Search Images Maps Play YouTube News Gmail Drive More »
Sign in
Screen reader users: click this link for accessible mode. Accessible mode has the same essential features but works better with your reader.

Patents

  1. Advanced Patent Search
Publication numberCN103610227 B
Publication typeGrant
Application numberCN 201310659839
Publication dateApr 15, 2015
Filing dateDec 9, 2013
Priority dateDec 9, 2013
Also published asCN103610227A
Publication number201310659839.1, CN 103610227 B, CN 103610227B, CN 201310659839, CN-B-103610227, CN103610227 B, CN103610227B, CN201310659839, CN201310659839.1
Inventors彭辉, 顾云峰, 王丹, 刘明月, 李立, 阮文杰, 魏吉敏, 肖玉娇
Applicant中南大学
Export CitationBiBTeX, EndNote, RefMan
External Links: SIPO, Espacenet
Cut tobacco dryer head and tail section process variable optimizing control method
CN 103610227 B
Abstract  translated from Chinese
本发明公开了一种烘丝机头尾段工艺变量优化控制方法,依据烘丝过程头尾段筒温、风温、排潮风门等工艺变量的历史数据,采用三次函数作为径向基函数的Cubic-RBF-ARX模型对烘丝动态特性进行建模;所建模型具有自调节能力,能反映不同模式下的入口流量以及入口水分的变化对出口水分的影响,可根据头尾段不同模式的入口流量及入口水分的变化来预测未来出口水分的变化情况;根据所建模型对各工艺变量进行优化设定,可使头尾段叶丝出口水分的控制达到较好的效果。 The present invention discloses a nose tail section Drying optimal control of process variables, historical data Drying process according to the head and tail sections barrel temperature, air temperature, moisture exhaust damper and other process variables, use three times as a function of radial basis function Cubic-RBF-ARX model tobacco drying modeling dynamic characteristics; the model has the ability to self-regulate, to reflect the impact of different modes of inlet flow rate and inlet water outlet water changes, according to the head and tail segments of different modes Changes inlet water inlet flow and to predict future changes in the export of water; according to the model for each process variables to optimize settings, leave the head and tail sections leaves wire outlet moisture control to achieve better results. 本发明方法综合考虑了来料量与各输入变量间的动态特性,可以更有效地克服来料流量和水分变化对烘丝过程头尾段的影响,适用于不同模式下叶丝入口流量与入口水分时的头尾段控制。 The method of the present invention to consider the quantity and the dynamic characteristics of the respective input variables, can be more effectively overcome the incoming flow and water Drying process changes on the head and tail sections for the inlet flow rate and inlet cut tobacco in different modes control of water when the head and tail sections.
Claims(3)  translated from Chinese
1. 一种烘丝机头尾段工艺变量优化控制方法,其特征在于,该方法为: 1) 根据烘丝机的运行流程,建立烘丝过程中叶丝入口流量、入口水分、筒温、风温、 排潮风门、出口水分的时序关系,同时根据烘丝过程干头阶段无叶丝出口水分检测值、 干尾阶段无叶丝入口流量与入口水分检测值的特点,采用三次函数作为径向基函数的Cubic-RBF-ARX模型,分别建立烘丝过程干头阶段与干尾阶段的Cubic-RBF-ARX模型; 2) 根据烘丝机头尾段的历史运行数据,采用结构化非线性参数优化方法分别优化烘丝过程干头阶段与干尾阶段的Cubic-RBF-ARX模型; 3) 依据优化的烘丝过程干头阶段与干尾阶段的Cubic-RBF-ARX模型,采用双S型函数描述干头阶段的排潮风门、风温、筒温的最优输入曲线;采用阶跃函数描述干头阶段的入口流量的最优输入曲线;采用指数函数描述干尾阶段排潮风门、风温、筒温和筒体电机频率的最优输入曲线; 4) 采用列维布格奈奎尔特方法,通过使优化的干头阶段与干尾阶段的Cubic-RBF-ARX 模型计算出的出口水分预测值与出口水分设定值的误差最小,寻找出烘丝过程干头阶段与干尾阶段的最优输入曲线的参数,以适应来料情况的变化,减少干尾阶段的干料量; 所述步骤1)中,烘丝机干头阶段Cubic-RBF-ARX模型为: A drying wire head Endian optimal control of process variables, characterized in that the method is: 1) Drying machine according to the operating procedures, the establishment of the mid-wire entrance Drying process flow, inlet water, barrel temperature, wind temperature, moisture exhaust damper, exit Moisture timing relationships while Drying process according to the first stage without leaves dry silk exports moisture detection value, dry end stage bladeless wire inlet water flow rate and inlet characteristics of the detected value, as a function of radial cubic Cubic-RBF-ARX model basis functions, namely the establishment of the first stage of dry Drying process with Cubic-RBF-ARX model tail dry stage; 2) Drying the nose tail section of historical operating data, structured nonlinear parameters Optimization Methods Drying process optimization dry first stage with Cubic-RBF-ARX model tail dry stage; 3) Drying process based on an optimized dry head stage with Cubic-RBF-ARX model tail dry stage, dual S-type function Description dry moisture exhaust damper head stage, air temperature, cylinder temperature optimum input curve; optimal input curve describes the step function using dry inlet flow head stage; the exponential function describes the dry end stage row tide damper, air temperature , mild cylinder motor cylinder optimal input frequency curve; 4) 列维布格奈奎 Stewart method, by optimizing the dry first stage with Cubic-RBF-ARX model to calculate the dry end stage outlet water forecast error value and the set value of the minimum outlet water, to find out the parameters dry head Drying process input stage and optimal curve dry end stages, in order to adapt to changes in incoming cases, reducing the dry end stage of dry quantity; the Step 1) Drying Dryer first stage Cubic-RBF-ARX model is:
Figure CN103610227BC00021
Figure CN103610227BC00031
其中,yH(tH)表示烘丝机干头阶段Cubic-RBF-ARX模型的出口水分; <(产),Oii),"f(〇, "fC <(产)分别表示干头阶段Cubic-RBF-ARX模型的排潮风门开度、风温、筒温、入口流量及入口水分;XH(t Hl)为入口流量和入口水分的状态变量;npH,nqH,d H和m H均表示干头阶段Cubic-RBF-ARX模型的阶次;$',Zf 分别为干头阶段Cubic-RBF-ARX模型输出项与输入项的RBF神经网络的中心; 为干头阶段Cubic_RBF_ARX模型的标量权系数;II · I If表示矩阵的Frobenius范数;ξ H(tH)是干头阶段Cubic-RBF-ARX模型的建模误差,为高斯白噪声;Tci11为烘丝机干头阶段Cubic-RBF-ARX模型建模采样时间,T 从有入口流量检测值到有入口水分检测值的时间,T2S从有入口水分检测值到有出口水分检测值的时间,T3为从有入口水分检测值到烘丝筒入口的时间,T 4为叶丝在烘丝筒烘干的时间; 所述步骤1)中,烘丝机干尾阶段Cubic-RBF-ARX模型为: Wherein, yH (tH) indicates outlet water Drying Dryer first stage Cubic-RBF-ARX model; <(yield), Oii), "f (square," fC <(production), respectively, the first stage of dry Cubic-RBF row tide throttle opening -ARX model, air temperature, cylinder temperature, inlet flow rate and inlet water; XH (t Hl) for the inlet water flow rate and inlet state variables; npH, nqH, d H and m H have said dry head Stage Cubic-RBF-ARX model order; $ ', Zf were dry first stage Cubic-RBF-ARX model output term RBF nerve center and the entry of the network; dry first stage scalar weighting coefficient Cubic_RBF_ARX model; II · I If the matrix represent Frobenius norm; ξ H (tH) is dry first stage modeling error Cubic-RBF-ARX model is Gaussian white noise; Tci11 to bake dry silk machine head stage Cubic-RBF-ARX model built mold sampling time, T flow from inlet to inlet water detection value detected value of time, T2S water from an inlet to an outlet moisture detection value detected value of time, T3 from the inlet moisture detection value to the drying wire tube inlet time, T 4 leaf silk yarn in the drying cylinder drying time; the step 1), Drying machine dry end stage Cubic-RBF-ARX model is:
Figure CN103610227BC00032
其中: among them:
Figure CN103610227BC00041
其中,yT(tT)表示烘丝机干尾阶段Cubic-RBF-ARX模型的出口水分; < (iK (〇,<(/), :),分别表示干尾阶段Cub i c-RBF-ARX模型的筒温、热风风温、排潮风门开度、入口流量、入口水分及筒体电机频率;XT(tT-l)为热风风温和筒体电机频率的状态变量;npT,nqT,dT和m %表示干尾阶段Cubic-RBF-ARX模型的阶次; ζί;ν,ζί;"分别为干尾阶段Cubic-RBF-ARX模型输出项与输入项的RBF神经网络的中心; ,<^(),' 6Clg 为干尾阶段Cubic-RBF_ARX 模型的标量权系数; f(tT)是干尾阶段Cubic-RBF-ARX模型建模误差,为高斯白噪声;IV为烘丝机干尾阶段Cubic-RBF-ARX模型建模采样时间; 所述步骤2)中,烘丝机干头阶段Cubic-RBF-ARX模型优化如下: Wherein, yT (tT) indicates outlet water Drying machine dry end stage Cubic-RBF-ARX model; <(iK (square, <(/), :), respectively, dry end stage Cub i c-RBF-ARX model The barrel temperature, hot air temperature, moisture exhaust throttle opening, inlet flow rate, inlet water and the cylinder motor frequency; XT (tT-l) for the state of mild hot wind cylinder motor frequency variable; npT, nqT, dT and m % represents dry end stage Cubic-RBF-ARX model order; ζί; ν, ζί; "RBF neural networks were central stem end stage Cubic-RBF-ARX model output items and input items; <^ () , '6Clg scalar weights of dry end stage Cubic-RBF_ARX model; f (tT) is a dry end stage Cubic-RBF-ARX modeling error, Gaussian white noise; IV for the drying wire machine dry end stage Cubic-RBF -ARX modeling sampling time; the step 2), Drying machine dry first stage Cubic-RBF-ARX model optimization as follows:
Figure CN103610227BC00042
其中,,(P)是烘丝机干头阶段出口水分的实际值,^V)是在实际输入作用下,由烘丝机干头阶段Cubic-RBF-ARX模型计算出的出口水分的预测值; 奸广Λ气二",<二4» 严=1,.·.,叫=1,,..,》产}为烘丝机干头II ,0 f!,_/ ,0 " J 阶段Cubic-RBF-ARX模型的线性参数;祀Z(,,为烘丝机干头阶段Cubic-RBF-ARX模型的非线性参数;Nh为烘丝机干头阶段Cubic-RBF-ARX模型建模数据长度; 烘丝机干尾阶段Cubic-RBF-ARX模型优化如下: Where ,, (P) is the actual value of the first phase Drying Dryer export water, ^ V) under the actual input role, the predicted value by the tobacco drying machine dry first stage Cubic-RBF-ARX model to calculate the export of water ; rape wide Λ air two "<= 4» Yan = 1, *, called = 1 ,, ..,. "production} is baked dry silk machine head II, 0 f, _ /, 0" J stage! Linear Parameter Cubic-RBF-ARX model; worship Z (,, nonlinear parameter Drying Dryer first stage Cubic-RBF-ARX model; Nh is baked dry silk machine head stage Cubic-RBF-ARX modeling data length; Drying machine dry end stage Cubic-RBF-ARX model optimization as follows:
Figure CN103610227BC00051
其中,,(〇是烘丝机干尾过程中出口水分的实际值;是在实际输入作用下,由烘丝机干尾阶段Cubic-RBF-ARX模型计算出的出口水分的预测值; Θ卜<|4'0,扮乂,气=1,…,n/7ry 阶段Cubic-RBF-ARX模型的线性参数,θ「ν = (Zf 'Zf卞=1,· ·,"/}为烘丝机干尾阶段Cubic-RBF-ARX模型的非线性参数,Nt为烘丝机干头阶段Cubic-RBF-ARX模型建模数据长度; 所述步骤3)中: 用于描述烘丝机干头阶段排潮风门、风温、筒温的最优输入曲线的双S型函数表达式为: Where ,, (square is the actual value of the tobacco drying machine dry end of the process water outlet; under actual input role, the predicted value by the tobacco drying machine dry end stage Cubic-RBF-ARX model to calculate the export of water; Θ BU <| Linear Parameter 4'0, play qe, gas = 1, ..., n / 7ry stage Cubic-RBF-ARX model, θ "ν = (Zf 'Zf Bian = 1, ·," /} is drying wire nonlinear parameters machine dry end stage Cubic-RBF-ARX model, Nt is Drying Dryer first stage Cubic-RBF-ARX modeling data length; the step 3): used to describe the first phase Drying Dryer row tide damper, air temperature, cylinder temperature optimum input curve double S-type function expression is:
Figure CN103610227BC00052
其中,ts为输入的时间,单位为s ; λ λ4, \5分别为双S型函数的起点、转折点及终点值;λ2, \6分别为双S型函数的两条对称轴中心位置;λ 3, λ 7分别为双S型函数上升或下降的速度;λ 3, λ 7大于〇时表示S型函数上升,λ 3, λ 7小于〇时表示S型函数下降;c = 1,2, 3, Usl (ts)是排潮风门的设定值;Us2 (ts)是风温的设定值;Us3(ts)是筒温的设定值; 用于描述烘丝机干头阶段入口流量的最优输入曲线的阶跃函数表达式为: Wherein, ts is the time input unit is s; λ λ4, \ 5 were starting double S-type function, turning point and end point value; λ2, \ 6 were two double S-type function symmetry axis center position; λ 3, λ 7 double S-shaped function respectively increase or decrease the speed; λ 3, λ 7 is greater than the square when said S-shaped function rising, λ 3, λ 7 when less than the square represents S type function decline; c = 1,2, 3, Usl (ts) is set to a value of moisture exhaust damper; Us2 (ts) is the setpoint air temperature; Us3 (ts) is the barrel temperature setpoint; Drying Dryer is used to describe the first phase inlet flow The optimal input step function curve expression:
Figure CN103610227BC00053
其中,tT为输入的时间,单位为S ; K K2, K3分别为阶跃函数的上升速度、上升时间与终值; 所述步骤3)中,用于描述干尾阶段排潮风门、风温、筒温和筒体电机频率的最优输入曲线的指数函数的表达式为: Wherein, tT to time input, the unit is S; K K2, K3 are rising speed of a step function, the rise time and the final value; the step 3), is used to describe the end of stage dry moisture exhaust damper, air temperature expression exponential curve cylinder mild optimal input frequency of the motor cylinder for:
Figure CN103610227BC00054
P = 1,2, 3, 4 式中Uzl (tz)、Uz2 (tz)、Uz3 (tz)、Uz4 (tz)分别表示干尾阶段排潮风门、风温、筒温和筒体电机频率的最优输入曲线。 P = 1,2, 3, 4 where Uzl (tz), Uz2 (tz), Uz3 (tz), Uz4 (tz), respectively, end stage dry moisture exhaust damper, air temperature, cylinder mild cylinder motor frequency most Excellent input curve.
2. 根据权利要求1所述的烘丝机头尾段工艺变量优化控制方法,其特征在于,所述步骤4)中,将烘丝机干头阶段优化设定曲线Mt s)、Ut (tT)代入所述烘丝机干头阶段Cubic-RBF-ARX模型的输入变量<(Π ,》f(0,(0, 4(0中,得到烘丝机干头阶段Cubic-RBF-ARX模型计算出的出口水分预测值产(Ο;通过使干头阶段Cubic-RBF-ARX模型计算出的出口水分预测值浐(O与出口水分设定值y srt(ta)的误差eH(ta)最小,即采用列M 维布格奈奎尔特方法求解优化问题ininJ = ,寻找出干头阶段排潮风门、风温、筒Av. κσ 温的输入曲线的参数λχ和入口流量输入曲线的参数KK 2,K 3;其中,X = 1,2,…,7 ;g =1,2, 3 ;M是干头阶段持续的时间。 Drying according to claim 1, wherein the head end section of process variables optimization control method, wherein said step 4), the first stage Drying Dryer optimal setting curve Mt s), Ut (tT ) into the wire machine dry head stage Cubic-RBF-ARX model of the drying of the input variables <(Π, "f (0, (0, 4 (0 to give Drying Dryer first stage Cubic-RBF-ARX model calculations the predictive value of the export of water production (Ο; dry head stage by Cubic-RBF-ARX model to calculate the predicted value of the export of water Chan (O and outlet water setpoint y srt (ta) error eH (ta) minimum, That takes a column M 维布格奈奎 Stewart method for solving optimization problems ininJ =, find out the first stage of dry moisture exhaust damper, parameters and inlet flow input curve λχ air temperature, cylinder Av. κσ temperature curve of the input parameters KK 2 , K 3; wherein, X = 1,2, ..., 7; g = 1,2, 3; M is the first phase duration dry.
3. 根据权利要求1所述的烘丝机头尾段工艺变量优化控制方法,其特征在于,所述步骤4)中,将烘丝机干尾阶段优化设定曲线Uzp(tz)代入所述烘丝机干尾阶段Cubic-RBF-ARX 模型的输入变量"「(广),"[(广),"【#),中,得到烘丝机干尾阶段的Cubic-RBF-ARX模型计算出的出口水分预测值尹Y);通过使干尾阶段的Cubic-RBF-ARX模型计算出的出口水分预测值/f)与出口水分设定值y' srt(tb)的误差eT(tb)最小,即采用列维布格奈奎尔特方M 法求解优化问题min./' = (^),寻找出干尾阶段排潮风门、风温、筒温和筒体电机频率aps k=\ 最优输入曲线的参数a pg;其中,g = 1,2, 3 ;M'是干尾阶段持续时间。 3. Drying process variables nose tail section 1 of the optimization control method as claimed in claim wherein said step 4), the drying wire machine dry end stage optimal setting curve Uzp (tz) is substituted into the input variables Drying Dryer Last stage Cubic-RBF-ARX model "" (Canton), "[(Canton)," [#), to give Cubic-RBF-ARX model tobacco drying machine dry end stage calculated The predictive value of the export of water Yin Y); by the end of stage dry Cubic-RBF-ARX model to calculate the predicted value of the export of water / f) with the outlet water setpoint y 'srt (tb) error eT (tb) Minimum , which uses 列维布格奈奎 M Stewart square method to solve the optimization problem min. / '= (^), to find out the dry end stage row tide damper, air temperature, cylinder mild cylinder motor frequency aps k = \ BEST Enter the curve parameters a pg; wherein, g = 1,2, 3; M 'is a dry end stage duration.
Description  translated from Chinese
一种烘丝机头尾段工艺变量优化控制方法 One kind of tobacco drying process variables nose Endian Optimal Control Method

技术领域 TECHNICAL FIELD

[0001] 本发明涉及烘丝机头尾段工艺变量优化控制方法。 [0001] The present invention relates to a drying wire head Endian optimal control of process variables.

背景技术 Background technique

[0002] 烘丝过程是香烟制丝生产中最重要的一道加工工序,它主要是通过对叶丝进行加热干燥,降低叶丝的含水率,使烘烤后叶丝的含水率、温度均匀一致,并控制在一定的数值范围内,以满足生产工艺要求。 [0002] Drying process of silk production of cigarettes is the most important one processing step, it is mainly through the leaves of the wire is heated and dried to reduce the moisture content of the leaf silk, silk leaves so after baking the moisture content, temperature uniform and controlled within a certain range of values, in order to meet the production process. 烘丝的工艺流程主要分为预热、干头、中间以及干尾过程四个部分。 Drying of the preheating process is divided into four parts, dry head, middle and end of the dry process. 在干头阶段,叶丝入口流量不断增加,但无叶丝出口水分的检测值,难以进行反馈控制,容易造成干头阶段出口水分控制品质差、干料多;在干尾阶段,由于叶丝入口流量骤然减少,而烘丝筒具有较大热容,筒壁内部温度难以按规定的速率下降等问题,也容易造成干尾阶段出口水分控制性能低且干料多。 In the first stage of dry leaves wire inlet flow is increasing, but there was no detection of leaf silk export value of water, it is difficult for feedback control, easily lead to dry first stage of export control poor quality water, dry material and more; in the dry end of the stage, because the leaf silk inlet flow suddenly decreased, while drying wire tube having a large heat capacity, the internal temperature of the cylinder wall is difficult to fall at the rate specified and other issues, it is likely to cause lower end stage dry moisture control export performance and more dry material. 因此,"干头干尾"是目前烘丝过程出口水分控制的难点所在。 Therefore, "dry head dry tail" is currently exporting Drying process Moisture control difficulty lies.

[0003] 现有的干头干尾过程控制方法主要有: [0003] The existing dry dry head end process control methods are:

[0004] (1)利用进入和输出烘丝机的物料和干燥介质作为热质平衡对象建立数学模型, 结合前馈PID调节筒温的控制方式。 [0004] (1) the use of incoming and outgoing materials and Drying machine drying medium heat and mass balance of the object as a mathematical model, combining feedforward PID regulator tube temperature control. 但前馈数学模型仅考虑了进料的含水率和流量,并没有考虑热风温度等其他对出口水分有重要影响的因素,不能完全反应真实过程,造成头尾段烘丝机出口水分波动大,需要操作人员进行人工干预,对于头尾段不同模式下的、具有不同入口流量和入口水分的来料难以获得满意的控制效果。 But feedforward mathematical models consider only feed moisture content and flow, and does not consider other important factors that impact on exports of hot air temperature, moisture, can not fully reflect the true process, resulting in head and tail section Drying outlet moisture fluctuation, require the operator manual intervention, at the head and tail sections for different modes with different inlet flow rate and inlet water control incoming difficult to obtain satisfactory results.

[0005] (2)在上述前馈控制的基础上,在头尾段增加蒸喷加湿装置对头尾料施加蒸汽水来提高头尾料的含水率,以降低干料量。 [0005] (2) On the basis of the above-mentioned feedforward control on the head and tail segment increased steam spray humidifying device for the head and tail material is applied to the head and tail of steam feed water to raise the moisture content to reduce the quantity of dry. 但此方法仅对叶丝表层进行加湿,仅提高了叶丝表层湿度,仍然会造成烟丝内在质量的降低,且增加了出口水分控制的难度与稳定性。 However, this method only humidify leaf silk surface, only to improve the wire surface humidity leaf, still cause reduce the intrinsic quality of the tobacco, and increase the difficulty of the outlet water control and stability.

[0006] (3)通过多次试验、寻求最佳头尾阶段热风温度值和调整排潮阀门开度等工艺参数来减少干料量。 [0006] (3) through several tests, looking for the best value and the beginning and end stage air temperature moisture exhaust valve opening adjustment of process parameters to reduce the quantity of dry. 此方法缺乏自调节能力,无法保证对于不同模式下具有不同入口流量和入口水分的来料时,该组工艺参数均为最优设定值; This method is the lack of self-regulation, can not be guaranteed for different modes with different inlet water flow rate and inlet incoming, the set of process parameters are optimal setpoint;

[0007] (4)在PID控制策略的基础上,将模糊控制的思想应用到烘丝机水分控制中。 [0007] (4) On the basis of PID control strategy, based on the ideas of fuzzy control applied to tobacco drying machine moisture control.

[0008] 仅仅用单纯的二维模糊控制器来解决烘丝过程出口水分的控制问题仍然无法获得最优的工艺参数设定值,而且对于不同模式下的入口流量与入口水分的变化,还需对模糊控制规则表进行调整,这对工业生产带来不便。 [0008] using only a simple two-dimensional fuzzy controller to solve control problems Drying process outlet water still can not get the optimum process parameter settings, and change the inlet flow rate and inlet water for different modes, the need to fuzzy control rule table adjusted industrial production this inconvenience.

发明内容 SUMMARY OF THE INVENTION

[0009] 本发明所要解决的技术问题是,针对现有技术不足,提供一种烘丝机头尾段工艺变量优化控制方法,使干头阶段叶丝出口水分尽可能快地上升、并快速到达稳定状态,使干尾阶段叶丝出口水分尽可能缓慢地下降,从而有效地减少头尾段的干料量,提高烘丝过程的控制性能;更有效地克服来料流量和水分变化对烘丝过程头尾段的影响,避免人工整定输入工艺变量参数的不便。 [0009] The technical problem to be solved by the present invention is for the deficiencies of the prior art, provides a tobacco drying head Endian optimal control of process variables, and the dry first stage leaf silk exports rose water as quickly as possible, and quickly reach steady state, and the dry end stage leaf silk exports declined as the water slowly, thus effectively reducing the head and tail sections of dry quantity, improve the control performance Drying process; more effectively overcome the incoming flow and water change on Drying head and tail sections influence the process to avoid manual tuning inconvenient process variable input parameters.

[0010] 为解决上述技术问题,本发明所采用的技术方案是:一种烘丝机头尾段工艺变量优化控制方法,该方法为: [0010] In order to solve the above problems, the technical aspect of the present invention is used is: a silk head Endian bake optimized control of process variables, the method is:

[0011] 1)根据烘丝机的运行流程,建立烘丝过程中叶丝入口流量、入口水分、筒温、风温、排潮风门、出口水分的时序关系,同时根据烘丝过程干头阶段无叶丝出口水分检测值、 干尾阶段无叶丝入口流量与入口水分检测值的特点,采用三次函数作为径向基函数的Cubic-RBF-ARX模型,分别建立烘丝过程干头阶段与干尾阶段的Cubic-RBF-ARX模型; [0011] 1) Drying machine according to the operating procedures, the establishment of the mid-wire entrance Drying process flow, inlet water, barrel temperature, air temperature, moisture exhaust damper, exit Moisture timing relationships while Drying process according to the first stage without dry Ye silk exports moisture detection value, dry end stage bladeless wire entrance to the inlet water flow characteristics of the detected value, cubic function as the radial basis function Cubic-RBF-ARX model, were established first stage dry Drying process and dry tail Cubic-RBF-ARX model stage;

[0012] 2)根据烘丝机头尾段的历史运行数据,采用结构化非线性参数优化方法分别优化烘丝过程干头阶段与干尾阶段的Cubic-RBF-ARX模型; [0012] 2) Drying the nose tail section of historical operating data, structured nonlinear parameter optimization method Drying process optimization were dry head stage with Cubic-RBF-ARX model tail dry stage;

[0013] 3)依据优化的烘丝过程干头阶段与干尾阶段的Cubic-RBF-ARX模型,采用双S型函数描述干头阶段的排潮风门、风温、筒温的最优输入曲线;采用阶跃函数描述干头阶段的入口流量的最优输入曲线;采用指数函数描述干尾阶段排潮风门、风温、筒温和筒体电机频率的最优输入曲线; [0013] 3) Drying process based on an optimized dry first stage with Cubic-RBF-ARX model tail dry stage, the use of moisture exhaust damper double S-type function descriptions dry first stage, the optimal input curves air temperature, cylinder temperature ; the use of a step function described in the first stage of dry inlet flow optimum input curve; exponential function describes the dry end stage row tide damper, air temperature, cylinder mild cylinder motor optimal input frequency curve;

[0014] 4)采用列维布格奈奎尔特方法,通过使优化的干头阶段与干尾阶段的Cubic-RBF-ARX模型计算出的出口水分预测值与出口水分设定值的误差最小,寻找出烘丝过程干头阶段与干尾阶段的最优输入曲线的参数,以适应来料情况的变化,减少干尾阶段的干料量。 [0014] 4) 列维布格奈奎 Holt method, error by Cubic-RBF-ARX model allows optimization of the dry end of the first stage and the dry stage of the calculated predicted value and export water outlet water setpoint minimum find out the parameters of the first stage of dry Drying process and optimal input curves dry end stages, in order to adapt to changes in incoming cases, reducing the dry end stage of dry quantity.

[0015] 所述步骤1)中,烘丝机干头阶段Cubic-RBF-ARX模型为: [0015] the step 1), the first stage of drying silk machine dry Cubic-RBF-ARX model is:

Figure CN103610227BD00081

Figure CN103610227BD00091

[0019] 其中,yH(tH)表示烘丝机干头阶段Cubic-RBF-ARX模型的出口水分; [0019] where, yH (tH) indicates outlet water Drying Dryer first stage Cubic-RBF-ARX model;

Figure CN103610227BD00092

分别表示干头阶段Cubic-RBF-ARX模型的排潮风门开度、风温、筒温、入口流量及入口水分;XH(tH-l)为入口流量和入口水分的状态变量;np H, nqH,dH和mH均表示干头阶段Cubic-RBF-ARX模型的阶次;Zf,Zf分别为干头阶段 Respectively ranked first dry tide throttle opening stage Cubic-RBF-ARX model, air temperature, cylinder temperature, inlet flow rate and inlet water; XH (tH-l) for the inlet water flow rate and inlet state variables; np H, nqH , dH and mH expressed order of dry head stage Cubic-RBF-ARX model; Zf, Zf were dry first stage

Figure CN103610227BD00093

Cub i c-RBF-ARX模型输出项与输入项的RBF神经网络的中. Cub i c-RBF-ARX model output term and the entry of RBF neural network.

Figure CN103610227BD00094

为干头阶段Cubic-RBF-ARX模型的标量权系数;M · I |F表示矩阵的Frobenius范数;|H(tH)是干头阶段Cubic-RBF-ARX模型的建模误差,为高斯白噪声;T QH为烘丝机干头阶段Cubic-RBF-ARX模型建模采样时间,T1为从有入口流量检测值到有入口水分检测值的时间,T 2为从有入口水分检测值到有出口水分检测值的时间,T 3为从有入口水分检测值到烘丝筒入口的时间,! Scalar weights of dry first stage Cubic-RBF-ARX model; M · I | F represents Frobenius matrix norm; | H (tH) is the first stage of the modeling error dry Cubic-RBF-ARX model is Gaussian white noise; T QH is baked dry silk machine head stage Cubic-RBF-ARX modeling sampling time, T1 flow from inlet to inlet water detection value detected value of time, T 2 from an inlet moisture detection value to have the export value of moisture detection time, T 3 from the detected value of the water inlet tube entrance Drying time! \为叶丝在烘丝筒烘干的时间。 \ Leaf silk yarn in the drying cylinder drying time.

[0020] 所述步骤1)中,烘丝机干尾阶段Cubic-RBF-ARX模型为: [0020] the step 1), Drying machine dry end stage Cubic-RBF-ARX model is:

[0021] [0021]

Figure CN103610227BD00101

「00241 其中,vT (tT)衷示烘玆机干尾阶段Cubic-RBF-ARX模型的出口水分; "00241 which, vT (tT) co shown bake hereby machine dry end stage Cubic-RBF-ARX model of export of water;

Figure CN103610227BD00102

分别表示干尾阶段Cub i c-RBF-ARX模型的筒温、热风风温、排潮风门开度、入口流量、入口水分及筒体电机频率;XT(tT_l)为热风风温和筒体电机频率的状态变量;npT,nqT,dT和m %表示干尾阶段Cubic-RBF-ARX模型的阶次; ,<;«分别为干尾阶段cub i c-RBF-ARX模型输出项与输入项的RBF神经网络的中心; Respectively cylinder temperature dry end stage Cub i c-RBF-ARX model, hot air temperature, moisture exhaust throttle opening, inlet flow rate, inlet water and the cylinder motor frequency; XT (tT_l) for the hot air temperature and the cylinder motor frequency The state variables; npT, nqT, dT and m% represents dry end stage Cubic-RBF-ARX model order; <; «were dry end stage cub i c-RBF-ARX model output term and the entry of RBF Centre neural network;

Figure CN103610227BD00103

为干尾阶段Cubic-RBF-ARX模型的标量权系数; f(tT)是干尾阶段Cubic-RBF-ARX模型建模误差,为高斯白噪声;IV为烘丝机干尾阶段Cubic-RBF-ARX模型建模采样时间。 Scalar weights of dry end stage Cubic-RBF-ARX model; f (tT) is a dry end stage Cubic-RBF-ARX modeling error, Gaussian white noise; IV for the drying wire machine dry end stage Cubic-RBF- ARX modeling sampling time.

[0025] 所述步骤2)中,烘丝机干头阶段Cubic-RBF-ARX模型优化如下: [0025] the step 2), Drying machine dry first stage Cubic-RBF-ARX model optimization as follows:

Figure CN103610227BD00111

[0027] 其中,严(严)是烘丝机干头阶段出口水分的实际值,产(严)是在实际输入作用下,由烘丝机干头阶段Cub i c-RBF-ARX模型计算出的出口水分的预测值; [0027] where Yan (Yan) is the actual value of the tobacco drying machine dry first stage outlet water yield (Yan) is a role in the actual input, calculated from the first stage Drying Dryer Cub i c-RBF-ARX model out The predictive value of exports of water;

Figure CN103610227BD00112

%烘丝机干头阶段Cubic-RBF-ARX模型的线性参数; % Drying Dryer first phase linear parameters Cubic-RBF-ARX model;

Figure CN103610227BD00113

为烘丝机干头阶段Cubic-RBF-ARX模型的非线性参数;Nh为烘丝机干头阶段Cubic-RBF-ARX模型建模数据长度。 Nonlinear parameter Drying Dryer first stage Cubic-RBF-ARX model; Nh is Drying Dryer first stage Cubic-RBF-ARX modeling data length.

[0028] 烘丝机干尾阶段Cubic-RBF-ARX模型优化如下: [0028] Drying machine dry end stage Cubic-RBF-ARX model optimization as follows:

Figure CN103610227BD00114

[0030] 其中,/(r)是烘丝机干尾过程中出口水分的实际值;r(r〇是在实际输入作用下,由烘丝机干尾阶段Cub i C-RBF-ARX模型计算出的出口水分的预测值; [0030] where, / (r) is the actual value of the tobacco drying machine dry end of the process water outlet; r (r〇 under actual input role, calculated by drying silk machine dry end stage Cub i C-RBF-ARX model export of water out of the predicted value;

Figure CN103610227BD00115

为烘丝机干尾阶段Cubic-RBF-ARX模型的线性参数, Drying machine for the dry end stage linear parameters Cubic-RBF-ARX model,

Figure CN103610227BD00116

为烘丝机干尾阶段Cubic-RBF-ARX模型的非线性参数;Nt为烘丝机干头阶段Cubic-RBF-ARX模型建模数据长度。 Nonlinear parameter Drying Dryer Last stage Cubic-RBF-ARX model; Nt is Drying Dryer first stage Cubic-RBF-ARX modeling data length.

[0031] 所述步骤3)中: [0031] The step 3):

[0032] 用于描述烘丝机干头阶段排潮风门、风温、筒温的最优输入曲线的双S型函数表达式为: Optimal input double S curve function expression [0032] is used to describe the first stage of drying silk machine dry moisture exhaust damper, air temperature, cylinder temperature is:

Figure CN103610227BD00117

[0034] 其中,ts为输入的时间,单位为s ; λ i,λ4, \5分别为双S型函数的起点、转折点及终点值;λ2, λ6分别为双S型函数的两条对称轴中心位置;λ 3, λ 7分别为双S型函数上升或下降的速度;λ 3, λ 7大于0时表示S型函数上升,λ 3, λ 7小于0时表示S型函数下降; c=l,2, 3, Usl(ts)是排潮风门的设定值;Us2(ts)是风温的设定值;U s3(ts)是筒温的设定值。 [0034] where, ts is the time input, the unit is s; λ i, λ4, \ 5, respectively, as a starting point the double S-type function, turning point and end point value; λ2, λ6 were two double S-type function symmetry axis center position; λ 3, λ 7 double S function respectively increase or decrease the speed; λ 3, λ 7 is greater than zero when said S-shaped function rising, λ 3, λ 7 is less than 0, said S-shaped function decline; c = l, 2, 3, Usl (ts) is set to a value of moisture exhaust damper; Us2 (ts) is the setpoint air temperature; U s3 (ts) is the barrel temperature setpoint.

[0035] 用于描述烘丝机干头阶段入口流量的最优输入曲线的阶跃函数表达式为: Step function expression [0035] is used to describe the optimum input curve Drying Dryer first stage inlet flow is as follows:

Figure CN103610227BD00118

[0037] 其中,tT为输入的时间,单位为s ; κ κ 2, κ 3分别为阶跃函数的上升速度、上升时间与终值。 [0037] where, tT to time input, the unit is s; κ κ 2, κ 3 respectively, the rising speed of a step function, the rise time and the final value.

[0038] 所述步骤4 )中,烘丝机干头阶段Cub i c-RBF-ARX模型计算出的出口水分预测值 [0038] said step 4), Drying machine dry first stage Cub i c-RBF-ARX model to calculate the predicted value of the export of water

Figure CN103610227BD00121

通过将烘丝机干头阶段各工艺变量的优化设定曲线代入所构建的干头阶段Cub i c-RBF-ARX的输入变量 By the first phase of the input variable dry Drying Dryer first stages of the process variables of optimal setting curve constructed by substituting Cub i c-RBF-ARX's

Figure CN103610227BD00122

中得到。 Obtained. 通过使干头阶段Cubic-RBF-ARX模型计算出的出口水分预测值与出口水分设定值ysrt(ta)的误差eH(ta)最小,即采用列维布格奈奎尔特方法求解优化问题 By dry head stage Cubic-RBF-ARX model to calculate the predicted values of the water outlet and the outlet water setpoint ysrt (ta) error eH (ta) minimum, which uses a method for solving optimization problems 列维布格奈奎 Technology

Figure CN103610227BD00123

.寻找出干头阶段排潮风门、风温、筒温的输入曲线的参数λχ和入口流量输入曲线的参数κ κ2, κ3;其中,^1,2,〜,7出=1,2,3"是干头阶段持续的时间。 Look for a dry first stage moisture exhaust damper, parameters air temperature, cylinder temperature input curve parameters λχ and inlet flow input curve κ κ2, κ3; where ^ 1,2, ..., 7 = 1,2,3 "it is the first stage of dry duration.

[0039] 所述步骤3)中,用于描述干尾阶段排潮风门、风温、筒温和筒体电机频率的最优输入曲线的指数函数的表达式为: [0039] the step 3), is used to describe the end of stage dry moisture exhaust damper, expression exponential curve optimal input air temperature, cylinder mild cylinder motor frequencies as follows:

Figure CN103610227BD00124

[0041] 式中Uzl(tz)、Uz2(tz)、Uz3(tz)、Uz4(tz)分别表示干尾阶段排潮风门、风温、筒温和筒体电机频率的最优输入曲线。 [0041] where Uzl (tz), Uz2 (tz), Uz3 (tz), Uz4 (tz), respectively, end stage dry moisture exhaust damper, optimal input air temperature curve, the cylinder barrel mild motor frequency.

[0042] 所述步骤4)中,干尾阶段的Cubic-RBF-ARX模型计算出的出口水分预测值 [0042] the step 4), Cubic-RBF-ARX model to calculate the dry end stage of the predicted value of the export of water

Figure CN103610227BD00125

通过将烘丝机干尾阶段各工艺变量的优化设定曲线代入所构建的干尾阶段Cub i c-RBF-ARX模型的输入变量》6V)中得到;通过使干尾阶段的Cubic-RBF-ARX模型计算出的出口水分预测值浐P)与出口水分设定值y'srt(tb)的误差eT(tb)最小,即采用列维布格奈奎尔特方法求解优化问题 By the end of phase input variable dry Drying machine dry end stages of the process variables of optimal setting curve constructed by substituting Cub i c-RBF-ARX model "6V) obtained; dry by the end of stage Cubic-RBF- ARX model to calculate the predicted value of the export of water Chan P) and outlet water setpoint y'srt (tb) error eT (tb) minimum, which uses a method for solving optimization problems 列维布格奈奎 Technology

Figure CN103610227BD00126

,寻找出干尾阶段排潮风门、风温、筒温和筒体电机频率最优输入曲线的参数apg;其中,g=l,2, 3 ;M'是干尾阶段持续时间。 Find out the dry end stage row tide damper, air temperature, cylinder mild cylinder motor frequency optimal input parameters apg curve; wherein, g = l, 2, 3; M 'is the dry end stage duration.

[0043] 与现有技术相比,本发明所具有的有益效果为:本发明方法可使干头阶段叶丝出口水分尽可能快地上升、并快速到达稳定状态,可使干尾阶段叶丝出口水分尽可能缓慢地下降,从而有效地减少头尾段的干料量,提高烘丝过程的控制性能,具有较大的经济价值; 本发明方法综合考虑了来料量与各输入变量间的动态特性,可以更有效地克服来料流量和水分变化对烘丝过程头尾段的影响,适用于不同模式下叶丝入口流量与入口水分时的头尾段控制;本发明方法基于辨识的模型优化出最优的输入设定曲线,避免了人工整定输入工艺变量参数的不便。 [0043] Compared with the prior art, the invention has the beneficial effects: the method of the present invention allows the first stage of dry leaves silk exports rose water as quickly as possible, and quickly reach a steady state, can end stage dry leaves wire outlet water decreases slowly as possible, so as to effectively reduce the head and tail sections of dry quantity, improve the control performance Drying process with great economic value; the method of the present invention to consider the quantity and between each of the input variables dynamic characteristics, can be more effectively overcome the impact of the incoming flow and water change on the head and tail sections Drying process leaves control head and tail pieces of yarn entrance to the inlet water flow when applied in different modes; the method of the present invention is based on identification of the model Optimization of the optimal input setting curve, avoiding the manual tuning process variable input parameters for the inconvenience.

附图说明 Brief Description

[0044] 图1为烘丝机工艺过程示意图。 [0044] FIG. 1 is a schematic diagram of the process drying wire machine.

具体实施方式 DETAILED DESCRIPTION

[0045] 烘丝机工艺过程如图1所示。 [0045] Drying machine process shown in Figure 1. 叶丝进入烘丝工序之前,首先检测叶丝的入口流量U4和入口水分U5。 Drying leaves wire before entering the process, the first detection lobe wire inlet water inlet flow U4 and U5. 经过T3时间,叶丝到达烘丝机入口处。 After time T3, leaf tobacco drying machine wire reach the entrance. 叶丝在烘丝筒烘干时,系统会定时采样筒体的排潮风门开度U 1、风温U2、筒温U3等工艺变量参数值。 When drying silk silk leaves drying cylinder, the system periodically sampling cylinder row tide throttle opening U 1, air temperature U2, U3 cylinder temperature and other process variables parameter value. 烘干过程持续T4时间, 烘干后的叶丝从烘丝筒出口倒出,并在出口处测量叶丝出口水分值y。 T4 time drying process continues, after drying silk leaves from drying wire tube outlet pour and measure leaf water content value y wire outlet at the exit. 从有入口流量检测值到有出口水分检测值需经历一段较长时间,例如某烘丝生产线大约需340s。 From an inlet to an outlet flow rate detection value detected value of water required to undergo a longer period of time, such as a drying wire production line takes about 340s. 另外,烘丝机的输入/输出变量间也具有较大的时滞。 In addition, between the input / output variables Drying machine also has a large time lag.

[0046] 当检测到有入口流量时,表明烘丝过程开始运行。 [0046] When it is detected with an inlet flow rate, indicating Drying process begins. 运行初期烘丝过程有叶丝入口流量与入口水分检测值,没有叶丝出口水分检测值,此时烘丝过程干头阶段开始。 Initial Operation Drying process leafy wire inlet flow rate and inlet moisture detection value, no leaf silk exports moisture detection value, then dry Drying process header stage begins. 根据烘丝过程干头阶段的特性,建立Cubic-RBF-ARX模型结构: Drying process according to the characteristics of dry first stage, the establishment of Cubic-RBF-ARX model structure:

Figure CN103610227BD00131

Figure CN103610227BD00141

[0050] 其中,yH(tH)表示烘丝机干头阶段CubiC-RBF-ARX模型的出口水分; [0050] where, yH (tH) indicates outlet water Drying Dryer first stage CubiC-RBF-ARX model;

Figure CN103610227BD00142

分别表示干头阶段Cubic-RBF-ARX模型的排潮风门开度、风温、筒温、入口流量及入口水分;XH(tH-l)为入口流量和入口水分的状态变量;np H, nqH,dH和mH均表示干头阶段Cubic-RBF-ARX模型的阶次; Respectively ranked first dry tide throttle opening stage Cubic-RBF-ARX model, air temperature, cylinder temperature, inlet flow rate and inlet water; XH (tH-l) for the inlet water flow rate and inlet state variables; np H, nqH , dH and dry first phase mH expressed Cubic-RBF-ARX model order;

Figure CN103610227BD00143

分别为干头阶段Cub i c-RBF-ARX模型输出项与输入项的RBF神经网络的中心 Centre stage were dry head Cub i c-RBF-ARX model output term and the entry of RBF Neural Networks

Figure CN103610227BD00144

Figure CN103610227BD00145

为干头阶段Cubic-RBF-ARX模型的标量权系数;M · I |F表示矩阵的Frobenius范数;|H(tH)是干头阶段Cubic-RBF-ARX模型的建模误差,为高斯白噪声;T QH为烘丝机干头阶段Cubic-RBF-ARX模型建模采样时间,T1为从有入口流量检测值到有入口水分检测值的时间,T 2为从有入口水分检测值到有出口水分检测值的时间,T 3为从有入口水分检测值到烘丝筒入口的时间,! Scalar weights of dry first stage Cubic-RBF-ARX model; M · I | F represents Frobenius matrix norm; | H (tH) is the first stage of the modeling error dry Cubic-RBF-ARX model is Gaussian white noise; T QH is baked dry silk machine head stage Cubic-RBF-ARX modeling sampling time, T1 flow from inlet to inlet water detection value detected value of time, T 2 from an inlet moisture detection value to have the export value of moisture detection time, T 3 from the detected value of the water inlet tube entrance Drying time! \为叶丝在烘丝筒烘干的时间。 \ Leaf silk yarn in the drying cylinder drying time.

[0051] 当入口流量由正常值变为0时,标志着干尾过程的开始,当出口水分下降到3%时, 标志着烘丝机整个烘丝过程的结束。 [0051] When the inlet flow rate from the normal value to 0, marking the beginning of the dry end of the process, when the outlet water down to 3%, marking the end of the whole machine Drying Drying process. 干尾过程中无入口流量检测值,但有出口水分检测值。 Dry end process without inlet flow detection value, but the value of exports of moisture detection. 根据烘丝机干尾过程段的特性,建立如下的Cubic-RBF-ARX模型: Drying machine according to the characteristics of dry tail section of the process to establish the following Cubic-RBF-ARX model:

[0052] [0052]

Figure CN103610227BD00151

[0055] 其中,yT(tT)表示烘丝机干尾阶段Cubic-RBF-ARX模型的出口水分; 分别表示干尾阶段Cub i c-RBF-ARX模型的筒温、热风风温、排潮风门开度、入口流量、入口水分及筒体电机频率;XT(tT_l)为热风风温和筒体电机频率的状态变量;npT,nqT,dT和m %表示干尾阶段Cubic-RBF-ARX模型的阶次; zP' 分别为干尾阶段Cubic-RBF-ARX模型输出项与输入项的RBF神经网络的中心; [0055] where, yT (tT) indicates outlet water Drying machine dry end stage Cubic-RBF-ARX model; denote a cylinder temperature of dry end stage Cub i c-RBF-ARX model, hot air temperature, moisture exhaust damper opening, the inlet flow rate, inlet water and the cylinder motor frequency; XT (tT_l) state of mild hot wind cylinder motor variable frequency; order npT, nqT, dT and m% represents dry end stage Cubic-RBF-ARX model times; zP 'center RBF neural networks were dry end stage Cubic-RBF-ARX model output items and input items;

Figure CN103610227BD00152

为干尾阶段Cubic-RBF-ARX模型的标量权系数; f(tT)是干尾阶段Cubic-RBF-ARX模型建模误差,为高斯白噪声;IV为烘丝机干尾阶段Cubic-RBF-ARX模型建模采样时间。 Scalar weights of dry end stage Cubic-RBF-ARX model; f (tT) is a dry end stage Cubic-RBF-ARX modeling error, Gaussian white noise; IV for the drying wire machine dry end stage Cubic-RBF- ARX modeling sampling time.

[0056] 本发明采用结构化非线性参数优化方法(SNPOM)方法对模型进行估计。 Optimization Method [0056] The present invention uses a structured nonlinear parameters (SNPOM) method to estimate the model. 为了使得上面所构造的Cubic-RBF-ARX模型能够描述烘丝过程头尾段的全局动态特性,我们首先采用SNPOM方法来优化模型的、一步预测误差最小情形下的参数,并以此参数作为长期预测优化目标下的模型参数初始值。 To make Cubic-RBF-ARX model constructed above the global dynamic characteristics can describe the head and tail sections Drying process, we first use SNPOM ways to optimize the model parameters step prediction error is minimized under the circumstances, and as long as the parameters Prediction optimization model parameters under initial value of the target. 然后,采用列维布格奈奎尔特方法(LMM)来进行长期预测性能最优的模型参数的优化。 Then, using 列维布格奈奎 Stewart method (LMM) to optimize the performance of the best long-term prediction model parameters.

[0057] 烘丝机干头阶段Cubic-RBF-ARX模型(1)的参数优化问题如下: [0057] Drying Dryer first stage Cubic-RBF-ARX model (1) the parameters of the optimization problem as follows:

Figure CN103610227BD00161

[0059] 其中,产(严)是烘丝机干头阶段出口水分的实际值,r (严)是在实际输入作用下,由烘丝机干头阶段Cub i c-RBF-ARX模型计算出的出口水分的预测值; [0059] where production (Yan) is the actual value of the first phase Drying Dryer export water, r (Yan) is a role in the actual input by Drying Dryer first stage Cub i c-RBF-ARX model to calculate the The predictive value of exports of water;

Figure CN103610227BD00162

为烘丝机干头阶段Cubic-RBF-ARX模型的线性参数 Linear parameters Drying Dryer first stage Cubic-RBF-ARX model

Figure CN103610227BD00163

为烘丝机干头阶段Cubic-RBF-ARX模型的非线性参数;Nh为烘丝机干头阶段Cubic-RBF-ARX模型建模数据长度。 Nonlinear parameter Drying Dryer first stage Cubic-RBF-ARX model; Nh is Drying Dryer first stage Cubic-RBF-ARX modeling data length.

[0060] 烘丝机干尾阶段Cubic-RBF-ARX模型(3)的参数优化问题如下: [0060] Drying machine dry end stage Cubic-RBF-ARX model (3) of the parameter optimization problem as follows:

Figure CN103610227BD00164

[0062] 其中,/(r)是烘丝机干尾过程中出口水分的实际值;r(r)是在实际输入作用下,由烘丝机干尾阶段Cub i c-RBF-ARX模型计算出的出口水分的预测值; [0062] where, / (r) is the actual value of the tobacco drying machine dry end of the process water outlet; r (r) is in effect the actual input, calculated by drying silk machine dry end stage Cub i c-RBF-ARX model export of water out of the predicted value;

Figure CN103610227BD00165

为烘丝机干尾阶段Cubic-RBF-ARX模型的线性参数 Linear parameters Drying machine dry end stage Cubic-RBF-ARX model

Figure CN103610227BD00166

为烘丝机干尾阶段Cubic-RBF-ARX模型的非线性参数;Nt为烘丝机干头阶段Cubic-RBF-ARX模型建模数据长度。 Nonlinear parameter Drying Dryer Last stage Cubic-RBF-ARX model; Nt is Drying Dryer first stage Cubic-RBF-ARX modeling data length.

[0063] 依据估计出的烘丝过程干头阶段Cubic-RBF-ARX模型来设计各工艺变量的最优输入曲线,以适应来料情况的变化,尽量减少干头阶段的干料量。 [0063] based on the estimated tobacco drying process dry head stage Cubic-RBF-ARX model to design optimal input curve for each process variables to accommodate changes in incoming cases to minimize the quantity of dry first stage of dry. 本发明采用双S型函数来描述干头阶段排潮风门、风温、筒温的最优输入曲线,采用阶跃型函数来描述入口流量的最优输入曲线。 The present invention uses double S-type function to describe the first stage of dry moisture exhaust damper, optimal input curve air temperature, cylinder temperature, using step-function input curve to describe the optimal inlet flow.

[0064] 双S型曲线公式如下: [0064] the double S-curve formula is as follows:

Figure CN103610227BD00167

[0066] 其中,ts为输入的时间,单位为s ; λ i,λ4, \5分别为双S型函数的起点、转折点及终点值;λ 2, λ 6分别为双S型函数的两条对称轴中心位置;λ 3, λ 7分别为双S型函数上升或下降的速度;λ 3, λ 7大于〇时表示S型函数上升,λ 3, λ 7小于〇时表示S型函数下降; c=l,2, 3, Usl(ts)是排潮风门的设定值;Us2(ts)是风温的设定值;U s3(ts)是筒温的设定值。 [0066] where, ts is the time input, the unit is s; λ i, λ4, \ 5, respectively, as a starting point the double S-type function, turning point and end point value; λ 2, λ 6 respectively, two double S-type function symmetry axis center position; λ 3, λ 7 double S function respectively increase or decrease the speed; λ 3, λ 7 is greater than the increase in square represents S type function, λ 3, λ 7 when less than the square represents S type function decline; c = l, 2, 3, Usl (ts) is set to a value of moisture exhaust damper; Us2 (ts) is the setpoint air temperature; U s3 (ts) is the barrel temperature setpoint.

[0067] 描述入口流量输入曲线的阶跃型函数公式如下: [0067] inlet flow input curve describes the step-function formula is as follows:

Figure CN103610227BD00171

[0069] 其中,tT为输入的时间,单位为s ; κ ρ κ 2, κ 3分别为阶跃函数的上升速度、上升时间与终值。 [0069] where, tT to time input, the unit is s; κ ρ κ 2, κ 3 respectively, the rising speed of a step function, the rise time and the final value.

[0070] 将各工艺变量的优化设定曲线(7-8)代入所构建的CubiC-RBF-ARX模型(1)的输入变量4 (〇,4(〇,4(〇,4(〇中,可得到干头阶段出口水分的预测值37 /?(〇: [0070] The optimal setting curve for each process variables (7-8) into the constructed CubiC-RBF-ARX model (1) of the input variable 4 (square, 4 (square, 4 (square, 4 (〇 in ? available dry first stage outlet water predictive value 37 / (○:

Figure CN103610227BD00172

[0072] 采用列维布格奈奎尔特(Levenberg-Marquardt Method,LMM)方法,通过使模型计算出的出口水分预测值与出口水分设定值的误差最小,寻找出干头阶段排潮风门、风温、筒温最优输入曲线的AiQ=Ij,"'?)参数和入口流量最优输入曲线的Kj(j=l,2,3)参数。 干头阶段出口水分设定值与基于干头动态模型预测值(9)之间的误差为: [0072] The 列维布格奈奎 Stewart (Levenberg-Marquardt Method, LMM) method, prediction error value and export water outlet water setpoint by model to calculate the minimum, look for a dry first stage moisture exhaust damper , air temperature, cylinder temperature optimum input curve AiQ = Ij, "'?) optimal input parameters and inlet flow curve Kj (j = l, 2,3) parameter outlet water setpoint and based on dry first stage error dry head dynamic model predictive value (9) is as follows:

Figure CN103610227BD00173

[0074] yset (ta)是出口水分设定值。 [0074] yset (ta) is the outlet water setpoint.

[0075] 干头阶段工艺变量最优设定的优化问题如下: [0075] Optimization dry first stage of the process variable optimal setting follows:

Figure CN103610227BD00174

[0077] M是干头阶段持续时间。 [0077] M is dry first phase duration. 通过求解上述优化问题可得到最优设定曲线的参数值,从而设计出烘丝机干头阶段各个工艺变量的最优输入曲线。 Can be obtained by solving the above optimization problem the optimal setting parameters of the curve, so that the optimal input design curve Drying Dryer first stage of each process variable.

[0078] 依据估计出的烘丝过程干尾阶段Cubic-RBF-ARX模型来设计各工艺变量的最优输入曲线,以适应来料情况的变化,尽量减少干尾阶段的干料量。 [0078] based on the estimated tobacco drying process dry end stage Cubic-RBF-ARX model to design optimal input curve for each process variables to accommodate changes in incoming cases to minimize the quantity of dry tail dry stage. 采用指数型函数来描述干尾阶段排潮风门、风温、筒温和筒体电机频率的最优输入曲线,该指数型曲线公式如下所示: Using exponential function to describe the dry end stage row tide damper, optimal input air temperature curve, the cylinder barrel mild motor frequency, the exponential curve equation is as follows:

Figure CN103610227BD00175

[0080] 式中Uzl(tz)、Uz2(tz)、U z3(tz)、Uz4(tz)分别表示干尾阶段排潮风门、风温、筒温和筒体电机频率的最优输入曲线。 [0080] where Uzl (tz), Uz2 (tz), U z3 (tz), Uz4 (tz), respectively, end stage dry moisture exhaust damper, optimal input air temperature curve, the cylinder barrel mild motor frequency. 将各工艺变量的优化设定曲线(12)代入所构建的Cubic-RBF-ARX模型(3)的输入变量《ff),中,可得到干尾阶段出口水分的预测值: The optimal setting curve for each process variables (12) into the constructed Cubic-RBF-ARX model (3) of the input variables "ff), the obtained dry end stage outlet water predictive value:

Figure CN103610227BD00176

[0082] 采用列维布格奈奎尔特(LMM)方法,通过使模型计算出的出口水分预测值与出口水分设定值的误差最小,寻找出干尾阶段排潮风门、风温、筒温和筒体电机频率最优输入曲线的apg;其中,g=l,2,3。 [0082] The 列维布格奈奎 Stewart (LMM) method, prediction error value and export water outlet water setpoint by model to calculate the minimum, look for a dry end stage row tide damper, air temperature, cylinder mild cylinder motor frequency optimal input apg curve; wherein, g = l, 2,3. 干尾阶段出口水分设定值与基于干尾动态模型预测值(13)之间的误差为: Dry end stage outlet water setpoint and based on the dry end of the dynamic model error predictive value (13) as follows:

Figure CN103610227BD00177

[0084] ysrt⑴是出口水分设定值。 [0084] ysrt⑴ export water setpoint.

[0085] 干尾阶段工艺变量最优设定的优化问题如下: [0085] End-stage process optimization dry optimal variable set as follows:

Figure CN103610227BD00181

[0087] M是干尾阶段持续时间。 [0087] M is a dry end stage duration. 通过求解上述优化问题可得到最优设定曲线的参数值,从而设计出烘丝机干尾阶段各个工艺变量的最优输入曲线。 Can be obtained by solving the above optimization problem the optimal setting parameters of the curve, so that the optimal input design curve Drying machine dry end stages of each process variable.

Patent Citations
Cited PatentFiling datePublication dateApplicantTitle
CN102090703A *Aug 26, 2010Jun 15, 2011龙岩烟草工业有限责任公司Method for improving parameter control effect in silk making technology
CN102090704A *Sep 21, 2010Jun 15, 2011龙岩烟草工业有限责任公司Method for improving batch procedure capability of tobacco shred making process
CN102090705A *Aug 26, 2010Jun 15, 2011龙岩烟草工业有限责任公司Method for improving tobacco-drying process capability
CN102488308A *Dec 14, 2011Jun 13, 2012东华大学Advanced coordinated control system for moisture in cut tobacco dryer
CN102871214A *Oct 8, 2012Jan 16, 2013秦皇岛烟草机械有限责任公司Model prediction based cut tobacco dryer outlet moisture control method
CN103238916A *May 30, 2013Aug 14, 2013龙岩烟草工业有限责任公司Method for reducing dry cut tobacco of drum cut tobacco dryer
CN103324089A *Mar 24, 2012Sep 25, 2013厦门烟草工业有限责任公司Thread making technological parameter dynamic quality stability characterization method
US3905123 *Oct 15, 1973Sep 16, 1975Industrial Nucleonics CorpMethod and apparatus for controlling a tobacco dryer
Classifications
International ClassificationA24B3/10
Legal Events
DateCodeEventDescription
Apr 2, 2014C10Entry into substantive examination
Apr 15, 2015C14Grant of patent or utility model