Multi-pipeline layout method based on niche co-evolution algorithm

Abstract

This invention provides a multi-pipeline layout method based on a niche co-evolutionary algorithm, belonging to the technical field of pipeline layout methods. It includes the following steps: A1: Establishing an environmental model of the pipeline layout space; A2: Applying the niche co-evolutionary algorithm to solve the multi-pipeline layout problem; A3: When the niche co-evolutionary algorithm meets the termination condition, outputting the optimal global solution as the final multi-pipeline layout scheme. Based on this, this invention introduces a niche layout pattern, an elite collaborator selection strategy, a multi-dimensional information sharing mechanism, and an adaptive evolutionary mechanism, overcoming the shortcomings of insufficient information interaction and premature convergence in existing co-evolutionary algorithms. It significantly improves the global search capability and solution quality for solving complex combinatorial optimization problems, thereby solving the problems of simple collaborator selection mechanisms, single and delayed shared information dimensions, and combinatorial explosion inherent in existing co-evolutionary algorithms applied to multi-pipeline layout.

Description

Technical Field

[0001] This invention relates to the field of pipeline layout methods, specifically a multi-pipeline layout method based on a niche co-evolutionary algorithm. Background Technology

[0002] In the design of large and complex systems such as floating production storage and offloading (FPSO) units, large ships, chemical plants, and aerospace vehicles, pipeline layout is a critical and highly challenging task. In particular, the multi-pipeline layout problem requires finding one or more optimal layout schemes while meeting multiple constraints such as process flow, safety regulations, operation and maintenance space, and equipment interference. The optimization objectives usually include the shortest total length, the fewest number of bends, the lowest construction cost, and the lowest pressure loss.

[0003] Traditional pipeline layout methods rely heavily on the experience of design engineers for manual or semi-automatic design, resulting in problems such as high workload, long design cycles, low automation, and difficulty in guaranteeing the global optimality of the solution. With the development of computer technology, various intelligent optimization algorithms have been applied to path planning, such as ant colony optimization and genetic algorithms. However, while these algorithms are effective for single-pipeline layout problems, when directly applied to multi-pipeline layout problems, they often ignore the cooperative relationships between pipelines because they typically adopt a sequential strategy of laying out each pipeline one by one, treating already laid pipelines as static obstacles before planning the next pipeline. This can easily lead to getting trapped in local optima due to improper layout order, making it difficult to obtain a high-quality global solution.

[0004] To address the aforementioned issues, a cooperative co-evolutionary algorithm has been introduced in this field to implement multi-pipeline layout. Specifically, the co-evolutionary algorithm is an optimization algorithm that simulates the evolutionary process of organisms in natural ecosystems. Its core idea is to decompose complex system problems into simpler subsystem problems through decomposition and coordination, that is, to decompose the multi-pipeline problem into multiple single-pipeline subproblems, and then seek the global optimal solution through co-evolution among populations. The cooperative relationship is a specific type of co-evolutionary algorithm, which mainly simulates the cooperative, synergistic, and mutually beneficial symbiotic relationship among populations. The solution steps include system problem decomposition, subproblem solving, and subproblem solution merging. However, existing co-evolutionary algorithms still have the following shortcomings when applied to multi-pipeline layout:

[0005] First, the collaborator selection mechanism is simplistic: it typically selects only a single representative or the best individual from other populations as the evolutionary environment for the current population. This limits the search space, resulting in insufficient pipeline traversal and poor interaction between pipelines, easily missing potentially better combinations. Second, the shared information dimension is singular and delayed: information interaction between populations is mostly limited to fitness value evaluation, and information sharing occurs after a complete iterative search, exhibiting a delay and leading to low efficiency in inter-population collaboration. Third, it is prone to combinatorial explosion: as the number of pipelines increases, the search space expands, and the possible combinations grow exponentially. Existing algorithms struggle to effectively explore the vast search space, reducing optimization efficiency and easily getting trapped in local optima.

[0006] Therefore, there is an urgent need for an optimization method that can overcome the above-mentioned defects, effectively handle the complex coordination relationships between pipelines, and efficiently and effectively solve multi-pipeline layout problems. Summary of the Invention

[0007] The purpose of this invention is to provide a multi-pipeline layout method based on a niche co-evolutionary algorithm, in order to solve the problems mentioned in the background art, such as the simple collaborator selection mechanism, the single and lagging shared information dimension, and the tendency to generate combinatorial explosion when existing co-evolutionary algorithms are applied to multi-pipeline layout.

[0008] This invention is achieved using the following technical solution:

[0009] A multi-pipeline layout method based on a niche co-evolutionary algorithm includes the following steps:

[0010] A1: Establish an environmental model of the pipeline layout space;

[0011] A2: Solve the multi-pipeline layout problem by applying the niche co-evolution algorithm;

[0012] A3: When the niche co-evolutionary algorithm meets the termination condition, it outputs the optimal global solution as the final multi-pipeline layout scheme.

[0013] The niche co-evolutionary algorithm includes the following steps:

[0014] S1: To address system issues, construct a niche layout;

[0015] S2: Conduct niche evolution, and implement elite collaborator selection strategy, multidimensional information sharing mechanism and adaptive evolution mechanism during the evolution process;

[0016] S3: Repeat S2 until the termination condition is met.

[0017] Furthermore:

[0018] S1 includes the following sub-steps:

[0019] S1-1: Decompose the system problem into multiple subproblems and create a corresponding population for each subproblem;

[0020] S1-2: Construct a niche for at least one representative combination of collaborative solutions to subproblems;

[0021] S2 includes the following sub-steps:

[0022] S2-1: Within the small population, evolution proceeds according to the adaptive evolution mechanism; for the currently evolving population, collaborators are selected based on the elite collaborator selection strategy;

[0023] S2-2: Constructing a co-evolutionary environment for the current population through a multi-dimensional information sharing mechanism;

[0024] S2-3: Update the global solution to the system problem based on the evolution of the niche.

[0025] Furthermore, in the microhabitat layout, multiple microhabitats are constructed, and all microhabitats evolve in parallel, jointly forming an ecosystem; the global solution to the system problem is the solution to the ecosystem; the elite collaborator selection strategy includes: selecting elite collaborator individuals from other populations located in the same microhabitat as the current evolutionary population, consisting of at least one optimal solution individual and / or a second-best solution individual; the multidimensional information sharing mechanism includes: constructing a co-evolutionary environment for the current population based on the elite collaborator individuals; specifically including fitness evaluation sharing and at least one type of environmental information sharing used to guide subsequent search processes.

[0026] Furthermore, the adaptive evolution mechanism includes a population evolution termination condition, namely, setting an evolutionary stagnation rule as one of the algorithm's termination conditions; the evolutionary stagnation rule refers to terminating population evolution if no evolution occurs within a specified number of iterations; specifically, it includes population evolutionary stagnation rules, niche evolutionary stagnation rules, and ecosystem evolutionary stagnation rules.

[0027] Furthermore, the fitness evaluation sharing includes: hierarchical fitness evaluation of individuals, populations, microhabitats, and ecosystems; wherein, the fitness of higher-level individuals is a comprehensive evaluation based on the fitness of the lower-level individuals it contains and their synergistic effects.

[0028] Furthermore, the environmental information sharing includes: modifying the quantization attributes of candidate solution elements in the solution space to create regions that have an attractive or repulsive effect on subsequent population searches, thereby guiding the search direction of subsequent populations; the quantization attributes include energy values ​​and / or pheromone concentrations.

[0029] Furthermore, in step A2, if the type of multi-pipeline layout problem is a parallel pipeline problem, then A2 specifically includes the following sub-steps:

[0030] A2-a1: Decompose the parallel pipeline problem into multiple single-pipeline subproblems. Each parallel pipeline is a single-pipeline subproblem, and each single-pipeline subproblem corresponds to a population. The populations evolve in a random order.

[0031] A2-a2: In the process of evolution, the pipeline solutions of the current population are regarded as obstacles, and through the environmental information sharing in the multidimensional information sharing mechanism, the nodes around the pipeline solutions are set as advantageous regions with an attractive effect, so as to guide the next population to seek optimization in the advantageous regions.

[0032] A2-a3: Solve multiple single-pipeline sub-problems step by step, complete the layout of the last parallel pipeline, and obtain the current optimal solution for the ecosystem;

[0033] A2-a4: Compare the current optimal solution with the global optimal solution. If the current optimal solution is better than the global optimal solution, then replace it; otherwise, leave it unchanged.

[0034] Furthermore, in step A2, if the type of multi-pipe layout problem is a branch pipe problem, then A2 specifically includes the following sub-steps:

[0035] A2-b1: Analyze the layout characteristics of branch pipeline problems, and classify branch pipeline problems according to whether the laying sequence and primary-secondary relationship need to be considered. The types include fixed branch pipeline problems that need to be laid according to primary-secondary and primary order and floating branch pipeline problems that do not need to be laid according to primary-secondary and primary order.

[0036] A2-b2: Different solution steps are performed depending on the type of branch pipeline problem;

[0037] If the branch pipe problem is a floating branch pipe problem, then A2-b2 includes the following sub-steps:

[0038] A2-b2-f1: Decompose the branch pipeline problem into multiple single-pipeline subproblems, that is, break down the branch pipeline into multiple single pipelines and determine the start and end positions of each single pipeline; each single-pipeline subproblem corresponds to a population, and the various populations evolve in a random order;

[0039] A2-b2-f2: During the evolutionary process, through the sharing of environmental information in the multidimensional information sharing mechanism, the current population's pipeline solution and surrounding nodes are set as advantageous regions with an attractive effect;

[0040] A2-b2-f3: Solve multiple single-pipeline subproblems step by step, complete the layout of the last single pipeline, and obtain the current optimal solution for the ecosystem;

[0041] A2-b2-f4: Compare the current optimal solution with the global optimal solution. If the current optimal solution is better than the global optimal solution, then replace it; otherwise, leave it unchanged.

[0042] Furthermore, in step A2, if the type of multi-pipe layout problem is a mixed-pipe problem, then A2 specifically includes the following sub-steps:

[0043] A2-c1: Classify and label the pipelines to be laid out, and determine the relationship between the pipelines through the synergy factor;

[0044] A2-c2: Assign importance to each pipeline and determine the pipeline layout order based on the importance;

[0045] A2-c3: Based on the classification labels of each pipeline, the corresponding solution method is called to perform the layout, that is:

[0046] If the classification label shows a parallel pipeline, then call steps A2-a1 to A2-a4;

[0047] If the classification label shows a branch pipeline, then call steps A2-b1 to A2-b2.

[0048] Furthermore, the objective function for parallel pipeline layout is:

[0049] ,

[0050] In the formula, for i The total length of a pipeline when it is broken down into individual pipelines; This refers to the length of the parallel section of the pipeline; for i The total number of bends when a pipeline is split into individual pipelines; This represents the number of elbows in the parallel section of the pipeline. for i The total energy value of the pipeline; For path length weights, Weights for path bends, The path energy value weight; This is the shortest total length; This represents the minimum total number of bends; This represents the maximum energy value of the pipeline node.

[0051] In the above multi-pipeline layout method based on niche co-evolutionary algorithm:

[0052] By introducing niche layout patterns and adaptive evolutionary mechanisms, complex system problems can be divided into multiple parallel evolutionary environments with different cooperative methods, forming a complete solution system that includes individual solutions, population solutions, niche solutions, and ecosystem solutions. This avoids blind exploration in a single huge search space, preserves population diversity, effectively avoids combinatorial explosion, and enables a more comprehensive exploration of the solution space and the discovery of higher-quality global optimal solutions.

[0053] By adopting an elite collaborator selection strategy, an elite set including the best and second-best individuals can be selected as the representative. This avoids the problem of prematurely falling into local optima caused by traditional methods that only select a single best individual. It provides a more diverse environment and richer cooperation possibilities for the subsequent evolution of the population, and significantly increases the probability of escaping local optima and finding better combination solutions.

[0054] By establishing a multi-dimensional information sharing mechanism, environmental information such as energy value and pheromone is introduced for sharing. The combination of environmental information sharing and fitness evaluation enables real-time and multi-level guidance of population coordination during the search process, thereby significantly improving coordination efficiency and layout quality (such as pipeline bundles).

[0055] In addition, this invention can unify different types of multi-pipeline layout problems (parallel, branching, and hybrid) into a solution framework for a set of single-pipeline subproblems. Through the flexible setting of synergy factors and information sharing mechanisms, it achieves a universal solution for various complex pipeline layout requirements. The method has strong versatility, and the solution quality and efficiency are significantly improved.

[0056] The beneficial effects achieved by this invention are:

[0057] A multi-pipeline layout method based on a niche co-evolutionary algorithm is proposed. This method introduces niche layout patterns, elite collaborator selection strategies, multi-dimensional information sharing mechanisms, and adaptive evolutionary mechanisms, overcoming the shortcomings of insufficient information interaction and premature convergence in existing co-evolutionary algorithms. It significantly improves the global search capability and solution quality for solving complex combinatorial optimization problems. Based on this, an automated, efficient, and universal engineering solution is provided, capable of solving different types of complex pipeline layout problems within a unified framework. The resulting layout scheme outperforms existing technologies in terms of pipeline length, number of bends, and bundle formation, while also effectively shortening the design cycle and reducing design costs. Attached Figure Description

[0058] Figure 1 This is a schematic diagram of a niche layout pattern;

[0059] Figure 2 This is a schematic diagram of the ecosystem solution architecture;

[0060] Figure 3This is a schematic diagram of the pipeline layout effect under the optimal individual method;

[0061] Figure 4 This is a schematic diagram illustrating the pipeline layout effect under the elite collaborator selection strategy;

[0062] Figure 5 This is a schematic diagram illustrating the solution method for parallel pipeline problems;

[0063] Figure 6 This is a schematic diagram illustrating the solution process and results of a fixed branch pipeline problem;

[0064] Figure 7 This is a schematic diagram illustrating the solution process and results of the floating branch pipeline problem;

[0065] Figure 8 This is a schematic diagram illustrating the solution method for mixed piping problems;

[0066] Figure 9 This is a schematic diagram of the optimal path obtained by the IOHA algorithm (in Simulation Experiment I of the Parallel Pipeline Problem example).

[0067] Figure 10 This is a schematic diagram of the optimal path obtained by the NCEA algorithm (in Simulation Experiment I of the Parallel Pipeline Problem example).

[0068] Figure 11 This is a schematic diagram of the optimal path obtained by the MACCCA-1 algorithm (in Simulation Experiment II of the Parallel Pipeline Problem Example).

[0069] Figure 12 This is a schematic diagram of the optimal path obtained by the MACCCA-2 algorithm (in Simulation Experiment II of the Parallel Pipeline Problem Example).

[0070] Figure 13 This is a schematic diagram of the optimal path obtained by the ACPRS algorithm (in Simulation Experiment II of the Parallel Pipeline Problem Example).

[0071] Figure 14 This is a schematic diagram of the optimal path obtained by the NCEA algorithm (in Simulation Experiment II of the Parallel Pipeline Problem example).

[0072] Figure 15 This is a schematic diagram of the optimal path obtained by the IOHA algorithm (in Simulation Experiment I of the Branch Pipeline Problem example);

[0073] Figure 16 This is a schematic diagram of the optimal path obtained by the NCEA algorithm (in Simulation Experiment I of the Branch Pipeline Problem example);

[0074] Figure 17 This is a schematic diagram of the optimal path obtained by the ACO-CEA algorithm (in Simulation Experiment II of the Branch Pipeline Problem Example).

[0075] Figure 18 This is a schematic diagram of the optimal path obtained by the GA-CEA algorithm (in Simulation Experiment II of the Branch Pipeline Problem Example).

[0076] Figure 19 This is a schematic diagram of the optimal path obtained by the NCEA algorithm (in Simulation Experiment II of the Branch Pipeline Problem example);

[0077] Figure 20 This is a schematic diagram of the optimal path obtained by the IOHA algorithm (in Simulation Experiment I of the Hybrid Pipeline Problem example);

[0078] Figure 21 This is a schematic diagram of the optimal path obtained by the NCEA algorithm (in Simulation Experiment I of the Hybrid Piping Problem example);

[0079] Figure 22 This is a schematic diagram of the optimal path obtained by the NCEA algorithm (in the simulation experiment II of the mixed pipeline problem example).

[0080] Figure 23 This is a second-view diagram of the optimal path obtained by the NCEA algorithm (in the simulation experiment II of the mixed pipeline problem example). Detailed Implementation

[0081] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0082] Example 1

[0083] This embodiment provides a niche co-evolutionary algorithm, comprising the following steps:

[0084] S1: To address system issues, construct a niche layout. Specifically:

[0085] A co-evolutionary algorithm is employed, introducing niches to provide an environment for information exchange, and using niche evolution to solve the overall problem. This includes the following sub-steps:

[0086] S1-1: Decompose the system problem into multiple subproblems and create a corresponding population for each subproblem. Collaborators in the current population influence the next population. For example, decompose a multi-pipe layout into multiple single-pipe layouts, with each pipe corresponding to a population. The already laid-out pipes are used as constraints for laying out other pipes.

[0087] S1-2: Please refer to Figure 1 Multiple subproblems can be combined in various ways. Based on this, we can construct a niche for collaborative solutions of multiple representative subproblems.

[0088] All niches evolve in parallel, collectively forming an ecosystem; please refer to... Figure 2 The global solution to a system problem is the solution to an ecosystem problem. Individual solutions, population solutions, microhabitat solutions, and ecosystem solutions constitute a complete solution system.

[0089] S2: It involves niche evolution, employing elite collaborator selection strategies, multidimensional information sharing mechanisms, and adaptive evolutionary mechanisms during the evolutionary process. Specifically:

[0090] S2 includes the following sub-steps:

[0091] S2-1: Within the small population, evolution proceeds according to the adaptive evolution mechanism; for the currently evolving population, collaborators are selected based on the elite collaborator selection strategy.

[0092] Regarding the strategy for selecting elite collaborators:

[0093] Within a niche population, cooperation primarily occurs through interactions between individual representatives. Representative selection includes random selection, selection of individuals with average fitness, and selection of the optimal individual. Random selection suffers from poor representativeness and overall poor solution performance. While selecting individuals with average fitness offers better solution distribution, its overall effectiveness is only average. The optimal individual selection approach tends towards a greedy strategy, only guaranteeing the satisfaction of the individual's optimal solution.

[0094] This embodiment proposes an elite collaborator selection strategy, which includes: selecting elite collaborator representatives from other populations located in the same small habitat as the current evolutionary population, consisting of at least one optimal solution individual and / or a second-best solution individual, to improve the effectiveness of the influence of the individual representatives and facilitate the selection of a better fitness solution as the overall optimal solution.

[0095] Please refer to Figure 3 and Figure 4 (Where S1 and S2 are start point 1 and start point 2, and T1 and T2 are end point 1 and end point 2), respectively, showing the pipeline layout effect under the optimal individual approach and the elite collaborator selection strategy. The pipeline layout effect under the elite collaborator selection strategy is significantly better than that under the optimal individual approach.

[0096] Regarding the adaptive evolution mechanism:

[0097] Adaptive evolutionary mechanisms include the population's evolutionary pattern, evolutionary order, and evolutionary termination conditions. Specifically:

[0098] The evolutionary pattern of a population includes four levels: individual evolution, population evolution, microhabitat evolution, and ecosystem evolution. Each level progresses in the order described above. Individual evolution is mainly reflected in the survival of the fittest among individuals through certain evolutionary rules, preserving individuals with better fitness. Elite collaborators are selected as representatives of the population, and different populations complete population evolution through information sharing mechanisms. Different microhabitats complete evolution, producing a better ecosystem set, which in turn promotes the evolution of the ecosystem.

[0099] The evolutionary order of the population adopts a nested cyclical approach to ensure the diversity of ecosystem solutions; the nested cyclical approach refers to the various populations taking turns as the dominant pathway, realizing the evolution of the microhabitat in a progressive manner.

[0100] The termination conditions for population evolution include evolutionary stagnation rules. Besides stopping the search when the algorithm reaches the maximum number of iterations K, evolutionary stagnation rules can serve as another termination condition adapted to niche algorithms. An evolutionary stagnation rule states that if no evolution occurs within a specified number of iterations, the population evolution terminates, and the globally optimal solution is output. Evolutionary stagnation rules include population evolutionary stagnation rules, niche evolutionary stagnation rules, and ecosystem evolutionary stagnation rules, as detailed below:

[0101] Population evolution stagnation rule: If the optimal solution remains unchanged when the number of iterations before evolution reaches a threshold, the population is considered to be stagnant, and the iterative search is terminated prematurely.

[0102] The niche evolution stagnation rule: various populations take turns acting as the dominant path to achieve niche evolution. When the number of iterations in a certain niche reaches a threshold, the population is considered to have stopped evolving, and the iterative search ends prematurely.

[0103] Ecosystem evolution stagnation rule: When all microhabitats have undergone a certain number of iterations and the fitness of the ecosystem no longer changes, the algorithm is considered to have stagnated and the pipeline layout result is output.

[0104] S2-2: Construct a co-evolutionary environment for the current population through a multi-dimensional information sharing mechanism.

[0105] Regarding the multi-dimensional information sharing mechanism:

[0106] Cooperation among populations within a niche is achieved through information sharing. Traditional cooperative algorithms only consider adaptation evaluation as a guiding factor influencing the results of multi-pipeline searches. The information sharing process occurs after the search is completed, which has delays and limitations.

[0107] In this embodiment, a multi-dimensional information sharing mechanism is introduced, which can construct the co-evolutionary environment of the current population based on elite collaborator individual representatives. This multi-dimensional information sharing mechanism includes fitness evaluation sharing and at least one type of environmental information sharing used to guide subsequent search processes. Specifically:

[0108] Fitness assessment sharing: This involves hierarchical fitness assessment of individuals, populations, microhabitats, and ecosystems. A fitness sharing mechanism is added to the microhabitat. Based on a solution system including individual solutions, population solutions, microhabitat solutions, and ecosystem solutions, a fitness system is obtained comprising individual fitness, population fitness, microhabitat fitness, and ecosystem fitness. Fitness assessment sharing influences each other in a progressive manner, assigning higher fitness to individuals that are beneficial to population coordination and lower fitness to individuals that are detrimental to population coordination, thus guiding populations towards a direction conducive to the coordinated development of the ecosystem. Higher-level fitness is a comprehensive evaluation based on the fitness of its included lower-level levels and their synergistic effects, as follows:

[0109] Individual fitness evaluation: Based on the collaborator representatives and the current environment, the best fitness is selected as the elite collaborator representatives; the other population collaborator representatives do not receive rewards or penalties during the fitness evaluation process, and only provide the environment for the current population.

[0110] Population fitness evaluation: Selected elite collaborators are combined with individuals in the current population to form a cooperative group, and the fitness evaluation of the population is calculated; each population completes the shared fitness evaluation calculation in turn.

[0111] Microhabitat fitness: After all populations have completed the fitness evaluation sharing in sequence, the fitness of the current multi-pipeline assemblage is calculated; then the fitness evaluation sharing calculation of other microhabitats is completed.

[0112] Ecosystem fitness: After all microhabitats have completed iterative evolution, the optimal pipeline layout is obtained by solving the problem.

[0113] Environmental information sharing: By modifying the quantization attributes of candidate solution elements in the solution space, regions with attractive or repulsive effects on subsequent population searches are created to guide the search direction of subsequent populations. The quantization attributes include energy values ​​and / or pheromone concentrations. Wherein:

[0114] Energy value sharing: For example, in the gridded space of pipeline layout, energy values ​​assign new attribute information to grid nodes. Grid nodes around general obstacles and parallel pipelines are designated as high-energy / dominant areas, and areas with high energy values ​​should be prioritized for layout. The energy value sharing mechanism treats the current population collaborator representative as an obstacle, sets the energy information of the grid nodes around it as high-energy areas, and superimposes the high-energy areas around the obstacle to form a new pipeline layout space with a new energy value distribution.

[0115] Pheromones sharing: For example, in the gridded space of pipeline layout, combined with the unique advantages of ant colony algorithm path optimization, pheromones are set as guiding and coordinating factors between multiple pipelines; elite ant individuals in the current population are selected as representatives, and then the grid nodes covered by the path information are given a high pheromone concentration as the initial condition for the next population path search. This can maximize the parallel layout of pipelines and avoid blind search in the early stage of computation.

[0116] S2-3: Update the global solution to the system problem based on the evolution of the niche.

[0117] Specifically, each niche will generate a combinatorial solution after evolution; the solutions of all niches are compared, and the optimal solution is used as the global optimal solution to update the ecosystem.

[0118] S3: Repeat S2 until the termination condition is met, and finally output the global optimal solution.

[0119] The second aspect of this embodiment provides a multi-pipeline layout method based on a niche co-evolutionary algorithm, which, when applied, includes the following steps:

[0120] A1: Establish an environmental model of the pipeline layout space; the pipeline layout area can be processed into a three-dimensional raster to establish an environmental model containing obstacle information.

[0121] A2: Solve the multi-pipeline layout problem by applying the niche co-evolution algorithm.

[0122] A3: Perform a loop search. When the niche co-evolutionary algorithm meets the termination condition, output the optimal global solution as the final multi-pipeline layout scheme.

[0123] Specifically, in step A2:

[0124] In multi-pipe layout problems, single-pipe layout is a sub-problem, where each pipe is both independent and interconnected. Independence means that each pipe should find the optimal layout under the constraints. Interconnection means that interference phenomena need to be considered during the layout process, as well as the coordination relationship between each pipe. Pipes should be bundled as much as possible to reduce the number of pipe supports and the space occupied by the pipes.

[0125] Therefore, in this step, the multi-pipe layout problem is regarded as a system problem in the niche co-evolution algorithm. Based on the type of the multi-pipe layout problem, it is decomposed into multiple single-pipe layout subproblems. Each single-pipe layout subproblem corresponds to a population, and the individuals in the population represent a candidate path of the single-pipe layout. The individual solution is the path solution of that pipe. The superposition of various populations forms an ecosystem, which can form a complete multi-pipe solution system to obtain the optimal layout scheme of the multi-pipe system.

[0126] The types of multi-pipe layout problems include parallel piping problems, branching piping problems, and mixed piping problems; multi-pipe layout should, under the condition of meeting the constraints of actual engineering rules, bundle different types of pipes as much as possible. Specifically:

[0127] Parallel multi-pipeline systems are multi-starting-point and multi-ending-point pipeline systems. The specific pipeline types are generally different. The key to the layout is how to guide different pipelines to be laid in bundles to meet the actual rules and constraints of the project.

[0128] The main difference between branch pipelines and parallel pipelines is that pipelines can overlap. Branch pipeline layout is a single-start-point to multiple-end-point layout problem, usually referring to the coordinated layout of pipelines of the same type, with the aim of maximizing path overlap. For branch pipeline layout problems, the branch points and pipeline laying sequence must be determined under the constraints of actual engineering rules. Due to the coupling and intersection between pipelines, the order of pipeline layout will have a significant impact on the overall layout result. Furthermore, when there are many branch points, determining their locations is quite difficult.

[0129] Mixed piping systems are complex systems containing multiple types of pipes. In mixed piping layouts, the intricate combinations of pipes mean that the order in which they are laid out significantly impacts the overall result. In practical engineering applications, mixed piping layouts must consider the priority order of pipe placement, such as laying larger pipes first, then smaller pipes; laying important equipment and special pipelines first, then conventional equipment pipelines; and prioritizing pipelines with high maintenance frequency.

[0130] Specifically:

[0131] If the multi-pipe layout problem is a parallel pipe problem, please refer to... Figure 5 A2 includes the following sub-steps:

[0132] A2-a1: The parallel pipeline problem is decomposed into multiple single-pipeline subproblems. Each parallel pipeline is a single-pipeline subproblem, and each single-pipeline subproblem corresponds to a population. The populations evolve in a random order to avoid getting trapped in local optima due to traditional fixed-order evolution.

[0133] For example, suppose that in the Nth iteration, the population that evolves preferentially... i That is, the firsti Optimal layout solution for pipeline path - i - N {}.

[0134] A2-a2: Parallel pipelines cannot interfere with each other. Therefore, during the evolutionary process, the pipeline solutions of the current population are regarded as obstacles. Through environmental information sharing in the multidimensional information sharing mechanism, the grid nodes around the pipeline solutions are set as advantageous regions with an attractive effect, that is, they are allocated higher energy values ​​and given higher pheromone concentrations to guide the next population to seek optimization in the advantageous regions.

[0135] For example, population i Elite ant individuals from the group join the next colony. j In the evolutionary environment; population j The evolutionary environment is determined by the population. i The resulting dominant habitat and its initial environment form a microhabitat for the population. j A current optimal solution is generated based on the niche. path - j - N {}, then the current solution for the ecosystem is path - ij - N {}.

[0136] A2-a3: Following the same steps, solve multiple single-pipeline sub-problems step by step to complete the layout of the last parallel pipeline and obtain the current optimal solution for the ecosystem.

[0137] A2-a4: Compare the current optimal solution with the global optimal solution. If the current optimal solution is better than the global optimal solution, then replace it; otherwise, leave it unchanged.

[0138] Parallel pipelines, based on single-pipeline systems, consider the economy of multiple pipelines and the coordination relationships between them. Therefore, in the layout of parallel pipelines, it is necessary to lay pipelines in bundles as much as possible to reduce the number of pipe supports, save space, and require a short total length, few bends, and many parallel paths. The objective function for parallel pipeline layout is normalized and transformed using the exponential transformation method. The objective function is to find its maximum value, expressed as follows:

[0139] ,

[0140] In the formula, for i The total length of a pipeline when it is broken down into individual pipelines; This refers to the length of the parallel section of the pipeline; for i The total number of bends when a pipeline is split into individual pipelines; This represents the number of elbows in the parallel section of the pipeline. for i The total energy value of the pipeline; For path length weights, Weights for path bends, The path energy value weight; This is the shortest total length; This represents the minimum total number of bends; This represents the maximum energy value of the pipeline node.

[0141] If the multi-pipe layout problem is a branch pipe problem, then A2 includes the following sub-steps:

[0142] A2-b1: Analyze the layout characteristics of branch pipeline problems, and classify branch pipeline problems according to whether the laying sequence and primary-secondary relationship need to be considered. The types include fixed branch pipeline problems, which require laying pipelines according to primary and secondary order, and floating branch pipeline problems, which do not require laying pipelines according to primary and secondary order.

[0143] A2-b2: Different solution steps are performed depending on the type of branch pipeline problem. Where:

[0144] If the branch piping problem is a fixed branch piping problem, then the layout order of the main and secondary piping needs to be considered. The main piping should be laid first, and then the branch piping should be coordinated and laid out. A2-b2 specifically includes the following sub-steps:

[0145] A2-b2-g1: Determine the main and branch pipelines of a multi-branch pipeline system, and find the start and end points of the main and branch pipelines;

[0146] A2-b2-g2: Layout the main pipeline, then take this pipeline as the target pipeline, and take the endpoints of other branch pipelines as the search starting point to optimize each branch pipeline; the intersection of the branch pipeline and the main pipeline is the pipeline branch point, thus obtaining a set of feasible solutions;

[0147] A2-b2-g3: Performs a cooperative iterative search. When the stopping condition is reached, the search ends and the optimal pipeline layout is output.

[0148] Please refer to Figure 6 This illustrates the specific solution process and results of the fixed branch pipeline problem: First, the pipeline types are marked. Assuming that the starting point S and the ending point T2 are the main pipelines, the other two are branch pipelines. Then, the main pipeline is laid out, and this pipeline is taken as the target pipeline. The ending points T1 and T3 of the other branch pipelines are taken as the search starting points. The algorithm is used to find the branch paths from the ending points T1 and T3 to the main pipeline. The intersection point is the branch point of the pipeline. Finally, the pipeline layout result containing the branch points is output.

[0149] If the branching pipeline problem is a floating branching pipeline problem, a decomposition strategy can be used to break down the branching pipelines into independent pipelines and transform it into a special parallel pipeline problem. This breaks the fixed pattern of determining the pipeline laying order before determining the branch points. During the solution process, a niche co-evolutionary algorithm is used, representing each branching pipeline with a population. The populations influence each other to complete the branching pipeline layout. A2-b2 specifically includes the following sub-steps:

[0150] A2-b2-f1: Decompose the branch pipeline problem into multiple single-pipeline subproblems, that is, break down the branch pipeline into multiple single pipelines and determine the start and end positions of each single pipeline; each single-pipeline subproblem corresponds to a population, and the various populations evolve in a random order;

[0151] A2-b2-f2: During the evolution process, considering the need for the branches to overlap as much as possible, and referring to the specific measures of parallel multi-pipeline layout, the optimal solution of the pipeline in the current population and the surrounding grid nodes are set as the dominant area with an attractive effect through the sharing of environmental information in the multi-dimensional information sharing mechanism.

[0152] A2-b2-f3: And so on, solve multiple single-pipeline sub-problems step by step, complete the layout of the last single pipeline, and obtain the current optimal solution for the ecosystem;

[0153] A2-b2-f4: Compare the current optimal solution with the global optimal solution. If the current optimal solution is better than the global optimal solution, then replace it; otherwise, leave it unchanged.

[0154] Please refer to Figure 7 This illustrates the specific solution process and results of the floating branch pipeline problem: First, a decomposition strategy is adopted, and three independent single pipelines are split according to the starting point S and the ending points T1, T2, and T3. Each pipeline is represented by a population. Then, the niche co-evolutionary algorithm is used for pipeline layout. Under the premise of ensuring short path distances and few bends, the paths between pipelines are made as overlapping as possible. Until the termination condition of the iterative search is met, the optimal pipeline layout is output.

[0155] The objective function for floating branch pipelines is the same as that for parallel pipelines, and its expression is as follows:

[0156] ;

[0157] In the formula, for i The total length of a pipeline when it is broken down into individual pipelines; This refers to the length of the overlapping section of the pipeline. for i The total number of bends when a pipeline is split into individual pipelines; This represents the number of elbows in the overlapping sections of the pipeline. for i The total energy value of the pipeline; For path length weights, Weights for path bends, The path energy value weight; This is the shortest total length; This represents the minimum total number of bends; This represents the maximum energy value of the pipeline node.

[0158] If the multi-pipe layout problem is a mixed-pipe problem, please refer to... Figure 8 A2 includes the following sub-steps:

[0159] A2-c1: Classify and label the pipelines to be laid out. Determine the relationships between pipelines using a synergy factor (whether they need to be close to or far apart), and determine the unit module layer position for the pipeline layout based on this. Wherein:

[0160] Classification-based labeling enables standardized pipeline layout. Specifically, complex coupling and intersections exist between mixed pipelines, necessitating the determination of pipeline laying order to guide the bundled laying of pipelines of the same type. To unify the issues of single pipelines, branch pipelines, and parallel pipelines, this embodiment proposes a digital identifier for mixed pipeline paths, expressed as follows:

[0161] ,

[0162] ,

[0163] ,

[0164] ;

[0165] In the formula, P is the set of pipe type labels, Pipe is a pipe, single is a single pipe type label, multi is a parallel pipe type label, and branch is a branch pipe type label.

[0166] Introducing a synergy factor can optimize the heuristic function. Specifically, the hybrid pipeline layout must comply with engineering constraints, considering the coordination relationships between multiple pipeline layouts. For special pipelines that do not require bundled laying, such as high-temperature oil pipes, the concept of a synergy factor can be introduced, which specifically includes an attraction factor and a repulsion factor. The synergy factor influences the multi-pipeline layout by setting a heuristic search mechanism and energy value. The attraction factor can satisfy the requirement of bundled laying between pipelines, while the repulsion factor can satisfy the special laying requirements of pipelines that need to be laid far apart from the main layout.

[0167] A2-c2: Assign importance to each pipeline and determine the pipeline layout order based on the importance.

[0168] In mixed pipeline layouts, the order of pipeline layout has a significant impact on the overall layout result. Therefore, this embodiment proposes a pipeline layout order based on importance allocation, using the analytic hierarchy process (AHP) and other methods to comprehensively consider multiple evaluation factors for importance allocation. Pipelines with high importance are prioritized for layout, while general pipelines are laid out using a nested evolutionary approach.

[0169] A2-c3: Based on the classification labels of each pipeline, the corresponding solution method is called to perform the layout, that is:

[0170] If the classification label indicates a parallel pipeline, steps A2-a1 to A2-a4 are called. Considering the synergy factor and multidimensional information sharing mechanism, when processing parallel pipelines, the optimal solution path of the previous pipeline is treated as an obstacle, and the surrounding grid is set as a dominant region to attract the next pipeline to be bundled as close as possible.

[0171] If the classification label shows a branch pipeline, then steps A2-b1 to A2-b2 are called. Here, the optimal layout result is selected by comparing the fixed and floating cases; when processing branch pipelines, path decomposition is performed, and the optimal solution path of the previous pipeline is used as the dominant region to attract the next pipeline to overlap as much as possible.

[0172] To verify the superior performance of the aforementioned niche co-evolutionary algorithm in solving multi-pipeline layout problems, a series of experimental tests were conducted in a three-dimensional model environment, comparing the niche co-evolutionary algorithm (NCEA) with other improved algorithms. For ease of understanding and correlation, the comparison algorithms involved in this embodiment are listed in Table 1.

[0173] Table 1: Comparison Algorithms and Their Abbreviations

[0174]

[0175] The specific experimental test calculations are as follows:

[0176] 1. Example of a parallel pipeline problem

[0177] (1) Simulation Experiment I

[0178] This section uses simulation experiments to test the performance of the NCEA algorithm in solving parallel pipeline problems, and compares it with the IOHA algorithm without the niche co-evolution concept. Specifically:

[0179] Both algorithms were set up with the same 3D simulation environment model (size 50×50×30). Parallel pipes 1, 2, and 3 were laid out in the model. The coordinates of each point in pipe 1 are: start point S1(2, 36,11), end point T1(40,28,24); the coordinates of each point in pipe 2 are: start point S2(2,24,1), end point T2(49,24,24); the coordinates of each point in pipe 3 are: start point S3(2,16,11), end point T3(40,22,24). The common parameters of both algorithms were set to be the same. Each algorithm was run independently 20 times, and the optimal paths obtained are as follows: Figure 9 , Figure 10 As shown in Table 2, the results of the operation are as follows.

[0180] Table 2: Simulation results of IOHA and NCEA algorithms (in the parallel pipeline problem example)

[0181]

[0182] Depend on Figure 9 , Figure 10 As shown, the sequential layout based on the IOHA algorithm yielded a feasible solution, but the dispersed arrangement of pipes hindered support sharing. The NCEA algorithm, however, enabled the coordinated layout of multiple pipes, with pipes bundled together, facilitating support sharing and significantly improving neatness, thus saving layout space. Therefore, this experiment effectively demonstrates that the niche co-evolutionary concept is applicable to parallel pipes and verifies the superiority of the NCEA algorithm in solving parallel pipe layout problems.

[0183] (1) Simulation Experiment II

[0184] To further verify the performance of the NCEA algorithm in solving parallel multi-pipeline problems, it is compared with three other improved algorithms. Specifically:

[0185] Set up a 3D spatial simulation environment model (50×50×30) identical to that used in Simulation Experiment I. Layout parallel pipelines 1, 2, and 3 within the model. The coordinates of each point in pipeline 1 are: start point S1(45,24,20), end point T1(4,36,10); the coordinates of each point in pipeline 2 are: start point S2(41,24,10), end point T2(4,10,5); the coordinates of each point in pipeline 3 are: start point S3(41,24,10), end point T3(4,39,5). Each algorithm is run independently 20 times, and the optimal paths obtained are as follows: Figure 11 , Figure 12 , Figure 13 , Figure 14 As shown in Table 3, the results of the operation are as follows.

[0186] Table 3: Simulation results of the NCEA algorithm and three other comparative algorithms (in the parallel pipeline problem example)

[0187]

[0188] The results show that the NCEA algorithm achieves superior results, with significant improvements in pipe length and the number of bends. Compared with the other three improved algorithms, the NCEA algorithm improves pipe length performance by 11.66%, 8.37%, and 15.17%, respectively, and improves bend performance by 60.87%, 25%, and 25%, respectively. Figure 11 , Figure 12 , Figure 13 , Figure 14 As shown, the NCEA algorithm better realizes the collaborative layout of parallel pipelines, in which three pipelines are laid in a bundle, which facilitates the construction of pipeline supports, saves layout space, and meets the requirements of aesthetics.

[0189] 2. Example of a branch pipeline problem

[0190] (1) Simulation Experiment I

[0191] This section tests the performance of the NCEA algorithm in resolving branch pipelines and compares it with the IOHA algorithm. Specifically:

[0192] Both algorithms were run using the same 3D simulation environment model (20×20×20). Pipelines 1 and 2 were laid out within the model. The coordinates of each point in pipe 1 are: start point S1(19,1,1), end point T1(1,19,19); the coordinates of each point in pipe 2 are: start point S2(9,1,1), end point T2(19,7,15). The parameters for both algorithms were set identically. Each algorithm was run independently 20 times, and the optimal paths obtained were as follows: Figure 15 , Figure 16 As shown in Table 4, the results of the operation are as follows.

[0193] Table 4: Simulation results of NCEA and IOHA algorithms (in the branch pipeline problem example)

[0194]

[0195] As shown in Table 4, the pipeline layout obtained by the IOHA algorithm is relatively scattered and occupies a large space; when the NCEA algorithm is used for the same type of pipeline layout, the pipeline length and elbow performance are significantly improved, with the pipeline length performance improved by 24% and the elbow performance improved by 44.45%.

[0196] (2) Simulation Experiment II

[0197] This section's simulation experiments further test the performance of the NCEA algorithm in resolving branch pipelines, comparing it with two other improved algorithms. Specifically:

[0198] The same 3D simulation environment model (50×50×50) was set up for the three algorithms. The model contained a single pipeline start point and multiple pipeline target points. The coordinates of the pipeline start point S were (4,10,10), and the coordinates of the pipeline target points T1, T2, and T3 were (15,30,20), (32,38,15), and (45, 30,30), respectively. Each algorithm was run independently 20 times, and the optimal paths obtained were as follows: Figure 17 , Figure 18 , Figure 19 As shown in Table 5, the results of the operation are as follows.

[0199] Table 5: Simulation results of the NCEA algorithm and two other comparative algorithms (in the branch pipeline problem example)

[0200]

[0201] Table 5 shows that the ACO-CEA algorithm yielded a total pipe length of 126, 5 elbows, and 2 tees; the GA-CEA algorithm yielded a total pipe length of 132, 6 elbows, and 2 tees; and the proposed NCEA algorithm yielded a total pipe length of 106, 3 elbows, and 2 tees. Therefore, the NCEA algorithm significantly improves performance compared to both the ACO-CEA and GA-CEA algorithms, with improvements of 15.87% and 19.69% in pipe length performance, and 40% and 50% in elbow performance, respectively. Figure 17 , Figure 18 , Figure 19 As shown, compared with the other two algorithms, the NCEA algorithm can achieve coordinated layout of branch pipelines, which can effectively save layout space and material consumption, and the layout is neat.

[0202] 3. Example of a mixed piping problem

[0203] (1) Simulation Experiment I

[0204] This section uses simulation experiments to test the performance of the NCEA algorithm in solving hybrid pipelines and compares it with the IOHA algorithm without the niche co-evolutionary concept. Specifically:

[0205] Both algorithms were set up with the same 3D spatial simulation environment model (size 20×20×20), and four different types of pipelines to be laid out were set in the model. Among them, pipeline 1 and pipeline 2 are branch pipelines, and it is desired that the two pipelines run in parallel as much as possible; pipeline 3 and pipeline 4 are single pipelines. The coordinates of each point in pipeline 1 are as follows: start point S1(1,1,1), end point T1-1(9,7,5), end point T1-2(3,5,1), end point T1-3(9,15,15), end point T1-4(19,19,19); the coordinates of each point in pipeline 2 are as follows: start point S2(1,1,3), end point T2-1(1,5,9), end point T2-2(1,17,9), end point T2-3(17,19,19); the coordinates of each point in pipeline 3 are as follows: start point S3(5,3,13), end point T3(11,13,15); the coordinates of each point in pipeline 4 are as follows: start point S4(11,1,10), end point T4(11,1,15). The two algorithms were run independently 20 times each, and the optimal paths obtained are as follows: Figure 20 , Figure 21 As shown in Table 6, the results of the operation are as follows.

[0206] Table 6: Simulation results of NCEA and IOHA algorithms (in the case of mixed piping problems)

[0207]

[0208] As shown in Table 6, compared with the layout results of the IOHA algorithm, the NCEA algorithm has better overall performance in pipe layout, with a 1.20% improvement in pipe length performance and a 31.25% improvement in elbow performance. Figure 20 , Figure 21 As shown, the NCEA algorithm can better guide the bundled laying of pipelines of the same type, thereby saving layout space and achieving a good effect of mixed pipeline collaborative layout.

[0209] (2) Simulation Experiment II

[0210] Further complex models were used to test the performance of the NCEA algorithm in solving mixed pipelines, specifically:

[0211] A three-dimensional spatial simulation environment model (100×100×100) was set up. This model was more complex than the one in Simulation Experiment I (mainly due to the increased number and layout of equipment). Two different types of pipelines were set up in the model. Pipeline 1 was a single pipeline, and pipeline 2 was a branch pipeline. The coordinates of each point in pipeline 1 were: start point S1(1,1,1), end point T1(9,7,5); the coordinates of each point in pipeline 2 were: start point S2(1,1,3), end point T2-1(1,5,9), end point T2-2(1,17,9), and end point T2-3(17,19,19). The NCEA algorithm was run independently 20 times, and the optimal paths obtained were as follows: Figure 22 , Figure 23 As shown.

[0212] Depend on Figure 22 , Figure 23 As shown, the NCEA algorithm can take into account the influence of the coordination relationship between multiple pipelines, realize the parallel multi-pipeline collaborative layout with branches in complex models, and lay pipelines in bundles.

[0213] In summary, this embodiment verifies the optimization capability of the NCEA algorithm in solving parallel pipeline, branch pipeline, and mixed pipeline problems by setting up three sets of experimental examples. Furthermore, the results compared with other improved algorithms show that the NCEA algorithm performs better in solving multi-pipeline layout problems, effectively improving pipeline layout quality and solution efficiency.

[0214] It should be noted that the parts of the above solutions that are not described in detail or in an elaborate manner are all prior art, and are not improvements made by this invention to the prior art, nor are they within the protection scope of the technical solutions of this invention. Therefore, they will not be elaborated on in this article.

[0215] Of course, the above description is only a preferred embodiment of the present invention and should not be considered as limiting the scope of the embodiments of the present invention. The present invention is also not limited to the above examples, and all equivalent changes and improvements made by those skilled in the art within the scope of the present invention should fall within the patent coverage of the present invention.

Claims

1. A multi-pipeline layout method based on a niche co-evolutionary algorithm, characterized in that, Includes the following steps: A1: Establish an environmental model of the pipeline layout space; A2: Solve the multi-pipeline layout problem by applying the niche co-evolution algorithm; A3: When the niche co-evolutionary algorithm meets the termination condition, it outputs the optimal global solution as the final multi-pipeline layout scheme. The niche co-evolutionary algorithm includes the following steps: S1: To address the system problem, construct a niche layout; S1 includes the following sub-steps: S1-1: Decompose the system problem into multiple subproblems and create a corresponding population for each subproblem; S1-2: Construct a niche for at least one representative combination of collaborative solutions to subproblems; S2: Conduct niche evolution, implementing elite collaborator selection strategies, multidimensional information sharing mechanisms, and adaptive evolutionary mechanisms during the evolutionary process; S2 includes the following sub-steps: S2-1: Within the small population, evolution proceeds according to the adaptive evolution mechanism; for the currently evolving population, collaborators are selected based on the elite collaborator selection strategy; S2-2: Constructing a co-evolutionary environment for the current population through a multi-dimensional information sharing mechanism; S2-3: Update the global solution to the system problem based on the evolution of the niche; S3: Repeat S2 until the termination condition is met; In the aforementioned microhabitat layout, multiple microhabitats are constructed, and all microhabitats evolve in parallel, together forming an ecosystem; The global solution to a system problem is the solution to an ecosystem problem; The elite collaborator selection strategy includes: selecting elite collaborator individuals from other populations located in the same small territory as the current evolutionary population, consisting of at least one optimal solution individual and / or a second-best solution individual; The multidimensional information sharing mechanism includes: constructing a co-evolutionary environment for the current population based on the individual representatives of the elite collaborators; specifically, it includes fitness evaluation sharing and at least one type of environmental information sharing used to guide the subsequent search process. The fitness evaluation sharing includes: hierarchical fitness evaluation of individuals, populations, microhabitats and ecosystems; wherein, the fitness of higher-level fitness is a comprehensive evaluation based on the fitness of the lower-level fitness it contains and its synergistic effect. The environmental information sharing includes: modifying the quantization attributes of candidate solution elements in the solution space to create regions that have an attractive or repulsive effect on subsequent population searches, thereby guiding the search direction of subsequent populations; the quantization attributes include energy values ​​and / or pheromone concentrations.

2. The multi-pipeline layout method based on niche co-evolutionary algorithm according to claim 1, characterized in that: The adaptive evolution mechanism includes a population evolution termination condition, namely, setting an evolutionary stagnation rule as one of the algorithm's termination conditions; the evolutionary stagnation rule means that if no evolution occurs within a specified number of iterations, the population evolution is terminated; specifically, it includes population evolutionary stagnation rules, niche evolutionary stagnation rules, and ecosystem evolutionary stagnation rules.

3. The multi-pipeline layout method based on niche co-evolutionary algorithm according to claim 1, characterized in that: In step A2, if the type of multi-pipe layout problem is a parallel pipe problem, then step A2 specifically includes the following sub-steps: A2-a1: Decompose the parallel pipeline problem into multiple single-pipeline subproblems. Each parallel pipeline is a single-pipeline subproblem, and each single-pipeline subproblem corresponds to a population. The populations evolve in a random order. A2-a2: In the process of evolution, the pipeline solutions of the current population are regarded as obstacles, and through the environmental information sharing in the multidimensional information sharing mechanism, the nodes around the pipeline solutions are set as advantageous regions with an attractive effect, so as to guide the next population to seek optimization in the advantageous regions. A2-a3: Solve multiple single-pipeline sub-problems step by step, complete the layout of the last parallel pipeline, and obtain the current optimal solution for the ecosystem; A2-a4: Compare the current optimal solution with the global optimal solution. If the current optimal solution is better than the global optimal solution, then replace it; otherwise, leave it unchanged.

4. The multi-pipeline layout method based on niche co-evolutionary algorithm according to claim 3, characterized in that: In step A2, if the type of multi-pipe layout problem is a branch pipe problem, then A2 specifically includes the following sub-steps: A2-b1: Analyze the layout characteristics of branch pipeline problems, and classify branch pipeline problems according to whether the laying sequence and primary-secondary relationship need to be considered. The types include fixed branch pipeline problems that need to be laid according to primary-secondary and primary order and floating branch pipeline problems that do not need to be laid according to primary-secondary and primary order. A2-b2: Different solution steps are performed depending on the type of branch pipeline problem; If the branch pipe problem is a floating branch pipe problem, then A2-b2 includes the following sub-steps: A2-b2-f1: Decompose the branch pipeline problem into multiple single-pipeline subproblems, that is, break down the branch pipeline into multiple single pipelines and determine the start and end positions of each single pipeline; each single-pipeline subproblem corresponds to a population, and the various populations evolve in a random order; A2-b2-f2: During the evolutionary process, through the sharing of environmental information in the multidimensional information sharing mechanism, the current population's pipeline solution and surrounding nodes are set as advantageous regions with an attractive effect; A2-b2-f3: Solve multiple single-pipeline subproblems step by step, complete the layout of the last single pipeline, and obtain the current optimal solution for the ecosystem; A2-b2-f4: Compare the current optimal solution with the global optimal solution. If the current optimal solution is better than the global optimal solution, then replace it; otherwise, leave it unchanged.

5. The multi-pipeline layout method based on niche co-evolutionary algorithm according to claim 4, characterized in that: In step A2, if the type of multi-pipe layout problem is a mixed-pipe problem, then A2 specifically includes the following sub-steps: A2-c1: Classify and label the pipelines to be laid out, and determine the relationship between the pipelines through the synergy factor; A2-c2: Assign importance to each pipeline and determine the pipeline layout order based on the importance; A2-c3: Based on the classification labels of each pipeline, the corresponding solution method is called to perform the layout, that is: If the classification label shows a parallel pipeline, then call steps A2-a1 to A2-a4; If the classification label shows a branch pipeline, then call steps A2-b1 to A2-b2.

6. The multi-pipeline layout method based on niche co-evolutionary algorithm according to claim 3, characterized in that: The objective function for parallel pipeline layout is: ， In the formula, for i The total length of a pipeline when it is broken down into individual pipelines; This refers to the length of the parallel section of the pipeline; for i The total number of bends when a pipeline is split into individual pipelines; This represents the number of elbows in the parallel section of the pipeline. for i The total energy value of the pipeline; For path length weights, Weights for path bends, The path energy value weight; This is the shortest total length; This represents the minimum total number of bends; This represents the maximum energy value of the pipeline node.

Images