CN101986144B - Soft measurement method of reducing power quality index in process of calcinating lithopone - Google Patents

Soft measurement method of reducing power quality index in process of calcinating lithopone Download PDF

Info

Publication number
CN101986144B
CN101986144B CN201010527552XA CN201010527552A CN101986144B CN 101986144 B CN101986144 B CN 101986144B CN 201010527552X A CN201010527552X A CN 201010527552XA CN 201010527552 A CN201010527552 A CN 201010527552A CN 101986144 B CN101986144 B CN 101986144B
Authority
CN
China
Prior art keywords
calcining
reducing power
data
sampling instant
energy
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.)
Expired - Fee Related
Application number
CN201010527552XA
Other languages
Chinese (zh)
Other versions
CN101986144A (en
Inventor
杜启亮
莫鸿强
田联房
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
South China University of Technology SCUT
Original Assignee
South China University of Technology SCUT
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by South China University of Technology SCUT filed Critical South China University of Technology SCUT
Priority to CN201010527552XA priority Critical patent/CN101986144B/en
Publication of CN101986144A publication Critical patent/CN101986144A/en
Application granted granted Critical
Publication of CN101986144B publication Critical patent/CN101986144B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Investigating Or Analyzing Materials Using Thermal Means (AREA)

Abstract

The invention provides a soft measurement method of a reducing power quality index in the process of calcinating lithopone, comprising the following steps: (1) acquiring the historical detection values of calcination temperature data, calcination speed data and reducing power; (2) establishing a soft measurement module; (3) calculating the unit energy of the lithopone at the discharging moment in the current lithopone calcination process; (4) extracting two historical detection values of the reducing power at two reducing power sampling moments before the discharging moment in the current lithopone calcination process from a data storage system in step (1) to respectively serve as a first historical detection value and a second historical detection value; and (5) constituting the unit energy obtained in step (3) and the first historical detection value and the second historical detection value which are obtained in step (4) into input vectors, and inputting the input vectors into the soft measurement module in step (2) to obtain a predicated value of the reducing power at the discharging moment in the current calcination process to complete soft measurement. The method has the advantages that the module structure is simplified and the like.

Description

The flexible measurement method of lithopone calcination process reducing power quality index
Technical field
The present invention relates to calcination process survey control technology in the process industry, particularly a kind of flexible measurement method of lithopone calcination process reducing power quality index.
Background technology
Owing to lack appropriate sensor, the quality index of many products can not on-line measurement in the process industry at present, and needs the laboratory determined off-line.This mode wastes time and energy, and the more important thing is, can not include product quality indicator in closed-loop system and control automatically, and the quality index of stable prod then is the important goal of most of process flow industry process.But existing way is key variables of selecting to influence some on-line measurements of product quality indicator to be controlled automatically, and then stabilized product quality indirectly.But in the practical application, the factor that influences product quality indicator is a lot, and the selection of key variables is carried out according to workman's experience often; The strictness that lacks on data or the mechanism proves; In addition, because the complicacy of processing procedure, some process mechanism also is not very distinct; And receive a lot of various interference in the processing procedure, so differing, this indirect control method obtains good control effect surely.
To this situation; People accelerate the development of corresponding hard instrument on the one hand, set about on the other hand studying from the angle of soft instrument, through each significant variable that can survey in the searching process and the relation of product quality indicator; Set up mathematical model, so that predict immesurable variable with the variable that can survey.Mathematical model can be set up from mechanism, also can directly set up with data.The lithopone calcination process is the physical/chemical process of a complicacy, and its mechanism is very complicated, and setting up mechanism model needs a large amount of assumed condition, lacks practical applications property.Therefore directly carry out the foundation of model, have more practicality, but do not see the report of soft sensor modeling of the product quality indicator of lithopone calcination process so far from data.
In setting up the process of data model; Because the characteristic of its large time delay (material in kiln residence time above 2 hours), material in kiln during this period of time in the various parameters variations of rotary kiln all possibly exert an influence to its calcining situation, and then influence last product quality; So the procedure parameter in this is long-time all will be considered as the input variable of data model; Thus one, it is very huge that this model will become, and its training and using all is restricted.With the lithopone calcination process is example; Calcining heat and calcination time (through the reflection of calcining rotating speed) are the main factors that influences reducing power; Other factors are through influencing calcining heat remote effect reducing power, so a comparison reasonable method is to choose calcining heat and the input of calcining rotating speed as model.Consider that both are accumulationes to the influence of reducing power; Promptly concerning certain a material; Its calcining effect not only receives the influence of discharging calcining heat constantly and calcining rotating speed; And receive its different calcining heats constantly and the influence of calcining rotating speed in the whole process of calcined by rotary kiln section, so both different sample data constantly in this section process all should be considered as the input variable of model.With material time between calcining zone in kiln be 1.5 hours, the sampling period of manufacturing variables is to calculate in 1 minute, should select calcining heat and the input of calcining rotating speed (amounting to 180 variablees) as model in the material preceding 1.5 hours constantly.A large amount of input dimension like this, one side has increased the complexity of model, has increased the susceptibility to the data noise on the other hand.In addition, because the calcining change in rotational speed, material can change the residence time between calcining zone thereupon.Therefore need a kind of model variable system of selection that can optimize, both matching process mechanism can reduce the dimension of input variable again, so that the reduced data structure of models makes it can drop into practical application.
Summary of the invention
The shortcoming that primary and foremost purpose of the present invention is to overcome above-mentioned prior art provides a kind of flexible measurement method of lithopone calcination process reducing power quality index of reduced data model with not enough.
For reaching above-mentioned purpose, the present invention adopts following technical scheme:
The flexible measurement method of lithopone calcination process reducing power comprises the steps:
(1) gathers the calcining heat data in the said lithopone calcination process through the thermopair in the data acquisition system (DAS); Gather the calcining rotary speed data in the said lithopone calcination process through the speed pickup in the data acquisition system (DAS), said calcining heat data, calcining rotary speed data are saved to the data store system in the controlling computer through transmitter, programmable logic controller (PLC); The historical detected value of reducing power is gathered in manual work, and imports the historical detected value of said reducing power to data store system;
(2) select the sample time period; Constitute input vector through unit energy between the corresponding calcining zone of sample each reducing power sampling instant in the time period and the corresponding corresponding historical detected value of reducing power of reducing power sampling instant two reducing power sampling instants before of each unit energy; Carry out the model parameter training, set up the soft-sensing model of the single output of three inputs;
(3) the discharging unit energy constantly of the current calcination process of the said lithopone of calculating;
(4) in the data store system of said step (1), extract the historical detected value of the reducing power of two reducing power sampling instants before the discharging constantly of said current calcination process, be respectively historical detected value one and historical detected value two;
(5) the historical detected value one of the unit energy of said step (3) gained and step (4) gained constitutes input vector with historical detected value two; In the soft-sensing model of said input vector input step (2); Obtain the predicted value of the discharging reducing power constantly of current calcination process, accomplish its soft measurement.
Said step (2) comprising:
(2-1) select the sample time period, the sample time period that the said sample time period constituted for historical detected value several time periods more than acceptance value by reducing power in the data store system;
(2-2) carry out the energy of activation parameter identification, and the energy of activation parameter is saved in the data store system;
The energy of activation parameter that (2-3) extraction step (2-2) obtains from data store system, and according to the unit energy between the calcining zone of the said sample of energy of activation calculation of parameter each reducing power sampling instant correspondence in the time period;
(2-4) from the data store system of step (1), choose the historical detected value of the corresponding reducing power of each reducing power sampling instant of each time period in the sample time period;
(2-5) carry out the model parameter training with each historical detected value of step (2-4) and each unit energy of step (2-3); Wherein per two historical detected values constitute an input vector with each unit energy, obtain the soft-sensing model of the single output of three inputs through the model parameter training; Said per two historical detected values are the historical detected value of the corresponding reducing power of two reducing power sampling instants before the corresponding reducing power sampling instant of each unit energy.
Said step (2-2) specifically comprises carrying out the energy of activation parameter identification:
(2-2-1) each calcining heat data and calcining rotary speed data the sample of step (2-1) gained extracts between the corresponding calcining zone of each reducing power sampling instant in the time period in;
(2-2-2) each calcining heat data is averaged, obtain with reference to calcining heat T RefEach calcining rotary speed data is averaged, obtain with reference to calcining rotating speed D Ref
(2-2-3) respectively calcine rotary speed data and constitute rotating speed sample data vector Y=[y 1, y 2, y 3..., y m] T, i rotating speed sample data y wherein iSatisfy: y i=ln D i, D iBe i the corresponding calcining rotating speed of calcining rotary speed data;
Each calcining heat data constitutes temperature samples data matrix X=[[x 1, 1] T, [x 2, 1] T, [x 3, 1] T..., [x m, 1] T] T, i temperature samples data x wherein iSatisfy: x i=(1/T Ref-1/T i), T RefFor with reference to calcining heat, T iBe the corresponding calcining heats of i calcining heat data;
M is the sum of calcining heat data or calcining rotary speed data;
The rotating speed sample data vector sum temperature samples data matrix that (2-2-4) said step (2-2-3) is obtained calculates through least square method, obtains result vector M; Said result vector M satisfies: M T=[X TX] -1[X TY], wherein Y is a rotating speed sample data vector, X is the temperature samples data matrix; First element of said result vector M is exactly the energy of activation parameter of being asked.
The calcining heat data of said step (2-2-1) or the calcining rotary speed data add up to 150~200.
Unit energy between the calcining zone of each reducing power sampling instant correspondence of said step (2-3) calculates and comprises:
(2-3-1) initialization: iterations is set to zero; The local variable of the raw material pace in the said lithopone calcination process is set to zero; The energy of reducing power sampling instant is set to zero, and the initial value of calcining rotating speed sampling instant or calcining heat sampling instant is the reducing power sampling instant;
(2-3-2) from data store system, extract corresponding calcining heat of this reducing power sampling instant and calcining rotating speed;
(2-3-3) the local variable l (n) of the raw material pace in the said lithopone calcination process of renewal; Local variable l (n) satisfies: l (n)=l (n-1)+V (t) * Δ t; Local variable when wherein l (n) is this iteration; Local variable when l (n-1) is last iteration, Δ t is calcining rotating speed sampling period or the calcining heat sampling period, when V (t) is constantly lithopone calcining of t along the radial velocity of rotary kiln;
T calcines the sampling instant of previous calcining rotating speed or the calcining heat sampling instant of rotating speed sampling instant or calcining heat sampling instant constantly for this; Satisfy: t=t0-n Δ t; Wherein, N is an iterations, and Δ t is calcining rotating speed sampling period or calcining heat sampling period, and t0 is this calcining rotating speed sampling instant or calcining heat sampling instant;
(2-3-4) calculate a calcining rotating speed sampling period or interior energy increment f (n) of calcining heat sampling period, energy increment f (n) satisfies:
f(n)=exp(R*(1/T ref-1/T(t)))*Δt,
Wherein R is the energy of activation parameter of step (2-2-4) gained, T RefFor step (2-2-2) gained with reference to calcining heat, T (t) is a t calcining heat constantly;
T calcines the sampling instant of previous calcining rotating speed or the calcining heat sampling instant of rotating speed sampling instant or calcining heat sampling instant constantly for this; Satisfy t=t0-n Δ t, wherein, n is an iterations; Δ t is the sampling period, and t0 is this calcining rotating speed sampling instant or calcining heat sampling instant;
(2-3-5) calculate the calcining energy UE (n) of this calcining rotating speed sampling instant or calcining heat sampling instant according to energy increment f (n); Calcining energy UE (n) satisfies: UE (n)=UE (n-1)+f (n); Wherein UE (n-1) is this calcining rotating speed sampling instant or the previous calcining rotating speed sampling instant of calcining heat sampling instant or the calcining energy of calcining heat sampling instant, and f (n) is an energy increment;
(2-3-6) judge whether to accomplish all calcining energy calculation between the corresponding calcining zone of this reducing power sampling instant, if then carry out next step, otherwise iterations n is from increasing 1, execution in step (2-3-3)~(2-3-5);
(2-3-7) confirm unit energy UE between the corresponding calcining zone of said each reducing power sampling instant, unit energy UE satisfies: UE=UE (n-1)/t Ref, wherein UE (n-1) is the UE (n-1) in the step (2-3-5), t RefFor with reference to calcination time;
Said with reference to calcination time t RefSatisfy: t Ref=L DS/ (k f* D Ref), L wherein DSBe the length between calcining zone; k fScale-up factor for calcining rotating speed and raw material pace; D RefBe reference calcining rotating speed.
The discharging unit energy constantly that said step (3) is calculated the current calcination process of said lithopone specifically comprises:
(3-1) the energy of activation parameter of extraction step (2-2) identification gained from data store system;
(3-2) said discharging is the reducing power sampling instant constantly, and execution in step (2-3-1)~(2-3-7) obtains this discharging unit energy constantly.
Said soft-sensing model is the least square method supporting vector machine model.
The said artificial sampling period of gathering the historical detected value of reducing power is 1~2 hour, and the collection period of the calcining heat data in the said lithopone calcination process, calcining rotary speed data is 0.5~5 minute.
Said data acquisition system (DAS) comprises thermopair, speed pickup, transmitter, programmable logic controller (PLC) and controlling computer; Be provided with data store system in the controlling computer; Said thermopair is external in the inwall of calcination rotary kiln; The drive motor of the external calcination rotary kiln of speed pickup, thermopair, speed pickup, transmitter are connected with programmable logic controller (PLC) respectively, and programmable logic controller (PLC) is connected with data store system in the controlling computer.
Compared with prior art, the present invention has following advantage and beneficial effect:
(1) cuts down the consumption of energy, improves burning quality:, solved the problem of the definite difficulty of being brought along with the calcining rotation speed change changes because of the variable that influences reducing power of model input variable through the soft sensor modeling of computer realization to the product reducing power quality index of lithopone calcination process; And then realize on-line prediction to reducing power numerical value, and reached the purpose of stable burning quality, to cutting down the consumption of energy, reduce environmental pollution simultaneously, also there is positive role the aspect of increasing economic efficiency.
(2) simplified model structure, raising speed: the energy changing information of utilizing material in the calcination process; Significantly reduce the input dimension of data model, and then in process, simplified model structure with the SVMs modeling; Improved the efficient of modeling; Accelerated training speed, improved the noise resisting ability of model, made model practical more.
Description of drawings
Fig. 1 is the overview flow chart of the inventive method.
Fig. 2 is the process flow diagram that the step (2) of method shown in Figure 1 is set up soft-sensing model.
Fig. 3 is the process flow diagram of step (2-2) the energy of activation parameter identification of method shown in Figure 2.
Fig. 4 is the process flow diagram that step (2-3) unit energy of method shown in Figure 2 calculates.
Embodiment
Below in conjunction with embodiment and accompanying drawing the present invention is described in further detail, but embodiment of the present invention is not limited thereto.
Embodiment 1
As shown in Figure 1, the flexible measurement method of this lithopone calcination process reducing power comprises the steps:
(1) gathers the calcining heat data in the said lithopone calcination process through the thermopair in the data acquisition system (DAS); Gather the calcining rotary speed data in the said lithopone calcination process through the speed pickup in the data acquisition system (DAS), said calcining heat data, calcining rotary speed data are saved to the data store system in the controlling computer through transmitter, programmable logic controller (PLC); The historical detected value of reducing power is gathered in manual work, and imports the historical detected value of said reducing power to data store system;
(2) select the sample time period; Constitute input vector through unit energy between the corresponding calcining zone of sample each reducing power sampling instant in the time period and the corresponding corresponding historical detected value of reducing power of reducing power sampling instant two reducing power sampling instants before of each unit energy; Carry out the model parameter training, set up the soft-sensing model of the single output of three inputs;
(3) the discharging unit energy constantly of the current calcination process of the said lithopone of calculating;
(4) in the data store system of said step (1), extract the historical detected value of the reducing power of two reducing power sampling instants before the discharging constantly of said current calcination process, be respectively historical detected value one and historical detected value two;
(5) the historical detected value one of the unit energy of said step (3) gained and step (4) gained constitutes input vector with historical detected value two; In the soft-sensing model of said input vector input step (2); Obtain the predicted value of the discharging reducing power constantly of current calcination process, accomplish its soft measurement.
As shown in Figure 2, said step (2) comprising:
(2-1) select the sample time period, the sample time period that the said sample time period constituted for historical detected value several time periods more than acceptance value by reducing power in the data store system;
(2-2) carry out the energy of activation parameter identification, and the energy of activation parameter is saved in the data store system;
The energy of activation parameter that (2-3) extraction step (2-2) obtains from data store system, and according to the unit energy between the calcining zone of the said sample of energy of activation calculation of parameter each reducing power sampling instant correspondence in the time period;
(2-4) from the data store system of step (1), choose the historical detected value of the corresponding reducing power of each reducing power sampling instant of each time period in the sample time period;
(2-5) carry out the model parameter training with each historical detected value of step (2-4) and each unit energy of step (2-3); Wherein per two historical detected values constitute an input vector with each unit energy, obtain the soft-sensing model of the single output of three inputs through the model parameter training; Said per two historical detected values are the historical detected value of the corresponding reducing power of two reducing power sampling instants before the corresponding reducing power sampling instant of each unit energy.
As shown in Figure 3, said step (2-2) specifically comprises carrying out the energy of activation parameter identification:
(2-2-1) each calcining heat data and calcining rotary speed data the sample of step (2-1) gained extracts between the corresponding calcining zone of each reducing power sampling instant in the time period in;
(2-2-2) each calcining heat data is averaged, obtain with reference to calcining heat T RefEach calcining rotary speed data is averaged, obtain with reference to calcining rotating speed D Ref
(2-2-3) respectively calcine rotary speed data and constitute rotating speed sample data vector Y=[y 1, y 2, y 3..., y m] T, i rotating speed sample data y wherein iSatisfy: y i=ln D i, D iBe i the corresponding calcining rotating speed of calcining rotary speed data;
Each calcining heat data constitutes temperature samples data matrix X=[[x 1, 1] T, [x 2, 1] T, [x 3, 1] T..., [x m, 1] T] T, i temperature samples data x wherein iSatisfy: x i=(1/T Ref-1/T i), T RefFor with reference to calcining heat, T iBe the corresponding calcining heats of i calcining heat data;
M is the sum of calcining heat data or calcining rotary speed data, and m gets 150;
The rotating speed sample data vector sum temperature samples data matrix that (2-2-4) said step (2-2-3) is obtained calculates through least square method, obtains result vector M; Said result vector M satisfies: M T=[X TX] -1[X TY], wherein Y is a rotating speed sample data vector, X is the temperature samples data matrix; First element of said result vector M is exactly the energy of activation parameter of being asked.
The calcining heat data of said step (2-2-1) or the calcining rotary speed data add up to 150.
As shown in Figure 4, the unit energy between the calcining zone of each reducing power sampling instant correspondence of said step (2-3) calculates and comprises:
(2-3-3) initialization: iterations is set to zero; The local variable of the raw material pace in the said lithopone calcination process is set to zero; The energy of reducing power sampling instant is set to zero, and the initial value of calcining rotating speed sampling instant or calcining heat sampling instant is the reducing power sampling instant; That is: n=0, l (0)=0, UE (0)=0;
(2-3-2) from data store system, extract corresponding calcining heat of this reducing power sampling instant and calcining rotating speed;
(2-3-3) the local variable l (n) of the raw material pace in the said lithopone calcination process of renewal; Local variable l (n) satisfies: l (n)=l (n-1)+V (t) * Δ t; Local variable when wherein l (n) is this iteration; Local variable when l (n-1) is last iteration, Δ t is calcining rotating speed sampling period or the calcining heat sampling period, when V (t) is constantly lithopone calcining of t along the radial velocity of rotary kiln;
T calcines the sampling instant of previous calcining rotating speed or the calcining heat sampling instant of rotating speed sampling instant or calcining heat sampling instant constantly for this; Satisfy: t=t0-n Δ t; Wherein, N is an iterations, and Δ t is calcining rotating speed sampling period or calcining heat sampling period, and t0 is this calcining rotating speed sampling instant or calcining heat sampling instant;
(2-3-4) calculate a calcining rotating speed sampling period or interior energy increment f (n) of calcining heat sampling period, energy increment f (n) satisfies:
f(n)=exp(R*(1/T ref-1/T(t)))*Δt,
Wherein R is the energy of activation parameter of step (2-2-4) gained, T RefFor step (2-2-2) gained with reference to calcining heat, T (t) is a t calcining heat constantly;
T calcines the sampling instant of previous calcining rotating speed or the calcining heat sampling instant of rotating speed sampling instant or calcining heat sampling instant constantly for this; Satisfy t=t0-n Δ t, wherein, n is an iterations; Δ t is the sampling period, and t0 is this calcining rotating speed sampling instant or calcining heat sampling instant;
(2-3-5) calculate the calcining energy UE (n) of this calcining rotating speed sampling instant or calcining heat sampling instant according to energy increment f (n); Calcining energy UE (n) satisfies: UE (n)=UE (n-1)+f (n); Wherein UE (n-1) is this calcining rotating speed sampling instant or the previous calcining rotating speed sampling instant of calcining heat sampling instant or the calcining energy of calcining heat sampling instant, and f (n) is an energy increment;
(2-3-6) judge whether to accomplish all calcining energy calculation between the corresponding calcining zone of this reducing power sampling instant, if (that is: local variable l (n) is greater than the length L between calcining zone DS), then carry out next step, otherwise iterations n is from increasing 1, execution in step (2-3-3)~(2-3-5);
(2-3-7) confirm unit energy UE between the corresponding calcining zone of said each reducing power sampling instant, unit energy UE satisfies: UE=UE (n-1)/t Ref, wherein UE (n-1) is the UE (n-1) in the step (2-3-5), t RefFor with reference to calcination time;
Said with reference to calcination time t RefSatisfy: t Ref=L DS/ (k f* D Ref), L wherein DSBe the length between calcining zone; k fScale-up factor for calcining rotating speed and raw material pace; D RefBe reference calcining rotating speed.
The discharging unit energy constantly that said step (3) is calculated the current calcination process of said lithopone specifically comprises:
(3-1) the energy of activation parameter of extraction step (2-2) identification gained from data store system;
(3-2) said discharging is the reducing power sampling instant constantly, and execution in step (2-3-1)~(2-3-7) obtains this discharging unit energy constantly.
Said soft-sensing model is the least square method supporting vector machine model.
The said artificial sampling period of gathering the historical detected value of reducing power is 1 hour, and the collection period of the calcining heat data in the said lithopone calcination process, calcining rotary speed data is 0.5 minute.
Said data acquisition system (DAS) comprises thermopair, speed pickup, transmitter, programmable logic controller (PLC) and controlling computer; Be provided with data store system in the controlling computer; Said thermopair is external in the inwall of calcination rotary kiln; The drive motor of the external calcination rotary kiln of speed pickup, thermopair, speed pickup, transmitter are connected with programmable logic controller (PLC) respectively, and programmable logic controller (PLC) is connected with data store system in the controlling computer.
Embodiment 2
Present embodiment except that following characteristics other characteristics with embodiment 1: the said artificial sampling period of gathering the historical detected value of reducing power is 2 hours, and the collection period of the calcining heat data in the said lithopone calcination process, calcining rotary speed data is 5 minutes.
In the said step (2-1), the sample time period that the said sample time period constituted for ten time periods of historical detected value more than acceptance value by reducing power in the data store system;
In the said step (2-2-3), m is the sum of calcining heat data or calcining rotary speed data, and m gets 200.
Embodiment 3
Present embodiment except that following characteristics other characteristics with embodiment 1: the said artificial sampling period of gathering the historical detected value of reducing power is 1.5 hours, and the collection period of the calcining heat data in the said lithopone calcination process, calcining rotary speed data is 3 minutes.
In the said step (2-1), the sample time period that the said sample time period constituted for 20 time periods of historical detected value more than acceptance value by reducing power in the data store system;
In the said step (2-2-3), m is the sum of calcining heat data or calcining rotary speed data, and m gets 160.
The foregoing description is a preferred implementation of the present invention; But embodiment of the present invention is not restricted to the described embodiments; Other any do not deviate from change, the modification done under spirit of the present invention and the principle, substitutes, combination, simplify; All should be the substitute mode of equivalence, be included within protection scope of the present invention.

Claims (6)

1. the flexible measurement method of lithopone calcination process reducing power is characterized in that, comprises the steps:
(1) gathers the calcining heat data in the said lithopone calcination process through the thermopair in the data acquisition system (DAS); Gather the calcining rotary speed data in the said lithopone calcination process through the speed pickup in the data acquisition system (DAS), said calcining heat data, calcining rotary speed data are saved to the data store system in the controlling computer through transmitter, programmable logic controller (PLC); The historical detected value of reducing power is gathered in manual work, and imports the historical detected value of said reducing power to data store system;
(2) select the sample time period; Constitute input vector through unit energy between the corresponding calcining zone of sample each reducing power sampling instant in the time period and the corresponding corresponding historical detected value of reducing power of reducing power sampling instant two reducing power sampling instants before of each unit energy; Carry out the model parameter training, set up the soft-sensing model of the single output of three inputs;
(3) the discharging unit energy constantly of the current calcination process of the said lithopone of calculating;
(4) in the data store system of said step (1), extract the historical detected value of the reducing power of two reducing power sampling instants before the discharging constantly of said current calcination process, be respectively historical detected value one and historical detected value two;
(5) the historical detected value one of the unit energy of said step (3) gained and step (4) gained constitutes input vector with historical detected value two; In the soft-sensing model of said input vector input step (2); Obtain the predicted value of the discharging reducing power constantly of current calcination process, accomplish its soft measurement;
Said step (2) comprising:
(2-1) select the sample time period, the sample time period that the said sample time period constituted for historical detected value several time periods more than acceptance value by reducing power in the data store system;
(2-2) carry out the energy of activation parameter identification, and the energy of activation parameter is saved in the data store system;
The energy of activation parameter that (2-3) extraction step (2-2) obtains from data store system, and according to the unit energy between the calcining zone of the said sample of energy of activation calculation of parameter each reducing power sampling instant correspondence in the time period;
(2-4) from the data store system of step (1), choose the historical detected value of the corresponding reducing power of each reducing power sampling instant of each time period in the sample time period;
(2-5) carry out the model parameter training with each historical detected value of step (2-4) and each unit energy of step (2-3); Wherein per two historical detected values constitute an input vector with each unit energy, obtain the soft-sensing model of the single output of three inputs through the model parameter training; Said per two historical detected values are the historical detected value of the corresponding reducing power of two reducing power sampling instants before the corresponding reducing power sampling instant of each unit energy;
Said step (2-2) specifically comprises carrying out the energy of activation parameter identification:
(2-2-1) each calcining heat data and calcining rotary speed data the sample of step (2-1) gained extracts between the corresponding calcining zone of each reducing power sampling instant in the time period in;
(2-2-2) each calcining heat data is averaged, obtain with reference to calcining heat T RefEach calcining rotary speed data is averaged, obtain with reference to calcining rotating speed D Ref
(2-2-3) respectively calcine rotary speed data and constitute rotating speed sample data vector Y=[y 1, y 2, y 3..., y m] T, i rotating speed sample data y wherein iSatisfy: y i=ln D i, D iBe i the corresponding calcining rotating speed of calcining rotary speed data;
Each calcining heat data constitutes temperature samples data matrix X=[[x 1, 1] T, [x 2, 1] T, [x 3, 1] T..., [x m, 1] T] T, i temperature samples data x wherein iSatisfy: x i=(1/T Ref-1/T i), T RefFor with reference to calcining heat, T iBe the corresponding calcining heats of i calcining heat data;
M is the sum of calcining heat data or calcining rotary speed data;
The rotating speed sample data vector sum temperature samples data matrix that (2-2-4) said step (2-2-3) is obtained calculates through least square method, obtains result vector M; Said result vector M satisfies: M T=[X TX] -1[X TY], wherein Y is a rotating speed sample data vector, X is the temperature samples data matrix; First element of said result vector M is exactly the energy of activation parameter of being asked;
Unit energy between the calcining zone of each reducing power sampling instant correspondence of said step (2-3) calculates and comprises:
(2-3-1) initialization: iterations is set to zero; The local variable of the raw material pace in the said lithopone calcination process is set to zero; The energy of reducing power sampling instant is set to zero, and the initial value of calcining rotating speed sampling instant or calcining heat sampling instant is the reducing power sampling instant;
(2-3-2) from data store system, extract corresponding calcining heat of this reducing power sampling instant and calcining rotating speed;
(2-3-3) the local variable l (n) of the raw material pace in the said lithopone calcination process of renewal; Local variable l (n) satisfies: l (n)=l (n-1)+V (t) * Δ t; Local variable when wherein l (n) is this iteration; Local variable when l (n-1) is last iteration, Δ t is calcining rotating speed sampling period or the calcining heat sampling period, when V (t) is constantly lithopone calcining of t along the radial velocity of rotary kiln;
T calcines the sampling instant of previous calcining rotating speed or the calcining heat sampling instant of rotating speed sampling instant or calcining heat sampling instant constantly for this; Satisfy: t=t0-n Δ t; Wherein, N is an iterations, and Δ t is calcining rotating speed sampling period or calcining heat sampling period, and t0 is this calcining rotating speed sampling instant or calcining heat sampling instant;
(2-3-4) calculate a calcining rotating speed sampling period or interior energy increment f (n) of calcining heat sampling period, energy increment f (n) satisfies:
f(n)=exp(R*(1/T ref-1/T(t)))*Δt,
Wherein R is the energy of activation parameter of step (2-2-4) gained, T RefFor step (2-2-2) gained with reference to calcining heat, T (t) is a t calcining heat constantly;
T calcines the sampling instant of previous calcining rotating speed or the calcining heat sampling instant of rotating speed sampling instant or calcining heat sampling instant constantly for this; Satisfy t=t0-n Δ t, wherein, n is an iterations; Δ t is the sampling period, and t0 is this calcining rotating speed sampling instant or calcining heat sampling instant;
(2-3-5) calculate the calcining energy UE (n) of this calcining rotating speed sampling instant or calcining heat sampling instant according to energy increment f (n); Calcining energy UE (n) satisfies: UE (n)=UE (n-1)+f (n); Wherein UE (n-1) is this calcining rotating speed sampling instant or the previous calcining rotating speed sampling instant of calcining heat sampling instant or the calcining energy of calcining heat sampling instant, and f (n) is an energy increment;
(2-3-6) judge whether to accomplish all calcining energy calculation between the corresponding calcining zone of this reducing power sampling instant, if then carry out next step, otherwise iterations n is from increasing 1, execution in step (2-3-3)~(2-3-5);
(2-3-7) confirm unit energy UE between the corresponding calcining zone of said each reducing power sampling instant, unit energy UE satisfies: UE=UE (n-1)/t Ref, wherein UE (n-1) is the UE (n-1) in the step (2-3-5), t RefFor with reference to calcination time;
Said with reference to calcination time t RefSatisfy: t Ref=L DS/ (k f* D Ref), L wherein DSBe the length between calcining zone; k fScale-up factor for calcining rotating speed and raw material pace; D RefBe reference calcining rotating speed.
2. the flexible measurement method of lithopone calcination process reducing power according to claim 1 is characterized in that, the calcining heat data of said step (2-2-1) or the calcining rotary speed data add up to 150~200.
3. the flexible measurement method of lithopone calcination process reducing power according to claim 1 is characterized in that, the discharging unit energy constantly that said step (3) is calculated the current calcination process of said lithopone comprises:
(3-1) the energy of activation parameter of extraction step (2-2) identification gained from data store system;
(3-2) said discharging is the reducing power sampling instant constantly, and execution in step (2-3-1)~(2-3-7) obtains this discharging unit energy constantly.
4. the flexible measurement method of lithopone calcination process reducing power according to claim 1 is characterized in that, said soft-sensing model is the least square method supporting vector machine model.
5. the flexible measurement method of lithopone calcination process reducing power according to claim 1; It is characterized in that; The said artificial sampling period of gathering the historical detected value of reducing power is 1~2 hour, and the collection period of the calcining heat data in the said lithopone calcination process, calcining rotary speed data is 0.5~5 minute.
6. the flexible measurement method of lithopone calcination process reducing power according to claim 1; It is characterized in that; Said data acquisition system (DAS) comprises thermopair, speed pickup, transmitter, programmable logic controller (PLC) and controlling computer; Be provided with data store system in the controlling computer, said thermopair is external in the inwall of calcination rotary kiln, the drive motor of the external calcination rotary kiln of speed pickup; Thermopair, speed pickup, transmitter are connected with programmable logic controller (PLC) respectively, and programmable logic controller (PLC) is connected with data store system in the controlling computer.
CN201010527552XA 2010-10-31 2010-10-31 Soft measurement method of reducing power quality index in process of calcinating lithopone Expired - Fee Related CN101986144B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201010527552XA CN101986144B (en) 2010-10-31 2010-10-31 Soft measurement method of reducing power quality index in process of calcinating lithopone

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201010527552XA CN101986144B (en) 2010-10-31 2010-10-31 Soft measurement method of reducing power quality index in process of calcinating lithopone

Publications (2)

Publication Number Publication Date
CN101986144A CN101986144A (en) 2011-03-16
CN101986144B true CN101986144B (en) 2012-07-18

Family

ID=43710509

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201010527552XA Expired - Fee Related CN101986144B (en) 2010-10-31 2010-10-31 Soft measurement method of reducing power quality index in process of calcinating lithopone

Country Status (1)

Country Link
CN (1) CN101986144B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108872308B (en) * 2018-09-28 2021-04-02 山东建筑大学 Detection and research method for ground heating wood floor heat slow-release rule

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1563902A (en) * 2004-04-08 2005-01-12 上海交通大学 Soft measuring meter moduling method based on supporting vector machine
CN101533265A (en) * 2009-04-10 2009-09-16 华南理工大学 Feedback control method for calcination process in rotary kiln

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1563902A (en) * 2004-04-08 2005-01-12 上海交通大学 Soft measuring meter moduling method based on supporting vector machine
CN101533265A (en) * 2009-04-10 2009-09-16 华南理工大学 Feedback control method for calcination process in rotary kiln

Also Published As

Publication number Publication date
CN101986144A (en) 2011-03-16

Similar Documents

Publication Publication Date Title
CN102629104B (en) Calcination predictive control system and method for rotary cement kiln
CN103336107B (en) Soft measurement method for f-CaO content of cement clinker
CN103217445B (en) Extruding and drawing device for in-situ structure detection in combination with scattering of X rays and experimental method thereof
CN113404502B (en) Shield hob abrasion monitoring device and method based on ballast piece morphology
CN102636516B (en) Circular stability testing system of multichannel solid-liquid phase change material
CN103399779A (en) Semi-physical simulation platform and simulation method thereof used in cement clinker calcining process
Sun et al. Optimized throughput improvement of assembly flow line with digital twin online analytics
CN101986144B (en) Soft measurement method of reducing power quality index in process of calcinating lithopone
CN201476905U (en) Neural network PID temperature controlled thermocouple automatic verification system
CN109583075A (en) Permanent magnet direct-drive wind-force machine military service quality evaluating method based on temperature parameter prediction
CN101458497A (en) Concurrency controlled cooking device by employing multi motor position rings and speed rings
CN105807711A (en) Experiment system of coal mining machine intelligent control for achieving multi-information fusion
CN107015541A (en) The flexible measurement method being combined based on mutual information and least square method supporting vector machine
CN108452895B (en) Energy-saving control method for ball mill
CN103900904B (en) A kind of neutron in-situ mechanical loading system
CN206256053U (en) A kind of turning-over device automatic control system based on touch-screen
CN107766880A (en) The flexible measurement method of Fermentation by Photosynthetic Bacteria process key parameter based on BA LSSVM
CN103971015A (en) Gypsum slurry stirring time analysis method
CN101339442A (en) Power cable conductor temperature automatic following and control method
CN108152629B (en) Method for drawing voltage tolerance curve of thermal power plant
CN103675010A (en) Supporting-vector-machine-based industrial melt index soft measuring meter and method
CN104699039A (en) Expert control method for burning double-chamber lime furnace
CN115774912A (en) Mechanism and data combined driving TBM hob abrasion prediction method
CN103439964A (en) On-line updating system and method for OCSVM monitoring model
CN105759604B (en) Dual-core servo drive system and servo motor online modeling method

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20120718

Termination date: 20141031

EXPY Termination of patent right or utility model