CN101551663A - Cracking reaction pre-warning method for cracking furnace - Google Patents
Cracking reaction pre-warning method for cracking furnace Download PDFInfo
- Publication number
- CN101551663A CN101551663A CNA2009100508605A CN200910050860A CN101551663A CN 101551663 A CN101551663 A CN 101551663A CN A2009100508605 A CNA2009100508605 A CN A2009100508605A CN 200910050860 A CN200910050860 A CN 200910050860A CN 101551663 A CN101551663 A CN 101551663A
- Authority
- CN
- China
- Prior art keywords
- cracking
- early warning
- neural network
- early
- warning
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Abstract
The invention relates to a cracking reaction pre-warning method for cracking furnace in order to predict the development of the cracking reaction, enable the operators to adjust in time, reduce abnormal working condition time, and increase production efficiency. The method in the invention comprises: training the cracking reaction pre-warning model established based on nerve network by acquiring, counting and analyzing the normal and abnormal historical data in production processes; determining the model structure and parameters; and utilizing in production process to implement pre-warning, that is, when new data is inputted into the model, performing working condition prediction, and judging whether the cracking process is normal, and whether the cracking depth is in the normal range, so as to the operator adjust correspondingly. The method in the invention can be realized by programming in distributed control system or field bus control system; compared with common cracked gas component on-line measuring instrument, the invention achieves the advantages of low cost, simple implementation, no maintenance and real time pre-warning.
Description
Technical field
The present invention relates to a kind of method for early warning of pyrolysis furnace cracking reaction, especially a kind of method for early warning of ethylene unit pyrolysis furnace cracking reaction.
Background technology
In process industry is produced, because production run complexity, operating procedure condition often change, be difficult to find out from detection variable whether production run occurs unusually intuitively, often to wait until that abnormal conditions have continued a period of time, when the stability of the quality of product or system being exerted an influence, situation just can note abnormalities.At present, the process industry process units adopts Distributed Control System (DCS) (DCS) monitoring or manually-operated mostly, DCS generally has panalarm, when the production run key parameter exceeds the warning limit value, will carry out sound, light warning or information indicating, but this moment, fault took place, and had caused certain negative effect to producing; And because alarm parameters is difficult for revising, system's false alarm is more, can't eliminate, and influences operating personnel's the state of mind to a certain extent.Turn to the transitional period of unusual service condition to pinpoint the problems if can be in from nominal situation, and give the alarm, just can in time find the problem and handle, loss is reduced to minimum at device.Early warning technology is exactly production run to be carried out under the situation of real-time analysis, the operating mode that in time notes abnormalities trend is also warned operating personnel, it is taken place or imminent unusual condition with the fastest speed cancellation element, to reduce the operation of equipment under " improper production status " largely, and the probability of reduction industrial accident generation, not only ensure the safety of production and process units, can also improve economic benefit of enterprises.
Ethene is the important foundation raw material of chemical industry, and its output becomes the outstanding feature of weighing a country petrochemical complex development.The scale of ethylene production, cost, production stability, product quality etc. all can produce significant impact to whole petroleum chemical industry, so ethylene unit just becomes core process units related to the overall situation.Faucet device as ethylene production, the quality of pyrolysis furnace ruuning situation, directly influence the yield of ethene and the carrying out of subsequent handling, if do not produce qualified pyrolysis gas at cracking zone, so follow-up rectified purified process also is difficult to the qualified ethylene product of output.
In process of production, after raw material enters into the pyrolysis furnace radiation section, reactions such as dehydrogenation, chain rupture take place owing to being subjected to the high temperature heating, resolve into multiple products such as multiple hydro carbons such as ethene, propylene and carbon, hydrogen, water, carbon monoxide, cracking reaction is carried out the online in real time analysis just can in time grasp the degree of cracking and the ratio of production key component, this is to improving output capacity, and it is vital increasing economic benefit.The production status of existing cracker process industry is as described above produced similar, shortage is to cracking severity alarm mechanism effectively and timely, though the in-line meter that is used for detecting the cracking reaction degree of depth is specially arranged, but generally all cost an arm and a leg, difficult in maintenance, poor stability, and have 20~30 minutes hysteresis, can't reflect the variation of the cracking reaction degree of depth in real time.
Summary of the invention
In order to solve the problems of the technologies described above, the method for early warning that the invention provides a kind of pyrolysis furnace cracking reaction is to reach following purpose: reduce improper operating mode by this early warning technology, increase the business economic benefit, the generation of minimizing accident, its principle of work is that production run is analyzed, the expertise that basis has had before accident takes place, to advising property of operating personnel suggestion, guiding operation personnel carry out control operation, and the steady production process is in time avoided loss.
The ultimate principle of technical solution of the present invention is as follows:
When cracking reaction occurs when unusual; variablees such as relative pyrolysis furnace outlet temperature, feed rate, dilution steam generation flow also can show corresponding variation; so statistics and analysis by the normal and exception history data in some production runes; can set up the Early-warning Model of cracking reaction; when new real time data is imported this model; go out this operating mode and belong to normal or unusual classification with regard to measurable; do not need to detect the yield of each product in the pyrolysis gas like this; just can judge whether cracking process normally carries out, and whether cracking severity is in normal range.
Pyrolysis furnace is the dynamical system of a complexity under running status, particularly under multiple faults source and unsteady state, require Early-warning Model to have self-adaptation and robustness, just require pattern classifier to have the ability of handling the modal distortion that causes by noise adaptively, can adjust assorting process according to the variation of equipment operation parameter.Nerual network technique possesses this advantage.Early warning technology based on neural network, exactly on the basis of the dynamic model of setting up monitored process, be input as the auxiliary variable of the easy measurement relevant with monitored variable, be output as the duty of system, use normal data and abnormal data training network respectively, by the learning ability of neural network, make model have oneself and judge the whether normal ability of operating mode, and can dope the abnormality of production run.Early warning for the cracking of ethylene reaction, by historical data neural network is trained, set up the neural network Early-warning Model of cracking reaction, this model can carry out real-time analysis to the variation of operating mode, ability with independent judgment working condition, when certain factor broke down and causes work condition abnormality, neural network can in time be made a response, before this fault also impacts cracking reaction, do not send early warning, be very helpful for the early detection tool of fault.
Method of the present invention is made of following steps.
1, the network structure of neural network
Because cracking of ethylene is a complex dynamic process, the present invention adopts dynamic recurrence network (Elman network) to set up Early-warning Model, the hidden layer unit of network adopts the Tansig function as excitation function, and output layer unit adopts the Logsig function as excitation function.The node number of input layer is 1~7, and the hidden layer node number is 1~30, and all there is a corresponding with it structural unit each unit.The embodiment of network structure as shown in Figure 1,7 nodes of this embodiment input layer, 8 nodes of hidden layer, 1 node of output layer.
2, the input of neural network, output variable
In fact determining of neural network input be exactly features extraction, for choosing of characteristic quantity, considers mainly whether it has clearer and more definite cause-effect relationship with the ruuning situation of installing, and promptly input variable is to reflect out of order measurable sign.Cracking severity is to reflect the whether normal index of operation of cracker, but since the value of cracking severity be not easy directly to record, therefore select that some are convenient to measure, with the related variable of cracking severity as input variable.The selectable input variable scope of the present invention is:
(1) pyrolysis furnace outlet temperature (x
1, ℃)
(2) dilution ratio (or vapour hydrocarbon ratio) (x
2)
(3) pyrolysis furnace feed loading (x
3, T/h)
(4) feedstock oil density (x
4, g/cm
3)
(5) waste heat boiler outlet temperature (x
5, ℃)
(6) pyrolysis furnace outlet temperature rate of change (x
6, ℃/m)
(7) pyrolysis furnace feed loading rate of change (x
7, T/hm)
When specifically choosing, can choose wherein one or more operating conditionss as input variable.
The input variable of network model utilizes formula (1) to carry out normalized:
(1) in the formula, x
iBe the actual measured value of i input variable, i
MaxBe input variable number, sx
iRepresent after i the input variable normalization value as the neural network input,
Expression collects the variation range of i input variable, and the variation range of input variable is [a, b] after the normalization.
The output variable y ' of neural network is a production status, and its value is that 0,1 or 2,0 expression operating modes are normal, the general early warning of 1 expression, the serious early warning of 2 expressions.
3, train according to training sample set pair neural network type
In general, training sample set not only should be contained the data of all fault modes comprehensively, also should have certain representativeness, must guarantee the validity of study simultaneously.The selection of test sample book collection should be satisfied the principle of " crosscheck ".
Collect n and organize representational commercial plant data, wherein every group of data comprise
After normalization be
Form training sample.To the cracking reaction Early-warning Model, with
As the input of network, corresponding y is as desired value, training network.Can adopt error backpropagation algorithm (BP algorithm) that the weights in the Elman network are trained, when reaching certain accuracy requirement, stop training, obtain cracking reaction process neural network Early-warning Model.Also can adopt intelligent optimization algorithms such as genetic algorithm that the weights in the Elman network are learnt, the process of study as shown in Figure 2.During study, the error of the actual condition y in the calculation training sample data and the output y ' of neural network with the inverse of the error fitness function as optimized Algorithm, is obtained the neural network weight of fitness function when maximum by the iteration of optimized Algorithm.
4, utilize the neural network model that trains to carry out the operating mode prediction
The network that trains can be told the degree whether system is in normal operating conditions or early warning.During prediction, with the network that the corresponding input variable substitution of cracking process trains, suitable classification thresholds c is selected in the output of calculating neural network
0And c
w, when exporting less than threshold value c
0The time, deterministic process is a nominal situation, when exporting greater than threshold value c
0And less than threshold value c
wThe time, deterministic process is general early warning; When exporting greater than threshold value c
wThe time, deterministic process is serious early warning.
The process flow diagram of said method as shown in Figure 3.
Description of drawings
Fig. 1 is the Early-warning Model block diagram of pyrolysis furnace heat scission reaction process;
Fig. 2 utilizes the Early-warning Model training module figure of optimized Algorithm to pyrolysis furnace heat scission reaction process;
Fig. 3 utilizes the method flow diagram of neural network to the cracking reaction early warning
Embodiment
Explanation by following most preferred embodiment will help to understand the present invention, but not limit content of the present invention.
The neural network of present embodiment method adopts structure shown in Figure 1, and promptly the input layer number is 7, and the hidden layer node number is 8, and output layer node number is 1.
At first, gather 300 groups of pyrolysis furnaces from on-the-spot and experimental provision at different pyrolysis furnace outlet temperature (x
1, ℃), dilution ratio (or vapour hydrocarbon than) (x
2), pyrolysis furnace feed loading (x
3, T/h), feedstock oil density (x
4, g/cm
3), waste heat boiler outlet temperature (x
5, ℃), pyrolysis furnace outlet temperature rate of change (x
6, ℃/m), pyrolysis furnace feed loading rate of change (x
7, T/hm) under, corresponding cracker operating mode (0: normal, 1: generally unusual, 2: severely subnormal) form sample data.
Then, utilize (1) formula, above-mentioned each independent variable is carried out normalized: x
1Variation range [840,860], x
2Variation range [0.6,0.9], x
3Variation range [22,27], x
4Variation range [0.64,0.65], x
5Variation range [400,410], x
6Variation range [10,10], x
7Variation range [1,1], get a=-1, b=1, carry out normalization and calculate:
With 300 groups of sample datas after the normalization is training sample, and present embodiment adopts the BP algorithm that network is trained, and training module figure as shown in Figure 2; When the error sum of squares of the output y ' of neural network and the measured data y of training sample during less than certain threshold value (accuracy requirement of choosing with Early-warning Model of this threshold value changes, and gets 0.1 in the present embodiment), the decision network convergence obtains following one group of weight w
Ij l(l=1,2,3) and amount of bias b
i(i=1,2 ..., 8), b
oW wherein
Ij 1(i=1,2 ..., 8; J=1,2 ..., 8) and be the weights of i structural unit to j node of hidden layer; w
Ij 2(i=1,2 ..., 7; J=1,2 ..., 8) and be the weights of i node of input layer to j node of hidden layer; w
Ij 3(i=1,2 ..., 8; J=1) be the weights of i node of hidden layer to output node.b
i(i=1,2 ..., 8) and be the bias of i node of hidden layer; b
oBias for output node.Pyrolysis furnace heat scission reaction process Early-warning Model is:
x
h(k)=f(w
1·x
c(k)+w
2·sx(k)+b)
x
c(k)=x
h(k-1)
y′(k)=g(w
3·x
h(k)+b
o)
Wherein,
x
c(k)=[x
c1(k)x
c2(k)x
c3(k)x
c4(k)x
c5(k)x
c6(k)x
c7(k)x
c8(k)]
x
h(k)=[x
h1(k)x
h2(k)x
h3(k)x
h4(k)x
h5(k)x
h6(k)x
h7(k)x
h8(k)]
sx(k)=[sx
1(k)sx
2(k)sx
3(k)sx
4(k)sx
5(k)sx
6(k)sx
7(k)]
Y ' is the predicted value of cracking reaction Early-warning Model (k), desirable classification thresholds c
0=0.5, c
w=1.5, when y ' (k)≤c
0The time, current is nominal situation; Work as c
0<y ' (k)≤c
wThe time, current is general early warning; When y ' (k)>c
wThe time, current is serious early warning.
Method for early warning of the present invention can be real by the method that programs in Distributed Control System or field bus control system Existing, compare with general cracking gas component on-line measurement instrument, have cost low, implement simple, need not safeguard, in real time pre-The advantages such as police.
Claims (10)
1, a kind of method for early warning of pyrolysis furnace cracking reaction is characterized in that,
(1) chooses the input variable of the operating conditions of pyrolysis furnace as cracking of ethylene course of reaction neural network Early-warning Model;
(2) choose the output variable of production status zone bit as model;
(3) utilize industrial pyrolysis furnace actual production data as training sample, set up ethane cracking furnace cracking reaction process neural network Early-warning Model, and judge current working according to the output of Early-warning Model, trend in time notes abnormalities, and operating personnel are warned, make its unusual condition, reduce the generation of industrial accident, improve the economic benefit of device with quick cancellation element.
2, early warning technology according to claim 1, it is characterized in that the operating conditions of choosing in the described step (1) is: the combination of the one or more conditions in pyrolysis furnace outlet temperature, dilution ratio (or vapour hydrocarbon ratio), pyrolysis furnace feed loading, feedstock oil density, waste heat boiler outlet temperature, pyrolysis furnace outlet temperature rate of change, the pyrolysis furnace feed loading rate of change.
3, early warning technology according to claim 1, it is characterized in that, described neural network Early-warning Model to the cracking of ethylene course of reaction, adopt the Dynamical Recurrent Neural Networks structure, the node number of input layer is 1~7, the hidden layer node number is 1~30, and all there is a corresponding with it structural unit each unit, and output layer node number is 1.
4, early warning technology according to claim 3 is characterized in that, described neural network model input variable is through normalized:
In the formula, x
iBe the actual measured value of i input variable, i
MaxBe input variable number, sx
iRepresent after i the input variable normalization value as the neural network input,
Expression collects the variation range of i input variable, and the variation range of input variable is [a, b] after the normalization.
5, early warning technology according to claim 4 is characterized in that, collects n and organizes representational commercial plant data, and wherein every group of data comprise
After normalization be
Form training sample, to the cracking reaction Early-warning Model, with
As the input of network, corresponding y is as desired value, training network.
6, early warning technology according to claim 5, it is characterized in that, adopt error backpropagation algorithm (BP algorithm) that the weights in the Elman network are trained, when the error sum of squares of neural network output and training sample desired value during less than threshold value, stop training, obtain cracking reaction process neural network Early-warning Model.
7, early warning technology according to claim 5, it is characterized in that, intelligent optimization algorithms such as employing genetic algorithm are learnt the weights in the Elman network, during study, the error of the actual condition y in the calculation training sample data and the output y ' of neural network, with the inverse of error fitness function, obtain the neural network weight of fitness function when maximum by the iteration of optimized Algorithm as optimized Algorithm.
8, early warning technology according to claim 1 is characterized in that, during prediction, brings the corresponding input variable of cracking process into train network, calculates the output of neural network, selects suitable classification thresholds c
0And c
w, when exporting less than threshold value c
0The time, deterministic process is a nominal situation, when exporting greater than threshold value c
0And less than threshold value c
wThe time, deterministic process is general early warning; When exporting greater than threshold value c
wThe time, deterministic process is serious early warning.
According to arbitrary described early warning technology among the claim 2-8, it is characterized in that 9, best input number of nodes is 7.
According to arbitrary described early warning technology among the claim 1-8, it is characterized in that 10, described early warning technology is based on that Distributed Control System (DCS) or field bus control system realize.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNA2009100508605A CN101551663A (en) | 2009-05-08 | 2009-05-08 | Cracking reaction pre-warning method for cracking furnace |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNA2009100508605A CN101551663A (en) | 2009-05-08 | 2009-05-08 | Cracking reaction pre-warning method for cracking furnace |
Publications (1)
Publication Number | Publication Date |
---|---|
CN101551663A true CN101551663A (en) | 2009-10-07 |
Family
ID=41155936
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CNA2009100508605A Pending CN101551663A (en) | 2009-05-08 | 2009-05-08 | Cracking reaction pre-warning method for cracking furnace |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN101551663A (en) |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101794119A (en) * | 2010-03-08 | 2010-08-04 | 浙江中控软件技术有限公司 | Gas system balancing and optimal scheduling method, gas balancing and optimal scheduling device and gas balancing and optimal scheduling system |
CN102053595A (en) * | 2009-10-30 | 2011-05-11 | 中国石油化工股份有限公司 | Method for controlling cracking depth of cracking furnace in ethylene device |
CN102087520A (en) * | 2009-12-08 | 2011-06-08 | 上海自动化仪表股份有限公司 | Time-sharing scanning method of I/O (Input/Output) module |
CN102937670A (en) * | 2012-11-17 | 2013-02-20 | 河南科技大学 | Method for identifying resistance parameters of rotors of induction motor on basis of Elman neural network |
CN103646168A (en) * | 2013-11-26 | 2014-03-19 | 浙江工商大学 | Tiny-oil ignition early warning method |
CN105241239A (en) * | 2015-09-10 | 2016-01-13 | 广西大学 | Intelligent optimal control method and device for sintered brick tunnel kiln roasting process |
CN107341444A (en) * | 2017-06-07 | 2017-11-10 | 北京星网锐捷网络技术有限公司 | Data exception method for early warning and device |
CN108681656A (en) * | 2018-04-27 | 2018-10-19 | 上海卓然工程技术股份有限公司 | A kind of process analysis method based on ethane cracking furnace operation data |
CN109389170A (en) * | 2018-10-10 | 2019-02-26 | 常州大学 | A kind of gradation type operating condition method for early warning based on 3D convolutional neural networks |
CN111394117A (en) * | 2020-03-23 | 2020-07-10 | 北京赛普泰克技术有限公司 | Treatment method and treatment system for gasoline adsorption desulfurization, electronic equipment and storage medium |
CN113219912A (en) * | 2021-03-31 | 2021-08-06 | 成都飞机工业(集团)有限责任公司 | Multi-Agent-based numerical control machining flexible manufacturing system machining process early warning method |
CN115386391A (en) * | 2022-08-01 | 2022-11-25 | 浙江中控技术股份有限公司 | Cracking furnace feeding flow monitoring and fault handling method and system |
-
2009
- 2009-05-08 CN CNA2009100508605A patent/CN101551663A/en active Pending
Cited By (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102053595A (en) * | 2009-10-30 | 2011-05-11 | 中国石油化工股份有限公司 | Method for controlling cracking depth of cracking furnace in ethylene device |
CN102087520A (en) * | 2009-12-08 | 2011-06-08 | 上海自动化仪表股份有限公司 | Time-sharing scanning method of I/O (Input/Output) module |
CN102087520B (en) * | 2009-12-08 | 2013-02-06 | 上海自动化仪表股份有限公司 | Time-sharing scanning method of I/O (Input/Output) module |
CN101794119A (en) * | 2010-03-08 | 2010-08-04 | 浙江中控软件技术有限公司 | Gas system balancing and optimal scheduling method, gas balancing and optimal scheduling device and gas balancing and optimal scheduling system |
CN101794119B (en) * | 2010-03-08 | 2012-05-23 | 浙江中控软件技术有限公司 | Gas system balancing and optimal scheduling method, gas balancing and optimal scheduling device and gas balancing and optimal scheduling system |
CN102937670A (en) * | 2012-11-17 | 2013-02-20 | 河南科技大学 | Method for identifying resistance parameters of rotors of induction motor on basis of Elman neural network |
CN102937670B (en) * | 2012-11-17 | 2015-05-27 | 河南科技大学 | Method for identifying resistance parameters of rotors of induction motor on basis of Elman neural network |
CN103646168A (en) * | 2013-11-26 | 2014-03-19 | 浙江工商大学 | Tiny-oil ignition early warning method |
CN105241239A (en) * | 2015-09-10 | 2016-01-13 | 广西大学 | Intelligent optimal control method and device for sintered brick tunnel kiln roasting process |
CN105241239B (en) * | 2015-09-10 | 2017-07-21 | 广西大学 | A kind of baked brick tunnel kiln roasting process intelligent optimized control method and device |
CN107341444A (en) * | 2017-06-07 | 2017-11-10 | 北京星网锐捷网络技术有限公司 | Data exception method for early warning and device |
CN108681656A (en) * | 2018-04-27 | 2018-10-19 | 上海卓然工程技术股份有限公司 | A kind of process analysis method based on ethane cracking furnace operation data |
CN108681656B (en) * | 2018-04-27 | 2023-09-01 | 上海卓然工程技术股份有限公司 | Process analysis method based on ethylene cracking furnace operation data |
CN109389170A (en) * | 2018-10-10 | 2019-02-26 | 常州大学 | A kind of gradation type operating condition method for early warning based on 3D convolutional neural networks |
CN109389170B (en) * | 2018-10-10 | 2021-11-02 | 常州大学 | Gradual change type working condition early warning method based on 3D convolutional neural network |
CN111394117A (en) * | 2020-03-23 | 2020-07-10 | 北京赛普泰克技术有限公司 | Treatment method and treatment system for gasoline adsorption desulfurization, electronic equipment and storage medium |
CN113219912A (en) * | 2021-03-31 | 2021-08-06 | 成都飞机工业(集团)有限责任公司 | Multi-Agent-based numerical control machining flexible manufacturing system machining process early warning method |
CN113219912B (en) * | 2021-03-31 | 2022-03-15 | 成都飞机工业(集团)有限责任公司 | Multi-Agent-based numerical control machining flexible manufacturing system machining process early warning method |
CN115386391A (en) * | 2022-08-01 | 2022-11-25 | 浙江中控技术股份有限公司 | Cracking furnace feeding flow monitoring and fault handling method and system |
CN115386391B (en) * | 2022-08-01 | 2023-10-13 | 浙江中控技术股份有限公司 | Method and system for monitoring feeding flow of cracking furnace and processing faults |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN101551663A (en) | Cracking reaction pre-warning method for cracking furnace | |
CN106872657B (en) | A kind of multivariable water quality parameter time series data accident detection method | |
CN103969066B (en) | The system and method for monitoring the process of operation | |
CN110045594B (en) | Intelligent management and control system and method for predicting state risk of four tubes of boiler | |
CN104573850A (en) | Method for evaluating state of thermal power plant equipment | |
CN106845796B (en) | One kind is hydrocracked flow product quality on-line prediction method | |
WO2021121182A1 (en) | Health diagnosis method for numerically controlled machine tool | |
CN103163389A (en) | Intelligent bus on-line monitoring and condition evaluation system | |
CN105758661B (en) | Boiler heating surface service life evaluation system and method | |
JP2002527581A (en) | On-line control method for polypropylene plant | |
CN111090939B (en) | Early warning method and system for abnormal working condition of petrochemical device | |
CN113762329A (en) | Method and system for constructing state prediction model of large rolling mill | |
CN116991130A (en) | Intelligent automatic control system and method for petrochemical production | |
CN108956041B (en) | Sintering machine air leakage fault online diagnosis method based on single-side detection method of frequency spectrum characteristics | |
CN111191855B (en) | Water quality abnormal event identification and early warning method based on pipe network multi-element water quality time sequence data | |
CN111783354A (en) | Flow-induced corrosion characteristic prediction and service life evaluation method of RBF neural network model | |
CN112990546A (en) | Chemical plant power transformer fault prediction method based on particle swarm and neural network | |
Li et al. | A two-layer fuzzy synthetic strategy for operational performance assessment of an industrial hydrocracking process | |
Jiang et al. | Design of early warning model based on time series data for production safety | |
CN117113166A (en) | Industrial boiler fault detection method based on improved integrated learning | |
CN107664690A (en) | A kind of method for predicting gas dissolved in oil of power trans-formers | |
CN102759602B (en) | Method for predicting faults in cascade polymerization reaction process of high density polyethylene (HDPE) | |
CN110532699B (en) | Fuzzy DCD-based fault diagnosis method for hydrometallurgy dense washing process | |
Shojaee et al. | Applying SVSSI sampling scheme to np-chart to decrease the time of detecting shifts using Markov chain approach and Monte Carlo simulation | |
KR20120117045A (en) | Method for real-time evaluating process in drinking water treatment facility using unit-process analysis model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C12 | Rejection of a patent application after its publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20091007 |