CN103838954A - Soft measurement system and method for relevant vector polypropylene melt index - Google Patents

Abstract

The invention discloses a soft measurement system and method for the relevant vector polypropylene melt index. The system comprises a standardized module, a model training module, a soft measurement module and a display module. The method includes the steps that key variables are collected from a database to serve as a training sample when the system is normal, a soft measurement model is established, and a corresponding melt index prediction value is obtained through a new sample subjected to standardized treatment. The soft measurement model of the melt index is established with the system and method, the melt index value can be predicted on line, and the calculation speed and the calculation precision are high.

Description

A kind of soft measuring system of associated vector polypropylene melt index and method

Technical field

The present invention relates to soft fields of measurement, especially, relate to the soft measuring system of a kind of polypropylene melt index and method.

Background technology

Polypropylene is a kind of thermoplastic resin being made by propylene polymerization, the most important downstream product of propylene, and 50% of World Propylene, 65% of China's propylene is all for polypropylene processed, is one of five large general-purpose plastics, closely related with our daily life.Melting index is that polypropylene product is determined one of important quality index of product grade, it has determined the different purposes of product, be an important step of production quality control during polypropylene is produced to the measurement of melting index, to producing and scientific research, have very important effect and directive significance.

But; the on-line analysis of melting index is measured and is difficult at present accomplish; being the shortage of online melting index analyser on the one hand, is that existing in-line analyzer is measured the inaccurate difficulty in caused use that even cannot normally use owing to often can stopping up on the other hand.Therefore, the measurement of MI in commercial production at present, mainly obtains by hand sampling, off-line assay, and can only analyze once for general every 2-4 hour, time lag is large, and the quality control of producing to propylene polymerization has brought difficulty, becomes a bottleneck problem being badly in need of solution in production.The online soft sensor instrument of polypropylene melt index and method research, thus forward position and the focus of academia and industry member become.

Summary of the invention

In order to overcome, the measuring accuracy of existing propylene polymerization production process is not high, the deficiency of the impact that is subject to human factor, the invention provides a kind of system and method that can effectively realize the soft measurement of polypropylene melt index.

The technical solution adopted for the present invention to solve the technical problems is:

The soft measuring system of a kind of associated vector polypropylene melt index, comprises the field intelligent instrument, database, standardized module, model training module, soft measurement module, the display module that are connected with propylene polymerization production process, wherein:

Standardized module, the key variables when normal from database acquisition system are as training sample, to training sample

carry out standardization, obtain input matrix X, adopt following process to complete:

$\stackrel{\‾}{X}=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}\tilde{X}---\left(1\right)$

${\mathrm{\σ}}^2=\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}(\tilde{X}-\stackrel{\‾}{X})---\left(2\right)$

$X=\frac{\tilde{X}-\stackrel{\‾}{X}}{\mathrm{\σ}}---\left(3\right)$

Wherein, X is the training sample after standardization,

for training sample, corresponding polypropylene melt index data are Y, and N is number of training,

for the average of training sample, the standard deviation that σ is training sample;

Model training module, for setting up soft-sensing model, its detailed process is as follows

$\underset{w,\mathrm{\β}}{\max }\left\{{\left(2\mathrm{\π}{\mathrm{\β}}^2\right)}^{-N/2}\exp (-\frac{1}{2{\mathrm{\β}}^2}{\left|\right|Y-\mathrm{\Φw}\left|\right|}^2)\right\}---\left(4\right)$

Solve above formula and can obtain soft-sensing model:

Wherein, w is regression coefficient,

(X) be kernel function row, Φ is nuclear matrix, and noise ε obeys that average is zero, variance is β ²gaussian distribution, subscript T representing matrix transposition;

Soft measurement module, for realizing the soft measurement of polypropylene melt index, by the new samples through standardization substitution formula (5), obtains corresponding melting index predicted value

Display module, for the demonstration of soft measurement result;

As preferred a kind of scheme: described key variables comprise three strands of propylene feed flows, major catalyst flow, cocatalyst flow, temperature in the kettle, pressure, liquid level, hydrogen volume concentration.

A kind of associated vector polypropylene melt index flexible measurement method, specific implementation step is as follows:

1) key variables when acquisition system is normal from database are as training sample, to training sample

carry out standardization, obtain input matrix X, adopt following process to complete:

$\stackrel{\‾}{X}=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}\tilde{X}---\left(1\right)$

${\mathrm{\σ}}^2=\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}(\tilde{X}-\stackrel{\‾}{X})---\left(2\right)$

$X=\frac{\tilde{X}-\stackrel{\‾}{X}}{\mathrm{\σ}}---\left(3\right)$

Wherein, X is the training sample after standardization,

for training sample, corresponding polypropylene melt index data are Y, and N is number of training, for the average of training sample, the standard deviation that σ is training sample;

2) set up soft-sensing model, its detailed process is as follows

$\underset{w,\mathrm{\β}}{\max }\left\{{\left(2\mathrm{\π}{\mathrm{\β}}^2\right)}^{-N/2}\exp (-\frac{1}{2{\mathrm{\β}}^2}{\left|\right|Y-\mathrm{\Φw}\left|\right|}^2)\right\}---\left(4\right)$

Solve above formula and can obtain soft-sensing model:

Wherein, w is regression coefficient,

(X) be kernel function row, Φ is nuclear matrix, and noise ε obeys that average is zero, variance is β ²gaussian distribution, subscript T representing matrix transposition;

3) by the new samples through standardization

substitution formula (5), obtains corresponding melting index predicted value

4) show soft measurement result.

As preferred a kind of scheme: the key variables of described method comprise three strands of propylene feed flows, major catalyst flow, cocatalyst flow, temperature in the kettle, pressure, liquid level, hydrogen volume concentration.

Technical conceive of the present invention is: the important quality index melting index to propylene polymerization production process is carried out online soft sensor, overcomes that existing polypropylene melting index measurement instrument measuring accuracy is not high, the deficiency of the impact that is subject to human factor.

Beneficial effect of the present invention is mainly manifested in: 1, set up the soft-sensing model of melting index, and can on-line prediction melt index values; 2, computing velocity is fast, and accuracy is good.

Accompanying drawing explanation

Fig. 1 is the structural drawing of system proposed by the invention;

Fig. 2 is the hardware structure diagram of embodiment in DCS system.

Embodiment

Below in conjunction with accompanying drawing, the invention will be further described.The embodiment of the present invention is used for the present invention that explains, rather than limits the invention, and in the protection domain of spirit of the present invention and claim, any modification and change that the present invention is made, all fall into protection scope of the present invention.

Embodiment 1

In the time that propylene polymerization production process has been furnished with DCS system, the real-time and historical data base of the detection of sample real-time dynamic data, memory by using DCS system, soft measurement function mainly completes on host computer.

With reference to Fig. 1, Fig. 2, the soft measuring system of a kind of associated vector polypropylene melt index, comprises field intelligent instrument 2, database 3 and host computer that propylene polymerization production process 1 connects, and described host computer comprises:

Standardized module 4, key variables when normal from database acquisition system are as training sample, described key variables comprise three strands of propylene feed flows, major catalyst flow, cocatalyst flow, temperature in the kettle, pressure, liquid level, and hydrogen volume concentration, to training sample

carry out standardization, obtain input matrix X, adopt following process to complete:

$\stackrel{\‾}{X}=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}\tilde{X}---\left(1\right)$

${\mathrm{\σ}}^2=\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}(\tilde{X}-\stackrel{\‾}{X})---\left(2\right)$

$X=\frac{\tilde{X}-\stackrel{\‾}{X}}{\mathrm{\σ}}---\left(3\right)$

Wherein, X is the training sample after standardization,

for training sample, corresponding polypropylene melt index data are Y, and N is number of training,

for the average of training sample, the standard deviation that σ is training sample;

Model training module 5, for setting up soft-sensing model, its detailed process is as follows:

$\underset{w,\mathrm{\β}}{\max }\left\{{\left(2\mathrm{\π}{\mathrm{\β}}^2\right)}^{-N/2}\exp (-\frac{1}{2{\mathrm{\β}}^2}{\left|\right|Y-\mathrm{\Φw}\left|\right|}^2)\right\}---\left(4\right)$

Solve above formula and can obtain soft-sensing model:

Wherein, w is regression coefficient,

(X) be kernel function row, Φ is nuclear matrix, and noise ε obeys that average is zero, variance is β ²gaussian distribution, subscript T representing matrix transposition;

Soft measurement module 6, for realizing the soft measurement of polypropylene melt index, by the new samples through standardization

substitution formula (5), obtains corresponding melting index predicted value

Display module 7, for the demonstration of soft measurement result.

In the time that production run is not equipped with DCS system, adopted data memory substitutes the data storage function of the real-time and historical data base of DCS system, and soft measuring system is manufactured and comprises I/O element, data-carrier store, program storage, arithmetical unit, several large members of display module one of the DCS system that do not rely on independently complete SOC (system on a chip), in the situation that whether being equipped with DCS regardless of production run, can both independently use, more be of value to and promoting the use of.

Embodiment 2

With reference to Fig. 1, Fig. 2, a kind of associated vector polypropylene melt index flexible measurement method, is characterized in that: described method specific implementation step is as follows:

1) key variables when acquisition system is normal from database 3 by module 4 are as training sample, described key variables comprise three strands of propylene feed flows, major catalyst flow, cocatalyst flow, temperature in the kettle, pressure, liquid level, hydrogen volume concentration, to training sample

carry out standardization, obtain input matrix X, adopt following process to complete:

$\stackrel{\‾}{X}=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}\tilde{X}---\left(1\right)$

${\mathrm{\σ}}^2=\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}(\tilde{X}-\stackrel{\‾}{X})---\left(2\right)$

$X=\frac{\tilde{X}-\stackrel{\‾}{X}}{\mathrm{\σ}}---\left(3\right)$

Wherein, X is the training sample after standardization, for training sample, corresponding polypropylene melt index data are Y, and N is number of training,

for the average of training sample, the standard deviation that σ is training sample;

2) module 5 is set up soft-sensing model, and its detailed process is as follows:

$\underset{w,\mathrm{\β}}{\max }\left\{{\left(2\mathrm{\π}{\mathrm{\β}}^2\right)}^{-N/2}\exp (-\frac{1}{2{\mathrm{\β}}^2}{\left|\right|Y-\mathrm{\Φw}\left|\right|}^2)\right\}---\left(4\right)$

Solve above formula and can obtain soft-sensing model:

Wherein, w is regression coefficient,

(X) be kernel function row, Φ is nuclear matrix, and noise ε obeys that average is zero, variance is β ²gaussian distribution, subscript T representing matrix transposition;

3) module 6 is by the new samples through standardization

substitution formula (5), obtains corresponding melting index predicted value 4) result is passed to DCS system by module 7, shows, and be delivered to operator station by DCS system and fieldbus and show at the control station of DCS.

Claims (2)

1. the soft measuring system of associated vector polypropylene melt index, comprises standardized module, model training module, soft measurement module, display module, it is characterized in that:

Standardized module, key variables when normal from database acquisition system are as training sample, described key variables comprise three strands of propylene feed flows, major catalyst flow, cocatalyst flow, temperature in the kettle, pressure, liquid level, and hydrogen volume concentration, to training sample

carry out standardization, obtain input matrix X, adopt following process to complete:

$\stackrel{\‾}{X}=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}\tilde{X}---\left(1\right)$

${\mathrm{\σ}}^2=\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}(\tilde{X}-\stackrel{\‾}{X})---\left(2\right)$

$X=\frac{\tilde{X}-\stackrel{\‾}{X}}{\mathrm{\σ}}---\left(3\right)$

Wherein, X is the training sample after standardization, for training sample, corresponding polypropylene melt index data are Y, and N is number of training,

for the average of training sample, the standard deviation that σ is training sample;

Model training module, for setting up soft-sensing model, its detailed process is as follows

$\underset{w,\mathrm{\β}}{\max }\left\{{\left(2\mathrm{\π}{\mathrm{\β}}^2\right)}^{-N/2}\exp (-\frac{1}{2{\mathrm{\β}}^2}{\left|\right|Y-\mathrm{\Φw}\left|\right|}^2)\right\}---\left(4\right)$

Solve above formula and can obtain soft-sensing model:

Wherein, w is regression coefficient,

(X) be kernel function row, Φ is nuclear matrix, and noise ε obeys that average is zero, variance is β ²gaussian distribution, subscript T representing matrix transposition;

Soft measurement module, for realizing the soft measurement of polypropylene melt index, by the new samples through standardization

substitution formula (5), obtains corresponding melting index predicted value

Display module, for the demonstration of soft measurement result.

2. an associated vector polypropylene melt index flexible measurement method, is characterized in that: described method specific implementation step is as follows:

1) key variables when acquisition system is normal from database are as training sample, and described key variables comprise three strands of propylene feed flows, major catalyst flow, cocatalyst flow, temperature in the kettle, pressure, liquid level, and hydrogen volume concentration, to training sample

carry out standardization, obtain input matrix X, adopt following process to complete:

$\stackrel{\‾}{X}=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}\tilde{X}---\left(1\right)$

${\mathrm{\σ}}^2=\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}(\tilde{X}-\stackrel{\‾}{X})---\left(2\right)$

$X=\frac{\tilde{X}-\stackrel{\‾}{X}}{\mathrm{\σ}}---\left(3\right)$

Wherein, X is the training sample after standardization,

for training sample, corresponding polypropylene melt index data are Y, and N is number of training,

for the average of training sample, the standard deviation that σ is training sample;

2) set up soft-sensing model, its detailed process is as follows

$\underset{w,\mathrm{\β}}{\max }\left\{{\left(2\mathrm{\π}{\mathrm{\β}}^2\right)}^{-N/2}\exp (-\frac{1}{2{\mathrm{\β}}^2}{\left|\right|Y-\mathrm{\Φw}\left|\right|}^2)\right\}---\left(4\right)$

Solve above formula and can obtain soft-sensing model:

Wherein, w is regression coefficient,

(X) be kernel function row, Φ is nuclear matrix, and noise ε obeys that average is zero, variance is β ²gaussian distribution, subscript T representing matrix transposition;

3) by the new samples through standardization

substitution formula (5), obtains corresponding melting index predicted value

4) show soft measurement result.

Images