CN101315557A - Propylene polymerization production process optimal soft survey instrument and method based on genetic algorithm optimization BP neural network - Google Patents

Abstract

A propylene polymerization production process optimal soft-measurement meter based on genetic algorithm optimized BP neural network comprises a propylene polymerization production process, a site intelligent meter, a control station, a DCS databank used for storing data, an optimal soft measurement model based on genetic algorithm optimized BP neural network, and a melting index soft-measurement value indicator. The site intelligent meter and the control station are connected with the propylene polymerization production process and the DCS databank; the optimal soft-measurement model is connected with the DCS databank and the soft-measurement value indicator. The optimal soft measurement model based on genetic algorithm optimized BP neural network comprises a data pre-processing module, an ICA dependent-component analysis module, a BP neural network modeling module and a genetic algorithm optimized BP neural network module. The invention also provides a soft measurement method adopting the soft measurement meter. The invention can realize on-line measurement and on-line automatic parameter optimization, with quick calculation, automatic model updating, strong anti-interference capability and high accuracy.

Description

Propylene polymerization production process optimal soft survey instrument and method based on genetic algorithm optimization BP neural network

Technical field

The present invention relates to optimal soft survey instrument and method, specifically is a kind of propylene polymerization production process optimal soft survey instrument and method based on genetic algorithm optimization BP neural network.

Background technology

Polypropylene is a kind of thermoplastic resin that is made by propylene polymerization, the most important downstream product of propylene, and 50% of world's propylene, 65% of China's propylene all is to be used for making polypropylene, is one of five big general-purpose plastics, and is closely related with our daily life.Polypropylene is fastest-rising in the world interchangeable heat plastic resin, and total amount only is only second to tygon and Polyvinylchloride.For making China's polypropylene product have the market competitiveness, exploitation rigidity, toughness, crushing-resistant copolymerization product, random copolymerization product, BOPP and CPP film material, fiber, nonwoven cloth and exploitation polypropylene that mobile balance is good all are important research project from now in the application of automobile and field of household appliances.Melting index is that polypropylene product is determined one of important quality index of product grade, and it has determined the different purposes of product, to the measurement of melting index, to producing and scientific research, important effect and directive significance is arranged all.Yet the on-line analysis of melting index is measured and is difficult at present accomplish, the in-line analyzer that lacks melting index is a main difficulty of production quality control during polypropylene is produced.MI can only obtain by hand sampling, off-line assay, and general every 2-4 hour intellectual analysis once, and time lag is big, is difficult to satisfy the requirement of producing real-time control.

Neural network development in recent years is used widely in fields such as economic, military affairs, commercial production and biomedicines, and has been produced far-reaching influence rapidly.What wherein be most widely used is feedforward neural network, especially the BP neural network.Neural network has the ability of very strong self-adaptation, self-organization, self study and the ability of large-scale parallel computing.But in actual applications, neural network has also exposed some self intrinsic defective: the initialization of weights is at random, easily is absorbed in local minimum; The selection of the interstitial content of hidden layer and other parameters can only rule of thumb be selected with experiment in the learning process; Convergence time is long, poor robustness etc.

By the genetic algorithm that Univ Michigan-Ann Arbor USA Holland professor and student thereof propose first, promptly GeneticAlgorithm is called for short GA, is a kind of parallel efficiently global search algorithm.This algorithm has good robustness, is applicable to parallel processing; Have good global search performance, reduced the risk that is absorbed in locally optimal solution.These advantages of genetic algorithm have remedied the shortcoming of traditional neural network just, therefore can adopt genetic algorithm to optimize neural network.The BP neural network is the most frequently used neural network algorithm, how to set up rational neural network model and be the core content that improves precision of prediction, therefore multiple optimization method occurred or improves one's methods.Soft measurement for melting index in the propylene polymerization production process, because production run has diversity and correlativity, set up its BP neural network optimal soft survey instrument and method, can access optimum soft-sensing model, improve soft measuring accuracy based on genetic algorithm.Therefore, have very important significance.

Summary of the invention

Not high for the measuring accuracy that overcomes existing propylene polymerization production process, as to be subject to artificial factor deficiency the invention provides a kind of on-line measurement, on-line parameter Automatic Optimal, computing velocity is fast, model upgrades automatically, antijamming capability is strong, precision is high propylene polymerization production process melting index optimal soft survey instrument and method based on genetic algorithm optimization BP neural network.

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

A kind of propylene polymerization production process optimal soft survey instrument based on genetic algorithm optimization BP neural network, comprise propylene polymerization production process, be used to measure the field intelligent instrument of easy survey variable, the control station that is used for the measuring operation variable, the DCS database of store data and melt index flexible measured value display instrument, described field intelligent instrument, control station is connected with propylene polymerization production process, described field intelligent instrument, control station is connected with the DCS database, described soft measuring instrument also comprises the optimal soft measurement model based on genetic algorithm optimization BP neural network, described DCS database is connected with the input end of described optimal soft measurement model based on genetic algorithm optimization BP neural network, the output terminal of described optimal soft measurement model based on genetic algorithm optimization BP neural network is connected with melt index flexible measured value display instrument, and described optimal soft measurement model based on genetic algorithm optimization BP neural network comprises:

Data preprocessing module is used for to the input variable centralization, promptly deducting the mean value of variable with carrying out pre-service from the model input variable of DCS database input; The input variable prewhitening being handled is the variable decorrelation again, and input variable is applied a linear transformation; By the independent component analysis method, from linear hybrid data, recover basic source signal through centralization and prewhitening processing;

BP neural network model module is used to adopt the BP neural network, establishes output layer k the neuronic actual y of being output as of BP neural network _k, be input as net _k, arbitrary neuron j is output as y in the hidden layer that layer is adjacent therewith _j, then have:

${\mathrm{net}}_k=\underset{i}{\mathrm{\Σ}}{w}_{\mathrm{kj}}\·y_j---\left(1\right)$

y _k＝f(net _k) (2)

In the formula, w _KjBe the connection weight between neuron k and the neuron j, f () is neuronic output function, is taken as the Sigmoid function usually, is expressed as:

$f\left({\mathrm{net}}_k\right)=1/(1+e^{-({\mathrm{net}}_k+h_k)/{\mathrm{\θ}}_0})---\left(3\right)$

In the formula, h _kBe the threshold value of neuron k, θ ₀Be the steepness parameter, in order to regulate the steepness of Sigmoid function;

Make that training sample is k, for any input pattern X _p, if k neuronic desired output O in the output layer should be arranged mutually _Pk, then the output variance of output layer is expressed as:

$E_p=\frac{1}{2}\underset{p}{\mathrm{\Σ}}{(O_{\mathrm{pk}}-y_{\mathrm{pk}})}^2---\left(4\right)$

In the formula, O _PkRepresent desired output, y _PkRepresent actual output; Oppositely error propagation the destination of study is to revise connection weight w value, makes E reach minimum value; Require connection weight w _Kj, w _JiShould be along E _pNegative gradient direction study; So w _KjCorrection be:

${\mathrm{\Δ}}_p{w}_{\mathrm{kj}}=-\mathrm{\β}\frac{\∂E_p}{\∂w_{\mathrm{kj}}}---\left(5\right)$

In the formula, β is that learning rate is adjusted the factor;

The genetic algorithm optimization BP neural network module is used to adopt the genetic algorithm optimization method that the BP neural network is optimized, and concrete steps are as follows:

1. determine the coded system of network weight, individual bit string length, chromosome adopts binary coding, each link weights of network all use 0/1 string list of certain-length to show, suppose that at first connection weight changes when coding in a certain predetermined scope, the actual value of connection weight and string table indicating value relation table are shown:

$W_t(i,j)=W_{\min }(i,j)+\frac{\mathrm{binrep}\left(t\right)}{2^t-1}[W_{\max }(i,j)-W_{\min }(i,j)]---\left(6\right)$

In the formula, binrep (t) is a bigit, [W _Max(i, j), W _Min(i, j)] be the variation range of each connection weight, the string of binary characters of all weights correspondences is cascaded, obtain a chromosome, represent a kind of weights combination of network.

2. the population scale n of She Dinging produces initial population at random;

3. the objective function of setting network is converted into fitness, and individual network weight is estimated; Error function by network calculates fitness function, and thinks that the big ideal adaptation degree of error is little, the fitness function f of individual i _iBe expressed as:

f _i＝1/(E _i+1) (7)

In the formula, E _iBe the error function of network, be expressed as:

$E_i=\frac{1}{2}\underset{k=1}{\overset{m}{\mathrm{\Σ}}}\underset{j=1}{\overset{n_3}{\mathrm{\Σ}}}{(c_j^k-c_j)}^2---\left(8\right)$

In the formula, m is the number of samples of training set, n ₃Be neuron number, c _j ^kBe the ideal output of unit j, c _jBe the actual output of unit j;

4. selected genetic manipulation is provided with genetic parameter and Adaptive adjusting algorithm, and crossover probability is made as P _c, the variation probability is made as P _m, the selecteed probability P of individual i is expressed as:

$P=f_i/\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{f}_j---\left(9\right)$

In the formula, n is a group size, f _iBe the fitness of individual i, f _iFitness for individual j;

5. carry out selection operation according to fitness in hereditary space;

6. according to selected intersection, variation and relevant algorithm, parameter, operate accordingly, obtain population of new generation;

7. judge whether to satisfy performance requirement, if, finish optimizing, obtain one group of weights of optimizing; Otherwise return step 5., continue the iteration optimizing.

As preferred a kind of scheme, described optimal soft measurement model based on genetic algorithm optimization BP neural network also comprises: the model modification module, be used for the online updating of model, and regularly the off-line analysis data is input in the training set, upgrade neural network model.

As preferred another scheme: in described data preprocessing module, adopt principal component analytical method to realize that prewhitening handles.

A kind of propylene polymerization production process optimal soft measuring method based on genetic algorithm optimization BP neural network, described flexible measurement method mainly may further comprise the steps:

1), to the propylene polymerization production process object, according to industrial analysis and Operations Analyst, selection operation variable and easily survey the input of variable, performance variable and easily survey variable and obtain by the DCS database as model;

2), sample data is carried out pre-service,, promptly deduct the mean value of variable to the input variable centralization; The input variable prewhitening being handled is the variable decorrelation again, and input variable is applied a linear transformation; By the independent component analysis method, from linear hybrid data, recover basic source signal through centralization and prewhitening processing;

3), set up initial neural network model, adopt the BP neural network, establish output layer k the neuronic actual y of being output as of BP neural network based on model input, output data _k, be input as net _k, arbitrary neuron j is output as y in the hidden layer that layer is adjacent therewith _j, then have:

${\mathrm{net}}_k=\underset{i}{\mathrm{\Σ}}{w}_{\mathrm{kj}}\·y_j---\left(1\right)$

y _k＝f(net _k) (2)

In the formula, w _KjBe the connection weight between neuron k and the neuron j, f () is neuronic output function, is taken as the Sigmoid function usually, is expressed as:

$f\left({\mathrm{net}}_k\right)=1/(1+e^{-({\mathrm{net}}_k+h_k)/{\mathrm{\θ}}_0})---\left(3\right)$

In the formula, h _kBe the threshold value of neuron k, θ ₀Be the steepness parameter, in order to regulate the steepness of Sigmoid function;

Make that training sample is k, for any input pattern X _p, if k neuronic desired output O in the output layer should be arranged mutually _Pk, then the output variance of output layer is expressed as:

$E_p=\frac{1}{2}\underset{p}{\mathrm{\Σ}}{(O_{\mathrm{pk}}-y_{\mathrm{pk}})}^2---\left(4\right)$

In the formula, O _PkRepresent desired output, y _PkRepresent actual output; Oppositely error propagation the destination of study is to revise connection weight w value, makes E reach minimum value; Require connection weight w _Kj, w _JiShould be along E _pNegative gradient direction study; So w _KjCorrection be:

${\mathrm{\Δ}}_p{w}_{\mathrm{kj}}=-\mathrm{\β}\frac{\∂E_p}{\∂w_{\mathrm{kj}}}---\left(5\right)$

In the formula, β is that learning rate is adjusted the factor;

4), adopt the genetic algorithm optimization method that the BP neural network is optimized, concrete steps are as follows:

1. determine the coded system of network weight, individual bit string length, chromosome adopts binary coding, each link weights of network all use 0/1 string list of certain-length to show, suppose that at first connection weight changes when coding in a certain predetermined scope, the actual value of connection weight and string table indicating value relation table are shown:

$W_t(i,j)=W_{\min }(i,j)+\frac{\mathrm{binrep}\left(t\right)}{2^t-1}[W_{\max }(i,j)-W_{\min }(i,j)]---\left(6\right)$

In the formula, binrep (t) is a bigit, [W _Max(i, j), W _Min(i, j)] be the variation range of each connection weight, the string of binary characters of all weights correspondences is cascaded, obtain a chromosome, represent a kind of weights combination of network.

2. the population scale n of She Dinging produces initial population at random;

3. the objective function of setting network is converted into fitness, and individual network weight is estimated; Error function by network calculates fitness function, and thinks that the big ideal adaptation degree of error is little, the fitness function f of individual i _iBe expressed as:

f _i＝1/(E _i+1) (7)

In the formula, E _iBe the error function of network, be expressed as:

$E_i=\frac{1}{2}\underset{k=1}{\overset{m}{\mathrm{\Σ}}}\underset{j=1}{\overset{n_3}{\mathrm{\Σ}}}{(c_j^k-c_j)}^2---\left(8\right)$

In the formula, m is the number of samples of training set, n ₃Be neuron number, c _j ^kBe the ideal output of unit j, c _jBe the actual output of unit j;

4. selected genetic manipulation is provided with genetic parameter and Adaptive adjusting algorithm, and crossover probability is made as P _c, the variation probability is made as P _m, the selecteed probability P of individual i is expressed as:

$P=f_i/\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{f}_j---\left(9\right)$

In the formula, n is a group size, f _iBe the fitness of individual i, f _iFitness for individual j;

5. carry out selection operation according to fitness in hereditary space;

6. according to selected intersection, variation and relevant algorithm, parameter, operate accordingly, obtain population of new generation;

7. judge whether to satisfy performance requirement, if, finish optimizing, obtain one group of weights of optimizing; Otherwise return step 5., continue the iteration optimizing.

As preferred a kind of scheme: described flexible measurement method is further comprising the steps of: 5), regularly the off-line analysis data is input in the training set, upgrade neural network model.

Further, in described step 2) in, adopt principal component analytical method to realize the prewhitening processing.

Technical conceive of the present invention is: the important quality index melting index of propylene polymerization production process is carried out the soft measurement of online optimum, overcome the deficiency that existing polypropylene melting index measurement instrument measuring accuracy is not high, be subject to artificial factor, introduce the genetic algorithm optimization method BP neural network weight is carried out Automatic Optimal, do not need artificial experience or repeatedly test adjust neural network, just can obtain optimum soft measurement result.Overcome the following shortcoming of traditional BP neural network: responsive unusually to initial weight vector, different initial weight vector value possibilities one cause diverse result; In concrete computation process, there be choosing of related parameter to determine with experience by experiment, in case value is improper, can cause the vibration of network again and can not restrain and promptly enable convergence and also can cause the training time long, or sink into local extremum and can not get best weight value and distribute because speed of convergence is slow.

Beneficial effect of the present invention mainly shows: 1, on-line measurement; 2, on-line parameter Automatic Optimal; 3, computing velocity is fast; 4, model upgrades automatically; 5, antijamming capability strong, 6, the precision height.

Description of drawings

Fig. 1 is based on the propylene polymerization production process optimal soft survey instrument of genetic algorithm optimization BP neural network and the basic structure synoptic diagram of method;

Fig. 2 is based on the optimal soft measurement model structural representation of genetic algorithm optimization BP neural network;

Fig. 3 is a propylene polymerization production process Hypol explained hereafter process flow diagram.

Embodiment

Below in conjunction with accompanying drawing the present invention is further described.The embodiment of the 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 to the present invention makes all fall into protection scope of the present invention.

Embodiment 1

With reference to Fig. 1, Fig. 2 and Fig. 3, a kind of propylene polymerization production process optimal soft survey instrument based on genetic algorithm optimization BP neural network, comprise propylene polymerization production process 1, be used to measure the field intelligent instrument 2 of easy survey variable, the control station 3 that is used for the measuring operation variable, the DCS database 4 of store data and melt index flexible measured value display instrument 6, described field intelligent instrument 2, control station 3 is connected with propylene polymerization production process 1, described field intelligent instrument 2, control station 3 is connected with DCS database 4, described soft measuring instrument also comprises the optimal soft measurement model 5 of genetic algorithm optimization BP neural network, described DCS database 4 is connected with the input end of described optimal soft measurement model 5 based on genetic algorithm optimization BP neural network, the output terminal of described optimal soft measurement model 5 based on genetic algorithm optimization BP neural network is connected with melt index flexible measured value display instrument 6, and described optimal soft measurement model based on genetic algorithm optimization BP neural network comprises:

Data preprocessing module is used for to the input variable centralization, promptly deducting the mean value of variable with carrying out pre-service from the model input variable of DCS database input; The input variable prewhitening being handled is the variable decorrelation again, and input variable is applied a linear transformation; By the independent component analysis method, from linear hybrid data, recover basic source signal through centralization and prewhitening processing;

BP neural network model module is used to adopt the BP neural network, establishes output layer k the neuronic actual y of being output as of BP neural network _k, be input as net _k, arbitrary neuron j is output as y in the hidden layer that layer is adjacent therewith _j, then have:

${\mathrm{net}}_k=\underset{i}{\mathrm{\Σ}}{w}_{\mathrm{kj}}\·y_j---\left(1\right)$

y _k＝f(net _k) (2)

In the formula, w _KjBe the connection weight between neuron k and the neuron j, f () is neuronic output function, is taken as the Sigmoid function usually, is expressed as:

$f\left({\mathrm{net}}_k\right)=1/(1+e^{-({\mathrm{net}}_k+h_k)/{\mathrm{\θ}}_0})---\left(3\right)$

In the formula, h _kBe the threshold value of neuron k, θ ₀Be the steepness parameter, in order to regulate the steepness of Sigmoid function;

Make that training sample is k, for any input pattern X _p, if k neuronic desired output O in the output layer should be arranged mutually _Pk, then the output variance of output layer is expressed as:

$E_p=\frac{1}{2}\underset{p}{\mathrm{\Σ}}{(O_{\mathrm{pk}}-y_{\mathrm{pk}})}^2---\left(4\right)$

In the formula, O _PkRepresent desired output, y _PkRepresent actual output; Oppositely error propagation the destination of study is to revise connection weight w value, makes E reach minimum value; Require connection weight w _Kj, w _JiShould be along E _pNegative gradient direction study; So w _KjCorrection be:

${\mathrm{\Δ}}_p{w}_{\mathrm{kj}}=-\mathrm{\β}\frac{\∂E_p}{\∂w_{\mathrm{kj}}}---\left(5\right)$

In the formula, β is that learning rate is adjusted the factor;

The genetic algorithm optimization BP neural network module is used to adopt the genetic algorithm optimization method that the BP neural network is optimized, and concrete steps are as follows:

1. determine the coded system of network weight, individual bit string length, chromosome adopts binary coding, each link weights of network all use 0/1 string list of certain-length to show, suppose that at first connection weight changes when coding in a certain predetermined scope, the actual value of connection weight and string table indicating value relation table are shown:

$W_t(i,j)=W_{\min }(i,j)+\frac{\mathrm{binrep}\left(t\right)}{2^t-1}[W_{\max }(i,j)-W_{\min }(i,j)]---\left(6\right)$

In the formula, binrep (t) is a bigit, [W _Max(i, j), W _Min(i, j)] be the variation range of each connection weight, the string of binary characters of all weights correspondences is cascaded, obtain a chromosome, represent a kind of weights combination of network.

2. the population scale n of She Dinging produces initial population at random;

3. the objective function of setting network is converted into fitness, and individual network weight is estimated; Error function by network calculates fitness function, and thinks that the big ideal adaptation degree of error is little, the fitness function f of individual i _iBe expressed as:

f _i＝1/(E _i+1) (7)

In the formula, E _iBe the error function of network, be expressed as:

$E_i=\frac{1}{2}\underset{k=1}{\overset{m}{\mathrm{\Σ}}}\underset{j=1}{\overset{n_3}{\mathrm{\Σ}}}{(c_j^k-c_j)}^2---\left(8\right)$

In the formula, m is the number of samples of training set, n ₃Be neuron number, c _j ^kBe the ideal output of unit j, c _jBe the actual output of unit j;

4. selected genetic manipulation is provided with genetic parameter and Adaptive adjusting algorithm, and crossover probability is made as P _c, the variation probability is made as P _m, the selecteed probability P of individual i is expressed as:

$P=f_i/\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{f}_j---\left(9\right)$

In the formula, n is a group size, f _iBe the fitness of individual i, f _iFitness for individual j;

5. carry out selection operation according to fitness in hereditary space;

6. according to selected intersection, variation and relevant algorithm, parameter, operate accordingly, obtain population of new generation;

7. judge whether to satisfy performance requirement, if, finish optimizing, obtain one group of weights of optimizing; Otherwise return step 5., continue the iteration optimizing.

Described optimal soft measurement model based on genetic algorithm optimization BP neural network also comprises: the model modification module, be used for the online updating of model, and regularly the off-line analysis data is input in the training set, upgrade neural network model.

In described data preprocessing module, adopt principal component analytical method to realize the prewhitening processing.

The propylene polymerization production process process flow diagram as shown in Figure 3, according to reaction mechanism and flow process analysis, consider the various factors that in the polypropylene production process melting index is exerted an influence, get nine performance variables commonly used in the actual production process and easily survey variable as the modeling variable, have: three strand of third rare feed flow rates, major catalyst flow rate, cocatalyst flow rate, hydrogen volume concentration in temperature in the kettle, pressure, the liquid level, still.Table 1 has been listed 9 modeling variablees importing as based on the optimal soft measurement model 5 of genetic algorithm optimization, is respectively liquid level (L) in temperature in the kettle (T), still internal pressure (p), the still, the interior hydrogen volume concentration (X of still _v), 3 bursts of propylene feed flow rates (first strand of third rare feed flow rates f1, second strand of third rare feed flow rates f2, the 3rd strand of third rare feed flow rates f3), 2 bursts of catalyst charge flow rates (major catalyst flow rate f4, cocatalyst flow rate f5).Polyreaction in the reactor is that reaction mass mixes back participation reaction repeatedly, so the model input variable relates to the mean value in preceding some moment of process variable employing of material.Last hour mean value of The data in this example.The conduct of melting index off-line laboratory values is based on the output variable of the optimal soft measurement model 5 of genetic algorithm optimization.Obtain by hand sampling, off-line assay, analyzed collection once in per 4 hours.

Field intelligent instrument 2 and control station 3 link to each other with propylene polymerization production process 1, link to each other with DCS database 4; Optimal soft measurement model 5 links to each other with DCS database and soft measured value display instrument 6.Field intelligent instrument 2 is measured the easy survey variable that propylene polymerization is produced object, will easily survey variable and be transferred to DCS database 4; Control station 3 control propylene polymerizations are produced the performance variable of object, and performance variable is transferred to DCS database 4.The conduct of the variable data of record is based on the input of the optimal soft measurement model 5 of genetic algorithm optimization BP neural network in the DCS database 4, soft measured value display instrument 6 is used to show the output based on the optimal soft measurement model 5 of genetic algorithm optimization BP neural network, promptly soft measured value.

Table 1 is based on the required modeling variable of the optimal soft measurement model of genetic algorithm optimization

Based on the optimal soft measurement model 5 of genetic algorithm optimization BP neural network, comprise following 4 parts:

1) data preprocessing module 7, at first to model input carrying out centralization and prewhitening.To the input variable centralization, deduct the mean value of variable exactly, making variable is the variable of zero-mean, thus shortcut calculation.It is the variable decorrelation that the input variable prewhitening is handled, and input variable is applied a linear transformation, makes between each component of variable after the conversion uncorrelatedly mutually, and its covariance matrix is a unit matrix simultaneously.By the independent component analysis method, from linear hybrid data, recover the method for basic source signal then through centralization and prewhitening processing.

BP neural network model module 8 is implemented as follows:

Typical B P neural network model has an input layer, an output layer and a hiding layer.In theory, for the number of plies of hiding layer without limits, but commonly used be one deck or two-layer.Can prove that in theory one three layers BP network can approach nonlinear system arbitrarily.The BP algorithm minimizes by error function and finishes a kind of height Nonlinear Mapping that is input to output, keeps topological invariance in the mapping.This network comes down to a kind of static network, and its output is the function of existing input, and irrelevant with inputing or outputing of past and future.By the BP neural network structure as can be known, arbitrary neuronic weighted sum that is output as the input pattern component in the input layer, the notion of this weighted sum are fit to all the other each layers equally.If the output layer k of BP neural network neuronic actual y that is output as _k, be input as net _k, arbitrary neuron j is output as y in the hidden layer that layer is adjacent therewith _j, then have:

${\mathrm{net}}_k=\underset{i}{\mathrm{\Σ}}{w}_{\mathrm{kj}}\·y_j---\left(1\right)$

y _k＝f(net _k) (2)

In the formula, w _KjBe the connection weight between neuron k and the neuron j, f () is neuronic output function, is taken as the Sigmoid function usually, is expressed as:

$f\left({\mathrm{net}}_k\right)=1/(1+e^{-({\mathrm{net}}_k+h_k)/{\mathrm{\θ}}_0})---\left(3\right)$

In the formula, h _kBe the threshold value of neuron k, θ ₀Be the steepness parameter, in order to regulate the steepness of Sigmoid function.Make that training sample is k.For any input pattern X _p, if k neuronic desired output O in the output layer should be arranged mutually _Pk, then the output variance of output layer can be expressed as:

$E_p=\frac{1}{2}\underset{p}{\mathrm{\Σ}}{(O_{\mathrm{pk}}-y_{\mathrm{pk}})}^2---\left(4\right)$

In the formula, O _PkRepresent desired output, y _PkRepresent actual output.Oppositely error propagation the destination of study is to revise connection weight w value, makes E reach minimum value.This just requires connection weight w _Kj, w _JiShould be along E _pNegative gradient direction study.

So w _KjCorrection be:

${\mathrm{\Δ}}_p{w}_{\mathrm{kj}}=-\mathrm{\β}\frac{\∂E_p}{\∂w_{\mathrm{kj}}}---\left(5\right)$

In the formula, β is that learning rate is adjusted the factor.

Genetic algorithm optimization module 9, adopt the genetic algorithm optimization method that the BP neural network is optimized, on the basis of basic BP neural network algorithm, optimize the weighting parameter of neural network by the global optimizing ability of genetic algorithm optimization, and carry out the study of neural network, thereby set up the BP neural network optimal soft measurement model that the genetic algorithm optimization of propylene polymerization melting index is optimized with this.Concrete steps with the genetic algorithm optimization network weight are as follows:

1. determine the coded system of network weight, individual bit string length.The relation of the individual bit string in one group of weights and hereditary space is determined by the coding mapping.Chromosome adopts binary coding, and each link weights of network all use 0/1 string list of certain-length to show.Suppose that at first connection weight changes when coding in a certain predetermined scope, the actual value of connection weight and string table indicating value relation can be expressed as:

$W_t(i,j)=W_{\min }(i,j)+\frac{\mathrm{binrep}\left(t\right)}{2^t-1}[W_{\max }(i,j)-W_{\min }(i,j)]---\left(6\right)$

In the formula, binrep (t) is a bigit, [W _Max(i, j), W _Min(i, j)] be the variation range of each connection weight.The string of binary characters of all weights correspondences is cascaded, obtains a chromosome, represent a kind of weights combination of network.

2. the population scale n of She Dinging produces initial population at random.

3. the objective function of setting network is converted into fitness, and individual network weight is estimated.Error function by network calculates fitness function, and thinks that the big ideal adaptation degree of error is little.The fitness function f of individual i _iCan be expressed as:

f _i＝1/(E _i+1) (7)

In the formula, E _iBe the error function of network, be expressed as:

$E_i=\frac{1}{2}\underset{k=1}{\overset{m}{\mathrm{\Σ}}}\underset{j=1}{\overset{n_3}{\mathrm{\Σ}}}{(c_j^k-c_j)}^2---\left(8\right)$

In the formula, m is the number of samples of training set, n ₃Be neuron number, c _j ^kBe the ideal output of unit j, c _jBe the actual output of unit j.

4. selected genetic manipulation is provided with genetic parameter and Adaptive adjusting algorithm etc.Crossover probability is made as P _c, the variation probability is made as P _mThe selecteed probability P of individual i is expressed as:

$P=f_i/\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{f}_j---\left(9\right)$

In the formula, n is a group size, f _iBe the fitness of individual i, f _iFitness for individual j.

5. carry out selection operation according to fitness in hereditary space.

6. according to selected intersection, variation and relevant algorithm, parameter, operate accordingly, obtain population of new generation.

7. judge whether to satisfy performance requirement, if, finish optimizing, obtain one group of weights of optimizing; Otherwise return step 5., continue the iteration optimizing.

Model modification module 10 is used for the online updating of model, regularly the off-line analysis data is input in the training set, upgrades neural network model.

Embodiment 2

With reference to Fig. 1, Fig. 2 and Fig. 3, a kind of propylene polymerization production process optimal soft measuring method based on genetic algorithm optimization BP neural network, described flexible measurement method mainly may further comprise the steps:

1), to the propylene polymerization production process object, according to industrial analysis and Operations Analyst, selection operation variable and easily survey the input of variable, performance variable and easily survey variable and obtain by the DCS database as model;

2), sample data is carried out pre-service,, promptly deduct the mean value of variable to the input variable centralization; The input variable prewhitening being handled is the variable decorrelation again, and input variable is applied a linear transformation; By the independent component analysis method, from linear hybrid data, recover basic source signal through centralization and prewhitening processing;

3), set up initial neural network model, adopt the BP neural network, establish output layer k the neuronic actual y of being output as of BP neural network based on model input, output data _k, be input as net _k, arbitrary neuron j is output as y in the hidden layer that layer is adjacent therewith _j, then have:

${\mathrm{net}}_k=\underset{i}{\mathrm{\Σ}}{w}_{\mathrm{kj}}\·y_j---\left(1\right)$

y _k＝f(net _k) (2)

In the formula, w _KjBe the connection weight between neuron k and the neuron j, f () is neuronic output function, is taken as the Sigmoid function usually, is expressed as:

$f\left({\mathrm{net}}_k\right)=1/(1+e^{-({\mathrm{net}}_k+h_k)/{\mathrm{\θ}}_0})---\left(3\right)$

In the formula, h _kBe the threshold value of neuron k, θ ₀Be the steepness parameter, in order to regulate the steepness of Sigmoid function;

Make that training sample is k, for any input pattern X _p, if k neuronic desired output O in the output layer should be arranged mutually _Pk, then the output variance of output layer is expressed as:

$E_p=\frac{1}{2}\underset{p}{\mathrm{\Σ}}{(O_{\mathrm{pk}}-y_{\mathrm{pk}})}^2---\left(4\right)$

In the formula, O _PkRepresent desired output, y _PkRepresent actual output; Oppositely error propagation the destination of study is to revise connection weight w value, makes E reach minimum value; Require connection weight w _Kj, w _JiShould be along E _pNegative gradient direction study; So w _KjCorrection be:

${\mathrm{\Δ}}_p{w}_{\mathrm{kj}}=-\mathrm{\β}\frac{\∂E_p}{\∂w_{\mathrm{kj}}}---\left(5\right)$

In the formula, β is that learning rate is adjusted the factor;

4), adopt the genetic algorithm optimization method that the BP neural network is optimized, concrete steps are as follows:

1. determine the coded system of network weight, individual bit string length, chromosome adopts binary coding, each link weights of network all use 0/1 string list of certain-length to show, suppose that at first connection weight changes when coding in a certain predetermined scope, the actual value of connection weight and string table indicating value relation table are shown:

$W_t(i,j)=W_{\min }(i,j)+\frac{\mathrm{binrep}\left(t\right)}{2^t-1}[W_{\max }(i,j)-W_{\min }(i,j)]---\left(6\right)$

In the formula, binrep (t) is a bigit, [W _Max(i, j), W _Min(i, j)] be the variation range of each connection weight, the string of binary characters of all weights correspondences is cascaded, obtain a chromosome, represent a kind of weights combination of network.

2. the population scale n of She Dinging produces initial population at random;

3. the objective function of setting network is converted into fitness, and individual network weight is estimated; Error function by network calculates fitness function, and thinks that the big ideal adaptation degree of error is little, the fitness function f of individual i _iBe expressed as:

f _i＝1/(E _i+1) (7)

In the formula, E _iBe the error function of network, be expressed as:

$E_i=\frac{1}{2}\underset{k=1}{\overset{m}{\mathrm{\Σ}}}\underset{j=1}{\overset{n_3}{\mathrm{\Σ}}}{(c_j^k-c_j)}^2---\left(8\right)$

In the formula, m is the number of samples of training set, n ₃Be neuron number, c _j ^kBe the ideal output of unit j, c _jBe the actual output of unit j;

4. selected genetic manipulation is provided with genetic parameter and Adaptive adjusting algorithm, and crossover probability is made as P _c, the variation probability is made as P _m, the selecteed probability P of individual i is expressed as:

$P=f_i/\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{f}_j---\left(9\right)$

In the formula, n is a group size, f _iBe the fitness of individual i, f _iFitness for individual j;

5. carry out selection operation according to fitness in hereditary space;

6. according to selected intersection, variation and relevant algorithm, parameter, operate accordingly, obtain population of new generation;

7. judge whether to satisfy performance requirement, if, finish optimizing, obtain one group of weights of optimizing; Otherwise return step 5., continue the iteration optimizing.

Described flexible measurement method is further comprising the steps of: 5), regularly the off-line analysis data is input in the training set, upgrade neural network model.

In described step 2) in, adopt principal component analytical method to realize the prewhitening processing.

The concrete implementation step of the method for present embodiment is as follows:

Step 1: to propylene polymerization production process object 1, according to industrial analysis and Operations Analyst, selection operation variable and easy input of surveying variable as model.Performance variable and the easy variable of surveying are obtained by DCS database 4.

Step 2: sample data is carried out pre-service, finish by data preprocessing module 7.

Step 3:, finish by ICA independent component analysis module 8 to carrying out independent component analysis through pretreated data.

Step 4: set up initial neural network model 9 based on model input, output data.The input data are as acquisition as described in the step 1, and output data is obtained by the off-line chemical examination.

Step 5: the weighting parameter of optimizing initial neural network 8 by genetic algorithm optimization BP neural network method 10.

Step 6: model modification module 11 regularly is input to the off-line analysis data in the training set, upgrades neural network model, sets up based on the optimal soft measurement model 5 of genetic algorithm optimization BP neural network and finishes.

Step 7: melt index flexible measured value display instrument 6 shows based on the output of the optimal soft measurement model 5 of genetic algorithm optimization BP neural network, finishes the demonstration to the optimum soft measurement of propylene polymerization production process melting index.

Claims (6)

1, a kind of propylene polymerization production process optimal soft survey instrument based on genetic algorithm optimization BP neural network, comprise propylene polymerization production process, be used to measure the field intelligent instrument of easy survey variable, the control station that is used for the measuring operation variable, the DCS database of store data and melt index flexible measured value display instrument, described field intelligent instrument, control station is connected with propylene polymerization production process, described field intelligent instrument, control station is connected with the DCS database, it is characterized in that: described soft measuring instrument also comprises the optimal soft measurement model based on genetic algorithm optimization BP neural network, described DCS database is connected with the input end of described optimal soft measurement model based on genetic algorithm optimization BP neural network, the output terminal of described optimal soft measurement model based on genetic algorithm optimization BP neural network is connected with melt index flexible measured value display instrument, and described optimal soft measurement model based on genetic algorithm optimization BP neural network comprises:

Data preprocessing module is used for to the input variable centralization, promptly deducting the mean value of variable with carrying out pre-service from the model input variable of DCS database input; The input variable prewhitening being handled is the variable decorrelation again, and input variable is applied a linear transformation;

By the independent component analysis method, from linear hybrid data, recover basic source signal through centralization and prewhitening processing;

BP neural network model module is used to adopt the BP neural network, establishes output layer k the neuronic actual y of being output as of BP neural network _k, be input as net _k, arbitrary neuron j is output as y in the hidden layer that layer is adjacent therewith _j, then have:

${\mathrm{net}}_k=\underset{i}{\mathrm{\Σ}}{w}_{\mathrm{kj}}\·y_j---\left(1\right)$

y _k＝f(net _k) (2)

In the formula, w _KjBe the connection weight between neuron k and the neuron j, f () is neuronic output function, is taken as the Sigmoid function usually, is expressed as:

$f\left({\mathrm{net}}_k\right)=1/(1+e^{-({\mathrm{net}}_k+h_k)/{\mathrm{\θ}}_0})---\left(3\right)$

In the formula, h _kBe the threshold value of neuron k, θ ₀Be the steepness parameter, in order to regulate the steepness of Sigmoid function;

Make that training sample is k, for any input pattern X _p, if k neuronic desired output O in the output layer should be arranged mutually _Pk, then the output variance of output layer is expressed as:

$E_p=\frac{1}{2}\underset{p}{\mathrm{\Σ}}{(O_{\mathrm{pk}}-y_{\mathrm{pk}})}^2---\left(4\right)$

In the formula, O _PkRepresent desired output, y _PkRepresent actual output; Oppositely error propagation the destination of study is to revise connection weight w value, makes E reach minimum value; Require connection weight w _Kj, w _JiShould be along E _pNegative gradient direction study; So w _KjCorrection be:

${\mathrm{\Δ}}_p{w}_{\mathrm{kj}}=-\mathrm{\β}\frac{\∂E_p}{{\∂w}_{\mathrm{kj}}}---\left(5\right)$

In the formula, β is that learning rate is adjusted the factor;

The genetic algorithm optimization BP neural network module is used to adopt the genetic algorithm optimization method that the BP neural network is optimized, and concrete steps are as follows:

1. determine the coded system of network weight, individual bit string length, chromosome adopts binary coding, each link weights of network all use 0/1 string list of certain-length to show, suppose that at first connection weight changes when coding in a certain predetermined scope, the actual value of connection weight and string table indicating value relation table are shown:

$W_t(i,j)=W_{\min }(i,j)+\frac{\mathrm{binrep}\left(t\right)}{2^t-1}[W_{\max }(i,j)-W_{\min }(i,j)]---\left(6\right)$

In the formula, binrep (t) is a bigit, [W _Max(i, j), W _Min(i, j)] be the variation range of each connection weight, the string of binary characters of all weights correspondences is cascaded, obtain a chromosome, represent a kind of weights combination of network.

2. the population scale n of She Dinging produces initial population at random;

3. the objective function of setting network is converted into fitness, and individual network weight is estimated; Error function by network calculates fitness function, and thinks that the big ideal adaptation degree of error is little, the fitness function f of individual i _iBe expressed as:

f _i＝1/(E _i+1) (7)

In the formula, E _iBe the error function of network, be expressed as:

$E_i=\frac{1}{2}\underset{k=1}{\overset{m}{\mathrm{\Σ}}}\underset{j=1}{\overset{n_3}{\mathrm{\Σ}}}{(c_j^k-c_j)}^2---\left(8\right)$

In the formula, m is the number of samples of training set, n ₃Be neuron number, c _j ^kBe the ideal output c of unit j _jBe the actual output of unit j;

4. selected genetic manipulation is provided with genetic parameter and Adaptive adjusting algorithm, and crossover probability is made as P _c, the variation probability is made as P _m, the selecteed probability P of individual i is expressed as:

$P=f_i/\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{f}_j---\left(9\right)$

In the formula, n is a group size, and fi is the fitness of individual i, f _jFitness for individual j;

5. carry out selection operation according to fitness in hereditary space;

6. according to selected intersection, variation and relevant algorithm, parameter, operate accordingly, obtain population of new generation;

7. judge whether to satisfy performance requirement, if, finish optimizing, obtain one group of weights of optimizing; Otherwise return step 5., continue the iteration optimizing.

2, the propylene polymerization production process optimal soft survey instrument based on genetic algorithm optimization BP neural network as claimed in claim 1 is characterized in that: described optimal soft measurement model based on genetic algorithm optimization BP neural network also comprises:

The model modification module is used for the online updating of model, regularly the off-line analysis data is input in the training set, upgrades neural network model.

3, the propylene polymerization production process optimal soft survey instrument based on genetic algorithm optimization BP neural network as claimed in claim 1 or 2 is characterized in that: in described data preprocessing module, adopt principal component analytical method to realize the prewhitening processing.

4, a kind of usefulness flexible measurement method of realizing based on the propylene polymerization production process optimal soft survey instrument of genetic algorithm optimization BP neural network as claimed in claim 1, it is characterized in that: described flexible measurement method mainly may further comprise the steps:

1), to the propylene polymerization production process object, according to industrial analysis and Operations Analyst, selection operation variable and easily survey the input of variable, performance variable and easily survey variable and obtain by the DCS database as model;

2), sample data is carried out pre-service,, promptly deduct the mean value of variable to the input variable centralization; The input variable prewhitening being handled is the variable decorrelation again, and input variable is applied a linear transformation; By the independent component analysis method, from linear hybrid data, recover basic source signal through centralization and prewhitening processing;

3), set up initial neural network model, adopt the BP neural network, establish output layer k the neuronic actual y of being output as of BP neural network based on model input, output data _k, be input as net _k, arbitrary neuron j is output as y in the hidden layer that layer is adjacent therewith _j, then have:

${\mathrm{net}}_k=\underset{i}{\mathrm{\Σ}}{w}_{\mathrm{kj}}\·y_j---\left(1\right)$

y _k＝f(net _k) (2)

In the formula, w _KjBe the connection weight between neuron k and the neuron j, f () is neuronic output function, is taken as the Sigmoid function usually, is expressed as:

$\left({\mathrm{net}}_k\right)=1/(1+e^{-({\mathrm{net}}_k+h_k)/{\mathrm{\θ}}_0})---\left(3\right)$

In the formula, h _kBe the threshold value of neuron k, θ ₀Be the steepness parameter, in order to regulate the steepness of Sigmoid function;

Make that training sample is k, for any input pattern X _p, if k neuronic desired output O in the output layer should be arranged mutually _Pk, then the output variance of output layer is expressed as:

$E_p=\frac{1}{2}\underset{p}{\mathrm{\Σ}}{(O_{\mathrm{pk}}-y_{\mathrm{pk}})}^2---\left(4\right)$

In the formula, O _PkRepresent desired output, y _PkRepresent actual output; Oppositely error propagation the destination of study is to revise connection weight w value, makes E reach minimum value; Require connection weight w _Kj, w _JiShould be along E _pNegative gradient direction study; So w _KjCorrection be:

${\mathrm{\Δ}}_p{w}_{\mathrm{kj}}=-\mathrm{\β}\frac{{\∂E}_p}{{\∂w}_{\mathrm{kj}}}---\left(5\right)$

In the formula, β is that learning rate is adjusted the factor;

4), adopt the genetic algorithm optimization method that the BP neural network is optimized, concrete steps are as follows:

1. determine the coded system of network weight, individual bit string length, chromosome adopts binary coding, each link weights of network all use 0/1 string list of certain-length to show, suppose that at first connection weight changes when coding in a certain predetermined scope, the actual value of connection weight and string table indicating value relation table are shown:

$W_t(i,j)=W_{\min }(i,j)+\frac{\mathrm{binrep}\left(t\right)}{2^t-1}[W_{\max }(i,j)-W_{\min }(i,j)]---\left(6\right)$

In the formula, binrep (t) is a bigit, [W _Max(i, j), W _Min(i, j)] be the variation range of each connection weight, the string of binary characters of all weights correspondences is cascaded, obtain a chromosome, represent a kind of weights combination of network.

2. the population scale n of She Dinging produces initial population at random;

3. the objective function of setting network is converted into fitness, and individual network weight is estimated; Error function by network calculates fitness function, and thinks that the big ideal adaptation degree of error is little, the fitness function f of individual i _iBe expressed as:

f _i＝1/(E _i+1) (7)

In the formula, E _iBe the error function of network, be expressed as:

$E_i=\frac{1}{2}\underset{k=1}{\overset{m}{\mathrm{\Σ}}}\underset{j=1}{\overset{n_3}{\mathrm{\Σ}}}{(c_j^k-c_j)}^2---\left(8\right)$

In the formula, m is the number of samples of training set, n ₃Be neuron number, c _j ^kBe the ideal output of unit j, c _jBe the actual output of unit j;

4. selected genetic manipulation is provided with genetic parameter and Adaptive adjusting algorithm, and crossover probability is made as P _c, the variation probability is made as P _m, the selecteed probability P of individual i is expressed as:

$P=f_i/\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{f}_j---\left(9\right)$

In the formula, n is a group size, f _iBe the fitness of individual i, f _iFitness for individual j;

5. carry out selection operation according to fitness in hereditary space;

6. according to selected intersection, variation and relevant algorithm, parameter, operate accordingly, obtain population of new generation;

7. judge whether to satisfy performance requirement, if, finish optimizing, obtain one group of weights of optimizing; Otherwise return step 5., continue the iteration optimizing.

5, flexible measurement method as claimed in claim 4 is characterized in that: described flexible measurement method is further comprising the steps of: 6), regularly the off-line analysis data is input in the training set, upgrade neural network model.

6, as claim 4 or 5 described flexible measurement methods, it is characterized in that: in described step 2) in, adopt principal component analytical method to realize the prewhitening processing.

Images