CN118035223A - Multi-scheme optimization method for complex system - Google Patents

Multi-scheme optimization method for complex system Download PDF

Info

Publication number
CN118035223A
CN118035223A CN202410213551.XA CN202410213551A CN118035223A CN 118035223 A CN118035223 A CN 118035223A CN 202410213551 A CN202410213551 A CN 202410213551A CN 118035223 A CN118035223 A CN 118035223A
Authority
CN
China
Prior art keywords
data
scheme
owl
model
csv
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202410213551.XA
Other languages
Chinese (zh)
Inventor
程鑫
范子贵
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xi'an Kongtian Simulation Technology Co ltd
Original Assignee
Xi'an Kongtian Simulation Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Xi'an Kongtian Simulation Technology Co ltd filed Critical Xi'an Kongtian Simulation Technology Co ltd
Priority to CN202410213551.XA priority Critical patent/CN118035223A/en
Publication of CN118035223A publication Critical patent/CN118035223A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/21—Design, administration or maintenance of databases
    • G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/28—Databases characterised by their database models, e.g. relational or object models
    • G06F16/284—Relational databases
    • G06F16/288—Entity relationship models
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/12—Computing arrangements based on biological models using genetic models
    • G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/01—Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Databases & Information Systems (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Biophysics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Evolutionary Biology (AREA)
  • Mathematical Physics (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computational Linguistics (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Genetics & Genomics (AREA)
  • Biomedical Technology (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Physiology (AREA)
  • Quality & Reliability (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a multi-scheme optimization method for a complex system, which comprises the following steps: defining data transmission based on a model diagram mode, and cleaning the OWL data by using GOPPRR model body files in an OWL format; CSV data are cleaned, CSV data after OWL data are cleaned are further cleaned, and model data for scheme generation are obtained; generating a scheme, namely generating a fitting function from CSV data for scheme optimization; optimizing definition, namely optimizing definition is carried out by selecting an algorithm based on an input fitting function so as to balance and optimize the scheme; and (3) analyzing results, displaying the results of the optimized schemes through a visual means, obtaining a more optimized scheme, and comparing the results of different schemes. The invention solves the problems that the prior multidisciplinary design optimization method optimizes and splits system design data, and cannot form top-down data chain transmission.

Description

Multi-scheme optimization method for complex system
Technical Field
The invention belongs to the technical field of computers, and particularly relates to a multi-scheme optimization method for a complex system.
Background
Modern engineering systems are increasingly large in scale, interactions between systems are finer and more complex, it has been difficult to apply conventional optimization methods, and to empirically coordinate coupling effects within the systems. In a multidisciplinary environment, there may be multiple schemes, each having different advantages and disadvantages in different fields. The system needs to support trade-off analysis of these different schemes to determine the final decision.
The application of multi-scheme trade-off in a real project needs to consider a plurality of factors, such as complex real system and higher corresponding model complexity, and in order to ensure the clarity and completeness of the system model logic, the real project needs to plan the organization mode of the system model in advance and carry out necessary adjustment in the modeling process along with the project evolution; the rules of variant configuration in real projects are more complex, and there are often some dependency and conflict relationships between each other, instead of simple element arrangement and combination.
With the rapid development and wide use of multidisciplinary design optimization techniques, the use of multidisciplinary design optimization techniques to solve multidisciplinary design optimization problems has become the mainstream of multidisciplinary design optimization technique applications. The general multidisciplinary design optimization method includes:
the resource scheduling server acquires a parameter setting instruction from a first user;
the resource scheduling server determines an optimization algorithm and optimization parameters corresponding to the multidisciplinary design optimization model based on the parameter setting instruction;
The resource scheduling server acquires a multidisciplinary design optimization calculation task from the first user side, wherein the multidisciplinary design optimization calculation task is used for requesting cloud computing equipment to perform iterative calculation on the optimization parameters based on the optimization algorithm to obtain a multidisciplinary design optimization result;
the resource scheduling server sends all sub-tasks included in the multidisciplinary design optimization computing task to the corresponding cloud computing equipment;
And the resource scheduling server receives the multidisciplinary design optimization result sent by the cloud computing equipment and sends the multidisciplinary design optimization result to the first user side.
The data of the scheme is configured by defining relevant parameters through a user instead of the information of the system design end, so that scheme optimization and system design data are split, data chain transmission from top to bottom cannot be formed, and data inconsistency is caused. In addition, general optimized information such as data transmission, flow steps and the like is not uniformly described and expressed, and is only connected through a general front-end interface, including introduction of references and the like, so that a third party cannot be more deeply informed when viewing the optimization of the invention again, incomplete and nonstandard information expression is caused, and communication and exchange difficulties of multiparty designers are caused.
Disclosure of Invention
Aiming at the problems that the existing multidisciplinary design optimization method optimizes and cuts system design data, and data chain transmission from top to bottom cannot be formed, the invention provides a multisystem optimization method aiming at a complex system.
In order to achieve the technical purpose, the invention adopts the following technical scheme:
A multi-scheme optimization method for a complex system, comprising the steps of:
Defining data transmission based on a model diagram mode, and cleaning the OWL data by using GOPPRR model body files in an OWL format;
CSV data are cleaned, CSV data after OWL data are cleaned are further cleaned, and model data for scheme generation are obtained;
Generating a scheme, namely generating a fitting function from CSV data for scheme optimization;
Optimizing definition, namely optimizing definition is carried out by selecting an algorithm based on an input fitting function so as to balance and optimize the scheme;
and (3) analyzing results, displaying the results of the optimized schemes through a visual means, obtaining a more optimized scheme, and comparing the results of different schemes.
Further, OWL data cleaning: extracting relevant scheme information in GOPPRR model ontology files of OWL to form CSV data, and carrying out multi-scheme optimization by adopting a solving algorithm;
OWL data cleansing mainly includes GOPPRR model information extraction and data compilation.
Further, GOPPRR model information extraction is carried out, and relevant scheme information in GOPPRR model ontology files of OWL is extracted, wherein the relevant scheme information comprises ontology class diagrams, objects, points, attributes, points, relations and roles;
And data collection, namely collecting the data screened from the ROPPRR model body files of the OWL, collecting all the data into a CSV table, and adding the screened data as data collection input into each row and each column in the CSV in a self-defined way.
Further, CSV data cleansing includes a screening row and a screening column;
the screening lines comprise a reserved line according to attribute values, an excluded line according to attribute values, a reserved line according to numbers and an excluded line according to numbers;
When the columns are screened, the user performs the hook or cancel the hook on all the columns, only the hooked columns are reserved, and the column data which is not hooked can be eliminated.
Further, the optimization definition algorithm comprises a simulated annealing algorithm and a genetic algorithm;
The simulated annealing algorithm starts from a higher initial temperature, gradually reduces the temperature until the temperature is reduced to meet the heat balance condition, performs n-round searching at each temperature, adds random disturbance to the old solution during each round of searching to generate a new solution, and receives the new solution according to a certain rule; the following settings were made when the simulated degradation algorithm was used: initial temperature, final temperature, cooling rate, iteration number, objective function gold and constraints;
genetic algorithms, for an optimization problem, a certain number of candidate solutions (called individuals) can be represented abstractly as chromosomes, which allow populations to evolve toward better solutions; the solution is represented in binary (i.e., a string of 0 and 1), with evolution starting from a population of perfectly random individuals, with the next generation occurring; evaluating fitness of the whole population in each generation, randomly selecting a plurality of individuals from the current population (based on their fitness), generating a new living population by natural selection and mutation, the population becoming the current population in the next iteration of the algorithm; the following settings were made when the genetic algorithm was used: population size, crossover rate, mutation rate, maximum number of iterations, objective function, and constraint.
Further, the visualization of the result analysis includes:
a vertical bar chart;
A transverse bar chart;
a scattered point statistical graph;
and (5) analyzing the graph by using the xy coordinate axis curve.
Compared with the prior art, the invention has the following beneficial effects:
The OWL file generated by the system architecture model is imported, a plurality of operations such as cleaning and extracting are carried out on the file data, the direct use of the model data is realized, a scheme optimization model development interface is provided, a scheme optimization rule is defined, the scheme optimization based on model driving is realized, better communication among a plurality of departments is realized, and therefore a plurality of schemes of the complex system are optimized.
The method and the device take the information from GOPPRRR bodies of the OWL files, and can not conveniently help a user to obtain all the information because the obtained data are relatively scattered to be suitable for most cases, and only the example attribute values of all the body types can be separately grasped for summarization. However, based on the situation, a specific extraction rule can be formulated according to the model construction rule, so that the aim of extracting a certain class of instantiation attribute values in batches is fulfilled, and the workload of a user is reduced.
Drawings
FIG. 1 is an overall flow chart of a multi-scheme optimization method for a complex system in an embodiment of the invention;
FIG. 2 is a flowchart illustrating GOPPRR model information extraction in accordance with an embodiment of the present invention;
FIG. 3 is a flow chart of a simulated annealing algorithm in an embodiment of the invention;
FIG. 4 is a flowchart of a genetic algorithm according to an embodiment of the present invention.
Detailed Description
The invention will be further described with reference to examples and drawings, to which reference is made, but which are not intended to limit the scope of the invention.
As shown in fig. 1, the present embodiment provides a multi-scheme optimization method for a complex system, including the steps of:
Defining data transmission based on a model diagram mode, and cleaning the OWL data by using GOPPRR model body files in an OWL format;
CSV data are cleaned, CSV data after OWL data are cleaned are further cleaned, and model data for scheme generation are obtained;
Generating a scheme, namely generating a fitting function from CSV data for scheme optimization;
Optimizing definition, namely optimizing definition is carried out by selecting an algorithm based on an input fitting function so as to balance and optimize the scheme;
and (3) analyzing results, displaying the results of the optimized schemes through a visual means, obtaining a more optimized scheme, and comparing the results of different schemes.
OWL data cleaning: extracting relevant scheme information in GOPPRR model ontology files of OWL to form CSV data, and carrying out multi-scheme optimization by adopting a solving algorithm;
OWL data cleansing mainly includes GOPPRR model information extraction and data compilation.
As shown in fig. 2, GOPPRR is a model information extraction, which is to extract related scheme information in the GOPPRR model ontology file of OWL, wherein the related scheme information includes ontology class diagrams, objects, points, attributes, points, relationships and roles;
And (3) graph screening:
the figure (graph) is composed of 5 other elements, and the system is described in one box. The graph screening mainly screens the instances of the graphs in the GOPPRR model ontology file of the OWL, so that attribute values in the instances can be obtained.
Object screening:
objects (objects) are basic elements in the model, and may exist alone or may be linked with other objects. The object screening mainly screens the object class instance from the GOPPRR model body file of the OWL, so that attribute values in the instance can be obtained.
Attribute screening:
Properties (properties) cannot exist alone, and are attached to other metamodels to represent their properties. The attribute screening mainly screens the instances of attribute classes from the GOPPRR model body file of the OWL, so that attribute values in the instances can be obtained.
Spot screening:
Points (ports) are typically attached to objects, representing ports to which the objects are connected. The point screening mainly screens out the instances of the point class in the GOPPRR model body file of the OWL, so that attribute values in the instances can be obtained.
Relationship screening:
Relationships (relationships) are connected to objects by roles, representing the manner in which the objects interact with each other. The relation screening is mainly to screen out the instance of the relation class from the GOPPRR model body file of the OWL, so that the attribute value in the instance can be obtained.
Role screening:
Roles connect objects at both ends of the relationship, indicating in what way or identity the objects are connected. The role screening is mainly to screen instances of role classes from GOPPRR model ontology files of OWL, so that attribute values in the instances can be obtained.
And data collection, namely collecting the data screened from the ROPPRR model body files of the OWL, collecting all the data into a CSV table, and adding the screened data as data collection input into each row and each column in the CSV in a self-defined way.
The CSV data cleaning in the invention can generate final data without one time, and can be carried out for a plurality of times. CSV data cleansing includes screening rows and screening columns;
the screening lines comprise a reserved line according to attribute values, an excluded line according to attribute values, a reserved line according to numbers and an excluded line according to numbers;
Preserving according to attribute values:
the rows are reserved by attribute values, and the user first selects a particular column and then defines the upper and lower boundaries of the attribute values, thereby reserving the range of data for the particular column.
Excluding rows by attribute value:
when excluding rows by attribute values, the user first selects a particular column and then defines the upper and lower boundaries of the attribute values, thereby excluding the range of data for the particular column.
The rows are reserved by number:
the rows are reserved by number, and the user needs to select the first row number and the last row number, so that only row data in the selected range is reserved.
Rows are excluded by number:
the rows are reserved by number, and the user needs to select the first row number and the last row number, so that row data in the selected range is excluded.
When the columns are screened, the user performs the hook or cancel the hook on all the columns, only the hooked columns are reserved, and the column data which is not hooked can be eliminated.
Further, the optimization definition algorithm comprises a simulated annealing algorithm and a genetic algorithm;
as shown in fig. 3, the simulated annealing algorithm starts from a higher initial temperature, gradually reduces the temperature until the temperature is reduced to meet the heat balance condition, performs n-round searching at each temperature, adds random disturbance to the old solution during each round of searching to generate a new solution, and receives the new solution according to a certain rule; the following settings were made when the simulated degradation algorithm was used: initial temperature, final temperature, cooling rate, iteration number, objective function gold and constraints;
As shown in fig. 4, the genetic algorithm, for an optimization problem, a certain number of candidate solutions (called individuals) can be represented abstractly as chromosomes, evolving the population to a better solution; the solution is represented in binary (i.e., a string of 0 and 1), with evolution starting from a population of perfectly random individuals, with the next generation occurring; evaluating fitness of the whole population in each generation, randomly selecting a plurality of individuals from the current population (based on their fitness), generating a new living population by natural selection and mutation, the population becoming the current population in the next iteration of the algorithm; the following settings were made when the genetic algorithm was used: population size, crossover rate, mutation rate, maximum number of iterations, objective function, and constraint.
Further, the visualization of the result analysis includes:
a vertical bar chart;
A transverse bar chart;
a scattered point statistical graph;
and (5) analyzing the graph by using the xy coordinate axis curve.
Compared with the prior art, the invention has the following beneficial effects:
The OWL file generated by the system architecture model is imported, a plurality of operations such as cleaning and extracting are carried out on the file data, the direct use of the model data is realized, a scheme optimization model development interface is provided, a scheme optimization rule is defined, the scheme optimization based on model driving is realized, better communication among a plurality of departments is realized, and therefore a plurality of schemes of the complex system are optimized.
The method and the device take the information from GOPPRRR bodies of the OWL files, and can not conveniently help a user to obtain all the information because the obtained data are relatively scattered to be suitable for most cases, and only the example attribute values of all the body types can be separately grasped for summarization. However, based on the situation, a specific extraction rule can be formulated according to the model construction rule, so that the aim of extracting a certain class of instantiation attribute values in batches is fulfilled, and the workload of a user is reduced.
The multi-scheme optimization method for the complex system is described in detail. The description of the specific embodiments is only intended to aid in understanding the method of the present application and its core ideas. It should be noted that it will be apparent to those skilled in the art that various modifications and adaptations of the application can be made without departing from the principles of the application and these modifications and adaptations are intended to be within the scope of the application as defined in the following claims.

Claims (6)

1. A multi-scheme optimization method for a complex system, comprising the steps of:
Defining data transmission based on a model diagram mode, and cleaning the OWL data by using GOPPRR model body files in an OWL format;
CSV data are cleaned, CSV data after OWL data are cleaned are further cleaned, and model data for scheme generation are obtained;
Generating a scheme, namely generating a fitting function from CSV data for scheme optimization;
Optimizing definition, namely optimizing definition is carried out by selecting an algorithm based on an input fitting function so as to balance and optimize the scheme;
and (3) analyzing results, displaying the results of the optimized schemes through a visual means, obtaining a more optimized scheme, and comparing the results of different schemes.
2. The multi-scheme optimization method for complex systems according to claim 1 wherein OWL data cleaning: extracting relevant scheme information in GOPPRR model ontology files of OWL to form CSV data, and carrying out multi-scheme optimization by adopting a solving algorithm;
OWL data cleansing mainly includes GOPPRR model information extraction and data compilation.
3. The multi-scheme optimizing method for the complex system according to claim 2, wherein GOPPRR model information extraction is performed to extract related scheme information in GOPPRR model ontology files of OWL, wherein the related scheme information comprises ontology class diagrams, objects, points, attributes, points, relations and roles;
And data collection, namely collecting the data screened from the ROPPRR model body files of the OWL, collecting all the data into a CSV table, and adding the screened data as data collection input into each row and each column in the CSV in a self-defined way.
4. A multi-scheme optimization method for complex systems according to claim 3 wherein CSV data cleansing comprises screening rows and screening columns;
the screening lines comprise a reserved line according to attribute values, an excluded line according to attribute values, a reserved line according to numbers and an excluded line according to numbers;
When the columns are screened, the user performs the hook or cancel the hook on all the columns, only the hooked columns are reserved, and the column data which is not hooked can be eliminated.
5. The multi-scheme optimization method for complex systems according to claim 4 wherein the optimization-defined algorithms include simulated annealing algorithms and genetic algorithms;
The simulated annealing algorithm starts from a higher initial temperature, gradually reduces the temperature until the temperature is reduced to meet the heat balance condition, performs n-round searching at each temperature, adds random disturbance to the old solution during each round of searching to generate a new solution, and receives the new solution according to a certain rule; the following settings were made when the simulated degradation algorithm was used: initial temperature, final temperature, cooling rate, iteration number, objective function gold and constraints;
genetic algorithms, for an optimization problem, a certain number of candidate solutions (called individuals) can be represented abstractly as chromosomes, which allow populations to evolve toward better solutions; the solution is represented in binary (i.e., a string of 0 and 1), with evolution starting from a population of perfectly random individuals, with the next generation occurring; evaluating fitness of the whole population in each generation, randomly selecting a plurality of individuals from the current population (based on their fitness), generating a new living population by natural selection and mutation, the population becoming the current population in the next iteration of the algorithm; the following settings were made when the genetic algorithm was used: population size, crossover rate, mutation rate, maximum number of iterations, objective function, and constraint.
6. The multi-scheme optimization method for complex systems according to claim 5 wherein the visualization of the result analysis comprises:
a vertical bar chart;
A transverse bar chart;
a scattered point statistical graph;
and (5) analyzing the graph by using the xy coordinate axis curve.
CN202410213551.XA 2024-02-27 2024-02-27 Multi-scheme optimization method for complex system Pending CN118035223A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202410213551.XA CN118035223A (en) 2024-02-27 2024-02-27 Multi-scheme optimization method for complex system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202410213551.XA CN118035223A (en) 2024-02-27 2024-02-27 Multi-scheme optimization method for complex system

Publications (1)

Publication Number Publication Date
CN118035223A true CN118035223A (en) 2024-05-14

Family

ID=90998281

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202410213551.XA Pending CN118035223A (en) 2024-02-27 2024-02-27 Multi-scheme optimization method for complex system

Country Status (1)

Country Link
CN (1) CN118035223A (en)

Similar Documents

Publication Publication Date Title
JP6758368B2 (en) Data discovery node
CN103853821B (en) Method for constructing multiuser collaboration oriented data mining platform
KR20060044772A (en) Using tables to learn trees
Fuertes et al. Chaotic genetic algorithm and the effects of entropy in performance optimization
Abdul-Rahman et al. An adaptive parameter binary-real coded genetic algorithm for constraint optimization problems: Performance analysis and estimation of optimal control parameters
CN110807026A (en) Automatic capture system for analyzing financial big data blood relationship
Wibig Dynamic programming and genetic algorithm for business processes optimisation
JP2020129222A (en) Model output program, model output method and model output device
Ostapchenko et al. Functional organization of system of support of decision-making of organizational management
Xu et al. IMVis: Visual analytics for influence maximization algorithm evaluation in hypergraphs
Gao et al. A regularity property-driven evolutionary algorithm for multiobjective optimization
CN118035223A (en) Multi-scheme optimization method for complex system
GB2589828A (en) Causal reasoning and counterfactual probabilistic programming framework using approximate inference
US20220179862A1 (en) Optimizing breakeven points for enhancing system performance
CN114881521A (en) Service evaluation method, device, electronic equipment and storage medium
CN113850454A (en) Method, device, equipment and medium for manufacturing parts based on industrial neural network
Hong Discrete optimization via simulation using coordinate search
Choi et al. Machine learning-based automated data visualization: A meta-feature engineering approach
CN111400050A (en) Method and device for allocating resources to execute tasks
Liu et al. Paths study on knowledge convergence and development in computational social science: Data metric analysis based on web of science
Kazakov et al. Developing a Method for Building Business Process Models Based on Graph Neural Networks in the Absence of Task Identifier Data
Jaiswal et al. Evocut: A new generalization of albert-barabasi model for evolution of complex networks
JP7533041B2 (en) Information processing device, information processing method, and program
Aguilar-Ruiz et al. Generation of management rules through system dynamics and evolutionary computation
Ammersdörfer et al. Access Logics for Situation-Appropriate Selection and Introduction of Methods in Engineering Design

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination