CN108446459B - Fuzzy semantic reasoning-based coking process heat consumption influence factor optimization method - Google Patents

Fuzzy semantic reasoning-based coking process heat consumption influence factor optimization method Download PDF

Info

Publication number
CN108446459B
CN108446459B CN201810170875.4A CN201810170875A CN108446459B CN 108446459 B CN108446459 B CN 108446459B CN 201810170875 A CN201810170875 A CN 201810170875A CN 108446459 B CN108446459 B CN 108446459B
Authority
CN
China
Prior art keywords
semantic
heat consumption
semantics
fuzzy
coking process
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
Application number
CN201810170875.4A
Other languages
Chinese (zh)
Other versions
CN108446459A (en
Inventor
甘健侯
周菊香
唐晓宁
文斌
王俊
邹伟
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.)
Yunnan Normal University
Original Assignee
Yunnan Normal University
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 Yunnan Normal University filed Critical Yunnan Normal University
Priority to CN201810170875.4A priority Critical patent/CN108446459B/en
Publication of CN108446459A publication Critical patent/CN108446459A/en
Application granted granted Critical
Publication of CN108446459B publication Critical patent/CN108446459B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/10—Numerical modelling
    • 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
    • G06F2119/08—Thermal analysis or thermal optimisation

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Evolutionary Computation (AREA)
  • Geometry (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention relates to a fuzzy semantic reasoning-based coking process heat consumption influence factor optimization method, and belongs to the technical field of metallurgical intelligent control and metallurgical automation. According to the invention, according to the AFS axiom fuzzy set theory, aiming at the influence factors of the heat consumption in the coking process, the internal fuzzy semantics are excavated, the key simple semantics of each influence factor are extracted through the evaluation of the semantics and the calculation of the importance factor, the parameter range of the heat consumption influence factor is set, and the optimization of the heat consumption influence factor in the coking process is completed. The invention optimizes the heat consumption influence factors in the coking process, and realizes the aim of ensuring the qualified coke quality and simultaneously ensuring lower heat consumption in the actual production process.

Description

Fuzzy semantic reasoning-based coking process heat consumption influence factor optimization method
Technical Field
The invention relates to a fuzzy semantic reasoning-based coking process heat consumption influence factor optimization method, and belongs to the technical field of metallurgical intelligent control and metallurgical automation.
Background
With the development of computer technology and artificial intelligence, advanced management technologies such as fuzzy logic, artificial neural networks, evolutionary computing and integrated intelligent models thereof have been introduced into various production links in the field of metallurgy, so as to realize automatic and intelligent management and control of metallurgy. As one of the important links in metallurgical production, the coking process includes a plurality of process links, each of which generates a plurality of technical parameters and data. Taking heat consumption as an example, in the actual production process, the heat consumption is expected to be lower as much as possible while the quality of coke is ensured to be qualified, so that the production cost is reduced and the production efficiency is improved. However, the heat consumption is influenced by a plurality of factors, and random and fuzzy uncertainties exist, so that an effective knowledge representation and discovery means is needed to accurately mine and describe the internal rules of the plurality of influencing factors of the heat consumption in the coking process and the relationship between the influencing factors and the heat consumption, and finally, an optimization goal of ensuring the low heat consumption while the coke quality is qualified is achieved.
The fuzzy semantic rule based system can effectively solve the problems of knowledge representation and reasoning, and represents knowledge through fuzzy rules and uses fuzzy logic to reason and apply the knowledge. The method based on the fuzzy semantic rule can solve the uncertainty of the data more effectively, the description of the data is more practical, and the method is high in analysis performance and easy to understand. To discuss how the information contained in the observed data is used to determine membership functions of Fuzzy concepts, and further, random and Fuzzy uncertainties can be handled by consistent mathematical methods, Axiomatic Fuzzy Set (AFS) theory is proposed to simulate the mechanism by which humans perceive and observe things and then form concepts and generate logic, discussing Fuzzy concepts and their logical operations from a more abstract and general level.
In view of the advantages of the AFS theory in data analysis, in the previous work, the idea of the AFS theory is introduced for the first time, fuzzy semantic representation is carried out on the heat consumption influence factors in the coking process, and relevant semantic rules are extracted from the fuzzy semantic representation, so that optimization of the heat consumption factors is realized. Because the quantity of the mined semantic rules is large, part of the semantic rules are finally selected by combining with expert experience, and the optimal combination of the heat consumption influence factors is obtained by analysis on the basis, so that the optimization of the heat consumption factors is preliminarily realized. However, the process lacks objectivity, still has certain limitations in practical application, and cannot completely achieve automatic optimization of heat consumption influence factors in the coking process.
Coking is an important process in ferrous metallurgy production, and the quality of produced coke directly influences the efficiency and quality of a subsequent iron-making process. The stable operation and the optimized management of the coking production process are key factors for determining important indexes such as the quality of coking products, the energy consumption of the process and the like. In order to more accurately mine and describe the internal rules of various influence factors of the heat consumption in the coking process, the relationship between the various influence factors and the heat consumption in the coking process is further provided, so that the optimization target of ensuring lower heat consumption when the coke quality is qualified is achieved; the invention provides a new heat consumption influence factor optimization algorithm.
Disclosure of Invention
The invention provides a fuzzy semantic reasoning-based optimization method for heat consumption influence factors in a coking process, which is used for solving the problem of optimization of the heat consumption influence factors in the coking process and achieving the aim of ensuring qualified coke quality and simultaneously reducing heat consumption in the actual production process.
The invention adopts the idea of AFS theory, describes the internal rules of various influence factors of the heat consumption in the coking process by fuzzy semantics, further excavates the relation between the various influence factors and the heat consumption in the coking process, defines a semantic evaluation function and an importance calculation method on the basis, designs a key semantic extraction algorithm of the heat consumption influence factors, finally completes the automatic optimization of the coking heat consumption influence factors, and realizes the aim of ensuring the qualified coke quality and simultaneously ensuring the lower heat consumption in the actual production process;
in order to achieve the aim, the production data of the actual coking process of a large iron and steel enterprise is taken as the basis, the average value of heat consumption is taken as a boundary, the data is divided into two types of data samples with higher and lower heat consumption, and the internal relation and the law between the heat consumption of the two types of samples and the influence factors thereof are mined through an AFS theory. The invention provides a coking process heat consumption influence factor optimization method based on fuzzy semantic reasoning on the basis of the existing work.
The specific technical scheme of the invention is as follows: the method comprises the steps of mining inherent fuzzy semantics aiming at the influence factors of heat consumption in the coking process according to an AFS axiom fuzzy set theory, extracting key simple semantics of each influence factor through semantic evaluation and importance factor calculation, setting the parameter range of the heat consumption influence factors, and finishing the optimization of the heat consumption influence factors in the coking process.
The optimization method comprises the following specific steps:
step1, extracting and preprocessing sample data;
the specific steps of Step1 are as follows:
step1.1, taking the average value of the heat consumption as a limit, and dividing the data into two types of data samples with higher heat consumption and lower heat consumption respectively according to the average value higher than and lower than the heat consumption;
step1.2, aiming at the two types of data samples, keeping main attributes from original data; according to the invention, heat consumption influence factors in the coking process are analyzed, and 9 main attributes are selected as research objects, including coking time, coal charging amount (single hole), coke oven gas main pipe flow, furnace temperature coefficient (uniformity coefficient), flue gas side suction, flue gas coke side suction, straight running machine side temperature, straight running coke side temperature and coal moisture;
and Step1.3, transforming and normalizing the attribute value of the sample space of each attribute in a linear transformation mode to finally form a sample data set.
Step2, constructing a sample semantic set according to an axiom fuzzy set theory;
the specific steps of Step2 are as follows:
step2.1, performing simple semantic representation on the preprocessed sample data;
the simple semantics (concepts) of the main influencing factors of the AFS-based heat consumption in the invention are expressed as follows:
assuming that the coking process production data are N samples, each sample having 9 influencing factors (attributes), an N × 9 matrix X ═ X can be usedi,j]To represent all samples, where xi,jRepresenting the jth attribute value of the ith sample. Let M be { M ═ Mj,rI 1 ≦ j ≦ 9,1 ≦ r ≦ 6} is a simple semantic set defined on the dataset X, where j ≦ 1,2, …,9 respectively denote the jth attribute, r ≦ 1 denotes "long/large/high", r ≦ 2 denotes "non-long/large/high", r ≦ 3 denotes "medium", r ≦ 4 denotes "non-medium", r ≦ 5 denotes "short/small/low", and r ≦ 6 denotes "non-short/small/low". For all fuzzy semantics M in any non-empty subset A of Mj,rReferred to as a simple semantic, e.g. m1,1I.e. the simple semantic meaning of coke formation is long.
The complex semantics (concepts) of the main influencing factors of the AFS-based heat consumption in the invention are expressed as follows:
on the basis of the definition of the simple semantics, a new fuzzy semantics, also called a complex concept or a complex semantics, can be generated by performing conjunction or disjunction operation on two or more simple semantics, namely logic operation "and" or ". For example A1＝m1,1and m9,3I.e. "long coking time and moderate coal moisture"; a. the2＝m2,5and m4,1Can express that the coal feeding amount is small and the uniformity coefficient is large; a. the3＝A1+A2It can be expressed as "the coking time is long and the moisture of the coal is moderate, or the coal feeding amount is small and the uniformity coefficient is large", or expressed as "m1,1m9,3+m2,5m4,1”。
Step2.2, constructing a semantic weight function;
step2.3, constructing a fuzzy semantic membership function;
(1) degree of semantic membership
The semantic concept describing a certain sample is defined according to the distribution of specific attribute values of the data sample and is embodied by the semantic membership degree. Suppose F is a fuzzy semantic set on dataset X, for
Figure BDA0001585777290000031
x∈X，Aτ(x) To the extent that sample x belongs to a, it is expressed as follows:
Figure BDA0001585777290000032
wherein m isj,rFor a simple concept in the set of F,
Figure BDA0001585777290000041
indicating that sample x belongs to concept mj,rDegree of (2)Less than or equal to sample y belongs to mj,rTo a degree of (A)τ(x) Is in accordance with
Figure BDA0001585777290000042
The set of all samples y of a condition is a subset of samples X.
(2) Semantic weight function
Definition of
Figure BDA0001585777290000043
Is a simple semantic mj,rWhen: 1)
Figure BDA0001585777290000044
how x does not belong to mj,rThen, then
Figure BDA0001585777290000045
If x belongs to mj,rGreater than the extent to which y belongs to mj,rTo the extent of
Figure BDA0001585777290000046
(3) Fuzzy semantic membership function (AFS membership function)
Concept of arbitrary ambiguity
Figure BDA0001585777290000047
Is given by the following formula:
Figure BDA0001585777290000048
wherein N isuThe number of times of observation of the sample is shown, and S is the number of A. Mu.sα(x) I.e. the degree of membership that can be said to the sample x belongs to the concept alpha.
Step2.4, calculating the membership degree of the sample belonging to each simple semantic through a membership function, screening a threshold value according to the membership degree and the simple semantic, and extracting a simple semantic set of the coking heat consumption;
for each sample X ∈ X, by the formula (2)Is used for calculating the membership degree of the sample x belonging to each simple semantic
Figure BDA0001585777290000049
Setting a screening threshold σ1Extracting simple semantic set of coking heat consumption
Figure BDA00015857772900000410
And Step2.5, constructing complex semantics according to the simple semantic set, calculating the membership degree of the sample belonging to each complex semantic through a membership function, screening a threshold value according to the membership degree and the complex semantics, and extracting the complex semantic set of the coking heat consumption.
Simple semantic set from sample x
Figure BDA00015857772900000411
Calculating each complex semantic by membership function of formula (2)
Figure BDA00015857772900000412
Degree of membership of
Figure BDA00015857772900000413
Setting a screening threshold σ2Extracting complex semantic set of coking heat consumption
Figure BDA00015857772900000414
Assuming sample data X is classified into class C, then class CkThe class semantic rule set is:
Figure BDA0001585777290000051
wherein
Figure BDA0001585777290000052
Is the CkNumber of samples of class.
Step3, performing semantic evaluation and importance factor calculation;
(1) evaluation of semantics
Defining an evaluation function omega (A)v) For semantic sets
Figure BDA0001585777290000053
Semantic of (A)vThe evaluation was carried out using the following specific calculation formula:
Figure BDA0001585777290000054
Figure BDA0001585777290000055
wherein
Figure BDA0001585777290000056
Is the CkNumber of samples of class, PiRepresenting semantics AvFor sample xiDegree of contribution of, ω (A)v) Representing semantics AvThe contribution degree of all samples is consistent, the higher the consistency is, the larger the evaluation value is, and the higher the contribution degree of the semantics is.
(2) Calculation of semantic importance factor
Defining semantic importance factors
Figure BDA0001585777290000057
Computing a semantic set
Figure BDA0001585777290000058
Semantic of (A)vThe calculation formula is as follows:
Figure BDA0001585777290000059
Figure BDA00015857772900000510
wherein the content of the first and second substances,
Figure BDA00015857772900000511
is shown in
Figure BDA00015857772900000512
Simple semantics m in a semantic setj,rThe number of times of occurrence of the event,
Figure BDA00015857772900000513
is shown in
Figure BDA00015857772900000514
The sum of the number of occurrences of all simple semantics in the semantic set,
Figure BDA00015857772900000515
is shown in
Figure BDA00015857772900000516
Simple semantics m in a semantic setj,rThe frequency of occurrence.
Step4, extracting key simple semantics of heat consumption influence factors in the coking process;
according to semantic AvOf importance
Figure BDA00015857772900000517
Sequencing from big to small and constructing a new semantic set
Figure BDA00015857772900000518
In turn from
Figure BDA00015857772900000519
Selecting semantic A with higher semantic importancevGo through AvAll simple semantics m inj,rSequentially extracting a key simple semantic of each attribute and forming a key simple semantic set Km。
Step5, setting parameter ranges of heat consumption influence factors in the coking process, and realizing optimization of the heat consumption influence factors in the coking process; and aiming at each sample attribute, sorting according to the attribute value, solving the membership degree of each sorted sample to the simple semantics according to the AFS membership degree function, and determining the parameter setting range according to the calculated membership degree distribution of all samples.
Hypothesis CkI.e. as a low heat consumption class, KmContains 9 key simple semantics mj,rWherein each heat consumption influence factor corresponds to a simple semantic. C is to bekAll samples are sorted according to the size of the attribute j, and the semantic m of all samples is calculated according to the formula (2)j,rTo determine m of the attribute jj,rAnd the semantic value range realizes the optimization of heat consumption influence factors in the coking process.
The key semantics of 9 main heat consumption influence factors are extracted, so that the value range of the coke oven control parameters in the coking process is obtained. However, the method of the present invention is not limited to the optimization of the listed 9 heat consumption factors, and can also optimize other key influencing factors in the actual production.
The invention has the beneficial effects that:
the invention optimizes the heat consumption influence factors in the coking process, realizes the aim of ensuring the qualified coke quality and simultaneously ensuring the lower heat consumption in the actual production process, can obtain a group of heat consumption factor optimization parameters which are more suitable for the actual production requirement and are more comprehensive and accurate along with the increase of production data, and provides decision support for the informatization and intellectualization of the actual production of the coking process.
Drawings
FIG. 1 is a triangular weighting function;
FIG. 2 is an AFS membership function.
Detailed Description
Example 1: as shown in fig. 1-2, the fuzzy semantic reasoning-based optimization method for heat consumption influence factors in a coking process takes 150 real production data (samples) in a coking production process as an example, and the optimization method specifically comprises the following steps:
step1, extracting and preprocessing sample data;
step1.1, calculate 150 samplesAnd classifying the samples into a lower heat consumption class C according to the mean value1Class C with (less than average) and higher heat consumption2(more than or equal to the mean value) two types of data, wherein the number of samples is respectively 80 and 70;
step1.2, only 9 main attributes (factors) are reserved from the original data aiming at each data in the two types, wherein the main attributes are respectively coking time, coal feeding amount (single hole), main coke oven gas pipe flow, furnace temperature coefficient (uniformity coefficient), flue gas machine side suction, flue gas coke side suction, straight line machine side temperature, straight line coke side temperature and coal moisture, and are marked as f1，f2，…，f9；
Step1.3, adopting a linear transformation mode to convert the attribute value of the sample space of each attribute
Figure BDA0001585777290000071
(j-th attribute value of i-th sample) transform normalized to [0, 1%]The interval of time is,
Figure BDA0001585777290000072
to preserve the probability distribution of the data. Finally, a sample data set X is formed, and the sample set X is a matrix of 150X 9.
Step2, constructing a sample semantic set according to an axiom fuzzy set theory;
step2.1, performing simple semantic representation on the preprocessed sample data; let M be { M }j,rJ is more than or equal to 1 and less than or equal to 9, r is more than or equal to 1 and less than or equal to 6, the simple semantic set is defined on the data set X, the corresponding semantics are shown in Table 1, namely M comprises 54 simple semantics;
TABLE 1
Figure BDA0001585777290000073
Step2.2, setting the semantic weight function as a trigonometric function, and setting the node values to be 0.2, 0.5 and 0.8 respectively as shown in fig. 1. Where ρ is1,ρ2,ρ3The weight functions are three semantic functions of short/small/low "," moderate "," long/large/high ", respectively. Suppose sample x is at some propertyThe value above is 0.5, then the sample belongs to the semantic "short/small/low" with a weight of 0, to the "medium" with a weight of 1, and to the "long/large/high" with a weight of 0. The weight of three semantics of 'non-short/small/low', 'non-moderate', 'non-long/large/high' is 1-0, 1-0.5 and 1-0 respectively;
step2.3, constructing a fuzzy semantic membership function; the function is as follows:
Figure BDA0001585777290000074
wherein N isuDenotes the number of observations of the sample, S is the number of A, μα(x) That is, the membership degree of the sample x belonging to the concept α;
step2.4, setting and screening simple semantic screening threshold value sigma for screening coking heat consumption factors1For each sample of each class, calculating the simple semantic m to which the sample x belongs by using the membership function formula (2) respectivelyj,rDegree of membership of
Figure BDA0001585777290000081
Extract to satisfy
Figure BDA0001585777290000082
Simple semantics of (2) constitute a simple semantic set of samples x
Figure BDA0001585777290000083
Step2.5, setting and screening complex semantic screening threshold value sigma for screening coking heat consumption factors2For each sample x, construct all slaves
Figure BDA0001585777290000084
Selecting 2 or more than 2 (length is not more than 4) simple semantics, and generating complex semantics A by' andvusing the degree of membership calculated by the formula (2)
Figure BDA0001585777290000085
Extract to satisfy
Figure BDA0001585777290000086
Of the complex semantics, constituting a complex semantic set of the sample x
Figure BDA0001585777290000087
The number of extracted fuzzy semantic sets and a threshold σ1、σ2The maximum value of the set semantic concept length is related to the number of samples and the attribute distribution.
Step2.6, the Complex semantics of all samples of each class
Figure BDA0001585777290000088
Connecting through disjunction operation to respectively construct two semantic sets
Figure BDA0001585777290000089
And
Figure BDA00015857772900000810
step3, performing semantic evaluation and importance factor calculation;
in order to achieve the aim of ensuring qualified coke quality and simultaneously reducing heat consumption in the actual production process, only semantic rule sets with low heat consumption are analyzed. Through the steps of 1 and 2, the semantic set with low heat consumption can be obtained
Figure BDA00015857772900000811
A complex set of 44 semantic concept (rule) extractions, with 14, 21 and 9 semantics of length 2, 3 and 4, respectively. These semantics are evaluated and the importance factors are calculated as follows:
step3.1, semantic evaluation; using an evaluation function
Figure BDA00015857772900000812
Calculate out
Figure BDA00015857772900000813
44 semantics A invEvaluation value ω (A) ofv)；
Wherein the content of the first and second substances,
Figure BDA00015857772900000814
is the CkNumber of samples of class, PiRepresenting semantics AvFor sample xiDegree of contribution of, ω (A)v) Representing semantics AvThe degree of consistency of the contribution degrees to all samples;
step3.2, calculating semantic importance factors; statistics of
Figure BDA00015857772900000815
Each simple semantic m of the 44 semantics in (1)j,rNumber of occurrences Nj,rCalculating the frequency of each simple semantic occurrence separately
Figure BDA00015857772900000816
Calculate each semantic AvSum of frequencies of all simple concepts present in
Figure BDA00015857772900000817
Step3.3, by formula
Figure BDA00015857772900000818
To calculate each semantic AvIs important factor
Figure BDA00015857772900000819
Wherein the content of the first and second substances,
Figure BDA0001585777290000091
Figure BDA0001585777290000092
is shown in
Figure BDA0001585777290000093
Simple semantics m in a semantic setj,rThe number of times of occurrence of the event,
Figure BDA0001585777290000094
is shown in
Figure BDA0001585777290000095
The sum of the number of occurrences of all simple semantics in the semantic set,
Figure BDA0001585777290000096
is shown in
Figure BDA0001585777290000097
Simple semantics m in a semantic setj,rThe frequency of occurrence.
Step4, extracting key simple semantics of heat consumption influence factors in the coking process;
step4.1 semantic-based importance factor
Figure BDA0001585777290000098
The semantics are sequenced from large to small, and a new semantic set is constructed
Figure BDA0001585777290000099
And initialize key simple semantic collections
Figure BDA00015857772900000910
Step4.2 from
Figure BDA00015857772900000911
In turn from the semantic A with higher importancevTraverse all simple semantics mj,rIf, if
Figure BDA00015857772900000912
And is
Figure BDA00015857772900000913
Q is more than or equal to 1 and less than or equal to 6, and r is an odd number (non-semantic), the simple semantic m is setj,rJoin to set KmIn (1). In short, only one key is chosen for each attributeSimple semantics of (2);
step4.3, repeat step Step4.2 until KmThe set comprises key simple semantics of all attributes;
step4.4, obtaining m by the above steps in sequence6,5，m2,5，m5,3，m4,1，m8,3，m9,5，m7,3，m1,3，m3,5Set K of simple semantic constituents of equal 9 factorsm。
And Step5, setting the parameter range of the heat consumption influence factors in the coking process, and finishing the optimization of the heat consumption influence factors in the coking process.
Set KmMiddle m1,3，m2,5，m3,5，m4,1，m5,3，m6,5，m7,3，m8,3，m9,5The semantics of the representation are respectively: moderate coking time, small coal feeding amount (single hole), small flow of coke oven gas main pipe, high furnace temperature coefficient (uniformity coefficient), moderate suction force of flue gas machine side, small suction force of flue gas coke side, moderate temperature of straight running machine side, moderate temperature of straight running coke side and small moisture of coal. In other words, to achieve the goal of ensuring the quality of coke in the actual production process and simultaneously reducing the heat consumption, the parameter values of the heat consumption factor in the above 9 should be controlled in the corresponding semantic range as much as possible. Next, we further determine the parameter setting range of each factor according to the above key semantics.
Step5.1 for each attribute, add coal amount f2For example, all samples are sorted according to the attribute value of the coal adding quantity, and then the semantic m of each sorted sample is calculated according to a formula (2) (AFS membership function)2,1(large amount of coal added) m2,3(moderate coal adding amount) m2,5(small coal charge) three semantic membership degrees, such as mf3, mf2 and mf1 shown in fig. 2.
As can be seen from fig. 2, the AFS theory constructs the membership function according to the actual attribute distribution of the features, instead of subjectively defining the membership function, so as to avoid inaccuracy due to subjective cognitive differences. The semantics constructed by the membership function can be used for logic rule operation, and the characteristic attributes can be more comprehensively described.
Step5.2, set KmExtracted coal addition amount f2The key semantic of an attribute is m2,5Then its value interval can be set as
Figure BDA0001585777290000101
Wherein
Figure BDA0001585777290000102
Value here indicates a desirable value of the attribute value, and satisfies
Figure BDA0001585777290000103
I.e. attribute value v belongs to semantic m2,5Is greater than the membership of other semantics. In fig. 2, it will be understood that,
Figure BDA0001585777290000104
the value of the sum of the values should be 0,
Figure BDA0001585777290000105
is the value of the abscissa corresponding to the intersection of the function mf1 and mf 2.
Step5.3, since the attribute values are normalized at the time of preprocessing, the following formula will be used
Figure BDA0001585777290000106
And
Figure BDA0001585777290000107
the value of (c) is mapped to the original value range of the attribute, and the coal charging amount f can be obtained2Value lower bound of attribute
Figure BDA0001585777290000108
And upper limit of
Figure BDA0001585777290000109
Figure BDA00015857772900001010
Figure BDA00015857772900001011
Step5.4, according to Step5.1 to Step5.3, the upper limit and the lower limit of other attributes can be obtained, and are respectively marked as
Figure BDA00015857772900001012
And
Figure BDA00015857772900001013
where j is 1,2, …, 9. The coke oven control parameter optimization data table and the parameter setting range in the coking process can be obtained, as shown in table 2.
TABLE 2
Figure BDA00015857772900001014
While the present invention has been described in detail with reference to the embodiments shown in the drawings, the present invention is not limited to the embodiments, and various changes can be made without departing from the spirit of the present invention within the knowledge of those skilled in the art.

Claims (4)

1. The coking process heat consumption influence factor optimization method based on fuzzy semantic reasoning is characterized by comprising the following steps: according to the AFS axiom fuzzy set theory, aiming at the influence factors of heat consumption in the coking process, the inherent fuzzy semantics are excavated, the key simple semantics of each influence factor are extracted through semantic evaluation and calculation of importance factors, the parameter range of the influence factors of the heat consumption is set, and the optimization of the influence factors of the heat consumption in the coking process is completed;
the optimization method comprises the following specific steps:
step1, extracting and preprocessing sample data;
step2, constructing a sample semantic set according to an axiom fuzzy set theory;
step3, performing semantic evaluation and importance factor calculation;
step4, extracting key simple semantics of heat consumption influence factors in the coking process;
step5, setting parameter ranges of heat consumption influence factors in the coking process, and finishing optimization of the heat consumption influence factors in the coking process;
in the Step3, an evaluation function is used
Figure FDA0003396022740000011
Evaluating the semantics by analyzing the consistency degree of the contribution degrees of the semantics to all samples;
wherein the content of the first and second substances,
Figure FDA0003396022740000012
is the CkNumber of samples of class, PiRepresenting semantics AvFor sample xiDegree of contribution of, ω (A)v) Representing semantics AvThe degree of consistency of the contribution degrees to all samples;
in Step3, the evaluation of the semantics and the frequency of the simple semantics appearing in the semantic set are calculated by formula
Figure FDA0003396022740000013
To calculate the importance factor of each semantic;
wherein the content of the first and second substances,
Figure FDA0003396022740000014
Figure FDA0003396022740000015
is shown in
Figure FDA0003396022740000016
Simple semantics m in a semantic setj,rThe number of times of occurrence of the event,
Figure FDA0003396022740000017
is shown in
Figure FDA0003396022740000018
The sum of the number of occurrences of all simple semantics in the semantic set,
Figure FDA0003396022740000019
is shown in
Figure FDA00033960227400000110
Simple semantics m in a semantic setj,rThe frequency of occurrence;
in Step4, extracting key simple semantics from the semantics with high importance factors in sequence according to the size of the semantic importance factors;
the specific steps of Step1 are as follows:
step1.1, taking the average value of the heat consumption as a limit, and dividing the data into two types of data samples with higher heat consumption and lower heat consumption respectively according to the average value higher than and lower than the heat consumption;
step1.2, aiming at the two types of data samples, keeping main attributes from original data; the method comprises the following steps of coking time, coal feeding amount, coke oven gas main pipe flow, oven temperature coefficient, flue gas coke side suction, straight running oven side temperature, straight running coke side temperature and coal moisture;
and Step1.3, transforming and normalizing the attribute value of the sample space of each attribute in a linear transformation mode to finally form a sample data set.
2. The fuzzy semantic reasoning based coking process heat consumption influence factor optimization method according to claim 1, characterized in that: the specific steps of Step2 are as follows:
step2.1, performing simple semantic representation on the preprocessed sample data;
step2.2, constructing a semantic weight function;
step2.3, constructing a fuzzy semantic membership function;
step2.4, calculating the membership degree of the sample belonging to each simple semantic through a membership function, screening a threshold value according to the membership degree and the simple semantic, and extracting a simple semantic set of the coking heat consumption;
and Step2.5, constructing complex semantics according to the simple semantic set, calculating the membership degree of the sample belonging to each complex semantic through a membership function, screening a threshold value according to the membership degree and the complex semantics, and extracting the complex semantic set of the coking heat consumption.
3. The fuzzy semantic reasoning based coking process heat consumption influence factor optimization method according to claim 2, characterized in that: the simple semantic screening threshold value is 0.8, and the complex semantic screening threshold value is 0.75;
extracting a simple semantic set of simple semantic composition samples, wherein the simple semantic set satisfies that the membership degree of the simple semantics is greater than or equal to a simple semantic screening threshold value 0.8;
and extracting a complex semantic set of the complex semantic composition sample which meets the complex semantic requirement and has the membership degree of more than or equal to the complex semantic screening threshold value of 0.75.
4. The fuzzy semantic reasoning based coking process heat consumption influence factor optimization method according to claim 1, characterized in that: in Step5, for each sample attribute, sorting according to the attribute value, solving the membership degree of each sorted sample to the simple semantics according to the AFS membership function, and determining the parameter setting range according to the calculated membership degree distribution of all samples.
CN201810170875.4A 2018-03-01 2018-03-01 Fuzzy semantic reasoning-based coking process heat consumption influence factor optimization method Active CN108446459B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810170875.4A CN108446459B (en) 2018-03-01 2018-03-01 Fuzzy semantic reasoning-based coking process heat consumption influence factor optimization method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810170875.4A CN108446459B (en) 2018-03-01 2018-03-01 Fuzzy semantic reasoning-based coking process heat consumption influence factor optimization method

Publications (2)

Publication Number Publication Date
CN108446459A CN108446459A (en) 2018-08-24
CN108446459B true CN108446459B (en) 2022-03-22

Family

ID=63193253

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810170875.4A Active CN108446459B (en) 2018-03-01 2018-03-01 Fuzzy semantic reasoning-based coking process heat consumption influence factor optimization method

Country Status (1)

Country Link
CN (1) CN108446459B (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1435781A (en) * 2003-02-24 2003-08-13 杨炳儒 Intelligent decision supporting configuration method based on information excavation
CN101674328A (en) * 2009-11-18 2010-03-17 南京理工大学 Subject service describing and matching method based on semantic and QoS constraint
CN102663122A (en) * 2012-04-20 2012-09-12 北京邮电大学 Semantic query expansion algorithm based on emergency ontology
CN104763999A (en) * 2015-03-04 2015-07-08 内蒙古瑞特优化科技股份有限公司 Power plant pulverized coal boiler combustion performance online optimizing method and system

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8301436B2 (en) * 2003-05-29 2012-10-30 Microsoft Corporation Semantic object synchronous understanding for highly interactive interface
CN102636991A (en) * 2012-04-18 2012-08-15 国电科学技术研究院 Method for optimizing running parameters of thermal power unit and based on fuzzy set association rule
US9436875B2 (en) * 2012-12-06 2016-09-06 Nokia Technologies Oy Method and apparatus for semantic extraction and video remix creation

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1435781A (en) * 2003-02-24 2003-08-13 杨炳儒 Intelligent decision supporting configuration method based on information excavation
CN101674328A (en) * 2009-11-18 2010-03-17 南京理工大学 Subject service describing and matching method based on semantic and QoS constraint
CN102663122A (en) * 2012-04-20 2012-09-12 北京邮电大学 Semantic query expansion algorithm based on emergency ontology
CN104763999A (en) * 2015-03-04 2015-07-08 内蒙古瑞特优化科技股份有限公司 Power plant pulverized coal boiler combustion performance online optimizing method and system

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
基于本体与物联网的冶金炼焦过程语义化及建模应用研究;甘健侯等;《中国博士学位论文全文数据库 工程科技Ⅰ辑》;20170415;第2017年卷(第04期);第6.4节 *

Also Published As

Publication number Publication date
CN108446459A (en) 2018-08-24

Similar Documents

Publication Publication Date Title
Ju et al. Extension of VIKOR method for multi-criteria group decision making problem with linguistic information
Wu et al. Case-based reasoning with optimized weight derived by particle swarm optimization for software effort estimation
CN111814871A (en) Image classification method based on reliable weight optimal transmission
CN110544011B (en) Intelligent system combat effectiveness evaluation and optimization method
Nguyen et al. Filter based backward elimination in wrapper based PSO for feature selection in classification
CN108763237A (en) A kind of knowledge mapping embedding grammar based on attention mechanism
CN108846526A (en) A kind of CO2 emissions prediction technique
CN106453293A (en) Network security situation prediction method based on improved BPNN (back propagation neural network)
CN109242188B (en) Long-term interval prediction and structure learning method for steel gas system
CN107045569B (en) Gear reducer optimization design method based on clustering multi-target distribution estimation algorithm
CN112182221B (en) Knowledge retrieval optimization method based on improved random forest
CN111507365A (en) Confidence rule automatic generation method based on fuzzy clustering
Tani et al. Fuzzy modeling by ID3 algorithm and its application to prediction of heater outlet temperature
Zhong et al. Construction project risk prediction model based on EW-FAHP and one dimensional convolution neural network
CN108446459B (en) Fuzzy semantic reasoning-based coking process heat consumption influence factor optimization method
Sun et al. Dynamic Intelligent Supply-Demand Adaptation Model Towards Intelligent Cloud Manufacturing.
CN114202086A (en) Multi-objective optimization method for mining scheme
Lu et al. Sudden risk predication model of construction supply chain based on data mining
Qiang et al. The combination prediction of BTP in sintering process based on Bayesian framework and LS-SVM
CN111080053B (en) EHF-WAS-based power grid development strategy efficiency and benefit evaluation method
Chen Optimization Design of Drilling Fluid Chemical Formula Based on Artificial Intelligence
Xu et al. A PSO-ANN integrated model of optimizing cut-off grade and grade of crude ore
Yin et al. An evidential reason approach for operational effectiveness evaluation of weapon system under uncertain
CN115392668A (en) Electric energy substitution project evaluation system and method based on machine learning
CN106777903A (en) A kind of spontaneous combustion of remaining coal influence factor importance determines method

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