CN111059896B - exergy model-based roller kiln system anomaly detection method - Google Patents

exergy model-based roller kiln system anomaly detection method Download PDF

Info

Publication number
CN111059896B
CN111059896B CN201911259474.7A CN201911259474A CN111059896B CN 111059896 B CN111059896 B CN 111059896B CN 201911259474 A CN201911259474 A CN 201911259474A CN 111059896 B CN111059896 B CN 111059896B
Authority
CN
China
Prior art keywords
fault
data
substance
temperature
input
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.)
Expired - Fee Related
Application number
CN201911259474.7A
Other languages
Chinese (zh)
Other versions
CN111059896A (en
Inventor
程明阳
杨海东
徐康康
朱成就
印四华
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.)
Guangdong University of Technology
Original Assignee
Guangdong University of Technology
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 Guangdong University of Technology filed Critical Guangdong University of Technology
Priority to CN201911259474.7A priority Critical patent/CN111059896B/en
Publication of CN111059896A publication Critical patent/CN111059896A/en
Application granted granted Critical
Publication of CN111059896B publication Critical patent/CN111059896B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F27—FURNACES; KILNS; OVENS; RETORTS
    • F27B—FURNACES, KILNS, OVENS, OR RETORTS IN GENERAL; OPEN SINTERING OR LIKE APPARATUS
    • F27B9/00—Furnaces through which the charge is moved mechanically, e.g. of tunnel type; Similar furnaces in which the charge moves by gravity
    • F27B9/30—Details, accessories, or equipment peculiar to furnaces of these types
    • F27B9/40—Arrangements of controlling or monitoring devices

Landscapes

  • Engineering & Computer Science (AREA)
  • Mechanical Engineering (AREA)
  • General Engineering & Computer Science (AREA)
  • Testing And Monitoring For Control Systems (AREA)
  • Furnace Details (AREA)

Abstract

The invention discloses a method based on
Figure DDA0002311215820000011
The method for detecting the abnormality of roller kiln system includes such steps as creating roller kiln system
Figure DDA0002311215820000012
An analysis framework for analyzing the type of fault present in the system by
Figure DDA0002311215820000013
Equilibrium analysis, construction
Figure DDA0002311215820000014
The fault vector table is used for detecting and diagnosing faults existing in the production operation system of the roller kiln; the invention realizes the fault detection of the roller kiln production operation system; by collecting the mass flow and temperature data of the input and output variables of the system and carrying out the process on the variables
Figure DDA0002311215820000015
Analysis of, constructed variables
Figure DDA0002311215820000016
The fault vector table is used for detecting and diagnosing faults existing in the production operation system of the roller kiln; different from the existing method for constructing an expert fault knowledge system, the method has small burden on enterprises and reduces the generation cost of the enterprises; in the aspect of practicability, the method is simple, easy and efficient, simple to operate, remarkable in effectiveness and convenient to apply.

Description

exergy model-based roller kiln system anomaly detection method
Technical Field
The invention relates to the technical field of roller kiln system detection, in particular to a roller kiln system detection method based on
Figure BDA0002311215800000012
An anomaly detection method for a roller kiln system of a model.
Background
The roller kiln is a main device applied to the ceramic industry production, and the problems of low production efficiency, poor product quality and the like exist in the production operation process of a roller kiln system, so that the energy utilization rate of the ceramic industry is seriously influenced; meanwhile, recent research shows that after the existing control scheme is improved, the production efficiency is improved by less than 5%, and the implementation of the advanced predictive maintenance scheme can improve the overall production efficiency by 20-40%; fault detection and diagnosis is a supporting technology for more advanced maintenance solutions.
Roller kilns are a typical comprehensive complex system, which makes roller kilns have great complexity in failure. At present, most ceramic enterprises have limited fault detection methods and means for roller kiln equipment, and have certain influence on the improvement of the overall production efficiency. Therefore, the simple and efficient abnormality detection method is of far-reaching and important significance in reducing the production cost of enterprises, improving the competitiveness of the enterprises and promoting the sustainable development of the ceramic industry.
In the prior art, the stone middle jade of the ceramics institute of Jingdezhen province is in 'the research of the ceramic kiln fault diagnosis expert system', and aiming at the fault characteristics of modern ceramic kilns, a comprehensive knowledge representation method combining a frame based on fault tree analysis and a generation formula is adopted, the expert system technology is introduced into the field of kiln fault diagnosis, and an expert system for ceramic kiln fault diagnosis is researched.
In the intelligent fault diagnosis technology of the complex process and the application research thereof in a large-scale industrial kiln, the Liu Xiao of the university of the China and the south, a method for detecting a fault signal of the complex process based on a fractal theory is provided by establishing an intelligent integrated fault diagnosis model framework based on fuzzy logic, a neural network and an expert system.
However, the existing technology relies on the construction of a fault knowledge system, and a complete system structure for acquiring expert experience knowledge covers knowledge acquisition processes and corresponding evaluation indexes of various objects, so that the fault knowledge system is difficult to construct for enterprises, and a large amount of manpower, material resources and time are required to be invested.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a simple and efficient base
Figure BDA0002311215800000021
An anomaly detection method for a roller kiln system of a model.
The purpose of the invention is realized by the following technical scheme:
based on
Figure BDA0002311215800000022
The method for detecting the abnormity of the roller kiln system of the model comprises the following steps:
firstly, analyzing input and output variables; the working mechanism of the roller kiln system is analyzed, and the system input variables are as follows: green bricks, natural gas, combustion-supporting gas and cooling gas; the output variables are: firing bricks, flue gas and cooling exhaust gas;
step two, data acquisition; the data acquisition mainly acquires data of input and output variables in a roller kiln system in real time, wherein the data comprises mass flow and temperature of each input and output variable; in addition, it needs to be collected
Figure BDA0002311215800000023
The basic data required for the analysis, including reference environmental conditions such as temperature and pressure, the mole-specific heat capacity of the input and output streams, and the composition of the streams and their standards
Figure BDA0002311215800000024
The data acquisition method mainly comprises the following steps:
(1) acquiring temperature data through a temperature sensor;
(2) collecting gas mass flow data including natural gas, combustion-supporting gas, flue gas, cooling gas and cooling waste gas through a flowmeter;
(3) collecting the feeding speed data of the ceramic tile through the speed of the conveyor belt;
(4) measuring smoke components through a smoke analyzer, and measuring components of fuel gas, combustion-supporting gas, cooling gas and cooling waste gas through an Ordovician gas analyzer;
(5) determining the composition of the green brick and the fired brick by x-ray diffraction;
(6) the basic environment state refers to an environment model of the Guishan-Jitian;
(7) the molar constant pressure heat capacity of the material flow is obtained by looking up the molar heat capacity table of the material;
(8) standard of material flow found by related books
Figure BDA0002311215800000037
Step three, performing a first step of cleaning the substrate,
Figure BDA0002311215800000032
analyzing; due to the environment condition will be right
Figure BDA0002311215800000033
The analysis has great influence on
Figure BDA0002311215800000034
In the analysis process, a reference environment state needs to be set; in addition, the production and operation system of the roller kiln is a very complex system, and in order to reduce the influence of secondary factors, some assumptions need to be made on the system, which are specifically as follows:
(1) ambient temperature T0300K, ambient pressure P0＝101.3kPa；
(2) The elements contained in the air take the corresponding composition gas of the air as a reference substance, and take the molar component of saturated humid air as the component of the reference substance;
(3) the other elements are based on the most stable pure substance (liquid or solid) containing the element, and the diffusion of the actual solid substance is considered
Figure BDA0002311215800000035
Difficult to use, adopt T0、P0Of pure solid reference substances under the conditions
Figure BDA0002311215800000036
A value of 0;
(4) assuming the system is a constant pressure, adiabatic, steady state operating open system;
(5) neglecting the kinetic energy and potential energy of the input flow and the output flow, and not considering the consumption of electric energy;
(6) assuming natural gas, air and flue gas as ideal gases;
(7) ignoring water vapor in the air;
of substances
Figure BDA0002311215800000047
The calculation of (a) is based on the reference environmental conditions and system assumptions, known as a steady flow open system, in which the molar quantity of substance x is determined
Figure BDA0002311215800000048
Is divided into physics
Figure BDA0002311215800000049
And chemistry
Figure BDA00023112158000000410
The following formula:
Ex＝Ex,ph+Ex,ch
in the above formula, ExIs the mole of substance x
Figure BDA00023112158000000411
The unit is kJ/mol; ex,phIs the physics of substance x
Figure BDA00023112158000000414
The unit is kJ/mol; ex,chIs the chemistry of substance x
Figure BDA00023112158000000412
The unit is kJ/mol;
input of substance x
Figure BDA00023112158000000413
Comprises the following steps:
Figure BDA0002311215800000041
in the above formula, the first and second carbon atoms are,
Figure BDA0002311215800000042
is the input of a substance x
Figure BDA00023112158000000415
The unit is kJ/s; m isxIs the mass flow of the substance x, with the unit being kg/s; mxIs the molar mass of substance x, in g/mol;
physics of physics
Figure BDA00023112158000000416
The formula of (1) is:
Figure BDA0002311215800000043
in the above formula, T0Is a reference environment temperature, T is a current environment temperature and the unit is K; p is a radical of0Is the reference environment pressure intensity, p is the current environment pressure intensity, and the unit is kPa; h is0For the substance m at a temperature T0The enthalpy of the substance m at the temperature T, h is the enthalpy of the substance m at the temperature T, and the unit is kJ/(mol.K); s0For the substance m at a temperature T0The entropy of time, s is the entropy of the substance m at the temperature T, and the unit is kJ/(mol.K); c. CpThe molar constant pressure heat capacity of a substance x is expressed in kJ/(mol. K); r is a general air constant of 8.3145 multiplied by 10-3kJ/(mol·K)；
The upper type
Figure BDA0002311215800000044
Is the thermal component of the physical congestion and,
Figure BDA0002311215800000045
is the pressure component of the physical congestion; if c ispIs constant or average specific heat, and neglecting the pressure component, then there are:
Figure BDA0002311215800000046
chemistry
Figure BDA0002311215800000054
The formula of (1) is:
Figure BDA0002311215800000051
in the above formula, xiIs the mole fraction of component i in substance x;
Figure BDA0002311215800000052
as standard for component i
Figure BDA0002311215800000055
The unit is kJ/mol;
step four, performing a first step of cleaning the substrate,
Figure BDA0002311215800000056
constructing a fault vector table; the complex structure and the working process of the modern roller kiln determine the complexity of the fault; by analysis, the fault types are mainly:
(1) abnormal failure of material flow;
(2) faults affecting the overall model;
(3) abnormal fluctuation fault of temperature in the kiln;
in order to generate a data set, mass flow and temperature data of each input/output variable when a fault occurs are obtained through fluent simulation and combined with mass flow and temperature data of each variable under normal working conditions, which are acquired by a sensor; the data obtained in the normal working condition is used for carrying out normalization processing on the whole data set, a unit-free quantity is generated, the unit-free quantity represents the size of deviation and can be used for constructing a fault vector;
wherein the normalization processing adopts a standard deviation normalization method; the standard deviation normalization method is to normalize the data through the mean value and the standard deviation of the sample data, and the processed data conforms to the standard normal distribution with the mean value of 0 and the standard deviation of 1; the calculation formula is as follows:
Figure BDA0002311215800000053
in the above formula, x*Is the corrected value; mu is the mean value of the sample data; σ is the standard deviation of the sample data;
to build up
Figure BDA0002311215800000057
Fault vector, applying threshold function to normalized physical of each variable
Figure BDA0002311215800000058
And chemistry
Figure BDA0002311215800000059
Data, making physical of variables
Figure BDA00023112158000000510
And chemistry
Figure BDA00023112158000000511
Converting into a qualitative vector; in order to eliminate the influence on rounding errors, a limit function f (x) epsilon { -1,0,1} is introduced to classify the normalized data x into { -1,0,1}, and the formula of the limit function is as follows:
x≤-M→f(x)＝-1
-M＜x＜M→f(x)＝0
x≥M→f(x)＝1
in the above formula, x is the normalized data, and M is the threshold function;
the threshold function is:
M＝3σ*
in the above formula, σ*The standard deviation for the normalized data is 1;
when a fault occurs, collecting the mass flow and temperature data of input and output variables, passing through
Figure BDA0002311215800000061
Analyzing the process to obtain the physics of the input and output variables
Figure BDA0002311215800000062
And chemistry
Figure BDA0002311215800000063
Then applying normalization method to input and output variable physics
Figure BDA0002311215800000064
And chemistry
Figure BDA0002311215800000065
Normalizing, introducing a limit function f (x) epsilon { -1,0,1} to normalize the physical property of the processed variable
Figure BDA0002311215800000066
And chemistry
Figure BDA0002311215800000067
The classification is { -1,0,1}, wherein 1 represents a positive vector, 0 represents a zero vector, and-1 represents a negative vector; if physics
Figure BDA0002311215800000068
And chemistry
Figure BDA0002311215800000069
If it is greater than the maximum threshold, it is a positive vector, indicated by an upward arrow, if physical
Figure BDA00023112158000000610
And chemistry
Figure BDA00023112158000000611
If the minimum threshold value is less than the minimum threshold value, the vector is a negative vector and is represented by a downward arrow;
step five, online abnormity diagnosis; inputting the mass flow and temperature data of the input and output variables obtained in the data acquisition step
Figure BDA00023112158000000613
In the fault vector construction step, the fault vector is obtainedCorresponding to
Figure BDA00023112158000000615
Fault vector, then with the established
Figure BDA00023112158000000612
Comparing the fault vector table; if present and
Figure BDA00023112158000000616
if the fault vector table shows the fault, the roller kiln system at the moment has the fault shown in the table, and if the fault does not exist, the fault vector table shows that the fault does not exist
Figure BDA00023112158000000614
And if the fault vector table is similar, the running state of the roller kiln system is good.
Compared with the prior art, the invention has the following beneficial effects:
the invention realizes the fault detection of the roller kiln production operation system; by collecting the mass flow and temperature data of the input and output variables of the system and carrying out the process on the variables
Figure BDA0002311215800000071
Analysis of, constructed variables
Figure BDA0002311215800000072
The fault vector table is used for detecting and diagnosing faults existing in the production operation system of the roller kiln; different from the existing method for constructing an expert fault knowledge system, the method has small burden on enterprises and reduces the generation cost of the enterprises; in the aspect of practicability, the method is simple, easy and efficient, simple to operate, remarkable in effectiveness and convenient to apply.
Drawings
FIG. 1 is a schematic flow diagram of the present invention;
FIG. 2 is a schematic view of an input/output frame of the roller kiln according to the present invention;
FIG. 3 is a schematic diagram of a fault vector generation process according to the present invention;
FIG. 4 shows the present invention
Figure BDA0002311215800000073
Failure vector table effect graph (burn stage).
Detailed Description
The present invention will be described in further detail with reference to examples and drawings, but the present invention is not limited thereto.
The invention discloses a method based on
Figure BDA0002311215800000074
The method for detecting the abnormality of roller kiln system includes such steps as creating roller kiln system
Figure BDA0002311215800000075
An analysis framework for analyzing the type of fault present in the system by
Figure BDA0002311215800000076
Equilibrium analysis, construction
Figure BDA0002311215800000077
And the fault vector table realizes detection and diagnosis of faults existing in the production operation system of the roller kiln.
Specifically, as shown in FIGS. 1-4, a method for manufacturing a semiconductor device
Figure BDA0002311215800000078
The method for detecting the abnormity of the roller kiln system of the model comprises the following steps:
firstly, analyzing input and output variables; analyzing the working mechanism of the roller kiln system to obtain an input and output frame diagram of the roller kiln, as shown in fig. 2, it can be known that the system input variables are: green bricks, natural gas, combustion-supporting gas and cooling gas; the output variables are: firing bricks, flue gas and cooling exhaust gas.
Step two, data acquisition; the data acquisition mainly acquires data of input and output variables in a roller kiln system in real time, wherein the data comprises mass flow and temperature of each input and output variable; in addition, it needs to be collected
Figure BDA0002311215800000082
The basic data required for the analysis, including reference environmental conditions such as temperature and pressure, the mole-specific heat capacity of the input and output streams, and the composition of the streams and their standards
Figure BDA0002311215800000083
The data acquisition method mainly comprises the following steps:
(1) acquiring temperature data through a temperature sensor;
(2) collecting gas mass flow data including natural gas, combustion-supporting gas, flue gas, cooling gas and cooling waste gas through a flowmeter;
(3) collecting the feeding speed data of the ceramic tile through the speed of the conveyor belt;
(4) measuring smoke components through a smoke analyzer, and measuring components of fuel gas, combustion-supporting gas, cooling gas and cooling waste gas through an Ordovician gas analyzer;
(5) determining the composition of the green brick and the fired brick by x-ray diffraction;
(6) the basic environment state refers to an environment model of the Guishan-Jitian;
(7) the molar constant pressure heat capacity of the material flow is obtained by looking up the molar heat capacity table of the material;
(8) through energy system
Figure BDA0002311215800000085
Analysis technical guide' and other related books to find the standard of material flow
Figure BDA0002311215800000084
The basic data (part) examined are shown in the following table:
TABLE 1 correlation data for natural gas
Figure BDA0002311215800000081
TABLE 2 correlation data of combustion supporting gas, cooling gas and cooling exhaust gas
Figure BDA0002311215800000093
TABLE 3 Smoke correlation data
Figure BDA0002311215800000091
TABLE 4 data relating to dried and fired bricks
Figure BDA0002311215800000092
Step three, performing a first step of cleaning the substrate,
Figure BDA0002311215800000094
analyzing; due to the environment condition will be right
Figure BDA0002311215800000095
The analysis has great influence on
Figure BDA0002311215800000096
In the analysis process, a reference environment state needs to be set; in addition, the production and operation system of the roller kiln is a very complex system, and in order to reduce the influence of secondary factors, some assumptions need to be made on the system, which are specifically as follows:
(1) ambient temperature T0300K, ambient pressure P0＝101.3kPa；
(2) The elements contained in the air take the corresponding composition gas of the air as a reference substance, and take the molar component of saturated humid air as the component of the reference substance;
(3) the other elements are based on the most stable pure substance (liquid or solid) containing the element, and the diffusion of the actual solid substance is considered
Figure BDA0002311215800000104
Difficult to use, adopt T0、P0Of pure solid reference substances under the conditions
Figure BDA0002311215800000105
A value of 0;
(4) assuming the system is a constant pressure, adiabatic, steady state operating open system;
(5) neglecting the kinetic energy and potential energy of the input flow and the output flow, and not considering the consumption of electric energy;
(6) assuming natural gas, air and flue gas as ideal gases;
(7) ignoring water vapor in the air;
of substances
Figure BDA0002311215800000106
The calculation of (a) is based on the reference environmental conditions and system assumptions, known as a steady flow open system, in which the molar quantity of substance x is determined
Figure BDA0002311215800000107
Is divided into physics
Figure BDA0002311215800000108
And chemistry
Figure BDA0002311215800000109
The following formula:
Ex＝Ex,ph+Ex,ch
in the above formula, ExIs the mole of substance x
Figure BDA00023112158000001010
The unit is kJ/mol; ex,phIs the physics of substance x
Figure BDA00023112158000001013
The unit is kJ/mol; ex,chIs the chemistry of substance x
Figure BDA00023112158000001011
The unit is kJ/mol;
input of substance x
Figure BDA00023112158000001012
Comprises the following steps:
Figure BDA0002311215800000101
in the above formula, the first and second carbon atoms are,
Figure BDA0002311215800000102
is the input of a substance x
Figure BDA00023112158000001014
The unit is kJ/s; m isxIs the mass flow of the substance x, with the unit being kg/s; mxIs the molar mass of substance x, in g/mol;
physics of physics
Figure BDA00023112158000001015
The formula of (1) is:
Figure BDA0002311215800000103
in the above formula, T0Is a reference environment temperature, T is a current environment temperature and the unit is K; p is a radical of0Is the reference environment pressure intensity, p is the current environment pressure intensity, and the unit is kPa; h is0For the substance m at a temperature T0The enthalpy of the substance m at the temperature T, h is the enthalpy of the substance m at the temperature T, and the unit is kJ/(mol.K); s0For the substance m at a temperature T0The entropy of time, s is the entropy of the substance m at the temperature T, and the unit is kJ/(mol.K); c. CpThe molar constant pressure heat capacity of a substance x is expressed in kJ/(mol. K); r is a general air constant of 8.3145 multiplied by 10-3kJ/(mol·K)；
The upper type
Figure BDA0002311215800000111
Is the thermal component of the physical congestion and,
Figure BDA0002311215800000112
is the pressure component of the physical congestion; if c ispIs constant or average specific heat, and neglecting the pressure component, then there are:
Figure BDA0002311215800000113
chemistry
Figure BDA0002311215800000116
The formula of (1) is:
Figure BDA0002311215800000114
in the above formula, xiIs the mole fraction of component i in substance x;
Figure BDA0002311215800000115
as standard for component i
Figure BDA0002311215800000117
The unit is kJ/mol.
Step four, performing a first step of cleaning the substrate,
Figure BDA0002311215800000118
constructing a fault vector table; the complex structure and the working process of the modern roller kiln determine the complexity of the fault; although the failure of the roller kiln is various, the failure can be mainly divided into two categories: one is a condition fault closely related to internal temperature, pressure, atmosphere, etc.; the other type is about the instrument faults of hardware equipment such as a motor, an actuating mechanism, a roller way and the like; by analysis, the fault types are mainly:
(1) abnormal failure of material flow;
(2) faults (e.g., leaks) that affect the overall model;
(3) abnormal fluctuation fault of temperature in the kiln;
the specific description of each type of failure (burn period) is shown in table 5 below:
TABLE 5 Fault types and their detailed description (firing stage)
Fault numbering Description of faults
1 Roller bed fault (Green brick mass flow reduced)
2 Roller bed fault (Green brick mass flow rate increasing)
3 Failure of combustion-supporting gas pipeline (mass flow of combustion-supporting gas is increased)
4 Failure of combustion-supporting gas pipeline (mass flow of combustion-supporting gas becomes small)
5 Natural gas pipeline failure (Natural gas mass flow rate increasing)
6 Natural gas pipeline failure (Natural gas mass flow rate decreasing)
7 Flue gas blower fault (flue gas mass flow reduced)
8 Flue gas fan failure (flue gas mass flow rate increase)
9 Nozzle failure (Natural gas leakage)
10 Temperature anomaly in kiln (inlet flue gas temperature rise)
11 Temperature anomaly in kiln (reduction of population smoke temperature)
In order to generate a data set, mass flow and temperature data of each input/output variable when a fault occurs, which are shown in table 1, are obtained through fluent simulation and are combined with mass flow and temperature data of each variable of a normal working condition, which are acquired by a sensor; the data obtained in the normal working condition is used for carrying out normalization processing on the whole data set, a unit-free quantity is generated, the unit-free quantity represents the size of deviation and can be used for constructing a fault vector; the generation of bright fault vectors is similar to that of fig. 3.
Wherein the normalization processing adopts a standard deviation normalization method; the standard deviation normalization method is to normalize the data through the mean value and the standard deviation of the sample data, and the processed data conforms to the standard normal distribution with the mean value of 0 and the standard deviation of 1; the calculation formula is as follows:
Figure BDA0002311215800000121
in the above formula, x*Is the corrected value; mu is the mean value of the sample data; σ is the standard deviation of the sample data;
in order to construct a bright failure vector, a threshold function is applied to the bright-physical and bright-chemical data of each normalized variable, which transforms each variable into a qualitative vector; in order to eliminate the influence on rounding errors, a limit function f (x) epsilon { -1,0,1} is introduced to classify the normalized data x into { -1,0,1}, and the formula of the limit function is as follows:
x≤-M→f(x)＝-1
-M＜x＜M→f(x)＝0
x≥M→f(x)＝1
in the above formula, x is the normalized data, and M is the threshold function;
the threshold function is:
M＝3σ*
in the above formula, σ*The standard deviation for the normalized data is 1;
when a fault occurs, collecting the mass flow and temperature data of input and output variables, passing through
Figure BDA0002311215800000131
Analyzing the process to obtain the physics of the input and output variables
Figure BDA0002311215800000132
And chemistry
Figure BDA00023112158000001319
Then applying normalization method to input and output variable physics
Figure BDA0002311215800000134
And chemistry
Figure BDA0002311215800000135
Normalizing, introducing a limit function f (x) epsilon { -1,0,1} to normalize the physical property of the processed variable
Figure BDA0002311215800000136
And chemistry
Figure BDA0002311215800000137
The classification is { -1,0,1}, wherein 1 represents a positive vector, 0 represents a zero vector, and-1 represents a negative vector; if physics
Figure BDA0002311215800000138
And chemistry
Figure BDA0002311215800000139
If it is greater than the maximum threshold, it is a positive vector, indicated by an upward arrow, if physical
Figure BDA00023112158000001310
And chemistry
Figure BDA00023112158000001311
If the minimum threshold value is less than the minimum threshold value, the vector is a negative vector and is represented by a downward arrow; constructed by
Figure BDA00023112158000001312
The failure vector table effect diagram (burn-in stage) is shown in fig. 4.
Step five, online abnormity diagnosis; inputting the mass flow and temperature data of the input and output variables obtained in the data acquisition step
Figure BDA00023112158000001313
In the step of constructing fault vectors, corresponding fault vectors are obtained
Figure BDA00023112158000001314
Fault vector, then with the established
Figure BDA00023112158000001315
Comparing the fault vector table; if present and
Figure BDA00023112158000001316
if the fault vector table shows the fault, the roller kiln system at the moment has the fault shown in the table, and if the fault does not exist, the fault vector table shows that the fault does not exist
Figure BDA00023112158000001317
And if the fault vector table is similar, the running state of the roller kiln system is good.
Figure BDA00023112158000001318
And (3) analysis: according to the collected mass flow and temperature data of the input and output variables, the related variables are carried out
Figure BDA0002311215800000141
Value calculation including physics of related variables
Figure BDA0002311215800000142
And chemistry
Figure BDA0002311215800000143
Figure BDA0002311215800000144
Constructing a fault vector table: analyzing and summarizing the fault type, the fault reason and the influence of the fault on the system in the roller kiln system; to a variable
Figure BDA0002311215800000145
The values are normalized to identify anomalies by determining a threshold
Figure BDA0002311215800000146
A value; according to the abnormality corresponding to each fault
Figure BDA0002311215800000147
Value, construct
Figure BDA0002311215800000148
A fault vector table.
The invention realizes the fault detection of the roller kiln production operation system; by collecting the mass flow and temperature data of the input and output variables of the system and carrying out the process on the variables
Figure BDA0002311215800000149
Analysis of, constructed variables
Figure BDA00023112158000001410
A fault vector table for realizing the generation of roller kilnsDetecting and diagnosing faults existing in the production and operation system; different from the existing method for constructing an expert fault knowledge system, the method has small burden on enterprises and reduces the generation cost of the enterprises; in the aspect of practicability, the method is simple, easy and efficient, simple to operate, remarkable in effectiveness and convenient to apply.
The present invention is not limited to the above embodiments, and any other changes, modifications, substitutions, combinations, and simplifications which do not depart from the spirit and principle of the present invention should be construed as equivalents and are included in the scope of the present invention.

Claims (1)

1. Based on
Figure FDA0003029853540000011
The method for detecting the abnormity of the roller kiln system of the model is characterized by comprising the following steps of:
firstly, analyzing input and output variables; the working mechanism of the roller kiln system is analyzed, and the system input variables are as follows: green bricks, natural gas, combustion-supporting gas and cooling gas; the output variables are: firing bricks, flue gas and cooling exhaust gas;
step two, data acquisition; acquiring data of input and output variables in a roller kiln system in real time by data acquisition, wherein the data comprises the mass flow and the temperature of each input and output variable; in addition, it needs to be collected
Figure FDA0003029853540000012
Basic data required for analysis including reference environmental conditions, temperature and pressure, heat capacity at molarity of input and output streams, and composition of streams and their standards
Figure FDA0003029853540000013
The data acquisition method mainly comprises the following steps:
(1) acquiring temperature data through a temperature sensor;
(2) collecting gas mass flow data including natural gas, combustion-supporting gas, flue gas, cooling gas and cooling waste gas through a flowmeter;
(3) collecting the feeding speed data of the ceramic tile through the speed of the conveyor belt;
(4) measuring smoke components by a smoke analyzer, and measuring natural gas, combustion-supporting gas, cooling gas and cooling waste gas components by an Ordovician gas analyzer;
(5) determining the composition of the green brick and the fired brick by x-ray diffraction;
(6) the basic environment state refers to an environment model of the Guishan-Jitian;
(7) the molar constant pressure heat capacity of the material flow is obtained by looking up the molar heat capacity table of the material;
(8) standard of material flow found by related books
Figure FDA0003029853540000014
Step three, performing a first step of cleaning the substrate,
Figure FDA0003029853540000015
analyzing; due to the environment condition will be right
Figure FDA0003029853540000016
The analysis has great influence on
Figure FDA0003029853540000017
In the analysis process, a reference environment state needs to be set; in addition, the production and operation system of the roller kiln is a very complex system, and in order to reduce the influence of secondary factors, some assumptions need to be made on the system, which are specifically as follows:
(1) ambient temperature T0300K, ambient pressure P0＝101.3kPa；
(2) The elements contained in the air take the corresponding composition gas of the air as a reference substance, and take the molar component of saturated humid air as the component of the reference substance;
(3) the other elements are based on the most stable pure substance containing the element, and the pure substance isThe state being liquid or solid, taking into account diffusion of the actual solid substance
Figure FDA0003029853540000024
Difficult to use, adopt T0、P0Of pure solid reference substances under the conditions
Figure FDA0003029853540000025
A value of 0;
(4) assuming the system is a constant pressure, adiabatic, steady state operating open system;
(5) neglecting the kinetic energy and potential energy of the input flow and the output flow, and not considering the consumption of electric energy;
(6) assuming natural gas, air and flue gas as ideal gases;
(7) ignoring water vapor in the air;
of substances
Figure FDA0003029853540000026
The calculation of (a) is based on the reference environmental conditions and system assumptions, known as a steady flow open system, in which the molar quantity of substance x is determined
Figure FDA0003029853540000027
Is divided into physics
Figure FDA0003029853540000028
And chemistry
Figure FDA0003029853540000029
The following formula:
Ex＝Ex，ph+Ex，ch
in the above formula, ExIs the mole of substance x
Figure FDA00030298535400000210
The unit is kJ/mol; ex，phIs the physics of substance x
Figure FDA00030298535400000211
The unit is kJ/mol; ex，chIs the chemistry of substance x
Figure FDA00030298535400000212
The unit is kJ/mol;
input of substance x
Figure FDA00030298535400000213
Comprises the following steps:
Figure FDA0003029853540000021
in the above formula, the first and second carbon atoms are,
Figure FDA0003029853540000022
is the input of a substance x
Figure FDA00030298535400000214
The unit is kJ/s; m isxIs the mass flow of the substance x, with the unit being kg/s; mxIs the molar mass of substance x, in g/mol;
physics of physics
Figure FDA00030298535400000215
The formula of (1) is:
Figure FDA0003029853540000023
in the above formula, T0Is a reference environment temperature, T is a current environment temperature and the unit is K; p is a radical of0Is the reference environment pressure intensity, p is the current environment pressure intensity, and the unit is kPa; h is0For the substance m at a temperature T0The enthalpy of the substance m at the temperature T, h is the enthalpy of the substance m at the temperature T, and the unit is kJ/(mol.K); s0For the substance m at a temperature T0The entropy of time, s is the entropy of the substance m at the temperature T, and the unit is kJ/(mol.K); c. CpThe molar constant pressure heat capacity of a substance x is expressed in kJ/(mol. K); r is a general air constant of 8.3145 multiplied by 10-3kJ/(mol·K)；
The upper type
Figure FDA0003029853540000031
Is the thermal component of the physical congestion and,
Figure FDA0003029853540000032
is the pressure component of the physical congestion; if c ispIs constant or average specific heat, and neglecting the pressure component, then there are:
Figure FDA0003029853540000033
chemistry
Figure FDA0003029853540000038
The formula of (1) is:
Figure FDA0003029853540000034
in the above formula, xiIs the mole fraction of component i in substance x;
Figure FDA0003029853540000035
as standard for component i
Figure FDA0003029853540000036
The unit is kJ/mol;
step four, performing a first step of cleaning the substrate,
Figure FDA0003029853540000037
constructing a fault vector table; the complex structure and the working process of the modern roller kiln determine the complexity of the fault; by analysis, the fault types are mainly:
(1) abnormal failure of material flow;
(2) faults affecting the overall model;
(3) abnormal fluctuation fault of temperature in the kiln;
in order to generate a data set, mass flow and temperature data of each input/output variable when a fault occurs are obtained through fluent simulation and combined with mass flow and temperature data of each variable under normal working conditions, which are acquired by a sensor; the data obtained in the normal working condition is used for carrying out normalization processing on the whole data set, a unit-free quantity is generated, the unit-free quantity represents the size of deviation and can be used for constructing a fault vector;
wherein the normalization processing adopts a standard deviation normalization method; the standard deviation normalization method is to normalize the data through the mean value and the standard deviation of the sample data, and the processed data conforms to the standard normal distribution with the mean value of 0 and the standard deviation of 1; the calculation formula is as follows:
Figure FDA0003029853540000041
in the above formula, x*Is the corrected value; mu is the mean value of the sample data; σ is the standard deviation of the sample data;
to build up
Figure FDA0003029853540000042
Fault vector, applying threshold function to normalized physical of each variable
Figure FDA0003029853540000043
And chemistry
Figure FDA0003029853540000044
Data, making physical of variables
Figure FDA0003029853540000045
And chemistry
Figure FDA0003029853540000046
Converting into a qualitative vector; in order to eliminate the influence on rounding errors, a limit function f (x) epsilon { -1,0,1} is introduced to classify the normalized data x into { -1,0,1}, and the formula of the limit function is as follows:
x≤-M→f(x)＝-1
-M＜x＜M→f(x)＝0
x≥M→f(x)＝1
in the above formula, x is the normalized data, and M is the threshold function;
the threshold function is:
M＝3σ*
in the above formula, σ*The standard deviation for the normalized data is 1;
when a fault occurs, collecting the mass flow and temperature data of input and output variables, passing through
Figure FDA0003029853540000051
Analyzing the process to obtain the physics of the input and output variables
Figure FDA0003029853540000052
And chemistry
Figure FDA0003029853540000053
Then applying normalization method to input and output variable physics
Figure FDA0003029853540000054
And chemistry
Figure FDA0003029853540000055
Normalizing, introducing a limit function f (x) epsilon { -1,0,1} to normalize the physical property of the processed variable
Figure FDA0003029853540000056
And chemistry
Figure FDA0003029853540000057
Is classified as { -1,0,1}, wherein 1 represents the forward directionQuantity, 0 represents a zero vector, -1 represents a negative vector; if physics
Figure FDA00030298535400000510
And chemistry
Figure FDA0003029853540000058
If it is greater than the maximum threshold, it is a positive vector, indicated by an upward arrow, if physical
Figure FDA0003029853540000059
And chemistry
Figure FDA00030298535400000511
If the minimum threshold value is less than the minimum threshold value, the vector is a negative vector and is represented by a downward arrow;
step five, online abnormity diagnosis; inputting the mass flow and temperature data of the input and output variables obtained in the data acquisition step
Figure FDA00030298535400000512
In the step of constructing fault vectors, corresponding fault vectors are obtained
Figure FDA00030298535400000513
Fault vector, then with the established
Figure FDA00030298535400000514
Comparing the fault vector table; if present and
Figure FDA00030298535400000515
if the fault vector table shows that the fault indicated in the table exists in the roller kiln system at the moment, if the fault vector table does not show that the fault vector table shows that the fault does not exist in the roller kiln system at the moment
Figure FDA00030298535400000516
And the condition in the fault vector table indicates that the running state of the roller kiln system is good at the moment.
CN201911259474.7A 2019-12-10 2019-12-10 exergy model-based roller kiln system anomaly detection method Expired - Fee Related CN111059896B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911259474.7A CN111059896B (en) 2019-12-10 2019-12-10 exergy model-based roller kiln system anomaly detection method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911259474.7A CN111059896B (en) 2019-12-10 2019-12-10 exergy model-based roller kiln system anomaly detection method

Publications (2)

Publication Number Publication Date
CN111059896A CN111059896A (en) 2020-04-24
CN111059896B true CN111059896B (en) 2021-08-24

Family

ID=70300295

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911259474.7A Expired - Fee Related CN111059896B (en) 2019-12-10 2019-12-10 exergy model-based roller kiln system anomaly detection method

Country Status (1)

Country Link
CN (1) CN111059896B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111931419B (en) * 2020-07-30 2022-07-26 广东工业大学 Improved particle swarm algorithm-based ceramic roller kiln process parameter optimization method
CN113532138B (en) * 2021-07-06 2023-07-28 广东工业大学 Roller kiln firing zone anomaly detection algorithm based on decision fusion frame

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104635724B (en) * 2014-12-25 2017-02-22 重庆科技学院 Abnormity detection method for natural gas purification process based on analysis of independent component of dynamic kernel
CN109407649B (en) * 2018-10-09 2021-01-05 宁波大学 Fault type matching method based on fault characteristic variable selection
CN109376778B (en) * 2018-10-09 2021-06-15 宁波大学 Fault classification diagnosis method based on characteristic variable weighting
CN209145663U (en) * 2018-11-07 2019-07-23 北京易泽动力科技有限公司 A kind of heat supply steam extraction optimization system based on small * damage Mixing Technology
CN110378036A (en) * 2019-07-23 2019-10-25 沈阳天眼智云信息科技有限公司 Fault Diagnosis for Chemical Process method based on transfer entropy

Also Published As

Publication number Publication date
CN111059896A (en) 2020-04-24

Similar Documents

Publication Publication Date Title
CN111059896B (en) exergy model-based roller kiln system anomaly detection method
CN109655488B (en) Gas calorific value soft measurement method based on mixed gas preheating combustion
CN109541168B (en) Coal powder economic fineness on-line monitoring and adjusting method
CN110003923B (en) Device and method for measuring coke burning loss in dry quenching furnace
CN108549792A (en) A kind of solid waste burning process dioxin emission concentration flexible measurement method based on latent structure mapping algorithm
WO2021159585A1 (en) Dioxin emission concentration prediction method
CN109615084A (en) Positive-pressure type medium-speed pulverizer fuel pulverizing plant fineness of pulverized coal real-time monitoring system
CN112131517B (en) Method for measuring and calculating lower calorific value of garbage in garbage incineration power plant
CN111308024A (en) System and method for gridding measurement of gaseous components in flue gas
CN104536396A (en) Soft measurement modeling method used in cement raw material decomposing process in decomposing furnace
CN106885876B (en) A kind of coke oven flue gas oxygen content and nitrous oxides concentration detection system and method
CN109086949B (en) Blast furnace gas generation amount and heat value prediction method based on gas component change
CN111505236B (en) Coal quality monitoring method for acquiring element analysis in real time based on coal quality industrial analysis
CN110888403A (en) Intelligent soot blowing closed-loop control system based on minimum loss boiler convection heating surface
CN113030002B (en) Water vapor quality abnormity determination method and monitoring equipment based on infrared spectroscopy
CN210012810U (en) Device for measuring coke burning loss in dry quenching furnace
CN109613059B (en) Metallurgical gas calorific value online measuring and calculating method based on combustion system operation parameters
CN107860763B (en) Online monitoring method and device for concentration of alkali metal and trace element in gas
CN113449954B (en) Method for measuring and calculating bottom air leakage rate of dry slag-discharging boiler
Lu et al. Online prediction method of cement clinker f-Cao based on K-ELM
CN109580711B (en) Soft measurement method for gas calorific value under condition of blast furnace gas and converter gas co-combustion
CN113419025A (en) Coke oven air excess coefficient real-time monitoring device and adjusting method
CN109632881B (en) Metallurgical gas calorific value soft measurement method based on gas preheating system heat exchange parameters
CN207992199U (en) One kind is based on14The mixed combustion of the biomass of C isotope on-line checkings is than monitoring system
CN109189029B (en) Energy-saving on-line monitoring system and method for low-temperature economizer of thermal power plant

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
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20210824

Termination date: 20211210

CF01 Termination of patent right due to non-payment of annual fee