CN106682312A - Industrial process soft-measurement modeling method of local weighing extreme learning machine model - Google Patents
Industrial process soft-measurement modeling method of local weighing extreme learning machine model Download PDFInfo
- Publication number
- CN106682312A CN106682312A CN201611234617.5A CN201611234617A CN106682312A CN 106682312 A CN106682312 A CN 106682312A CN 201611234617 A CN201611234617 A CN 201611234617A CN 106682312 A CN106682312 A CN 106682312A
- Authority
- CN
- China
- Prior art keywords
- sample
- learning machine
- extreme learning
- modeling
- local
- 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
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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Abstract
The invention discloses an industrial process soft-measurement modeling method of a local weighing extreme learning machine model. The method based on a local weighing extreme learning machine has good precision and high operating speed for the nonlinear process, but the data volume of the industrial process often produces the problem of poor generalization performance during application of the extreme learning machine. The applied improved method adopts an instant learning thought, and local weighing is conducted on samples in spatial dimension according to the dynamic characteristics of the industrial process. The industrial process soft-measurement modeling method combines a local weighing method adopting the instant learning thought and the advantages of the weighing extreme learning machine, overcomes respective main disadvantages and is high in speed and precision.
Description
Technical field
The invention belongs to industrial process prediction and control field, more particularly to a kind of soft survey of local weighted extreme learning machine
Amount modeling method.
Background technology
In traditional industrial process, exist and many for improve product quality and ensure that safety has and most important make used
Parameter such as reaction rate, product composition content etc., but be much often difficult to or directly cannot be measured with sensor.
Using needing the on-line analyses instrument of great amount of investment to be detected, often have larger delayed and cause to adjust not prompt enough, from
And make product quality be difficult to be guaranteed.We are called leading variable for industrial process has the variable of important function,
It is other some with the variable of measurement we term it auxiliary variable.Hard measurement essence thing by set up industrial process variable it
Between mathematical model, realize by auxiliary variable prediction leading variable technical method.In recent years, hard measurement is in industrial process
In apply many neural network algorithms, but due to neural network algorithm, to still suffer from convergence rate slower, is easily trapped into local
The shortcomings of optimal solution and poor Generalization Capability.
For the problems referred to above for overcoming neural network algorithm to exist, extreme learning machine is used as a kind of single hidden layer stochastic neural net
Network, the convergence rate that the inverse iteration of traditional neural network algorithm can be overcome to produce is slow, is absorbed in locally optimal solution problem.But due to
The randomness and hidden node number of its model sacrifices Generalization Capability commonly greater than other neutral nets.Ask to solve this
Topic, it is proposed that local weighted extreme learning machine model, using the thought of instant learning, by local weighted method, improves
Extreme learning machine model is applied to the poor shortcoming of industrial process extreme learning machine generalization ability, pin as a kind of nonlinear method
Problem to industrial process data dynamic, improves the performance that model is set up and predicted.
The content of the invention
Present invention aims to the deficiencies in the prior art, propose a kind of hard measurement of local weighted extreme learning machine
Modeling method.
The purpose of the present invention is achieved through the following technical solutions：It is a kind of based on the soft of local weighted extreme learning machine
The foundation of measurement model, mainly including following step.
(1) using system under collecting and distributing control and off-line monitoring method, the data of industrial processes are chronological
Training sample set Xtr∈RN×nAnd Ytrain∈RN×mWith test sample collection Xte∈RK×nAnd Ytest∈RN×m.Wherein N is training sample
Length, n is training sample dimension, and K is test sample collection length.Test sample collection is that these data are stored in into historical data base.
And pre-treatment carried out according to training sample set to training sample set and test sample and normalization makes training sample set its average
For 0 and variance be 1.
(2) by test sample XtestRow is taken successively as query sample qi(i=1,2 ..., K), and correspondence test specimens reality
Real-valued yi(i=1,2 ..., K).Afterwards local weighted extreme learning machine modeling is carried out respectively by each query sample.
(3) query sample average is carried out to query sample and training set.This part goes the value of average modeling and to obtain
To after modeling result again on final result reduction go inquire about average data deviation.
(4) weight parameter w is determined apart from d according to query sample and historical sample (training set sample), to historical sample
Weighting, obtains new training sample XwThe sample space to local weightedization.
(5) extreme learning machine modeling is carried out to the sample space of local weightedization, obtains hard measurement result.
(6) repeat to have obtained whole test sample after the vectorial modeling of all of query sample and hard measurement result
Hard measurement result.
(7) using derived above the data of industrial process are modeled based on local weighted extreme learning machine method,
Realize the hard measurement of process.
The invention has the beneficial effects as follows：The present invention adopts instant learning thought, and using local weighted method localized mode is set up
Type, improves extreme learning machine as a kind of high speed high precision nonlinear method in industrial processes because sample space has limited production
The not enough problem of raw Generalization Capability.Extreme learning machine possesses good modeling speed and regression accuracy, but often in industrial mistake
In journey to sample space deficiency in the case of Generalization Capability it is very sensitive, partial model is set up by local weighted method, can be with
Effectively strengthen the Generalization Capability of algorithm, and local weighted method also can to a certain extent solve the limit in cohesive process
Learning machine does not have a problem of noise reduction process, and can reduce and unstability of the extreme learning machine in real process.Obtain
The hard measurement device of high-precision industrial process nonlinear feature at high speed.
Description of the drawings
Fig. 1 is extreme learning machine debutanizing tower hard measurement Error Graph；
Fig. 2 is local weighted extreme learning machine debutanizing tower hard measurement Error Graph.
Specific embodiment
The present invention be directed to industrial process nonlinear problem, using instant learning thought by local weighted side spatially
Method simplifies the problem of non-linear property, solves that process is non-linear to ask as a kind of nonlinear algorithm in conjunction with extreme learning machine algorithm
Topic.Local weighted method can with the generalization ability of lift scheme, and can reduce to a certain extent noise should around with reduce pole
Limit learning machine unstability.This method can either realize extraction and modeling of the local weighted method to industrial process dynamic,
Extreme learning machine can be embodied as a kind of nonlinear regression algo high precision, fireballing advantage.
Below in conjunction with the accompanying drawings the present invention is described in detail with instantiation.
The technical solution used in the present invention key step is as follows：
The first step is using system and off-line monitoring method under collecting and distributing control, the data composition modeling of industrial processes
Training sample set and test sample collection.Wherein, training sample set includes Xtr∈RN×nAnd Ytrain∈RN×m, wherein, N is instruction
Practice sample length, n is training sample dimension.Test sample collection is Xte∈RK×nAnd Ytest∈RN×mThese data are stored in into history
Data base, wherein, K is test sample length, and n is training sample dimension.And to training sample set and test sample according to instruction
Practice that sample set carries out pre-treatment and normalization makes that training sample set its average is 0 and variance is 1.
Second step is by test sample XtestRow is taken successively as query sample, as (q1,q2,…qi…qK), wherein qiFor
Query sample, n dimension row vectors.Test sample actual value y is corresponded to respectivelyi(i=1,2 ..., K).From the first row (i=1) to finally
A line carries out respectively local weighted extreme learning machine modeling.
3rd step is q when query sampleiWhen (i=1,2 ..., K), query sample is carried out to query sample and training set
Average.This part go average value will model and after obtaining modeling result again on final result the number for inquiring about average is removed in reduction
According to deviation.
Training sample is concentrated X by the 4th steptrainCapable amount form is decomposed into, willIn each xjAs history sample
This (j=1,2 ..., N), you can to each historical sample and query sample qiSimilarity or the distance of point measuredThen weight can be formulated to similar sampleWhereinFor the local of setting
Weighting parameters.Obtain all weight W=[w1, w2,...wN].The input value of input neutral net is replaced byIt
Afterwards operating limit learning machine is modeled.
5th step extreme learning machine is a kind of single hidden layer probabilistic neural metanetwork, and C hidden neuron node g is generated at random
All parameter (a of (ai, bi)1,b1)(a2,b2)…(aC,bC) it is generated after, according to the output formula of monolayer neural networksThis formula also can be write as：
H β=T, wherein β are that neuron exports weight matrix, and T is neutral net output result,
Error | T-Y | is caused to pursue2Minimum, is obtained weight matrix computing formula β=H+T.The neutral net for obtaining
Weight beta random node parameter a for generating plus beforei, biIt is exactly the neural network parameter of whole extreme learning machine.
6th step repeats above step, and the predictive value of the local weighted limit study of query sample is obtained every time, repeats
Carry out having obtained the hard measurement result of whole test sample after the vectorial modeling of all of query sample and hard measurement result.
7th step is built based on local weighted extreme learning machine method using derived above to the data of industrial process
Mould, realizes the hard measurement of process.
Effectiveness of the invention is illustrated below in conjunction with a specific debutanizing tower example.For the process, adopt altogether
2394 data are collected, wherein 1197 data are used for modeling, and off-line analysiss and mark have been carried out to its corresponding butane content value
Note.In addition 1197 data samples of collection are used for verifying the effectiveness of soft-sensing model.In this process, 7 are have chosen altogether
Individual process variable carries out soft sensor modeling to the butane content of the process, and this 7 process variables are respectively tower top temperature, tower top pressure
Power, return flow, next stage flow, sensitive plate temperature, column bottom temperature and tower bottom pressure.Next the detailed process is combined to this
The implementation steps of invention are set forth in：
1. data pre-processing, to the process variable and butane content in 2394 modeling samples pretreatment and normalizing are carried out
Change so that the average of each process variable and key variables is zero, and variance is 1, obtains new modeling data matrix.
2. the soft sensor modeling of local weighted extreme learning machine is based on
Using choose procedure variable composition data matrix as soft-sensing model input, butane content data square
Battle array, according to the method detailed provided in implementation steps, sets up related as the output of soft-sensing model to each test sample point
Local weighted extreme learning machine soft-sensing model.
3. the data of pair whole sample sets are modeled and predict
In order to test the effectiveness of new method, we carry out locally fine point to 2394 test samples and hard measurement is pre-
Survey, obtain comparing and studying after hard measurement result.
4. the soft sensor modeling prediction of butane content is compared
Online soft sensor is carried out to 2394 checking samples, corresponding On-line Estimation value is obtained.Limits of application learning machine side
For the online soft sensor of 2394 checking samples, mean error is 0.130686 to method.Using the local weighted limit of the present invention
The hard measurement that learning machine soft-sensing model is carried out to same sample, mean error is 0.0890488.It can be seen that local weighted pole
Limit learning machine reduces the error of prediction, improves the precision of soft-side face model.
Above-described embodiment limits the invention again illustrating the present invention, the present invention spirit and
In scope of the claims, to initial people of the invention and modifications and changes, protection scope of the present invention is both fallen within.
Claims (3)
1. a kind of industrial process soft-measuring modeling method based on local weighted extreme learning machine, it is characterised in that including following
Step:
(1) using system under collecting and distributing control and off-line monitoring method, the chronological training of the data of industrial processes
Sample set Xtr∈RN×nAnd Ytrain∈RN×mWith test sample collection Xte∈RK×nAnd Ytest∈RN×m.Wherein N is training sample length,
N is training sample dimension, and K is test sample collection length.Test sample collection is that these data are stored in into historical data base.And it is right
Training sample set and test sample carry out pre-treatment and normalization according to training sample set makes its average of training sample set for 0 and side
Difference is 1.
(2) by test sample XtestRow is taken successively as query sample qi(i=1,2 ..., K), and correspondence test sample actual value
yi(i=1,2 ..., K).Afterwards local weighted extreme learning machine modeling is carried out respectively by each query sample.
(3) query sample average is carried out to query sample and training set.This part goes the value of average modeling and to be built
After mould result again on final result reduction go inquire about average data deviation.
(4) determine weight parameter w apart from d according to query sample and historical sample (training set sample), historical sample weighted,
Obtain new training sample XwThe sample space to local weightedization.
(5) extreme learning machine modeling is carried out to the sample space of local weightedization, obtains hard measurement result.
(6) the soft survey that whole test sample has been obtained after the vectorial modeling of all of query sample and hard measurement result is repeated
Amount result.
(7) using derived above the data of industrial process are modeled based on local weighted extreme learning machine method, are realized
The hard measurement of process.
2. a kind of industrial process soft sensor modeling of local weighted extreme learning machine model according to right 1, it is characterised in that
In the step 2, in query sample q that the test set of model data is divided into test set pointsi(i=1,2 ..., K), each
Individually modeling is predicted query sample during the foundation of ensuing model.
3. a kind of industrial process soft sensor modeling of local weighted extreme learning machine model according to right 1, it is characterised in that
In the step 5, according to query sample qi (i=1,2 ..., K) and the similarity of historical sample dj (j=1,2 ..., N)The distance between i.e. determine weightWhereinFor the local weighted ginseng of setting
Number.By local weighted dataCarry out as the local weighted model training sample input of extreme learning machine local weighted
Modeling.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611234617.5A CN106682312A (en) | 2016-12-28 | 2016-12-28 | Industrial process soft-measurement modeling method of local weighing extreme learning machine model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611234617.5A CN106682312A (en) | 2016-12-28 | 2016-12-28 | Industrial process soft-measurement modeling method of local weighing extreme learning machine model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106682312A true CN106682312A (en) | 2017-05-17 |
Family
ID=58872251
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611234617.5A Pending CN106682312A (en) | 2016-12-28 | 2016-12-28 | Industrial process soft-measurement modeling method of local weighing extreme learning machine model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106682312A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107463994A (en) * | 2017-07-07 | 2017-12-12 | 浙江大学 | Semi-supervised flexible measurement method based on coorinated training extreme learning machine model |
| CN109840362A (en) * | 2019-01-16 | 2019-06-04 | 昆明理工大学 | A kind of integrated instant learning industrial process soft-measuring modeling method based on multiple-objection optimization |
| CN111291020A (en) * | 2019-11-11 | 2020-06-16 | 中国计量大学 | Dynamic process soft measurement modeling method based on local weighted linear dynamic system |
| CN115310561A (en) * | 2022-09-29 | 2022-11-08 | 中国空气动力研究与发展中心设备设计与测试技术研究所 | Electromagnetic valve fault monitoring method based on integrated instant learning |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2003330977A (en) * | 2002-05-17 | 2003-11-21 | Nec Corp | Electromagnetic field analyzing device and electromagnetic field analyzing method |
| CN105868164A (en) * | 2016-03-19 | 2016-08-17 | 浙江大学 | Soft measurement modeling method based on monitored linear dynamic system model |
-
2016
- 2016-12-28 CN CN201611234617.5A patent/CN106682312A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2003330977A (en) * | 2002-05-17 | 2003-11-21 | Nec Corp | Electromagnetic field analyzing device and electromagnetic field analyzing method |
| CN105868164A (en) * | 2016-03-19 | 2016-08-17 | 浙江大学 | Soft measurement modeling method based on monitored linear dynamic system model |
Non-Patent Citations (4)
| Title |
|---|
| ZAMPROGNA E ETAL.: "Optimal selection of soft sensor inputs for batch distillation columns using principal component analysis", 《JOURNAL OF PROCESS CONTROL》 * |
| 孙茂伟 等: "局部加权混合核偏最小二乘算法及其在软测量中的应用", 《信息与控制》 * |
| 袁小锋 等: "基于时间差分和局部加权偏最小二乘算法的过程自适应软测量建模", 《化工学报》 * |
| 郭斐斐 等: "基于距离度量的局部加权朴素贝叶斯算法改进", 《电脑开发与应用》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107463994A (en) * | 2017-07-07 | 2017-12-12 | 浙江大学 | Semi-supervised flexible measurement method based on coorinated training extreme learning machine model |
| CN109840362A (en) * | 2019-01-16 | 2019-06-04 | 昆明理工大学 | A kind of integrated instant learning industrial process soft-measuring modeling method based on multiple-objection optimization |
| CN109840362B (en) * | 2019-01-16 | 2022-06-14 | 昆明理工大学 | Multi-objective optimization-based integrated just-in-time learning industrial process soft measurement modeling method |
| CN111291020A (en) * | 2019-11-11 | 2020-06-16 | 中国计量大学 | Dynamic process soft measurement modeling method based on local weighted linear dynamic system |
| CN115310561A (en) * | 2022-09-29 | 2022-11-08 | 中国空气动力研究与发展中心设备设计与测试技术研究所 | Electromagnetic valve fault monitoring method based on integrated instant learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109659933B (en) | Electric energy quality prediction method for power distribution network with distributed power supply based on deep learning model | |
| CN104537415B (en) | A kind of non-linear process industry failure predication and recognition methodss based on compressed sensing and DROS ELM | |
| CN108897286B (en) | Fault detection method based on distributed nonlinear dynamic relation model | |
| CN108197648A (en) | A kind of Fault Diagnosis Method of Hydro-generating Unit and system based on LSTM deep learning models | |
| CN103927412B (en) | Instant learning debutanizing tower soft-measuring modeling method based on gauss hybrid models | |
| CN111191191B (en) | Construction method of combined model for accurately predicting deformation effect of concrete dam | |
| CN106682312A (en) | Industrial process soft-measurement modeling method of local weighing extreme learning machine model | |
| CN109635245A (en) | A kind of robust width learning system | |
| CN112101480A (en) | Multivariate clustering and fused time sequence combined prediction method | |
| Tian et al. | Short-term wind power prediction based on empirical mode decomposition and improved extreme learning machine | |
| CN110928187B (en) | Sewage treatment process fault monitoring method based on fuzzy width self-adaptive learning model | |
| CN106127300A (en) | A kind of rotating machinery health status Forecasting Methodology | |
| CN113379164B (en) | Load prediction method and system based on deep self-attention network | |
| Zhang et al. | Short-term load forecasting using recurrent neural networks with input attention mechanism and hidden connection mechanism | |
| CN107993012A (en) | A kind of adaptive electric system on-line transient stability appraisal procedure of time | |
| CN104050547A (en) | Non-linear optimization decision-making method of planning schemes for oilfield development | |
| CN107966600A (en) | A kind of electricity anti-theft system and its electricity anti-theft method based on deep learning algorithm | |
| CN103544527A (en) | Method of predicting water level of oxygen extractor in ultra-supercritical steam turbine under FCB conditions | |
| CN106709570A (en) | Time dimension expansion and local weighting extreme learning machine-based soft measurement modeling method | |
| Ji et al. | Application of the improved the ELM algorithm for prediction of blast furnace gas utilization rate | |
| CN103279030A (en) | Bayesian framework-based dynamic soft measurement modeling method and device | |
| CN104182854A (en) | Mixed energy consumption measuring method for enterprise energy management system | |
| CN106773697A (en) | A kind of time dimension expands the industrial process soft-measuring modeling method of extreme learning machine model | |
| CN109408896B (en) | Multi-element intelligent real-time monitoring method for anaerobic sewage treatment gas production | |
| CN116629126A (en) | Soft measurement modeling method based on dynamic multi-head attention mechanism |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170517 |