through which the cable cannot pass in the layout space, and then generate a grid map.

3. The FPSO deck parallel cable layout method according to claim 1, characterized in that, The calculation formula in S2-2-1 is as follows: next = current + dir × step; dir = goalpost - current; step = max(gridheight, gridwidth) / (Dim + 1); In the formula, next is the next path point, current is the current node, goalpos is the end point corresponding to the cable, gridheight is the number of rows of the grid map, gridwidth is the number of columns of the grid map, Dim is the solution dimension, dir is the direction vector, and step is the moving distance. The formula for obtaining the solution based on linear interpolation in S2-2-2 and S2-2-3 is as follows: next=startpos+ (goalpos-startpos); = (j-1) / (Dim+1), j=2: Dim+1; In the formula, startpos is the cable starting point. Here, j is the interpolation control variable, and j is the interpolation dimension. After adding random perturbations in S2-2-3, the calculation formula for the solution is: next = current + rand · min(gridheight, gridwidth) / 5; In the formula, rand is a random number with a value in [0, 1].

4. The FPSO deck parallel cable layout method according to claim 1, characterized in that, The simplification strategy in S4-2 is as follows: Virtually connect a straight line between the start and end points of the curve to be processed. Calculate the distances between all intermediate points and the straight line, and find the maximum distance value dmax. Compare dmax with the thinning threshold threshold: If dmax < threshold, all intermediate points on this curve are discarded; if dmax ≥ threshold, then take this point as the boundary, divide the curve into two parts, and repeat the above process for these two parts of the curve until all points are processed.

5. The FPSO deck parallel cable layout method according to claim 1, characterized in that, The smoothing calculation formula of the smoothing strategy in S4-2 is as follows: newpos=(1-a) 3 p0+3·(1-a) 2 ·a·p1+3·(1-a)·a·p2+a 3 ·p3; where newpos is the path point after smoothing, a is the control parameter, a = 0.5, and p0, p1, p2, p3 are four consecutive points on the original cable path.

Images