Ship cabin layout method and system based on dynamic fusion of three-way knowledge gene layers

Abstract

The present invention belongs to the field of ship engineering design, and discloses a ship cabin layout method and system based on the dynamic fusion of three-way knowledge gene layers. The system includes a cabin layout reasoning module based on prior knowledge, an intelligent cabin layout optimization algorithm module based on computational knowledge, an online design module for the cabin layout scheme of human intelligence knowledge, and a three-way knowledge fusion and solution cabin layout design module. The three-way knowledge individuals of the prior knowledge scheme, the algorithm knowledge scheme, and the human intelligence knowledge scheme obtained respectively are dynamically fused at the gene level of the evolutionary algorithm to form an evolutionary population, and are operated through a replacement method, a crossover method, and a combination method of crossover and multi-objective function trade-off method to obtain a cabin layout scheme. The present invention can take into account the optimization objectives and constraint conditions that are difficult to express in mathematical language during the cabin layout process during the algorithm solving process, and obtain a more practical ship cabin layout scheme.

Description

Technical Field

[0001] The present invention belongs to the field of ship engineering design, and particularly relates to a ship cabin layout method and system based on dynamic fusion of three-way knowledge gene layers. Background Art

[0002] During the shipbuilding cycle, ship design takes a large amount of time. Ship design is divided into preliminary design, detailed design, and production design. The overall ship layout design is a key step in the detailed ship design, and the cabin layout design is an important part of the overall layout design. In the intelligent layout design of ship cabins (including cabins, internal equipment, pipelines, etc.), there are many types of objects to be arranged, a large number of them, complex coupling relationships, and complicated constraint conditions, with multiple optimization objectives. It has computational complexity, engineering complexity, and engineering practical complexity. The traditional cabin layout design is carried out by means of interactive methods such as the mouse and keyboard provided by the CAD system, and it is a process of continuous repetition and spiral improvement. In the design, often a local problem causes the overall plan to be readjusted, which not only increases the workload, but also prolongs the design cycle, and may even cause economic losses. With the development of computer technology, it is imperative to improve the intelligence and automation of cabin layout.

[0003] In recent years, the existing technologies have conducted a large number of studies using theories and methods such as intelligent algorithms, knowledge reasoning, and expert systems. However, most of them are based on the main technical route of "machine intelligence". The optimization process lacks human participation, does not fully consider the parallel and effective utilization of prior knowledge and the professional knowledge and experience of designers, the optimization process lacks effective guidance, the obtained design scheme lacks innovation, and there is still a large gap from actual engineering applications. Therefore, it is necessary to establish a new layout design method based on intelligent algorithms that can effectively utilize prior knowledge in parallel and give full play to the intelligence of designers, enabling better cooperation between humans and machines, which can fundamentally solve the problem of intelligent layout optimization design of ship cabins and improve the intelligence and practicality of ship cabin layout design schemes. Summary of the Invention

[0004] To overcome the problems existing in the related technologies, the disclosed embodiments of the present invention provide a ship cabin layout method and system based on dynamic fusion of three-way knowledge gene layers, specifically related to an intelligent layout design method for ship cabins based on dynamic fusion of three-way knowledge gene layers.

[0005] The technical solution is as follows: A ship cabin layout system based on dynamic fusion of three-way knowledge gene layers, the system includes:

[0006] A cabin layout reasoning module based on prior knowledge, used to sort out the prior knowledge of cabin layout design, and perform knowledge representation and knowledge reasoning using the method of knowledge engineering reasoning expressions to obtain a prior knowledge scheme;

[0007] The intelligent cabin layout optimization algorithm module based on computational knowledge uses an improved ant colony-genetic algorithm for cabin layout to obtain an algorithmic knowledge solution, which is used for obtaining the computational knowledge solution and realizing the three-way layout knowledge fusion.

[0008] The human intelligence knowledge cabin layout scheme online design module is used for the online design of a three-dimensional visualization solution for the cabin layout scheme to obtain a human intelligence knowledge solution.

[0009] The three-way knowledge fusion solving cabin layout design module dynamically fuses the three-way knowledge individuals of the prior knowledge solution, algorithmic knowledge solution, and human intelligence knowledge solution obtained respectively at the gene level of the improved ant colony-genetic algorithm to form an evolutionary population, and through operations such as replacement, crossover, and the combination of crossover and multi-objective function trade-off method, obtains a cabin layout scheme based on three-way knowledge fusion solving; the three-way knowledge individuals include prior knowledge individuals, artificial individuals, and algorithm individuals.

[0010] Furthermore, in the cabin layout reasoning module based on prior knowledge, the prior knowledge includes excellent design examples, professional knowledge, and rule data information; the solution to the problem of the warship cabin layout to be solved is obtained through the acquisition and reuse of prior knowledge based on knowledge engineering, and the solution to the problem is in the form of a coded string as a prior knowledge individual.

[0011] In the intelligent cabin layout optimization algorithm module based on computational knowledge, the computational knowledge includes the knowledge obtained by the improved ant colony-genetic algorithm calculation; for the problem of the warship cabin layout to be solved, the solution to the problem is automatically generated, and the solution to the problem is in the form of a coded string as an algorithm individual.

[0012] In the human intelligence knowledge cabin layout scheme online design module, the human intelligence knowledge includes the professional knowledge and experience knowledge provided by designers; using human intelligence knowledge, combining engineering reality and calculation situations, and using the interaction information provided by the human intelligence knowledge scheme online design interface, the solution to a certain cabin layout problem is obtained, and the solution to the problem is in the form of a coded string called an artificial individual.

[0013] In the human intelligence knowledge cabin layout scheme online design module, the three-dimensional visualization of the cabin layout scheme is carried out through the human intelligence knowledge scheme online design interface. Specifically, the online designer obtains the coded string of the cabin layout scheme to be modified from the layout scheme library and generates the corresponding three-dimensional layout scheme, and after modifying the coded string, the three-dimensional scheme is updated in real time to determine whether to further modify or save the scheme.

[0014] The online design interface of the human intelligence knowledge solution includes: a cabin sequence modification module, a function module, and a layout scheme update module. The online design interface of the human intelligence knowledge solution contains a cabin sequence modification module, a function module, and a layout scheme update module. The cabin sequence modification module refers to the three-dimensional layout scheme corresponding to the cabin layout scheme coding string, and optimizes the cabin layout coding string based on the knowledge and experience of the online designer to achieve the optimization and adjustment of the cabin sequence coding string; the function module can save, display, and switch the cabin sequence coding string, and is responsible for obtaining the cabin layout scheme sequence coding string in Excel to the online design interface, and then saving it back to the Excel solution library after modification to form a solution set of artificial individuals different from prior knowledge and algorithm individuals; the layout scheme update module generates a three-dimensional cabin layout scheme according to the cabin layout scheme coding string, which is convenient for designers to intuitively understand the layout situation and can quickly update and generate a new scheme after modification.

[0015] In the three-way knowledge fusion and solution cabin layout design module, the prior knowledge individuals, artificial individuals, and algorithm individuals are uniformly expressed by coding strings, and it is determined whether to add them to the improved ant colony-genetic algorithm population according to their fitness values and diversity values; during the evolution process, three operation methods are adopted for the fusion of prior knowledge individuals, artificial individuals, and algorithm individuals to complete continuous and dynamic information fusion; the three operation methods include the replacement method, the crossover method, and the method combining crossover and multi-objective function trade-off.

[0016] The replacement method includes: adding artificial individuals to the algorithm population and randomly replacing a certain number of algorithm solutions, which are randomly selected from the algorithm population, or the algorithm solutions are the poorer solutions in the algorithm population, and then performing an evolution operation on the algorithm population;

[0017] The crossover method includes: randomly selecting algorithm solutions in the algorithm population, performing crossover operations with a certain artificial individual respectively, and randomly selecting one of the two offspring individuals generated after each crossover operation and putting it back into the algorithm population, so that the gene fragments of the artificial individual and the algorithm solution can fully exchange and fuse information;

[0018] The method combining crossover and multi-objective function trade-off includes: first, according to the functional requirements of the cabin layout and the running state of the algorithm, adjusting the relative importance between each objective function by changing the weight coefficients of each objective function, and secondly, forcing each individual in the algorithm to perform a crossover operation with a prior knowledge individual or an artificial individual in the current evolution generation, so that the individuals in the algorithm population inherit part of the genes of the prior knowledge individual or the artificial individual, thereby playing a guiding role in the evolution of the entire algorithm population.

[0019] Another object of the present invention is to provide a ship cabin layout method based on three-way knowledge gene layer dynamic fusion, including:

[0020] S1. Organize the prior knowledge of the cabin layout design, use the method of knowledge engineering inference expression for knowledge representation and knowledge reasoning to obtain the prior knowledge scheme; use the improved ant colony-genetic algorithm to obtain the algorithm knowledge scheme and algorithm individuals; and perform three-dimensional visualization of the online designed cabin layout scheme to obtain the human-intelligent knowledge scheme and artificial individuals.

[0021] S2. Dynamically fuse the three-way knowledge individuals of the prior knowledge scheme, algorithm knowledge scheme, and human-intelligent knowledge scheme obtained respectively at the gene level of the improved ant colony-genetic algorithm to form an evolutionary population, and perform operations through replacement methods, crossover methods, and the combination method of crossover and multi-objective function trade-off method to obtain the cabin layout scheme based on three-way knowledge fusion solution.

[0022] S3. Apply it to the layout design of multiple cabins to be arranged on the boat deck based on the obtained cabin layout scheme.

[0023] Furthermore, in step S1, the obtained prior knowledge scheme includes:

[0024] In step S1, the obtained prior knowledge scheme includes:

[0025] Use the built-in vba editor and Automation technology of CATIA to transfer the parameter values between the layout boundary conditions and the inference rules. Among them, the layout boundary conditions are fixed parameters stored in the CATIA model, and the inference rules, as the rules of the knowledge engineering module, are also stored in the model accordingly. Establish a multi-objective inference system based on the cabin layout requirements in the comprehensive specification convention and the requirements for lighting, vibration noise, and adjacency of the cabins. Specifically, first, based on the layout rules, lighting, and vibration noise in the specification, complete the inference of the cabin layout area on the deck, and then based on the adjacency and circulation between the cabins, complete the inference of the specific layout position of the cabins in the area to obtain the prior knowledge scheme. Taking the inference of the cabin layout for lighting as an example, starting from the requirements, cabins with high lighting requirements such as living cabins and recreational rooms are preferentially considered to be arranged in the outer area of the deck. At the same time, by analyzing the architectural structure of the deck and the hull shape, determine which areas can obtain better natural lighting. For example, cabins on the ship's side usually have better lighting than cabins near the ship's center.

[0026] Using the improved ant colony-genetic algorithm to obtain the algorithm knowledge scheme includes:

[0027] First, use the ant colony algorithm to initialize the cabin coding and then perform area allocation to randomly generate the cabin layout scheme. Then, use the roulette wheel method to select the cabin layout scheme with excellent fitness value to update the pheromone to guide the iteration of the algorithm. The pheromone formula is as follows:

[0028] ;

[0029] In the formula, is the pheromone concentration between the cabin and the cabin ; the variable represents the evaporation rate of the pheromone concentration; represents the th scheme, and the increased pheromone between the cabin and ; is the th scheme, and its value range is ;

[0030] Aiming at the fact that the pheromone mechanism is prone to falling into local optimum, the crossover and mutation operators of the genetic algorithm are introduced to increase the diversity of the solution population of the scheme, guide the solution of the algorithm, and obtain the specific algorithm flow of the algorithm knowledge scheme as Figure 6 shown;

[0031] The obtained human intelligence knowledge scheme includes:

[0032] Using the GUI function provided by the built-in VB editor in CATIA, adding the EXCEl function library through the VB editor to call the EXCEL command, obtaining the cabin layout scheme stored in EXCEL in the form of a real number coding string, each real number in the coding string represents a corresponding cabin, and the position of the cabin in the coding string represents the layout order of the cabin. Then, the cabin layout is carried out in the corresponding deck area through the strategy of recommended layout area of the cabin and evenly distributing the remaining area.

[0033] The designer modifies the position of the corresponding real number in the coding string through his own experience in the online design interface of the human intelligence knowledge scheme, combines the deck boundary to be arranged and the area information of the cabin, determines the width of each cabin according to the aisle position and then obtains the length value of the cabin; for the situation and modification suggestions of the cabin layout scheme, the position and size parameters of the cabin are updated in real time according to the designer's modification of the layout scheme. The update is realized by first determining the coding of the cabin, then determining the layout area of the cabin, and then carrying out area allocation to realize the real-time update of the three-dimensional layout scheme; finally, with the help of the GUI function provided by the VB editor, an online operation interface for the human intelligence knowledge scheme is built.

[0034] In step S2, the obtained cabin layout scheme based on the three-way knowledge fusion solution includes:

[0035] First, acquisition methods based on prior knowledge, human intelligence knowledge, and computational knowledge are established respectively. The acquisition methods of the three kinds of knowledge have been described before:

[0036] In the cabin layout reasoning module based on prior knowledge, the layout knowledge includes: excellent design examples, professional knowledge, rules, and regulations; for the problem of ship cabin layout to be solved, the solution of the problem is obtained through the acquisition and reuse of prior knowledge based on knowledge engineering, and the solution of the problem in the form of a coded string is a prior knowledge individual.

[0037] In the intelligent cabin layout optimization algorithm module based on computational knowledge, the computational knowledge includes the knowledge obtained by the improved ant colony-genetic algorithm; for the problem of ship cabin layout to be solved, the solution of the problem is automatically generated, and the solution of the problem in the form of a coded string is an algorithm individual.

[0038] In the online design module for the human intelligence knowledge cabin layout scheme, the human intelligence knowledge includes the professional knowledge and experience mastered by designers; for a specific problem of a cabin layout, based on human intelligence knowledge, combined with the actual engineering and calculation situations, and the interactive information provided by the online design interface of the human intelligence knowledge scheme, the designer provides the solution of the problem, and the solution of the problem in the form of a coded string is called an artificial individual.

[0039] Then, select some excellent prior knowledge scheme solutions and human intelligence knowledge scheme solutions to form a hybrid scheme solution set, and represent it in the form of a coded string; when the improved ant colony-genetic algorithm runs to a certain stage, add the hybrid scheme set to the population of the improved ant colony-genetic algorithm, and in each iterative solution of the improved ant colony-genetic algorithm, realize the fusion and dynamic optimization of three-way knowledge; finally, through operations such as replacement and crossover, adjust the generation number and quantity of the above hybrid scheme set added through multiple simulation experiments to obtain the best timing, quantity, and individual quality of the hybrid knowledge population added, and obtain the optimal three-way knowledge fusion solution mechanism;

[0040] Perform weighted processing on each sub-objective function in the above multi-objective function trade-off method to obtain a cabin layout scheme with excellent comprehensive performance.

[0041] Combined with all the above technical solutions, the beneficial effects of the present invention are as follows: In the process of optimizing the design of ship cabin layout based on intelligent algorithms, the present invention gives full play to the guiding role of prior knowledge and human intelligence knowledge in the algorithm, gives full play to the role of various layout knowledge, and forms an evolutionary population through the dynamic fusion of prior knowledge, human intelligence knowledge, and computational knowledge (three-way knowledge) at the gene level of the evolutionary algorithm. Through operations such as selection, crossover, and mutation, a cabin layout scheme with high intelligence and strong engineering practicability is automatically obtained. For the problem of optimizing the design of ship cabin layout, the present invention mainly includes methods for obtaining prior knowledge, computational knowledge, and human intelligence knowledge of cabin layout, the dynamic fusion theory of three-way knowledge at the gene level, and effective fusion mechanisms, etc. The present invention can take into account the optimization objectives and constraints that are difficult to express in mathematical language during the cabin layout process during the algorithm solving process, and obtain a more practical ship cabin layout scheme. Description of the Drawings

[0042] The accompanying drawings herein are incorporated into the specification and form a part of this specification, showing embodiments consistent with the present disclosure, and are used together with the specification to explain the principles of the present disclosure;

[0043] Figure 1 It is a schematic diagram of a ship cabin layout system based on the dynamic fusion of three-way knowledge gene layers provided by an embodiment of the present invention;

[0044] Figure 2 It is a schematic diagram of the principle of a ship cabin layout system based on the dynamic fusion of three-way knowledge gene layers provided by an embodiment of the present invention;

[0045] Figure 3 It is a flowchart of a ship cabin layout method based on the dynamic fusion of three-way knowledge gene layers provided by an embodiment of the present invention;

[0046] Figure 4 It is a schematic diagram of the principle of a ship cabin layout method based on the dynamic fusion of three-way knowledge gene layers provided by an embodiment of the present invention;

[0047] Figure 5 It is a schematic diagram of the principle of a cabin layout design fusion mechanism for three-way knowledge fusion provided by an embodiment of the present invention;

[0048] Figure 6 It is a flowchart of an improved ant colony-genetic optimization algorithm provided by an embodiment of the present invention;

[0049] Figure 7 It is a schematic diagram of an online design interface based on the construction of GUI functions using a VB editor provided by an embodiment of the present invention;

[0050] Figure 8 It is a two-dimensional effect diagram of the cabin layout obtained by an embodiment of the present invention;

[0051] Figure 9 It is a schematic diagram of the obtained cabin layout and three-dimensional effect provided by an embodiment of the present invention. Detailed Embodiments

[0052] To make the above objects, features, and advantages of the present invention more apparent and understandable, the following provides a detailed description of the specific embodiments of the present invention with reference to the accompanying drawings. Many specific details are set forth in the following description to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0053] The innovation points of the present invention are as follows: The present invention integrates various knowledge sources, including prior knowledge, human intelligence knowledge, and computational knowledge, to provide a rich material basis for the cabin layout scheme. The present invention proposes a layout design mechanism and implementation method for the fusion of prior, computational, and human intelligence knowledge, and verifies its superiority in the intelligent layout design of cabins through simulation and comparison calculations. The present invention systematically discusses the acquisition methods of prior knowledge schemes, computational knowledge schemes, and human intelligence knowledge schemes for multi-layer decks, and verifies the feasibility and efficiency of this method in the layout design of multi-layer decks. These innovation points promote the intelligence and automation of ship cabin layout design, and improve the overall design efficiency and quality of ships.

[0054] Example 1. In the ship cabin layout system based on the dynamic fusion of three-way knowledge gene layers provided by the embodiment of the present invention, the ship cabin includes cabins and internal equipment, pipelines, etc.; an evolutionary population is formed through the dynamic fusion of prior knowledge, human intelligence knowledge, and computational knowledge (three-way knowledge) at the gene level of the evolutionary algorithm, and through operations such as selection, crossover, and mutation, a cabin layout scheme with high intelligence and strong engineering practicability is automatically obtained.

[0055] As Figure 1 shown, it includes: a cabin layout reasoning module based on prior knowledge, an intelligent cabin layout optimization algorithm module based on computational knowledge, an online design module for the cabin layout scheme of human intelligence knowledge, and a three-way knowledge fusion solution module for cabin layout design.

[0056] The cabin layout reasoning module based on prior knowledge is used to sort out the prior knowledge of cabin layout design, and uses the method of knowledge engineering expression for knowledge representation and knowledge reasoning to obtain the prior knowledge scheme. The basic form of the knowledge engineering expression is the "IF-THEN" structure. The cabin layout knowledge obtained in the knowledge acquisition stage is represented using the production expression method, thereby forming an inference rule code library.

[0057] The intelligent cabin layout optimization algorithm module based on computational knowledge uses an improved ant colony-genetic algorithm for cabin layout to obtain an algorithm knowledge scheme, which is used for the acquisition of computational knowledge schemes and the realization of the fusion of three-way layout knowledge.

[0058] The online design module for the cabin layout scheme of human intelligence knowledge is used to online design a three-dimensional visualization solution for the cabin layout scheme to obtain the human intelligence knowledge scheme;

[0059] Exemplarily, a three-dimensional visualization solution for the cabin layout scheme is online designed through the online design interface of the human intelligence knowledge scheme. Specifically, the online designer obtains the encoded string of the cabin layout scheme to be modified from the layout scheme library and generates the corresponding three-dimensional layout scheme. After modifying the encoded string, the three-dimensional scheme is updated in real time to determine whether to further modify or save the scheme;

[0060] The online design interface of the human intelligence knowledge solution includes: a cabin sequence modification module, a function module, and a layout scheme update module;

[0061] The cabin sequence modification module refers to the three-dimensional layout scheme corresponding to the cabin layout scheme coding string, and optimizes the cabin layout coding string by virtue of the knowledge and experience of the online designer to realize the optimized adjustment of the cabin sequence coding string; the function module can save, display, and switch the cabin sequence coding string, and is responsible for obtaining the cabin layout scheme sequence coding string in Excel to the online design interface; after modification, it is saved back to the Excel scheme library to form a solution set of artificial individuals different from prior knowledge and algorithm individuals; the layout scheme update module generates a three-dimensional cabin layout scheme according to the cabin layout scheme coding string, which is convenient for designers to intuitively understand the layout situation and can quickly update and generate a new scheme after modification.

[0062] Among them, human intelligence knowledge includes professional knowledge and experience knowledge provided by designers; by using human intelligence knowledge, combining engineering practice and calculation conditions, and using the interaction information provided by the online design interface of the human intelligence knowledge solution, the solution to a certain cabin layout problem is obtained, and the solution coding string form of the problem is called an artificial individual.

[0063] The three-way knowledge fusion solution cabin layout design module is used to dynamically fuse the obtained three-way knowledge individuals (solutions) at the gene level of the improved ant colony-genetic algorithm to form an evolutionary population. Through operations such as selection, crossover, and mutation, strengths are complemented, and the roles of various layout knowledge are fully exerted to obtain a highly intelligent and practical cabin layout scheme based on three-way knowledge fusion solution.

[0064] Exemplarily, the present invention proposes a research idea for intelligent layout design of ship cabins by integrating three-way cabin layout knowledge of prior knowledge, calculation knowledge, and human intelligence knowledge, and a method for obtaining prior knowledge, calculation knowledge, and human intelligence knowledge solutions of cabin layout.

[0065] The present invention proposes a three-way knowledge fusion solution mechanism and implementation method, and studies the influence law of important parameters of three-way knowledge fusion on the cabin layout design result through comparative experiments, determines the three-way knowledge fusion parameters, and determines the layout design principle based on three-way knowledge fusion.

[0066] Based on the research of knowledge engineering reasoning layout, the present invention proposes a research idea of combining knowledge engineering and parametric design, which can realize reasoning layout for multiple layout scenarios. At the same time, during the reasoning process, the vibration noise, lighting, and adjacency conditions of the cabins are taken into account, and more comprehensive and reasonable reasoning rules are established for the cabins to be arranged, improving the quality of the design scheme.

[0067] Based on the method of secondary development of CATIA, the present invention constructs an online design interface for human-intelligent knowledge solutions, enabling online designers to obtain the corresponding solution conditions more intuitively and improving the convenience and quality of human-intelligent knowledge solution design.

[0068] Figure 2 It is the schematic diagram of the ship cabin layout system based on the dynamic fusion of three-way knowledge gene layers provided by the embodiment of the present invention.

[0069] As can be seen from the above embodiments, the technical solution of the present invention can improve the design efficiency after transformation, quickly generate the cabin layout plan, complete the design project in a shorter time, so as to undertake more orders, increase the business volume and income. At the same time, it can also optimize the layout quality, take into account the optimization objectives and constraints that are difficult to express in mathematical language, and the obtained ship cabin layout plan is more practical.

[0070] The method based on the dynamic fusion of three-way knowledge gene layers proposed by the present invention is different from the traditional methods in the field of ship cabin layout. It does not focus on a single knowledge source or design method, but dynamically fuses the prior knowledge solution, algorithmic knowledge solution and human-intelligent knowledge solution at the genetic level of the evolutionary algorithm, filling the technical gap in using multiple knowledge sources for cabin layout design.

[0071] The technical solution of the present invention, through the way of three-way knowledge fusion, solves to a certain extent the two technical problems that have long plagued ship engineering designers: the constraint processing that is difficult to quantify and the need for multi-knowledge source fusion in ship cabin layout.

[0072] The existing technology in the field of ship cabin layout has an excessive dependence on traditional single knowledge sources, such as simply relying on computational knowledge for layout optimization, which limits the designers' exploration of other methods. The present invention adopts the three-way knowledge fusion method, breaks this dependence, demonstrates the advantages of integrating multiple knowledge sources, and overcomes this bias. At the same time, there used to be a bias that only factors that can be accurately expressed mathematically can be incorporated into algorithm design, but the present invention attaches importance to and successfully processes layout objectives and constraints that are difficult to quantify, overcoming the problem of one-sided emphasis on quantifiable factors, and providing a more comprehensive and scientific idea for ship cabin layout design.

[0073] Example 2, as Figure 3 shown, the ship cabin layout method based on the dynamic fusion of three-way knowledge gene layers provided by the embodiment of the present invention includes:

[0074] S1, sort out the prior knowledge of cabin layout design, use the method of knowledge engineering inference expression for knowledge representation and knowledge reasoning to obtain the prior knowledge solution; use the improved ant colony-genetic algorithm to obtain the algorithmic knowledge solution and algorithm individuals; and perform three-dimensional visualization of the online designed cabin layout plan to obtain the human-intelligent knowledge solution and artificial individuals;

[0075] S2. Dynamically fuse the three-way knowledge individuals of the prior knowledge scheme, algorithm knowledge scheme, and human intelligence knowledge scheme obtained respectively at the gene level of the improved ant colony-genetic algorithm to form an evolutionary population. Through operations such as replacement, crossover, and the combination of crossover and multi-objective function trade-off method, obtain a cabin layout scheme based on three-way knowledge fusion solution;

[0076] S3. Apply it to the layout design of multiple cabins to be arranged on the boat deck based on the obtained cabin layout scheme.

[0077] Exemplarily, in step S1, the obtained prior knowledge scheme includes:

[0078] Use the built-in vba editor and Automation technology of CATIA to transfer parameter values between layout boundary conditions and inference rules. Among them, the boundary conditions of the layout are fixed parameters stored in the CATIA model, and the inference rules, as the rules of the knowledge engineering module, are also stored in the model accordingly. Establish a multi-objective inference system based on the cabin layout requirements in the comprehensive specification convention and the requirements for lighting, vibration noise, and adjacency of cabins. Specifically, first complete the inference of the layout area of cabins on the deck based on the layout rules, lighting, and vibration noise in the specification, and then complete the inference of the specific layout positions of cabins within the area based on the adjacency and circulation between cabins to obtain the prior knowledge scheme. Taking the inference of cabin layout for lighting as an example, starting from the requirements, cabins with high lighting requirements such as living cabins and recreational rooms are preferentially considered to be arranged in the outer areas of the deck. At the same time, by analyzing the building structure of the deck and the hull shape, determine which areas can obtain better natural lighting. For example, cabins on the ship's side usually have better lighting than those close to the ship's center;

[0079] Obtaining the algorithm knowledge scheme using the improved ant colony-genetic algorithm includes:

[0080] First use the ant colony algorithm to initialize the cabin coding and then perform area allocation to randomly generate a cabin layout scheme. Then use the roulette wheel method to select excellent fitness value cabin layout schemes to update the pheromone to guide the iteration of the algorithm. The pheromone formula is as follows:

[0081] ;

[0082] In the formula, is the pheromone concentration between cabin and cabin ; the variable represents the evaporation rate of the pheromone concentration; represents the th scheme at cabin and The pheromone added in between; For the th solution, The value range of ;

[0083] Aiming at the problem that the pheromone mechanism is prone to falling into local optimum, the crossover and mutation operators of the genetic algorithm are introduced to increase the diversity of the solution population of the scheme, guide the solution of the algorithm, and obtain the specific algorithm flow of the algorithm knowledge scheme as Figure 6 shown;

[0084] The obtained human intelligence knowledge scheme includes:

[0085] Using the GUI function provided by the built-in VB editor in CATIA, adding the EXCEl function library through the VB editor to call the EXCEL command, obtaining the cabin layout scheme stored in EXCEL in the form of a real number coding string, where each real number in the coding string represents a corresponding cabin, and the position of the cabin in the coding string represents the layout order of the cabin. Then, the cabins are arranged in the corresponding deck area through the strategy of recommended layout area of the cabin and evenly distributing the remaining area;

[0086] The designer modifies the position of the corresponding real number in the coding string in the online design interface of the human intelligence knowledge scheme to obtain different cabin layout schemes. Combining the deck boundary to be arranged and the area information of the cabins, the width of each cabin is determined according to the aisle position and then the length value of the cabin is obtained; for the situation and modification suggestions of the cabin layout scheme, the position and size parameters of the cabins are updated in real time according to the designer's modification of the layout scheme. The real-time update of the three-dimensional layout scheme is realized by the method of first determining the coding of the cabin to determine the layout area of the cabin and then performing area allocation; finally, with the help of the GUI function provided by the VB editor, an online operation interface for the human intelligence knowledge scheme is built.

[0087] In step S2, the obtained cabin layout scheme based on the three-way knowledge fusion solution includes:

[0088] First, the acquisition methods based on prior knowledge, human intelligence knowledge, and computational knowledge are established respectively. Then, some excellent prior knowledge scheme solutions and human intelligence knowledge scheme solutions are selected to form a hybrid scheme solution set, which is represented in the form of a coding string; when the ant colony-genetic algorithm runs to a certain stage, the hybrid scheme set is added to the population of the improved ant colony-genetic algorithm. In each iterative solution of the improved ant colony-genetic algorithm, the three-way knowledge fusion and dynamic optimization are realized; finally, through replacement and crossover operations, multiple simulation experiments are carried out to adjust the generation number and quantity of the above hybrid scheme set added, so as to obtain the best timing, quantity, and individual quality of the hybrid knowledge population added, and the optimal three-way knowledge fusion solution mechanism is obtained; each sub-objective function in the above multi-objective function trade-off method is weighted to obtain a cabin layout scheme with excellent comprehensive performance.

[0089] Example 3, as another implementation manner of the present invention, as Figure 4 shown, the ship cabin layout method based on the dynamic fusion of three-way knowledge gene layers provided by the embodiment of the present invention is a ship cabin intelligent layout optimization design method based on the dynamic fusion of gene layers of prior knowledge, human intelligence knowledge, and computational knowledge (three-way knowledge). During the process of the cabin intelligent layout optimization algorithm, it can give full play to the guiding role of the prior knowledge and human intelligence knowledge of the cabin layout in the intelligent algorithm, give full play to the respective advantages of humans and machines, and adopt the strategy of effectively using in parallel the excellent design examples, rules, regulations, etc. (prior knowledge) of the ship cabin layout design, the professional knowledge and experience of designers (human intelligence knowledge), and the improved ant colony-genetic algorithm (computational knowledge) in the ship cabin intelligent layout design. It can take into account the optimization objectives and constraints that are difficult to express in mathematical language during the cabin layout process during the algorithm solving process, and obtain a more practical ship cabin layout plan.

[0090] Prior knowledge and prior knowledge individuals: Prior knowledge mainly includes excellent design examples, professional knowledge, rule data information, etc. The solution to the ship cabin layout problem to be solved is obtained through the acquisition and reuse of prior knowledge based on knowledge engineering, and the solution of the problem in the form of a coded string is a prior knowledge individual (prior knowledge solution).

[0091] Human intelligence knowledge and artificial individuals: Human intelligence knowledge mainly refers to the professional knowledge and experience knowledge mastered by designers. For a specific problem of a cabin layout, based on human intelligence knowledge, considering the engineering reality and calculation situation, as well as the interactive information provided by the algorithm, etc., the solution to the problem provided by the designer, the form of its coded string is called an artificial individual (artificial individual).

[0092] Computational knowledge and algorithm individuals: Computational knowledge mainly refers to the knowledge obtained by calculating the improved ant colony-genetic algorithm. For the ship cabin layout problem to be solved, without human participation, the solution to the problem completely automatically generated by the algorithm, the form of its coded string is called an algorithm individual (algorithm solution).

[0093] Exemplarily, the present invention constructs a ship cabin intelligent layout design method that forms an evolutionary population through the dynamic fusion of prior knowledge, human intelligence knowledge, and computational knowledge (three-way knowledge) at the gene level of the evolutionary algorithm, and performs operations such as selection, crossover, and mutation for evolutionary optimization. The prior knowledge cabin layout reasoning module based on prior knowledge is used to obtain the prior knowledge solution through knowledge engineering and parametric design cabin layout reasoning. In view of the characteristics of the ship cabin layout design, a method for constructing prior knowledge of the cabin layout design based on ontology is proposed, a complete layout design knowledge representation model based on ontology is established, the sharing and reuse of the prior knowledge of the layout design are realized, and prior knowledge individuals (prior knowledge solutions) are provided for the three-way knowledge fusion.

[0094] The layout optimization algorithm module based on computational knowledge is used to obtain a computational knowledge solution and a fusion algorithm as the three-way cabin layout knowledge through calculation with an improved ant colony-genetic algorithm. The computational knowledge is mainly obtained through an efficient improved ant colony-genetic algorithm constructed for the cabin layout, providing algorithm individuals (algorithm solutions) for the three-way knowledge fusion.

[0095] Using the human intelligence knowledge solution online design module, designers use their professional knowledge and experience to obtain the human intelligence knowledge solution through online design. Based on the diversity of feasible solutions for warship cabin layout, a certain number of artificial design solutions are obtained through the online design of designers, manual modification of prior knowledge individuals, and manual modification of algorithm individuals, providing artificial individuals (artificial individuals) for the three-way knowledge fusion.

[0096] Increase the diversity of prior knowledge individuals and artificial individuals: The prior knowledge individuals and artificial individuals added to the algorithm will affect the evolution of algorithm individuals to a certain extent. For example, when the performance of the added artificial individuals is overly superior to that of the algorithm individuals, the excellent genes of the algorithm individuals are easily submerged during the iteration process, resulting in the underutilization of the performance of the algorithm individuals and causing the algorithm to converge to a common solution containing only the excellent genes of the artificial individuals. Therefore, the diversity of prior knowledge individuals and artificial individuals is expanded through human intelligence or automation technology to obtain prior knowledge individuals and artificial individuals with different performances, and their diversity is evaluated to increase the probability of effective fusion of prior knowledge individuals, artificial individuals, and algorithm individuals in subsequent research and to continuously and effectively combine with the algorithm.

[0097] The cabin layout design module using three-way knowledge fusion solution is used to realize the dynamic fusion of prior knowledge, computational knowledge, and human intelligence knowledge at the algorithm gene level for the design of the cabin layout plan.

[0098] And a method of combining prior knowledge individuals, artificial individuals, and computational individuals at the gene level of the evolutionary algorithm is adopted for in-depth three-way knowledge fusion. The principle of the fusion mechanism of the cabin layout design with three-way knowledge fusion is as Figure 5 shown, including:

[0099] Unify the expression of prior knowledge individuals, artificial individuals, and algorithm individuals generated by the evolutionary algorithm into a coding string, and decide whether to add them to the evolutionary algorithm population based on their fitness values and diversity values. During the evolution process, the fusion of prior knowledge individuals, artificial individuals, and algorithm individuals mainly adopts three operation methods to achieve a continuous and dynamic information fusion: ① "Replacement" method. For example: Add an artificial individual to the algorithm population and randomly replace a certain number of algorithm solutions. These algorithm solutions can be randomly selected from the algorithm population or the poorer solutions in the algorithm population, and then perform evolutionary operations on the algorithm population; ② "Crossover" method. For example: Randomly select algorithm solutions in the algorithm population and perform crossover operations with a certain artificial individual respectively, and randomly select one of the two offspring individuals generated after each crossover operation and put it back into the algorithm population, so that the gene segments of the artificial individual and the algorithm solution can fully communicate and fuse; ③ The method of combining "crossover" with the multi-objective function trade-off method. First, according to the functional requirements of the cabin layout and the running state of the algorithm, adjust the relative importance between each objective function by changing the weight coefficients of each objective function. Secondly, in the current evolutionary generation, force each individual in the algorithm to perform crossover operations with prior knowledge individuals or artificial individuals, so that the individuals in the algorithm population inherit part of the genes of the prior knowledge individuals or artificial individuals, thus playing a guiding role in the evolution of the entire algorithm population.

[0100] Another exemplary aspect, the cabin layout reasoning module designed by the present invention proposes a cabin layout reasoning method that combines knowledge engineering and parametric design. By means of the built-in editor of CATIA and Automation technology, the parameter value transfer between the layout boundary conditions and the reasoning rules is realized, which improves the generality of the layout reasoning rules. A multi-objective reasoning system is established by comprehensively considering the cabin layout requirements in the codes and conventions and the layout requirements such as daylighting, vibration noise, and adjacency of the cabins. That is, first, based on the layout rules, daylighting, and vibration noise in the codes, the reasoning of the layout area of the cabins on the deck is completed, and then based on the adjacency and circulation between the cabins, the reasoning of the specific layout positions of the cabins within the area is completed.

[0101] The cabin layout optimization algorithm module designed by the present invention based on computational knowledge establishes a cabin layout algorithm mathematical model, and proposes an improved ant colony-genetic cabin layout optimization algorithm. The algorithm is mainly based on the idea of ​​solving the traveling salesman problem by the ant colony algorithm, and selects a sequential coding method. Each cabin is represented by a natural number, and a fixed-length real number code string is used to represent the layout plan of the cabin on the deck. First, the cabin layout plan is randomly generated, and then the cabin layout plan with excellent fitness value is selected to update the pheromone to guide the iteration of the algorithm. Since the pheromone mechanism is prone to fall into the local optimum, the crossover and mutation operators of the genetic algorithm are introduced to increase the diversity of the solution population, and the pheromone mechanism of the algorithm is better used to guide the solution of the algorithm. The ant colony algorithm, the genetic algorithm and the algorithm are used to solve the cabin layout design problem, respectively, and the superiority of the algorithm in the ship cabin layout design is verified.

[0102] in, Figure 6 The main flow chart of the improved ant colony-genetic optimization algorithm provided by the embodiment of the present invention.

[0103] The human intelligence knowledge scheme online design module designed by the present invention uses the GUI function provided by the built-in VB editor in CATIA, adds the EXCEl function library through the VB editor to call the EXCEL command, obtains the cabin layout scheme stored in the EXCEL in the form of a real number code string, each real number in the code string represents a corresponding cabin, and the position of the cabin in the code string represents the layout order of the cabin. Then, the corresponding deck area and layout method are clarified, and the cabin layout of the deck can be obtained. Designers can obtain different cabin layout schemes by modifying the position of the corresponding real number in the code string, and the length, width and other parameter values ​​of each cabin can be calculated by combining the area information of the deck boundary and the cabin to be arranged. In order to facilitate designers to better understand the situation of the cabin layout scheme and put forward reasonable modification suggestions, the software functions and commands provided in the Automation technology provided by CATIA are used to obtain the corresponding cabin model, and then the length and width parameter information of the cabin are assigned, so that the position and size of the cabin and other parameters can be updated in real time according to the designer's modification of the layout scheme, and the real-time update of the three-dimensional layout scheme is realized, so that designers can provide human intelligence knowledge schemes more intuitively and reasonably. Finally, with the help of the GUI function provided by the VB editor, the above development codes were classified and organized, and the online operation interface of the human intelligence knowledge solution was built.

[0104] The cabin layout design module for three-way knowledge fusion solving designed in the present invention proposes a theory and implementation method for cabin layout design based on the dynamic fusion of prior knowledge, human intelligence knowledge, and computational knowledge (three-way knowledge) at the gene layer, and establishes a dynamic optimization and evaluation model for warship cabin layout based on three-way knowledge fusion. That is, first, acquisition methods based on prior knowledge, human intelligence knowledge, and computational knowledge are established respectively. Then, a partial set of excellent prior knowledge solutions and human intelligence knowledge solutions are selected to form a mixed solution set, which is represented in the form of algorithm coding strings. When the algorithm runs to a certain stage, the mixed solution set is added to the population of the algorithm. In each iterative solution of the algorithm, the fusion and dynamic optimization of three-way knowledge can be achieved. Finally, the optimal three-way knowledge fusion solving mechanism is obtained by exploring the best timing, quantity, and individual quality of the addition of the mixed knowledge population. As Figure 5 shown.

[0105] The determination of three-way knowledge fusion parameters is as follows: In the mixed knowledge population composed of prior knowledge solutions and human intelligence knowledge solutions, it is better that prior knowledge solutions and human intelligence knowledge solutions each account for 50%; the quantity of the mixed knowledge population is about 20% of the algorithm population size; the range of the quality of solution individuals in the mixed knowledge population is 0.46 - 0.56; the timing of adding the mixed knowledge population is after 20 generations of iterative solution of the fusion algorithm. The above provides an important reference for three-way knowledge fusion layout design.

[0106] Exemplarily, the dynamic optimization model should not only consider the complex coupling relationship, complicated constraint conditions, and requirements of multiple optimization objectives of warship cabin layout, but also consider the needs of three-way knowledge fusion. It can perform dynamic fusion of three-way knowledge at different time points such as the start, middle, and local optimum of the algorithm. When establishing the objective function to be solved, factors such as adjacency, circulation, vibration noise, lighting conditions, and ergonomics are comprehensively considered. Among them, adjacency and circulation are evaluated for the relationship between each cabin to be arranged and all other cabins to be arranged. During the algorithm solving process, each of the above sub-objective functions is weighted to obtain a cabin layout scheme with excellent comprehensive performance.

[0107] Take the layout design of 28 cabins to be arranged on the deck of a teaching practice ship as an example.

[0108] In the specific implementation process, as Figure 2 shown, layout knowledge such as the standard layout rules, cabin vibration noise requirements, cabin lighting requirements, adjacency between cabins, cabin area, and furniture allocation collected and sorted from the cabin layout design specifications.

[0109] The knowledge acquisition, knowledge representation, and knowledge reasoning for the cabins to be arranged are integrated in the form of parameters and added in the form of reasoning rules to the parameterized model of the cabin and the layout reasoning rule code, forming a corresponding cabin parameterized model library and layout rule code library. Then, according toFigure 2 Perform deck configuration reasoning, layout area reasoning of cabins on the deck, and specific position reasoning of cabins within the layout area for the shown reasoning layout process, and obtain the corresponding prior knowledge layout design scheme.

[0110] In the specific implementation process, the main process of the improved ant colony-genetic algorithm proposed by the present invention is as Figure 6 shown. After determining the population of cabin layout schemes according to the solution mechanism of the ant colony algorithm, the crossover and mutation operators of the genetic algorithm are introduced, and the corresponding poorly quality schemes in the original scheme population are replaced with new layout schemes, effectively improving the layout search space of the algorithm. For the 28 cabins to be arranged in the case, a real number coding method is adopted, and they are respectively represented by the numbers from 1 to 28. The coding string of the cabin layout scheme is as follows: [2 4 14 16 3 20 18 9 10 11 19 5 1 6 8 15 13 12 177 25 23 22 24 28 27 26 21]. The algorithm objective function uses the weighted sum of adjacency, circulation, vibration noise, daylighting, and ergonomics.

[0111] In the specific implementation process, the online design interface of the human intelligence knowledge scheme proposed by the present invention is as Figure 7 shown and is mainly divided into three modules: the cabin sequence modification module, the function module, and the layout scheme update module. The online designer can obtain the coding string of the cabin layout scheme to be modified from the layout scheme library and generate the corresponding 3D layout scheme. After modifying the coding string, the 3D scheme can be updated in real time to determine whether to further modify or save the scheme. This design module enables the designer to intuitively obtain the layout situation of the scheme and put forward reasonable suggestions. Figure 7 The coding string on the left in

[0112] In the specific implementation process, the cabin layout design with three-way knowledge fusion solution proposed by the present invention is as Figure 5 shown. The improved-genetic algorithm proposed in the present invention is selected as the fusion algorithm for three-way knowledge. According to the important fusion parameters determined by the comparative experiment exploration, the prior knowledge scheme and the human intelligence knowledge scheme are added during the solution process of the algorithm to achieve the effective fusion of three-way layout knowledge. The 2D scheme and 3D effect schematic diagram of the cabin layout obtained for the 28 cabins to be arranged on the deck of a certain teaching practice boat in the case are as Figure 8 , Figure 9 shown. This design scheme comprehensively considers factors such as the adjacency and circulation between cabins and the daylighting of cabins, has excellent comprehensive performance, and has strong practicability.

[0113] As described above, it is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, any modifications, equivalent substitutions and improvements made within the spirit and principle of the present invention shall be covered within the protection scope of the present invention.

Claims

1. A ship cabin layout system based on dynamic fusion of three-way knowledge gene layer, characterized in that: The system includes: The cabin layout reasoning module based on prior knowledge is used to organize the prior knowledge of cabin layout design, and use the knowledge engineering reasoning expression method to perform knowledge representation and knowledge reasoning to obtain the prior knowledge scheme. The knowledge engineering expression is an IF-THEN structure. The knowledge about cabin layout acquired in the knowledge acquisition stage is represented by the production expression method to form a reasoning rule code library; The intelligent cabin layout optimization algorithm module based on computational knowledge uses the improved ant colony-genetic algorithm to perform cabin layout and obtain the algorithm knowledge solution, which is used to obtain the computational knowledge solution and realize the three-way layout knowledge fusion; The human intelligence knowledge cabin layout plan online design module is used to design the cabin layout plan online with a three-dimensional visualization solution to obtain the human intelligence knowledge plan; The three-way knowledge fusion solution cabin layout design module dynamically fuses the three-way knowledge individuals of the prior knowledge scheme, algorithm knowledge scheme, and human intelligence knowledge scheme obtained respectively at the gene level of the improved ant colony-genetic algorithm to form an evolutionary group, and obtains the cabin layout scheme based on the three-way knowledge fusion solution through replacement, crossover, and a combination of crossover and multi-objective function trade-off method; the three-way knowledge individuals include prior knowledge individuals, artificial individuals, and algorithm individuals; In the cabin layout reasoning module based on prior knowledge, prior knowledge includes excellent design examples, professional knowledge, and rule data information; the solution to the ship cabin layout problem to be solved is obtained by acquiring and reusing prior knowledge based on knowledge engineering, and the solution code string of the problem is in the form of prior knowledge individuals; In the intelligent cabin layout optimization algorithm module based on computational knowledge, the computational knowledge includes knowledge obtained by improved ant colony-genetic algorithm calculation; for the ship cabin layout problem to be solved, the solution of the problem is automatically generated, and the solution code string form of the problem is the algorithm individual; In the online design module of the human intelligence knowledge cabin layout solution, human intelligence knowledge includes the professional knowledge and experiential knowledge provided by the designer; by using human intelligence knowledge, combining engineering practice and calculation conditions, and using the interactive information provided by the online design interface of the human intelligence knowledge solution, the solution to the ship cabin layout problem is obtained. The solution string form of the problem is called an artificial individual.

2. The ship cabin layout system based on three-way knowledge gene layer dynamic fusion according to claim 1 is characterized in that: In the online design module of cabin layout scheme of human intelligence knowledge, a three-dimensional visualization solution of cabin layout scheme is designed online through the online design interface of human intelligence knowledge scheme, which specifically includes that the online designer obtains the cabin layout scheme code string to be modified from the layout scheme library and generates the corresponding three-dimensional layout scheme, and updates the three-dimensional scheme in real time after modifying the code string to determine further modification or saving the scheme; The online design interface of the human intelligence knowledge solution includes: a cabin sequence modification module, a function module, and a layout solution update module; The cabin sequence modification module refers to the three-dimensional layout scheme corresponding to the cabin layout scheme code string, and optimizes the cabin layout code string to achieve optimized adjustment of the cabin sequence code string; The functional module is used to save, display, and switch cabin sequence code strings, and is responsible for obtaining cabin layout scheme sequence code strings in Excel to the online design interface, and then saving them back to the Excel scheme library after modification to form a solution set of artificial individuals that are different from prior knowledge and algorithm individuals; The layout scheme updating module generates a three-dimensional cabin layout scheme according to the cabin layout scheme coding string, which is used for designers to understand the layout situation and update and generate a new scheme after modification.

3. The ship cabin layout system based on three-way knowledge gene layer dynamic fusion according to claim 1 is characterized in that: In the three-way knowledge fusion solution for cabin layout design module, prior knowledge individuals, artificial individuals and algorithm individuals are expressed in a unified coding string, and whether to join the improved ant colony-genetic algorithm group is determined based on the fitness value and diversity value; in the evolutionary process, the fusion of prior knowledge individuals, artificial individuals and algorithm individuals adopts three operation modes to complete continuous dynamic information fusion; the three operation modes include replacement mode, crossover mode, and a combination of crossover and multi-objective function trade-off method.

4. The ship cabin layout system based on three-way knowledge gene layer dynamic fusion according to claim 3 is characterized in that: The replacement method includes: adding artificial individuals to the algorithm population and randomly replacing a certain number of algorithm solutions, where the algorithm solution is randomly selected from the algorithm population, or the algorithm solution is a poor solution of the algorithm population, and then performing an evolution operation on the algorithm population; The crossover method includes: randomly selecting an algorithm solution from the algorithm population, performing a crossover operation with an artificial individual respectively, and randomly selecting one of the two sub-individuals generated after each crossover operation, and returning it to the algorithm population, so that the gene fragments of the artificial individual and the gene fragments of the algorithm solution can fully exchange information and merge; The combination of the crossover and multi-objective function trade-off method includes: first, adjusting the relative importance of each objective function by changing the weight coefficient of each objective function according to the functional requirements of the cabin layout and the algorithm operation status; second, forcing each individual in the algorithm to perform a crossover operation with the prior knowledge individual or the artificial individual in the current evolutionary generation, so that the individuals in the algorithm group inherit part of the genes of the prior knowledge individual or the artificial individual, thereby guiding the evolution of the entire algorithm group.

5. A ship cabin layout method based on dynamic fusion of three-way knowledge gene layer, characterized in that: The method is implemented in the ship cabin layout system based on three-way knowledge gene layer dynamic fusion as claimed in any one of claims 1 to 4, and the method comprises: S1, organize the prior knowledge of cabin layout design, use the knowledge engineering reasoning expression method to perform knowledge representation and knowledge reasoning to obtain the prior knowledge scheme; use the improved ant colony-genetic algorithm to obtain the algorithm knowledge scheme and algorithm individuals; and perform three-dimensional visualization of the online design of cabin layout scheme to obtain the human intelligence knowledge scheme and artificial individuals; S2, the three-way knowledge individuals of the prior knowledge scheme, algorithm knowledge scheme and human wisdom knowledge scheme obtained respectively are dynamically integrated at the gene level of the improved ant colony-genetic algorithm to form an evolutionary group, and the cabin layout scheme based on the three-way knowledge fusion solution is obtained through replacement, crossover, and the combination of crossover and multi-objective function trade-off method; S3, applying the obtained cabin layout scheme to the layout design of multiple cabins to be arranged on the deck of the ship.

6. The method for ship cabin layout based on three-way knowledge gene layer dynamic fusion according to claim 5 is characterized in that: In step S1, obtaining a priori knowledge solution includes: The built-in VBA editor and Automation technology of CATIA are used to transfer parameter values ​​between layout boundary conditions and reasoning rules. The boundary conditions of the layout are stored in the CATIA model as fixed parameters, and the reasoning rules are also stored in the model as the rules of the knowledge engineering module. A multi-objective reasoning system is established by integrating the cabin layout requirements in the specification conventions and the cabin lighting, vibration noise and adjacency layout requirements. Specifically, the reasoning of the cabin layout area on the deck is completed based on the layout rules, lighting, vibration noise in the specifications, and then the reasoning of the specific layout position of the cabin in the area is completed based on the adjacency and circulation between the cabins to obtain a priori knowledge plan. The use of improved ant colony-genetic algorithm to obtain algorithm knowledge solutions includes: The ant colony algorithm is used to first initialize the cabin code and allocate the area to randomly generate the cabin layout plan. Then the roulette method is used to select the cabin layout plan with excellent fitness value to update the iteration of the pheromone guidance algorithm. The pheromone formula is as follows: ； In the formula, For the cabin With cabin Pheromone concentration between; variable The volatilization rate indicating the pheromone concentration; Indicates Plans in the cabin and Increased pheromones between; For the A plan, The value range is ; As the pheromone mechanism is prone to fall into local optimality, the crossover and mutation operators of the genetic algorithm are introduced to increase the diversity of the solution population, guide the solution of the algorithm, and obtain the algorithm knowledge solution; The scheme for obtaining human intelligence knowledge includes: By using the GUI function provided by the built-in VBA editor in CATIA, the Excel function library is added through the VBA editor to call the Excel command, and the cabin layout plan stored in Excel in the form of a real number code string is obtained. Each real number in the code string represents a corresponding cabin, and the position of the cabin in the code string represents the layout order of the cabin. Then, the cabin layout is carried out in the corresponding deck area according to the recommended layout area of ​​the cabin and the strategy of evenly distributing the remaining area. Different cabin layout schemes are obtained by modifying the position of the corresponding real number in the code string in the online design interface of the human intelligence knowledge solution. Combined with the area information of the deck boundary and the cabin to be arranged, the width of each cabin is determined according to the aisle position and then the length value of the cabin is obtained; according to the situation of the cabin layout plan and modification suggestions, the position and size parameters of the cabin are updated in real time according to the modified layout plan. The modified layout plan includes: first determining the cabin code to determine the layout area of ​​the cabin, and then allocating the area to achieve the modification of the three-dimensional layout plan; using the GUI function provided by the VBA editor to build the online operation interface of the human intelligence knowledge solution.

7. The method for ship cabin layout based on three-way knowledge gene layer dynamic fusion according to claim 5 is characterized in that: In step S2, the cabin layout solution based on the three-way knowledge fusion solution includes: Firstly, the acquisition methods based on prior knowledge, human intelligence knowledge and computational knowledge are established respectively. Then, some excellent prior knowledge solutions and human intelligence knowledge solutions are selected to form a hybrid solution set, which is represented in the form of a coding string. When the algorithm runs to a certain stage, the hybrid solution set is added to the population of the improved ant colony-genetic algorithm. In each iterative solution of the improved ant colony-genetic algorithm, the fusion and dynamic optimization of three-way knowledge are realized. Finally, through replacement and crossover operations, multiple simulation experiments are conducted to adjust the number and quantity of the hybrid solution set added, so as to obtain the best time, quantity and individual quality of the hybrid knowledge population, and obtain the optimal three-way knowledge fusion solution mechanism. Each sub-objective function in the above multi-objective function trade-off method is weighted to obtain a cabin layout solution with excellent comprehensive performance.

Images