CN113343589B - Genetic-random constant-based acidic natural gas hydrate generation condition prediction method adopting genetic expression programming - Google Patents
Genetic-random constant-based acidic natural gas hydrate generation condition prediction method adopting genetic expression programming Download PDFInfo
- Publication number
- CN113343589B CN113343589B CN202110765785.1A CN202110765785A CN113343589B CN 113343589 B CN113343589 B CN 113343589B CN 202110765785 A CN202110765785 A CN 202110765785A CN 113343589 B CN113343589 B CN 113343589B
- Authority
- CN
- China
- Prior art keywords
- genetic
- natural gas
- algorithm
- gene expression
- constant
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/02—Reliability analysis or reliability optimisation; Failure analysis, e.g. worst case scenario performance, failure mode and effects analysis [FMEA]
Abstract
The invention discloses a genetic-random constant-based acidic natural gas hydrate generation condition prediction method programmed by a gene expression. The method comprises the following steps: acquiring data of components, pressure and temperature generated by the acidic natural gas hydrate; calculating the molar mass and the acid factor contribution rate; dividing original data into a training set and a test set; programming and establishing a dynamic formula prediction model of the generation temperature and the influence factors of the acidic natural gas hydrate based on a gene expression combined with a random constant; introducing correction parameters into the formula model; and optimizing the correction parameters based on the genetic algorithm. The method has the beneficial effects that the defects of slow solving speed and slightly poor prediction precision limited by a single tree structure of the traditional genetic programming algorithm are overcome by the novel method of the genetic expression programming algorithm based on the genetic-random constant. The method can accurately predict the generation conditions of different acidic natural gas hydrates with the pressure of 15MPa and the pressure of 300K, and ensure the ordered production of the acid-containing gas field gathering and transportation system.
Description
Technical Field
The invention belongs to the field of flow guarantee of natural gas gathering and transportation systems, and particularly relates to a genetic-random constant-based acidic natural gas generation condition prediction method based on genetic expression programming.
Background
At present, a large number of acid gas fields exist at home and abroad and are urgently to be developed, however, the developed acid natural gas is usually generated in links such as throttling of a gathering and transportation system due to changes of pressure and temperature. Therefore, the prediction of the generation conditions of the acid natural gas hydrate is a hot problem at present.
The existing means for predicting the natural gas hydrate generating conditions commonly found on site are mainly commercial software such as HYSYS and the like, but the commercial software is not ideal for predicting the acid natural gas hydrate generating conditions, because a plurality of emerging methods are proposed one by one. Currently, the most interesting methods are machine learning algorithms, including neural network algorithms, decision tree algorithms, genetic programming algorithms, and the like. The genetic programming algorithm capable of generating the tree-shaped explicit expression is an algorithm with strong interpretability, but the calculation speed, the single tree-shaped structure and the accuracy of the algorithm still need to be improved.
Therefore, it is necessary to research an explicit expression calculation method capable of improving the calculation speed and accuracy of the genetic programming algorithm and accurately predicting the acid natural gas hydrate.
Disclosure of Invention
The invention mainly aims to design a genetic-random constant-based acidic natural gas hydrate generation condition prediction method programmed by a gene expression. The genetic expression algorithm integrates the advantages of genetic code (GP) and Genetic Algorithm (GA), and both the genetic expression algorithm and genetic programming have flexible tree structures on the expression. Both the gene expression algorithm and the genetic algorithm have relatively simple fixed-length linear coding in the expression form. Therefore, the gene expression algorithm integrates the advantages of the genetic algorithm and the genetic programming algorithm, and the calculation speed of the gene expression algorithm is 2-4 orders of magnitude faster than that of the common genetic programming when the expression optimization for solving the complex problem is carried out. And the gene expression algorithm GEP-RNC combined with the random constant has higher precision.
In order to solve the problems of low solving speed, single tree structure and poor accuracy of a common genetic programming algorithm, the invention establishes a genetic-random constant-based genetic expression programming-based acidic natural gas hydrate generation condition prediction method. Provides a solution for the problem of the generation of acid natural gas hydrate in a large number of acid gas field gathering and transportation systems at home and abroad. The overall structure of the invention is shown in figure 1. The technical scheme adopted by the invention is as follows:
providing a genetic-random constant based genetic expression programmed acidic natural gas hydrate formation condition prediction method, the method comprising the steps of:
collecting basic data of temperature, pressure and the like of hydrates generated by acidic natural gas under different components;
preprocessing the data, performing characteristic construction operation on the original data, calculating the contribution rate C and molar mass M of new variable acid factors, and dividing the original data into a training set and a test set;
setting basic parameters of a gene expression programming regression model combined with a random constant, inputting a training set to perform modeling training, searching a relational expression of the generation temperature of the acid natural gas hydrate, finally obtaining the relational expression with the best fitness, and successfully establishing the relation between the key parameters and the generation temperature of the hydrate;
step four, introducing correction parameters into the trained gene expression formula model combining the random constants in the step three;
optimizing the explicit expression model introduced with the correction coefficient by using a genetic algorithm, and obtaining a correction parameter with the best fitness after iterative heredity for multiple times;
and step six, inputting the parameters of the acidic natural gas to be predicted into an explicit expression model with correction coefficients introduced, and predicting the generation conditions of the hydrate.
The acidic hydrate formation condition data are experimental data published in domestic and foreign literatures.
The gas molar components refer to different molar fractions of the methane component and other gas components such as carbon dioxide and hydrogen sulfide contained in the sour natural gas.
The calculation formula of the acid factor contribution rate is as follows:
in the formula x 1 -the molar fraction of methane in the gas,%;
x 2 -the carbon dioxide mole fraction in the gas,%;
x 3 -the molar fraction of hydrogen sulphide in the gas,%;
the molar mass calculation formula is as follows:
M＝16.043*x 1 +44.0095*x 2 +34.076*x 3 (2)
in the formula x 1 -the mole fraction of methane in the gas,%;
x 2 -carbon dioxide mole fraction in gas,%;
x 3 -the mole fraction of hydrogen sulphide in the gas,%;
the genetic expression programming algorithm is an explicit expression generation algorithm with strong interpretability, is a novel self-adaptive evolution algorithm based on biological gene structure and function, is developed from a genetic algorithm and a genetic programming algorithm, and overcomes the defects of the genetic algorithm and the genetic programming algorithm while absorbing the advantages of the genetic algorithm and the genetic programming algorithm.
The gene expression algorithm comprises the following steps:
1) initializing a population: randomly generating a plurality of individuals according to a set function set, initializing a population, and setting a termination set and containing a random constant string;
2) evaluation: evaluating the fitness of all individuals in the population according to a set fitness function;
3) selecting: the fitness obtained by calculating the fitness function is adopted, all individuals are arranged, and two individuals needing genetic operation are selected from the individuals;
4) mutation: in two selected individuals, each position of the gene was randomly tested and if the mutation probability is met, the position would be encoded with a newly generated variation. If a mutation occurs at the head of a gene, all elements in the function and terminator can be selected; if an ectopic position occurs at the tail of a gene, only the elements in the terminator may be selected;
5) inserting strings: firstly, selecting a substring from any position of a gene, then randomly inserting the substring into the head of the gene (except for the first position), and pushing elements of the head backwards and forwards to cut off the elements exceeding the length of the head;
6) and (3) recombination: firstly, selecting the position of a whole gene, and then interchanging the correspondingly selected gene in two parent chromosomes;
7) and (4) terminating: and (5) repeating the steps 2-5 until the termination condition is met.
The gene expression algorithm combined with the random constant overcomes the defect that the originally generated explicit expression of the tree structure has no constant on the basis of the gene expression. It adds a Dc domain with length equal to tail length t and symbolized corresponding constant at the tail of the gene of chromosome generated by gene expression algorithm, and the array stores a set of candidate random constants with set number. The GEP-RNC algorithm has an additional set of genetic operators including RNC variation, Dc inverse string, DcIS run-in, Dc permutation, constant correction, constant run-in, constant range lookup. The content in the set of Dc domain and candidate constants of the GEP-RNC algorithm can also participate in various genetic operations of the GEP, and the candidate constants are evolved along with the evolution process, so that the random constants can effectively generate proper diversity during operation, and the diversity is maintained in the later evolution process. The GEP-RNC algorithm does not need to make hypothesis on the function form, overcomes the defect that the genetic programming algorithm needs to make a hypothesis, only needs to set a function symbol set and a terminal symbol set, and obtains the function relation through continuous evolution through operations such as selection, variation, string insertion, recombination and the like.
The genetic algorithm is used as an optimization algorithm and is used for further improving the precision of the explicit expression generated by the genetic expression programming algorithm combined with the random constant.
The genetic-random constant gene expression programming algorithm is a high-precision combination method, the steps are shown in figure 2, an explicit expression structure is generated by means of the random constant gene expression algorithm, values of all items are corrected through correction coefficients, and prediction precision can be remarkably improved.
The combination method comprises the steps that a gene expression algorithm with random constants generates a basic explicit expression structure, the genetic algorithm optimizes correction parameters for introducing explicit expressions, and the combination method has higher accuracy than a common gene expression programming algorithm.
The basic function set of the gene expression algorithm comprises { +, -, ×, -, exp, sqrt, arctan, tanh, pow, ^ 1, ^2 }. The basic parameters to be set by the gene expression combined with the random constant comprise a population size N, a gene head length h, a function maximum operand N, a gene number k, an adaptability value maxf for terminating iteration, a maximum iteration tree maxI and an exchange probability P tr Probability of mutation P mu Recombination probability P re The range, length, etc. of the fitness and constant terms.
The population size, the base factor, the head length, the string insertion probability, the variation probability and the recombination probability of the gene expression algorithm are respectively 150, 7, 8, 0.05, 0.02 and 0.1, gene expression trees are connected in an addition mode and are coded by adopting a mixed constant, the population fitness function is the root mean square error, the constant term range is 0-100, and the constant term length is 7.
Fitness function is root mean square error RMSE:
n represents the number of samples;
HFT real,i experimental data of acid natural gas hydrate formation temperature, ° c;
HFT prediction,i -prediction data of acid natural gas hydrate formation temperature, ° c;
the individual of the optimal gene expression is the optimal explicit expression structure, and the structure is simplified to obtain:
in the formula x 1 -the molar fraction of methane in the gas,%;
x 2 -the carbon dioxide mole fraction in the gas,%;
x 3 -the mole fraction of hydrogen sulphide in the gas,%;
p-pressure, MPa;
c-acidity factor contribution,%;
m is gas molar mass, kg/mol;
a is constant term, kg/mol;
constant term a 1 -a 12 Comprises the following steps:
TABLE 1 Gene expression constants
The introducing of the correction parameter k to the explicit expression structure specifically includes:
HFT＝k 1 *y 1 +k 2 *y 2 +k 3 *y 3 +k 4 *y 4 +k 5 *y 5 +k 6 *y 6 +k 7 *y 7 +b (5)
the optimized variables of the genetic algorithm are correction parameters, the optimization function is an explicit expression obtained by a gene expression programming algorithm, the population fitness function is root mean square error, and the population size, the variation probability and the cross probability are respectively 150, 0.05 and 0.1.
The correction parameters after the genetic algorithm training optimization are as follows:
TABLE 2 explicit expression correction parameters
Compared with the prior art, the invention has the following advantages:
the invention designs a model for predicting the generation condition of the acid natural gas hydrate, and the model combines a gene expression programming algorithm combined with a random constant and a genetic algorithm, so that the combined method not only can generate an explicit expression in the form of a multi-gene multi-tree structure, but also overcomes the defects of low solving speed of the genetic programming algorithm and slightly poor prediction precision caused by single tree structure limitation, and can provide the basis of flow guarantee analysis for a gathering and transportation system of an acid natural gas field.
Drawings
FIG. 1 is a flow chart of acidic natural gas hydrate generation condition prediction of genetic expression programming algorithm based on genetic-random constants
FIG. 2 is a flow chart of a gene expression programming algorithm incorporating random constants
The method is based on genetic-random constant gene expression programming algorithm, and realizes accurate prediction of acid natural gas hydrate generation conditions. In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As shown in fig. 1, the present invention provides a method for predicting sour natural gas hydrate formation conditions, the method comprising the steps of:
acquiring basic data such as detailed molar components and pressure of natural gas;
secondly, preprocessing data, performing characteristic construction operation on the original data, calculating an acidic contribution factor C and a molar mass M, and dividing a training set and a test set;
step three, taking the methane molar component, the carbon dioxide molar component, the hydrogen sulfide molar component, the pressure, the acidic contribution factor and the molar mass as input variables, taking the hydrate generation temperature as a target variable, training a gene expression algorithm combined with a random constant, and testing by using a test set to obtain an explicit expression structure with optimal individual fitness;
introducing correction parameters to an explicit expression structure;
step five, optimizing the correction parameters by adopting a genetic algorithm to determine the optimal correction parameters;
step six, obtaining an optimal explicit expression formula;
in step one, the basic parameters to be collected include the molar composition of methane
Molar component of carbon dioxide
Molar composition of hydrogen sulfide
Pressure P, temperature HFT;
in the second step, the calculation formula of the acidity contribution factor C and the molar mass M is as follows:
M＝16*x 1 +44*x 2 +38*x 3 (7)
in the third step, a gene expression algorithm is adopted for training, and the original data are divided into a training set and a testing set, which respectively occupy 70 percent and 30 percent. The obtained explicit expression of the individual with the optimal fitness is as follows:
in the fourth step, a correction parameter is introduced into the explicit expression generated in the third step, and the structure is as follows:
y＝k 1 *y 1 +k 2 *y 2 +k 3 *y 3 +k 4 *y 4 +k 5 *y 5 +k 6 *y 6 +k 7 *y 7 +b (9)
in the fifth step, the correction parameters obtained by genetic algorithm optimization are as follows:
TABLE 3 explicit expression correction parameters
The application of the present invention is further described below with reference to specific examples.
Example (c): the gathering and transportation system of a certain acid natural gas field contains acid gas, and the generation temperature of the natural gas hydrate is obtained under the condition of known acid natural gas components and pressure.
The first step is as follows: the composition data and pressure data for natural gas are known as shown in table 4:
TABLE 4 acid gas base data
The second step is that: calculating the acid natural gas hydrate generation temperature using a formula generated by genetic expression of genetic-random constants: the results of the formula calculations were compared with the results of the experiments, as shown in table 5.
TABLE 5 HFT comparison of hydrate formation temperatures
As can be seen from Table 3, the calculation result of the formula is very similar to the simulation result, and the relative errors are small and within 5%. The example proves that the genetic expression of the genetic-random constant provided by the design can realize high-precision prediction of the generation condition of the natural gas hydrate.
The invention provides a method for predicting the generation condition of an acidic natural gas hydrate. The method solves the defects that the common genetic programming algorithm needs an assumed form in the prediction of the generation condition of the natural gas hydrate, the solving speed is low, and the prediction precision is slightly poor due to the limitation of a single tree structure, and the obtained calculation formula can provide a basis for flow guarantee analysis for a gathering and transportation system capable of becoming an acid natural gas field.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.
Claims (4)
1. A method for predicting the generation condition of an acidic natural gas hydrate based on genetic-random constant gene expression programming is characterized by comprising the following steps:
step one, collecting mole fractions of mole components of acidic natural gas including methane, carbon dioxide and hydrogen sulfide under a stable condition of generating hydrate, pressure and temperature data;
step two, constructing characteristics, introducing a new characteristic acid factor contribution rate C and a gas molar mass M, and dividing original data into a training set and a testing set;
setting basic parameters of a gene expression programming algorithm combined with a random constant, taking the parameters in the first step and the second step as input variables, taking the hydrate generation temperature as an output variable to construct a training function, and performing learning regression by using the gene expression programming algorithm to obtain a hydrate generation condition prediction explicit expression of the gene expression programming:
HFT＝y 1 +y 2 +y 3 +y 4 +y 5 +y 6 +y 7 +y 8 +y 9 (2)
in the formula x 1 -the mole fraction of methane in the gas,%;
x 2 -carbon dioxide mole fraction in gas,%;
x 3 -the mole fraction of hydrogen sulphide in the gas,%;
p-pressure, MPa;
c-acid factor contribution,%;
m-gas molar mass, kg/mol;
a-constant term;
TABLE 1 Gene expression constants
The above table is a constant term a 1 -a 12 ；
Introducing correction parameters into the hydrate generation condition formula model obtained in the step three, and training and learning the correction parameters by adopting a genetic algorithm to obtain optimal correction parameters;
and step five, testing by adopting the test set to determine the optimal explicit expression.
2. The method for predicting acid natural gas hydrate formation conditions programmed based on genetic-random constants according to claim 1, wherein step two introduces a new characteristic acid factor contribution rate C:
in the formula x 1 -the mole fraction of methane in the gas,%;
x 2 -carbon dioxide mole fraction in gas,%;
x 3 -the molar fraction of hydrogen sulphide in the gas,%.
3. The method of predicting acidic natural gas hydrate generation conditions based on genetic-random constant gene expression programming of claim 1, comprising generating a basic explicit expression structure by a random constant gene expression algorithm;
the basic function set by the gene expression algorithm comprises { +, -, ×, - +, exp, sqrt, arctan, tanh, pow, ^ (-1), ^2}, a termination set comprises constants and independent variables, the population size, the basis factor, the head length, the string insertion probability, the variation probability and the recombination probability are respectively 150, 7, 8, 0.05, 0.02 and 0.1, a gene expression tree is connected by addition, and the population fitness function is root mean square error RMSE;
random constants are combined, so that constant terms can be generated in an original gene expression algorithm tree structure, mixed constant coding is adopted, the range of the constant terms is 0-100, and the length of the constant terms is 7;
the selection operation adopts an improved wheel algorithm, firstly, the fitness of each individual is calculated, then, the selection probability of each individual in a selected set is calculated, after the probability of each alternative individual is determined, a random number Pc between [0,1] is randomly generated, the cumulative probability individual corresponding to the random number Pc is the selected individual, the method can ensure that the individual with high fitness has higher selected probability, and excellent genes are inherited.
4. The method of predicting acidic natural gas hydrate formation conditions programmed by a genetic-random constant-based gene expression according to claim 1, wherein step four introduces correction parameters into the explicit expression:
HFT＝k 1 *y 1 +k 2 *y 2 +k 3 *y 3 +k 4 *y 4 +k 5 *y 5 +k 6 *y 6 +k 7 *y 7 +b (4)
step four, optimizing the correction parameters by adopting a genetic algorithm, so that the accuracy of the combination method can be further improved;
TABLE 2 explicit expression correction parameters
The combination method can accurately predict the generation conditions of different acidic natural gas hydrates with the pressure of 15MPa and the pressure of 300K below.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110765785.1A CN113343589B (en) | 2021-06-30 | 2021-06-30 | Genetic-random constant-based acidic natural gas hydrate generation condition prediction method adopting genetic expression programming |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110765785.1A CN113343589B (en) | 2021-06-30 | 2021-06-30 | Genetic-random constant-based acidic natural gas hydrate generation condition prediction method adopting genetic expression programming |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113343589A CN113343589A (en) | 2021-09-03 |
| CN113343589B true CN113343589B (en) | 2022-07-26 |
Family
ID=77482832
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110765785.1A Active CN113343589B (en) | 2021-06-30 | 2021-06-30 | Genetic-random constant-based acidic natural gas hydrate generation condition prediction method adopting genetic expression programming |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113343589B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017205634A1 (en) * | 2016-05-27 | 2017-11-30 | Ionic Material, Inc. | ELECTROCHEMICAL CELL HAVING SOLID lONICALLY CONDUCTING POLYMER MATERIAL |
| CN110378075A (en) * | 2019-08-19 | 2019-10-25 | 西南石油大学 | A kind of acid and nonacid natural gas water content amount prediction technique |
| CN112733307A (en) * | 2021-01-12 | 2021-04-30 | 北京市燃气集团有限责任公司 | Urban natural gas load prediction method and device in heating season |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2009151511A1 (en) * | 2008-04-29 | 2009-12-17 | Therasis, Inc. | Systems and methods for identifying combinations of compounds of therapeutic interest |
| US9526784B2 (en) * | 2013-09-06 | 2016-12-27 | President And Fellows Of Harvard College | Delivery system for functional nucleases |
| CN107229972A (en) * | 2017-03-10 | 2017-10-03 | 东莞理工学院 | A kind of global optimization based on Lamarch inheritance of acquired characters principle, search and machine learning method |
-
2021
- 2021-06-30 CN CN202110765785.1A patent/CN113343589B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017205634A1 (en) * | 2016-05-27 | 2017-11-30 | Ionic Material, Inc. | ELECTROCHEMICAL CELL HAVING SOLID lONICALLY CONDUCTING POLYMER MATERIAL |
| CN110378075A (en) * | 2019-08-19 | 2019-10-25 | 西南石油大学 | A kind of acid and nonacid natural gas water content amount prediction technique |
| CN112733307A (en) * | 2021-01-12 | 2021-04-30 | 北京市燃气集团有限责任公司 | Urban natural gas load prediction method and device in heating season |
Non-Patent Citations (5)
| Title |
|---|
| "Integrating Support Vector Regression with Genetic Algorithm for Hydrate Formation Condition Prediction";Jie Cao等;《processes》;20200427;第8卷(第5期);第1-11页 * |
| "Prediction of hydrate formation temperature using gene expression programming";Menad Nait Amar;《Journal of Natural Gas Science and Engineering》;20210223;第89卷;第1-12页 * |
| "基于支持向量机结合遗传算法的天然气水合物相平衡研究";马贵阳等;《天然气工业》;20170531;第37卷(第5期);第46-52页 * |
| 中俄东线天然气管道投产初期水合物预测及防治模拟;张科嘉等;《油气储运》;20200408;第39卷(第07期);第821-826页 * |
| 基于基因表达式编程算法的磨料射流切割深度预测模型;王瑞和等;《中国石油大学学报(自然科学版)》;20150228;第39卷(第01期);第60-65页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113343589A (en) | 2021-09-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111832101B (en) | Construction method of cement strength prediction model and cement strength prediction method | |
| CN105335752A (en) | Principal component analysis multivariable decision-making tree-based connection manner identification method | |
| CN102331966A (en) | Software test data evolution generation system facing path | |
| CN111553117B (en) | Polyester intrinsic viscosity control method based on stacked ensemble learning of genetic algorithm | |
| CN102521654B (en) | Supercritical water oxidation reaction kinetic model parameter estimation method employing RNA (Ribonucleic Acid) genetic algorithm | |
| CN112488315A (en) | Batch scheduling optimization method based on deep reinforcement learning and genetic algorithm | |
| CN112733417A (en) | Abnormal load data detection and correction method and system based on model optimization | |
| CN111259953A (en) | Equipment defect time prediction method based on capacitive equipment defect data | |
| CN113449930A (en) | Optical fiber preform preparation quality prediction method based on BP neural network | |
| CN113222165A (en) | Quantum line optimization method based on genetic algorithm | |
| CN104732067A (en) | Industrial process modeling forecasting method oriented at flow object | |
| CN112488147A (en) | Redundancy removal active learning method based on countermeasure network | |
| CN114707692A (en) | Wetland effluent ammonia nitrogen concentration prediction method and system based on hybrid neural network | |
| CN113343589B (en) | Genetic-random constant-based acidic natural gas hydrate generation condition prediction method adopting genetic expression programming | |
| CN110879778B (en) | Novel dynamic feedback and improved patch evaluation software automatic repair method | |
| CN115620046A (en) | Multi-target neural architecture searching method based on semi-supervised performance predictor | |
| CN115114838A (en) | Spectral characteristic wavelength selection method based on particle swarm algorithm thought and simulated annealing strategy | |
| CN114926075A (en) | Mechanical part production scheduling method based on man-hour prediction | |
| CN111178627B (en) | Neural network hybrid optimization prediction method based on SPCA | |
| CN112348275A (en) | Regional ecological environment change prediction method based on online incremental learning | |
| Wang et al. | Machine learning and ensemble learning for transcriptome data: principles and advances | |
| CN105069323A (en) | Memetic algorithm based microbial fermentation control and optimization method | |
| CN113326665B (en) | Genetic programming-based acidic natural gas hydrate generation temperature prediction method | |
| CN116468167A (en) | Biogas yield prediction method and system based on improved GA-BP neural network | |
| CN114184695A (en) | Parameter optimization-based method and system for predicting gas concentration in random forest oil |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |