CN108804851A - A kind of high-precision propylene polymerization production process optimal soft survey instrument of chaos gunz optimizing - Google Patents
A kind of high-precision propylene polymerization production process optimal soft survey instrument of chaos gunz optimizing Download PDFInfo
- Publication number
- CN108804851A CN108804851A CN201810689702.3A CN201810689702A CN108804851A CN 108804851 A CN108804851 A CN 108804851A CN 201810689702 A CN201810689702 A CN 201810689702A CN 108804851 A CN108804851 A CN 108804851A
- Authority
- CN
- China
- Prior art keywords
- particle
- chaos
- module
- propylene polymerization
- production process
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Complex Calculations (AREA)
Abstract
The invention discloses a kind of high-precision propylene polymerization production process optimal soft survey instruments of chaos gunz optimizing, including propylene polymerization production process, field intelligent instrument, control station, the DCS database for storing data, optimal soft measurement model and melt index flexible measured value display instrument based on improvement gravitation search algorithm optimization chaos least square method supporting vector machine, field intelligent instrument and control station are connected with propylene polymerization production process, are connected with DCS database；Optimal soft measurement model is connected with DCS database and hard measurement value display instrument.Described includes data preprocessing module, chaos analysis module, least square method supporting vector machine module, model modification module, improvement gravitation search algorithm optimization module based on improving gravitation search algorithm to optimize the optimal soft measurement model of chaos least square method supporting vector machine.The propylene polymerization production process optimal soft survey instrument of the present invention realizes gunz optimizing, the forecast of high-precision chaos.
Description
Technical field
It is specifically a kind of to be based on the present invention relates to a kind of high-precision optimal soft survey instrument of chaos gunz optimizing and method
Improve the propylene polymerization processes optimal soft survey instrument of gravitation search algorithm optimization chaos least square method supporting vector machine.
Background technology
A kind of thermoplastic resin of prepared by polypropylene is by propylene polymerization, the most important downstream product of propylene, the world third
The 50% of alkene, the 65% of China's propylene is all for polypropylene processed, is one of five big general-purpose plastics, close with our daily life
Cut phase is closed.Polypropylene is that fastest-rising general thermoplastic resin, total amount are only only second to polyethylene and polyvinyl chloride in the world.For
Make China's polypropylene product that there is the market competitiveness, exploitation rigidity toughness, to flow the good crushing-resistant copolymerization product of sexual balance, is random
Copolymerized product, BOPP and CPP film materials, fiber, nonwoven cloth, and application of the polypropylene in automobile and field of household appliances is developed, all
It is important from now on research topic.
Melt index is that polypropylene product determines one of important quality index of product grade, it determines the difference of product
Purposes, the measurement to melt index are an important links of control of product quality in polypropylene production, to production and scientific research, all
There are very important effect and directive significance.
However, it is difficult to accomplish at present that the on-line analysis of melt index, which measures, on the one hand it is online melt index analyzer
Lack, be on the other hand existing in-line analyzer measured often blocking it is inaccurate even can not be caused by normal use
Use on difficulty.Therefore, at present in industrial production MI measurement, mainly obtained by manual sampling, offline assay
, and it is general primary per that can only analyze within 2-4 hours, time lag is big, is brought to the quality control of propylene polymerization production tired
Difficulty becomes a bottleneck problem for being badly in need of solving in production.The online forecasting system and method for polypropylene melt index is studied, from
And as a forward position and the hot spot of academia and industrial quarters.
Invention content
In order to overcome the influence of not high, the artificial selection parameter of measurement accuracy of existing propylene polymerization production process at present
Deficiency, the purpose of the present invention is to provide a kind of forecast of chaos, gunz optimizing, high-precision propylene polymerization production process optimal softs
Measuring instrumentss.
The technical solution adopted by the present invention to solve the technical problems is：A kind of high-precision propylene of chaos gunz optimizing is poly-
Process optimum soft measuring instrument is produced in symphysis, described soft for carrying out hard measurement to the melt index in propylene polymerization production process
Measuring unit includes the data preprocessing module being sequentially connected, chaos analysis module, least square method supporting vector machine module and changes
Into gravitation search algorithm optimization module, wherein：
Data preprocessing module：The mode input variable inputted from DCS database is pre-processed, with specific reference to following formula
It is standardized：
Wherein, mean indicates that the arithmetic mean of instantaneous value of each variable, std indicate the standard deviation of each variable,Indicate input variable
Value, subscript i indicates that ith detection, j indicate that jth ties up variable, x respectivelyijIndicate that the input variable after standardization, S indicate model
Input variable.
Chaos analysis module：According to chaology to sample X={ xijCarry out chaotic Property Analysis.It first has to determine mutually empty
Between two important parameters reconstructing, i.e. delay time and Embedded dimensions.If the maximum Lyapunov exponent of system is more than 0,
The system must be chaos, it is possible to the chaotic characteristic of system is judged according to maximum Lyapunov exponent.
(1) mutual information method determines the delay time T of reconstruct, is under nonlinear case, to sometime t and another moment
Relationship between the information content of t+ τ is analyzed.If A, B two systems, ar、bkIt is the measurement result of a certain physical quantity, PA
(ar)、PB(bk) it is a respectivelyrWith bkProbability, PAB(ar,bk) it is expressed as the joint probability of A and B.Then mutual information is defined as
Average is defined as follows：
If A and B are uncorrelated, IAB(ar,bk)=0.If A represents time series y (t), B represents time series y (t+
τ), then Average is：
τ~I (τ) curve is drawn, delay of the first minimum point corresponding time of curve as phase space reconfiguration is taken
Time τ.
(2) Cao methods determine the Embedded dimensions of reconstruct, and in m ties up phase space, i-th of phase point vector is denoted as
Xm(i)=[x (i), x (i+ τ) ..., x (i+ (m-1) τ)]T (5)
It defines first
Wherein,For Xm(i) nearest neighbor point, Xm+1(i) andIt is phase point Xm(i) andIt is tieed up in m+1
The continuation of phase space.Range formula uses maximum norm, i.e.,
Remember a2(i, m) is about the mean value of i
For the change situation of research E (m), definition
If time series is generated by attractor, when m is more than some m0When, E1(m) it no longer changes, then
m0It is exactly the minimum embedding dimension number of phase space reconstruction.
(3) small data sets arithmetic is used to calculate maximum Lyapunov exponent, process is as follows：
1. carrying out Fast Fourier Transform (FFT) to time series { x (t) | t=1,2 ..., n }, its average period of P is calculated；
2. the delay time T and Embedded dimensions m of time series are determined, if it is { X to reconstruct later phase spacei| i=1,
2 ..., K }, wherein K=n- (m-1) τ represents the number of phase point in phase space；
3. finding phase point XiThe corresponding nearest neighbor point in phase spaceSimultaneously it is limited to detach in short-term in phase space：
4. for any phase point X in phase spacei, acquire it and correspond to the distance d after j discrete time of adjoint point pairi(j)
For：
5. to each j, all ln (d are found outi(j)) average y (j), i.e.,：
Wherein, q is record di(j) ≠ 0 number.With least square fitting, then when the slope of regression straight line is exactly this
Between sequence maximum Lyapunov exponent λ1。
(4) phase space reconfiguration is carried out to sample：
Wherein, N=n- (m-1) τ is phase point sum.
Least square method supporting vector machine module：For establishing soft-sensing model, modeling process is as follows：
In least square method supporting vector machine, it is assumed that given set of data samples is combined into (x1,y1),…(xn,yn),x∈Rn,
y∈R.Under nonlinear situation, transformation is introducedSample is mapped to a high-dimensional feature space from the input space.Find letter
Several parameters equivalents is in the following quadratic programming problem of solution：
Wherein, ξiFor slack variable, C is penalty factor.Introduce Lagrange multipliers αi, the Lagrange letters that are defined as follows
Number：
Enable L to variable w, b, αi, ξiPartial derivative be equal to zero：
Optimization problem is converted into solution linear equation：
Wherein y=[y1,...,yn]T, 1v=[1 ..., 1]T, α=[α1,...,αn]T, K=[kij]n×n,I is unit battle array.Finally obtain nonlinear model：
Improve gravitation search algorithm optimization module：Gravitation search algorithm is improved to least square method supporting vector machine for using
The parameter of module optimizes, and realization is as follows：
(1) algorithm initialization, all particles of random initializtion, each particle represent a candidate solution of problem.In a D
In the search space of dimension, it is assumed that have NP particle, the position for defining i-th of particle is
Set iteration termination condition, i.e. maximum iteration itermax。
(2) it in certain t moment, defines j-th of particle and acts on the gravitation size on i-th of particleFor：
Wherein, Maj(t) and Mpi(t) it is respectively the inertia mass for acting on particle j and the inertia mass for being applied particle i, Rij
(t) it is Euclidean distance between i-th of particle and j-th of particle, ε is the constant of a very little, and G (t) is the gravitation in t moment
Constant：
Wherein, α is descent coefficient, G0It is initial gravitational constant, itermaxIt is maximum iteration.
(3) inertia mass of particle is calculated according to the size of its fitness value, and inertia mass shows that more greatly it is closer
Optimal value, while meaning that the attraction of the particle is bigger, but its movement speed is slower.Assuming that gravitational mass and inertia mass
Equal, the quality of particle can go to update by operation rule appropriate, and more new algorithm is as follows：
Mai=Mpi=Mii=Mi, i=1,2 ..., NP (22)
Wherein, fiti(t) size in the fitness value of i-th of particle of t moment is represented.To solving minimum problems,
Best (t) and worst (t) are defined as follows：
To solving max problem, best (t) and worst (t) are defined as follows：
(4) assume that t moment acts on the total force on i-th of particle in d dimensionsEqual to other all particles
To its sum of active force, calculation formula is as follows：
Wherein, randjIt is random number of the range in [0,1], Kbest is the preceding K grain for having at the beginning optimal adaptation degree
The set of son.
According to Newton's second law, acceleration of the t moment particle i in d dimensionsFor：
Wherein, Mi(t) be i-th of particle inertia mass.
(5) in next iteration, the new speed of particle is the summation of part present speed and its acceleration.Therefore, GSA
During interative computation each time, particle can all update its speed and position according to following formula：
vi(t+1)=ω vi(t)+c1ri1ai(t)+c2ri2(gbest-xi(t)) (31)
xi(t+1)=xi(t)+vi(t+1) (32)
Wherein, vi(t) it is speed of the particle i in the t times iteration, xi(t) it is particle i in the position of the t times iteration, ai(t)
It is acceleration of the particle i in the t times iteration, gbest is current optimal solution, ri1And ri2It is two between [0,1] random
Number；The inertial factor that ω is gradually reduced, c1And c2It is self adaptable acceleration coefficient, calculation formula is as follows：
Wherein, t and itermaxIt is current iteration number and maximum iteration respectively；c1i、c1f、c2iAnd c2fIt is constant, makes
Obtain c10.5, c is gradually decreased to from 2.522.5 are progressively increased to from 0.5.
(6) above step is repeated until reaching maximum iteration, selection fitness value is optimal to be solved as algorithm most
Excellent solution terminates algorithm and returns.
The high-precision propylene polymerization production process optimal soft survey instrument of chaos gunz optimizing further includes model modification
Module is used for the online updating of model, periodically offline analysis data is input in training set, updates least square supporting vector
Machine model.
Beneficial effects of the present invention are mainly manifested in：The present invention melts the important quality index of propylene polymerization production process
Index carries out online optimal hard measurement, overcomes that existing polypropylene melt index measuring instrumentss measurement accuracy is high, artificial selection
The deficiency of the influence of parameter is firstly introduced into chaology and analyzes melt index time series, will by phase space reconfiguration
It is extended to hyperspace from one, can probe into the more information that sequential is included；Secondly it introduces and improves gravitation search algorithm
Optimization module carries out Automatic Optimal to the parameter of least square method supporting vector machine, does not need artificial experience or repeatedly tests to adjust
Least square method supporting vector machine, to obtain the optimal soft survey instrument for the melt index forecast function of having optimal.
Description of the drawings
Fig. 1 is the high-precision propylene polymerization production process optimal soft survey instrument of chaos gunz optimizing and the basic knot of method
Structure schematic diagram；
Fig. 2 is based on the optimal soft measurement model knot for improving gravitation search algorithm optimization chaos least square method supporting vector machine
Structure schematic diagram；
Fig. 3 is propylene polymerization production process Hypol technique productions flow charts.
Specific implementation mode
The present invention is illustrated below according to attached drawing.
Referring to Fig.1, a kind of high-precision propylene polymerization production process optimal soft survey instrument of chaos gunz optimizing, including third
Alkene polymerization production process 1, the control station 3 for measuring performance variable, is deposited the field intelligent instrument 2 for measuring easy survey variable
Put the DCS database 4 and melt index flexible measured value display instrument 6 of data, the field intelligent instrument 2, control station 3 and propylene
Polymerization production process 1 connects, and the field intelligent instrument 2, control station 3 are connect with DCS database 4, and the soft measuring instrument is also
Include the optimal soft measurement model 5 for optimizing chaos least square least square method supporting vector machine based on improvement gravitation search algorithm,
The DCS database 4 optimizes chaos least square least square method supporting vector machine with described based on improvement gravitation search algorithm
The input terminal of optimal soft measurement model 5 connects, described to optimize chaos least square least square based on improvement gravitation search algorithm
The output end of the optimal soft measurement model 5 of support vector machines is connect with melt index flexible measured value display instrument 6.
It with reference to Fig. 3, is analyzed according to reaction mechanism and flow process, takes common nine behaviour in propylene polymerization production process
Make variable and easily surveys variable as modeling variable, including：Three bursts of propylene feed flow rates, major catalyst flow rate, cocatalyst flow rate,
Temperature in the kettle, pressure, liquid level, hydrogen volume concentration in kettle, as shown in table 1, respectively temperature in the kettle (T), pressure (p) in kettle,
Liquid level (L) in kettle, hydrogen volume concentration (X in kettlev), 3 bursts of propylene feed flow rates (first gang of propylene feed flow rates f1, second strand
Propylene feed flow rates f2, third stock propylene feed flow rates f3), 2 bursts of catalyst charge flow rates (major catalyst flow rate f4, auxiliary catalysis
Agent flow rate f5).Input variable of the variable data of propylene polymerization processes as optimal soft measurement model 5, melt index are changed offline
Output variable of the value as optimal soft measurement model 5 is tested, is obtained by manual sampling, offline assay, analysis in every 4 hours is adopted
Collection is primary.
The variable of 1 propylene polymerization processes of table
| Variable symbol | Variable meaning | Variable symbol | Variable meaning |
| T | Temperature in the kettle | f1 | First burst of propylene feed flow rates |
| p | Pressure in kettle | f2 | Second burst of propylene feed flow rates |
| L | Liquid level in kettle | f3 | Third stock propylene feed flow rates |
| Xv | Hydrogen volume concentration in kettle | f4 | Major catalyst flow rate |
| f5 | Cocatalyst flow rate |
It is described that chaos least square least square method supporting vector machine is optimized based on improvement gravitation search algorithm with reference to Fig. 2
Optimal soft measurement model 5 further includes：
Data preprocessing module 7：The mode input variable inputted from DCS database is pre-processed, with specific reference under
Formula is standardized：
Wherein, mean indicates that the arithmetic mean of instantaneous value of each variable, std indicate the standard deviation of each variable,Indicate input variable
Value, subscript i indicates that ith detection, j indicate that jth ties up variable, x respectivelyijIndicate that the input variable after standardization, S indicate model
Input variable.
Chaos analysis module 8：According to chaology to sample X={ xijCarry out chaotic Property Analysis.It first has to determine phase
Two important parameters of Space Reconstruction, i.e. delay time and Embedded dimensions.If the maximum Lyapunov exponent of system is more than 0,
Then the system must be chaos, it is possible to the chaotic characteristic of system is judged according to maximum Lyapunov exponent.
(1) mutual information method determines the delay time T of reconstruct, is under nonlinear case, to sometime t and another moment
Relationship between the information content of t+ τ is analyzed.If A, B two systems, ar、bkIt is the measurement result of a certain physical quantity, PA
(ar)、PB(bk) it is a respectivelyrWith bkProbability, PAB(ar,bk) it is expressed as the joint probability of A and B.Then mutual information is defined as
Average is defined as follows：
If A and B are uncorrelated, IAB(ar,bk)=0.If A represents time series y (t), B represents time series y (t+
τ), then Average is：
τ~I (τ) curve is drawn, delay of the first minimum point corresponding time of curve as phase space reconfiguration is taken
Time τ.
(2) Cao methods determine the Embedded dimensions of reconstruct, and in m ties up phase space, i-th of phase point vector is denoted as
Xm(i)=[x (i), x (i+ τ) ..., x (i+ (m-1) τ)]T (5)
It defines first
Wherein,For Xm(i) nearest neighbor point, Xm+1(i) andIt is phase point Xm(i) andIt is tieed up in m+1
The continuation of phase space.Range formula uses maximum norm, i.e.,
Remember a2(i, m) is about the mean value of i
For the change situation of research E (m), definition
If time series is generated by attractor, when m is more than some m0When, E1(m) it no longer changes, then
m0It is exactly the minimum embedding dimension number of phase space reconstruction.
(3) small data sets arithmetic is used to calculate maximum Lyapunov exponent, process is as follows：
1. carrying out Fast Fourier Transform (FFT) to time series { x (t) | t=1,2 ..., n }, its average period of P is calculated；
2. the delay time T and Embedded dimensions m of time series are determined, if it is { X to reconstruct later phase spacei| i=1,
2 ..., K }, wherein K=n- (m-1) τ represents the number of phase point in phase space；
3. finding phase point XiThe corresponding nearest neighbor point in phase spaceSimultaneously it is limited to detach in short-term in phase space：
4. for any phase point X in phase spacei, acquire it and correspond to the distance d after j discrete time of adjoint point pairi(j)
For：
5. to each j, all ln (d are found outi(j)) average y (j), i.e.,：
Wherein, q is record di(j) ≠ 0 number.With least square fitting, then when the slope of regression straight line is exactly this
Between sequence maximum Lyapunov exponent λ1。
(4) phase space reconfiguration is carried out to sample：
Wherein, N=n- (m-1) τ is phase point sum.
Least square method supporting vector machine module 9：For establishing soft-sensing model：
In least square method supporting vector machine, it is assumed that given set of data samples is combined into (x1,y1),…(xn,yn),x∈Rn,
y∈R.Under nonlinear situation, transformation is introducedSample is mapped to a high-dimensional feature space from the input space.Find letter
Several parameters equivalents is in the following quadratic programming problem of solution：
Wherein, ξiFor slack variable, C is penalty factor.Introduce Lagrange multipliers αi, the Lagrange letters that are defined as follows
Number：
Enable L to variable w, b, αi, ξiPartial derivative be equal to zero：
Optimization problem is converted into solution linear equation：
Wherein y=[y1,...,yn]T, 1v=[1 ..., 1]T, α=[α1,...,αn]T, K=[kij]n×n,I is unit battle array.Finally obtain nonlinear model：
Improve gravitation search algorithm optimization module 10：Gravitation search algorithm is improved to Method Using Relevance Vector Machine module for using
Nuclear parameter σ1It optimizes, realization is as follows：
(1) algorithm initialization, all particles of random initializtion, each particle represent a candidate solution of problem.In a D
In the search space of dimension, it is assumed that have NP particle, the position for defining i-th of particle is
Set iteration termination condition, i.e. maximum iteration itermax。
(2) it in certain t moment, defines j-th of particle and acts on the gravitation size on i-th of particleFor：
Wherein, Maj(t) and Mpi(t) it is respectively the inertia mass for acting on particle j and the inertia mass for being applied particle i, Rij
(t) it is Euclidean distance between i-th of particle and j-th of particle, ε is the constant of a very little, and G (t) is the gravitation in t moment
Constant：
Wherein, α is descent coefficient, G0It is initial gravitational constant, itermaxIt is maximum iteration.
(3) inertia mass of particle is calculated according to the size of its fitness value, and inertia mass shows that more greatly it is closer
Optimal value, while meaning that the attraction of the particle is bigger, but its movement speed is slower.Assuming that gravitational mass and inertia mass
Equal, the quality of particle can go to update by operation rule appropriate, and more new algorithm is as follows：
Mai=Mpi=Mii=Mi, i=1,2 ..., NP (22)
Wherein, fiti(t) size in the fitness value of i-th of particle of t moment is represented.To solving minimum problems,
Best (t) and worst (t) are defined as follows：
To solving max problem, best (t) and worst (t) are defined as follows：
(4) assume that t moment acts on the total force F on i-th of particle in d dimensionsi d(t) it is equal to other all particles
To its sum of active force, calculation formula is as follows：
Wherein, randjIt is random number of the range in [0,1], Kbest is the preceding K grain for having at the beginning optimal adaptation degree
The set of son.
According to Newton's second law, acceleration of the t moment particle i in d dimensionsFor：
Wherein, Mi(t) be i-th of particle inertia mass.
(5) in next iteration, the new speed of particle is the summation of part present speed and its acceleration.Therefore, GSA
During interative computation each time, particle can all update its speed and position according to following formula：
vi(t+1)=ω vi(t)+c1ri1ai(t)+c2ri2(gbest-xi(t)) (31)
xi(t+1)=xi(t)+vi(t+1) (32)
Wherein, vi(t) it is speed of the particle i in the t times iteration, xi(t) it is particle i in the position of the t times iteration, ai(t)
It is acceleration of the particle i in the t times iteration, gbest is current optimal solution, ri1And ri2It is two between [0,1] random
Number；The inertial factor that ω is gradually reduced, c1And c2It is self adaptable acceleration coefficient, calculation formula is as follows：
Wherein, t and itermaxIt is current iteration number and maximum iteration respectively；c1i、c1f、c2iAnd c2fIt is constant, makes
Obtain c10.5, c is gradually decreased to from 2.522.5 are progressively increased to from 0.5.
(6) above step is repeated until reaching maximum iteration, selection fitness value is optimal to be solved as algorithm most
Excellent solution terminates algorithm and returns.
Model modification module 11 is used for the online updating of model, periodically offline analysis data is input in training set, more
New least square method supporting vector machine model.
The embodiment of the present invention is used for illustrating the present invention, rather than limits the invention, in the spirit of the present invention
In scope of the claims, to any modifications and changes that the present invention makes, protection scope of the present invention is both fallen within.
Claims (5)
1. a kind of high-precision propylene polymerization production process optimal soft survey instrument of chaos gunz optimizing, for being given birth to propylene polymerization
Melt index during production carries out hard measurement, which is characterized in that including field intelligent instrument, control station, DCS database, soft
Measuring unit, display instrument；Field intelligent instrument and control station measure easy survey variable and behaviour in propylene polymerization production process respectively
After making variable, it is stored in DCS database, after hard measurement unit carries out hard measurement processing to the data in DCS database, output
To display instrument；The hard measurement unit includes the data preprocessing module being sequentially connected, chaos analysis module, least square support
Vector machine module and improvement gravitation search algorithm optimization module, the mode input variable of DCS database input is by data prediction
After module is pre-processed, chaos analysis is carried out in chaos analysis module, then least square method supporting vector machine module into
Row modeling, finally optimizes model parameter by improvement gravitation search algorithm optimization module.
2. the high-precision propylene polymerization production process optimal soft survey instrument of chaos gunz optimizing according to claim 1,
It being characterized in that, the data preprocessing module pre-processes the mode input variable inputted from DCS database, with specific reference to
Following formula is standardized：
Wherein, mean indicates that the arithmetic mean of instantaneous value of each variable, std indicate the standard deviation of each variable,Indicate the value of input variable,
Subscript i indicates that ith detection, j indicate that jth ties up variable, x respectivelyijIndicate that the input variable after standardization, S indicate mode input
Variable.
3. the high-precision propylene polymerization production process optimal soft survey instrument of chaos gunz optimizing according to claim 1,
It is characterized in that, the chaos analysis module is according to chaology to sample X={ xijCarry out chaotic Property Analysis.It first has to determine
Two important parameters of phase space reconfiguration, i.e. delay time and Embedded dimensions.If the maximum Lyapunov exponent of system is more than
0, then the system must be chaos, it is possible to the chaotic characteristic of system is judged according to maximum Lyapunov exponent.
(1) mutual information method determines the delay time T of reconstruct, is under nonlinear case, to sometime t and another moment t+ τ
Information content between relationship analyzed.If A, B two systems, ar、bkIt is the measurement result of a certain physical quantity, PA(ar)、PB
(bk) it is a respectivelyrWith bkProbability, PAB(ar, bk) it is expressed as the joint probability of A and B.Then mutual information is defined as
Average is defined as follows：
If A and B are uncorrelated, IAB(ar, bk)=0.If A represents time series y (t), B represents time series y (t+ τ), then puts down
Equal mutual information is：
τ~I (τ) curve is drawn, delay time of the first minimum point corresponding time of curve as phase space reconfiguration is taken
τ。
(2) Cao methods determine the Embedded dimensions of reconstruct, and in m ties up phase space, i-th of phase point vector is denoted as
Xm(i)=[x (i), x (i+ τ) ..., x (i+ (m-1) τ)]T (5)
It defines first
Wherein,For Xm(i) nearest neighbor point, Xm+1(i) andIt is phase point Xm(i) andIt is tieed up in m+1 mutually empty
Between continuation.Range formula uses maximum norm, i.e.,
Remember a2(i, m) is about the mean value of i
For the change situation of research E (m), definition
If time series is generated by attractor, when m is more than some m0When, E1(m) it no longer changes, then m0Just
It is the minimum embedding dimension number of phase space reconstruction.
(3) small data sets arithmetic is used to calculate maximum Lyapunov exponent, process is as follows：
1. carrying out Fast Fourier Transform (FFT) to time series { x (t) | t=1,2 ..., n }, its average period of P is calculated；
2. the delay time T and Embedded dimensions m of time series are determined, if it is { X to reconstruct later phase spacei| i=1,2 ..., K },
Wherein K=n- (m-1) τ represents the number of phase point in phase space；
3. finding phase point XiThe corresponding nearest neighbor point in phase spaceSimultaneously it is limited to detach in short-term in phase space：
4. for any phase point X in phase spacei, acquire it and correspond to the distance d after j discrete time of adjoint point pairi(j) it is：
5. to each j, all ln (d are found outi(j)) average y (j), i.e.,：
Wherein, q is record di(j) ≠ 0 number.With least square fitting, then the slope of regression straight line is exactly the time series
Maximum Lyapunov exponent λ1。
(4) phase space reconfiguration is carried out to sample：
Wherein, N=n- (m-1) τ is phase point sum.
4. the high-precision propylene polymerization production process optimal soft survey instrument of chaos gunz optimizing according to claim 1,
It is characterized in that, the least square method supporting vector machine module is for establishing soft-sensing model.
In least square method supporting vector machine, it is assumed that given set of data samples is combined into (x1, y1) ... (xn, yn), x ∈ Rn, y ∈
R.Under nonlinear situation, transformation is introducedSample is mapped to a high-dimensional feature space from the input space.Find function
Parameters equivalent in solving following quadratic programming problem：
Wherein, ξiFor slack variable, C is penalty factor.Introduce Lagrange multipliers αi, the Lagrange functions that are defined as follows：
Enable L to variable w, b, αi, ξiPartial derivative be equal to zero：
Optimization problem is converted into solution linear equation：
Wherein y=[y1..., yn]T, 1v=[1 ..., 1]T, α=[α1..., αn]T, K=[kij]n×n,
I is unit battle array.Finally obtain nonlinear model：
5. the high-precision propylene polymerization production process optimal soft survey instrument of chaos gunz optimizing according to claim 1,
Be characterized in that, the improvement gravitation search algorithm optimization module be used for using improve gravitation search algorithm to least square support to
The parameter of amount machine module optimizes, and realization is as follows：
(1) algorithm initialization, all particles of random initializtion, each particle represent a candidate solution of problem.It is tieed up in a D
In search space, it is assumed that have NP particle, the position for defining i-th of particle is
Set iteration termination condition, i.e. maximum iteration itermax。
(2) it in certain t moment, defines j-th of particle and acts on the gravitation size on i-th of particleFor：
Wherein, Maj(t) and Mpi(t) it is respectively the inertia mass for acting on particle j and the inertia mass for being applied particle i, Rij(t)
The Euclidean distance between i-th of particle and j-th of particle, ε is the constant of a very little, G (t) be t moment gravitation it is normal
Number：
Wherein, α is descent coefficient, G0It is initial gravitational constant, itermaxIt is maximum iteration.
(3) inertia mass of particle is calculated according to the size of its fitness value, and inertia mass shows more greatly it closer to optimal
Value, while meaning that the attraction of the particle is bigger, but its movement speed is slower.Assuming that gravitational mass and inertia mass phase
Deng the quality of particle can go to update by operation rule appropriate, and more new algorithm is as follows：
Mai=Mpi=Mii=Mi, i=1,2 ..., NP (22)
Wherein, fiti(t) size in the fitness value of i-th of particle of t moment is represented.To solving minimum problems, best (t)
It is defined as follows with worst (t)：
To solving max problem, best (t) and worst (t) are defined as follows：
(4) assume that t moment acts on the total force F on i-th of particle in d dimensionsi d(t) it is equal to other all particles to it
The sum of active force, calculation formula is as follows：
Wherein, randjIt is random number of the range in [0,1], Kbest is the collection of the preceding K particle with optimal adaptation degree at the beginning
It closes.
According to Newton's second law, acceleration of the t moment particle i in d dimensionsFor：
Wherein, Mi(t) be i-th of particle inertia mass.
(5) in next iteration, the new speed of particle is the summation of part present speed and its acceleration.Therefore, GSA is every
In an iteration calculating process, particle can all update its speed and position according to following formula：
vi(t+1)=ω vi(t)+c1ri1ai(t)+c2ri2(gbest-xi(t)) (31)
xi(t+1)=xi(t)+vi(t+1) (32)
Wherein, vi(t) it is speed of the particle i in the t times iteration, xi(t) it is particle i in the position of the t times iteration, ai(t) it is grain
For sub- i in the acceleration of the t times iteration, gbest is current optimal solution, ri1And ri2It is two random numbers between [0,1]；ω
The inertial factor being gradually reduced, c1And c2It is self adaptable acceleration coefficient, calculation formula is as follows：
Wherein, t and itermaxIt is current iteration number and maximum iteration respectively；c1i、c1f、c2iAnd c2fIt is constant so that c1
0.5, c is gradually decreased to from 2.522.5 are progressively increased to from 0.5.
(6) above step is repeated until reaching maximum iteration, choose optimal solution of the optimal solution of fitness value as algorithm,
Terminate algorithm and returns.
The high-precision propylene polymerization production process optimal soft survey instrument of chaos gunz optimizing further includes model modification module,
For the online updating of model, periodically offline analysis data is input in training set, updates least square method supporting vector machine mould
Type.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810689702.3A CN108804851A (en) | 2018-06-28 | 2018-06-28 | A kind of high-precision propylene polymerization production process optimal soft survey instrument of chaos gunz optimizing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810689702.3A CN108804851A (en) | 2018-06-28 | 2018-06-28 | A kind of high-precision propylene polymerization production process optimal soft survey instrument of chaos gunz optimizing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108804851A true CN108804851A (en) | 2018-11-13 |
Family
ID=64072354
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810689702.3A Pending CN108804851A (en) | 2018-06-28 | 2018-06-28 | A kind of high-precision propylene polymerization production process optimal soft survey instrument of chaos gunz optimizing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108804851A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109634117A (en) * | 2018-12-14 | 2019-04-16 | 华北电力大学（保定） | A kind of information physical emerging system and its control method for denitration control |
| CN109711628A (en) * | 2018-12-28 | 2019-05-03 | 浙江大学 | A kind of propylene polymerization production process self-adapting multi-dimension forecast system of on-line correction |
| CN109726474A (en) * | 2018-12-28 | 2019-05-07 | 浙江大学 | A kind of multiple dimensioned forecast system of the propylene polymerization production process of on-line correction |
| CN109800888A (en) * | 2019-01-08 | 2019-05-24 | 浙江大学 | A kind of coalcutter online system failure diagnosis based on colony intelligence machine learning |
| CN109871865A (en) * | 2019-01-08 | 2019-06-11 | 浙江大学 | A kind of coalcutter online system failure diagnosis based on colony intelligence optimizing |
| CN113689922A (en) * | 2021-07-19 | 2021-11-23 | 南通大学 | Activated sludge process identification method based on improved chaotic gravity search algorithm |
| CN113759834A (en) * | 2020-06-05 | 2021-12-07 | 中国石油天然气股份有限公司 | Chaos multi-scale intelligent optimal propylene polymerization process measuring instrument |
| CN113764048A (en) * | 2020-06-05 | 2021-12-07 | 中国石油天然气股份有限公司 | Polypropylene production quality on-line measuring system |
| CN113764047A (en) * | 2020-06-05 | 2021-12-07 | 中国石油天然气股份有限公司 | Propylene polymerization quality on-line measuring system |
| CN113764047B (en) * | 2020-06-05 | 2024-06-28 | 中国石油天然气股份有限公司 | Propylene polymerization quality on-line measuring system |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103675011A (en) * | 2013-09-22 | 2014-03-26 | 浙江大学 | Soft industrial melt index measurement instrument and method of optimal support vector machine |
| CN103675006A (en) * | 2013-09-22 | 2014-03-26 | 浙江大学 | Least-squares-based industrial melt index soft measuring meter and method |
-
2018
- 2018-06-28 CN CN201810689702.3A patent/CN108804851A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103675011A (en) * | 2013-09-22 | 2014-03-26 | 浙江大学 | Soft industrial melt index measurement instrument and method of optimal support vector machine |
| CN103675006A (en) * | 2013-09-22 | 2014-03-26 | 浙江大学 | Least-squares-based industrial melt index soft measuring meter and method |
Non-Patent Citations (3)
| Title |
|---|
| MIAO ZHANG ET AL.: "Predicting industrial polymer melt index via incorporating chaotic characters into Chou"s general PseAAC", 《CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS》 * |
| ZHANG MIAO ET AL.: "A soft sensor for industrial melt index prediction based on evolutionary extreme learning machine", 《中国化学工程学报 英文版》 * |
| 胡茑庆: "《转子碰摩非线性行为与故障辨识的研究》", 30 April 2005, 国防科技大学出版社 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109634117A (en) * | 2018-12-14 | 2019-04-16 | 华北电力大学（保定） | A kind of information physical emerging system and its control method for denitration control |
| CN109711628A (en) * | 2018-12-28 | 2019-05-03 | 浙江大学 | A kind of propylene polymerization production process self-adapting multi-dimension forecast system of on-line correction |
| CN109726474A (en) * | 2018-12-28 | 2019-05-07 | 浙江大学 | A kind of multiple dimensioned forecast system of the propylene polymerization production process of on-line correction |
| CN109726474B (en) * | 2018-12-28 | 2021-04-09 | 浙江大学 | Online-correction multi-scale forecasting system for propylene polymerization production process |
| CN109800888A (en) * | 2019-01-08 | 2019-05-24 | 浙江大学 | A kind of coalcutter online system failure diagnosis based on colony intelligence machine learning |
| CN109871865A (en) * | 2019-01-08 | 2019-06-11 | 浙江大学 | A kind of coalcutter online system failure diagnosis based on colony intelligence optimizing |
| CN113759834A (en) * | 2020-06-05 | 2021-12-07 | 中国石油天然气股份有限公司 | Chaos multi-scale intelligent optimal propylene polymerization process measuring instrument |
| CN113764048A (en) * | 2020-06-05 | 2021-12-07 | 中国石油天然气股份有限公司 | Polypropylene production quality on-line measuring system |
| CN113764047A (en) * | 2020-06-05 | 2021-12-07 | 中国石油天然气股份有限公司 | Propylene polymerization quality on-line measuring system |
| CN113764047B (en) * | 2020-06-05 | 2024-06-28 | 中国石油天然气股份有限公司 | Propylene polymerization quality on-line measuring system |
| CN113689922A (en) * | 2021-07-19 | 2021-11-23 | 南通大学 | Activated sludge process identification method based on improved chaotic gravity search algorithm |
| CN113689922B (en) * | 2021-07-19 | 2022-05-17 | 南通大学 | Activated sludge process identification method based on improved chaotic gravitation search algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108804851A (en) | A kind of high-precision propylene polymerization production process optimal soft survey instrument of chaos gunz optimizing | |
| CN101315557B (en) | Propylene polymerization production process optimal soft survey instrument and method based on genetic algorithm optimization BP neural network | |
| CN101382801B (en) | Optimum soft measuring instrument based on EGA-optimized polymerization of propylene production process and method | |
| CN108803525A (en) | A kind of quick high-precision propylene polymerization production process optimal soft survey instrument of chaos | |
| CN108873838A (en) | A kind of propylene polymerization production process optimal soft survey instrument of gunz optimizing | |
| CN102393881B (en) | A kind of high-precision detecting method of real-time many sensing temperatures data fusion | |
| CN108982576A (en) | A kind of high-precision propylene polymerization production process optimal soft survey instrument of chaos | |
| CN109507888B (en) | Propylene polymerization production process optimal online forecasting system based on ensemble learning | |
| CN106991285A (en) | A kind of short-term wind speed multistep forecasting method and device | |
| CN103674778A (en) | Industrial melt index soft measuring meter and method based on RBF (radial basis function) particle swarm optimization | |
| CN103823430A (en) | Intelligent weighing propylene polymerization production process optimal soft measurement system and method | |
| CN103675011A (en) | Soft industrial melt index measurement instrument and method of optimal support vector machine | |
| CN103675006A (en) | Least-squares-based industrial melt index soft measuring meter and method | |
| CN108958181A (en) | A kind of propylene polymerization production process optimal soft survey instrument of agility | |
| CN103838142B (en) | Based on propylene polymerization production process optimal soft measuring system and the method for mixing optimizing | |
| CN109829197A (en) | The propylene polymerization production process optimal soft survey instrument of improved cuckoo optimizing | |
| CN109445398B (en) | Propylene polymerization production process optimal online forecasting system based on weighted combination learning | |
| CN109507889A (en) | The propylene polymerization production process optimal online forecasting system of convolutional neural networks | |
| CN103838955A (en) | Optimum soft measurement instrument and method for optimum mixing in propylene polymerization production process | |
| CN109856971A (en) | Propylene polymerization production process optimal online forecasting system based on gunz optimizing | |
| CN103839103A (en) | BP optimal forecasting system and method in propylene polymerization production process | |
| CN103955170A (en) | Propylene polymerization production process online forecasting system and method based on group intelligent optimization | |
| CN103838206A (en) | Optimal soft measurement meter and method in optimal BP multi-mode propylene polymerization production process | |
| CN103838209B (en) | Propylene polymerization production process adaptive optimal forecast system and method | |
| CN103675005A (en) | Soft industrial melt index measurement instrument and method for optimal fuzzy network |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181113 |
|
| RJ01 | Rejection of invention patent application after publication |