PEI Yuling

Department of Automation, Chongqing Industry Polytechnic College, Chongqing 401120, China

IntelligenceBasedFusionControlStrategyforComplexVulcanizingProcess

PEI Yuling*

DepartmentofAutomation,ChongqingIndustryPolytechnicCollege,Chongqing401120,China

Thevulcanizationofplastic&rubberproductsisatypicalcomplexcontrolprocesswithuncertainty.Aimedatthepuzzlethattheshapingprocessofplastic&rubberproductsisdifficulttocontrolresultedfromfuzzificationanduncertainty,andpoorinstabilityofproductqualityintheprocessofvulcanization,thepaperproposedasortofintelligencebasedcontrolstrategybasedonhumansimulatedintelligentcontroller.Inthepaper,itsummedupthecontrolpuzzlesintheprocessofcomplexvulcanization,exploredthecyberneticscharacteristicofvulcanizingprocess,researchedonthecontrolstrategyofcomplexprocesswithuncertainty,proposedasortofintelligencebasedcontrolstrategy,discussedcontrolmodel,andconstructedthecontrolalgorithmbasedonhumansimulatedintelligence.Itmadethesimulationexperiment,andthecurveofalgorithmresponsedemonstratedthatitcouldbestrongerinrobustness,andhigherincontrolprecisioncomparedwithPIDcontroller.Theresearchresultshowsthatitisfeasibleandeffectivetotheproposedcontrolstrategybasedonhumansimulatedintelligenceinactualizingcontrolforcomplexvulcanizingprocesswithuncertainty.

vulcanizingprocess，uncertainty，equivalentvulcanizing，humansimulatedintelligence，intelligencefusionbasedcontrolstrategy

The plastic & rubber products play an important role in the national economy and national defense construction. Currently the control method of not a few production equipments is relatively backward in China, and the production process is suffered from the influence of subjective factors such as experiences and skills of the operators, and therefore it is leaded to be difficult to actualize the optimizing quantitative control for the process of complex vulcanization in plastic & rubber products. The product features such as rigidity and endurance and so on not only are related to the factors of formulations of materials & supplies and structure size and so on, but also the physical and chemical properties of products mainly depend on the control effect of physical shaping in the vulcanizing phase. Despite the control precision and control level and so on of vulcanizing equipments have obviously made the progress after technical reform of the devices, but the stability of product quality still appears numerous problems, and therefore it is necessary to explore the control strategy for vulcanizing process.

1.Control puzzle and analysis of cybernetics characteristic

1.1.Controlpuzzleofvulcanizingprocess

The plastic & rubber product is the outcome of chemical reaction of cross-linking characteristics occurrence for a certain formulation of plastic & rubber materials, which is produced in the mould by means of a variety of assistants under specified temperature and pressure within stipulated time period. The mechanism of physicochemistry is very complex in the process of product shaping, and it is the puzzle of control engineering[1-2].

1) Complexity of materials and supplies.The plastic cement belongs to the polymer materials, and maybe the difference of its chemistry molecular weight is over hundreds of times even thousands of times, and therefore it owns grievous uncertainty, decentrality and unknown, and leads to be difficult to hold the characteristic after formulation of plastic & rubber materials. Due to the influence of uncertainty, after chemical reaction resulted in occurring cross-linking between sizing material and various assistants, the stability of qualified product performance may be unsatisfactory, and also its variance distribution is a larger, therefore how to avoid the influence of uncertainty is one of control puzzles in the vulcanizing process.

2) Characteristic of large pure delay and great inertia .Essentially the vulcanizing is a heating process under the condition of a certain pressure. The exchange and transmission of the heat is implemented by means of thermal radiation and convection, and the sizing material itself is not a better heat conductor, but its distribution of temperature field is also heterogeneous. Obviously there exists large pure delay in heat transfer, it owns the characteristic of great inertia, and therefore it is the bottleneck problem in control engineering.

3) Strongly nonlinear characteristics.The relationship of heat transfer described by heat transfer equation does not satisfy the condition of homogeneity and additivity. The heat transfer is strongly nonlinear, and currently there is still not any mature control method to follow and use for reference in strongly nonlinear process control.

4) Uncertainty and complexity of vulcanizing process.For instance, the unknown and time-varying of time lag suffer from influence of temperature hysteresis effect. It makes the inner temperature be lower than the outer temperature obviously, and different layer of plastic cement makes vulcanizing be under different temperature. Along with increase of specifications and dimensions in product, it is hardly possible to make the inner and outer temperature of the plastic cement reach the optimum cure temperature at the same time. Just about the influence of uncertainty and complexity, it makes quantitative control became very difficult.

5) Being intangible in best curing degree.It results in best curing degree to be difficult to determine for the non-uniform of sizing material distribution, the unknown, variety and randomness of the control parameters.

1.2.Analysisofvulcanizingprocess

The key factors of determining quality stability are depended on pressure, temperature and time of vulcanizing stage in the production process of sizing material product.

1) Pressure of vulcanization.Except small structure size or thinning plastic cement product, it generally needs to make vulcanizing under a certain pressure. The high or low pressure in the vulcanizing process is varied with product structure, sizing material property and technology condition and so on. In the shaping process of sizing material, if there exists in pore then it has important influence for physical property of sizing material, specially abrasion and wear as well as aging properties. The action of pressure lies in that it compels streaming of sizing material in the mould, and makes sizing material fill with mould chamber so as to eliminate the pin hole. It can ensure the compactness of product in complex shaping structure, and enhances the adhesive force between bond line and metal layer or the cloth layer as well as flex resistance of product and so on. Generally it is relatively large to the vulcanizing pressure for the product of complex shaping, poor liquidity of sizing material and large structure size. Therefore it is necessary to control the pressure of each stage of vulcanization accurately, so that it can ensure the mechanical properties such as compactness, adhesive force and flex resistance of product and so on.

2) Temperature of vulcanization.The vulcanizing temperature is so called as plateau cure temperature. The plateau cure is a process, and all the main physical and mechanical properties can reach or verge on optimal value in the process. Its corresponding temperature and time is respectively called as the plateau cure temperature and the plateau cure time. In actual operation, it should aim at the cases such as specific formulation of sizing material, technology condition, performance characteristic and product structure and so on to design the plateau cure temperature. If it is too long to the time then it would result in excessive vulcanizing phenomenon, and if it too short to the time then it would result in less vulcanizing phenomenon. The length of vulcanizing time is carried with vulcanizing temperature, and its technical key rests with actualizing equivalent vulcanization for vulcanizing process. According to the change relationship between vulcanizing time and vulcanizing temperature, it can actualize accurate control of vulcanizing time in real time and ensure the physical and chemical properties of product to reach optimal index.

3) vulcanizing time and equivalent vulcanization principle.When the temperature of vulcanization occurs in fluctuating, the effect of equivalent vulcanization can be expressed as the following.

In which,Kis the coefficient of vulcanizing temperature,Iis the vulcanizing intensity,τ1，τ2is respectively the start and end time of vulcanization. The equivalent vulcanization shows that its vulcanization is equal to above integral result. When Δτis very small, it can be considered that the coefficientKand timetare basically invariant in the range of Δτ, and therefore it can make the approximate calculation of equivalent vulcanization.

For convenience of expression, assume that if the standard temperature of vulcanization ist0, and theE0is the vulcanizing effect reached tot0, and selectingt2=t0, then the formula can be simplified asEi=Δτ。According to the precision requirement of actual control, when the sampling time period of temperature is Δtit can computeEiin terms of actual measure temperatureti. When ∑Eireaches toE0it can be considered as that the optimal vulcanization has been completed, and the vulcanizing process should be ended.

1.3.Cyberneticscharacteristicofvulcanizingpro-cess

The selection of control strategy and control algorithm is determined by cybernetics characteristic in the vulcanizing process, and therefore it is necessary to research on cybernetics characteristic of vulcanizing process. From the summarizing of above control puzzles and the analysis of vulcanizing process characteristic, it can be summed up that the vulcanizing process is a complex process with uncertainty, and the cybernetics characteristic mainly would be expressed as the following.

1) Randomness, distribution, unknown and time varying of parameters in vulcanizing process. In fact, the process parameter is always changing.

2) Large pure lag and great inertia property in vulcanizing process, because of the materials and supplies of vulcanization being not a better conductor.

3) The complexity of heating transfer relation determines that the vulcanizing process owns the strong nonlinearity property obviously.

4) The attribute of high polymer in materials and supplies results in unknown and time carrying of time lag parameter.

5) There are the unknown, randomness and variety of interfering in the outer environment.

The cybernetics characteristic of vulcanizing process mentioned above reminds us that if the conventional control strategy is applied to control vulcanizing process then it is not advisable, and therefore it is necessary to explore the new control strategy so as to satisfy the control demand of vulcanizing process in plastic cement.

2.Selection of control strategy in vulcanizing process

From the cybernetics characteristic mentioned above it can be seen that it is difficult to construct the mathematical model of vulcanizing process, and therefore whether the method and technique based on classical control theory or modern control theory all are impossible to obtain expected vulcanizing grade of optimum. Because of its control algorithm being based on the strict mathematic model, so the routine control strategy such as PID control strategy is not advisable. From summarizing up the experiences of control engineering it can be seen that it plays an important role to the professional knowledge, actual experience, operating skill and adjustment technique of operator in the general process control. For instance, the fuzzy control technology always adopts fuzzy language to describe control process, through the fuzzification processing it can make corresponding control decision by means of fuzzy inference method, and then it can actualize the defuzzification to transform as the actual control quantity so as to implement control for complex process. Currently it has been lots of successful application example of case[3]. But there are some puzzles such as the selection of membership function and so on for complex process with uncertainty and large pure lag as well as great inertia, and it is difficult to reflect the influence completely in the control algorithm for professional knowledge, actual experience, operating skill and adjustment technique of the operator. It would lead to be low in precision and poor in robustness, and therefore it is not advisable to adopt fuzzy control for vulcanizing process. The expert control system can deal with each kind of qualitative, quantitative, accurate and fuzzy information, and it can fully summarize the knowledge, experience and technique of the operator and bring them into the control program. For the control of vulcanizing process, there are some puzzles such as how to obtain and express knowledge and so on in constructing perfect knowledge base, and it is ill-advised to adopt expert system control roundly in the vulcanizing process, but its basic technique has meaningful to use for reference. It is worth to pay attention to HSIC (Human Simulated Intelligent Controller), and it is effectively a sort of homomorphism transformation of human body control system[4]. In the actual operating process, the operator can timely make corresponding control action on the basis of the control effect and operating information of previous a moment. There are lots of information in a continual operating control process, and which is useful to control. The operator can select to make memory aiming at some information of having characteristic meaning, namely the characteristic information, and it can avoid large information storage space resulted in conventional controller because of blind memory information. The experience shows that the characteristic memory has the implication relationship with control output, taking this it is available to eliminate the process deviation, and make it not to produce the integral saturation phenomenon so as to reduce the lag resulted in integral. By means of HSIC, as long as the process change exists in deviation, then the characteristic model can make the perception for its change. It can find its corresponding control pattern through the recognition of judgement, and it does not consider the reason resulted in what cause at al. For vulcanizing process control influenced by lots of uncertainty factors, the strategy of HSIC should be considered as a better selection. As a result of characteristic information having reflected the dynamic characteristic of whole vulcanizing process, and according to the characteristic reflected by characteristic model, it can find corresponding control pattern. In terms of corresponding control pattern, it can construct the control algorithm that is suitable for vulcanization, and thereby it can ensure the process to achieve the demand of control quality of optimal vulcanizing degree in dynamic and steady state. Aiming at the features of vulcanizing process, based on the technical fusion between human simulated intelligent and expert control technology, in this paper it adopts intelligence fusion control strategy[5-6] so as to ensure that it is high in precision, short in transient process time, strong in robustness as well as no overshoot for vulcanizing process.

3.Control model and control algorithm

3.1.Controlmodelofvulcanizingprocess

It selects generalized control model as the control model of vulcanizing process in the control strategy of intelligence fusion, and it is shown as in Fig.1. The model makes the fusion between knowledge model and mechanism, and it is considered as an integrated control model. The superiority of control model rests with that it is unnecessary to know more priori knowledge for uncertain vulcanizing control process, and as long as it knows the process deviation and its change rate then it can actualize effective control at once for vulcanizing process. Fig.2 shows the structure of general control system, and in which,r(t) is the process input,y(t) is the process output,e(t) is the process deviation, ande(t)=r(t)-y(t). The process deviation, change rate of process deviation and the time can construct the solving information space of control problem. In the control problem of HSIC, the controller of generalized control model is based on characteristic recognition of solving information space, for instance such as big or small in process deviation, change direction of process deviation, and change trend and so on, and if it knows the feature then it can select corresponding control pattern, and therefore it could construct corresponding control algorithm so as to actualize high accurate and strong robust control.

Fig.1 Generalized control model based on HSIC

Fig.2 Structure of generalized control process

The generalized control model based on HSIC can fuse the technology of HSIC and expert system technique to make control system structure be simplified. It can easily fuse the control rule and inference organization as well as knowledge base into a intelligent controller, and also it is convenient to adopt the production rule of “If 〈condition〉 Then 〈action〉” to establish knowledge set. Its outstanding advantages are that each piece of rule can independently be added, deleted and revised, it is better in modularity and naturality, no direct relation among the rules, and it owns strong self-adaptive for environment change. By means of basic characteristic combined open-loop and closed-loop control in HSIC, it can easily actualize nimble multi-modal control, and it strengthens the capability of judgement and inference[7]. Therefore the strategy is suitable for feature of vulcanizing process control.

3.2.Basiccontrolalgorithm

4.Simulation experiment

The vulcanization of plastic cement is a complex uncertainty process with large pure lag and great inertia. For not losing the generality, it can transfer the control problem as that it examines the robustness of control algorithm. The control algorithm with strong robustness owns very strong adaptation capacity, and when the control parameter changes the influence is very small, and it influences hardly the control quality of the system. Therefore it takes familiar two-order control model with lag as an example, and assumes the model of controlled vulcanizing process as the following.

In which,K0= 4.134,T1= 1.0 s,T2=2.0 s,τ=10 s. It adapts respectively PID control algorithm and intelligence fusion control algorithm (simplified as HSIC) to actualize the control for vulcanizing process in the experiment, and then investigates the response of its process. At timet=15 s, it joins a pulse signal to be as the outer interference, the pulse amplitude is 1.0, and the pulse width is 0.2 s. The response curve of two sorts of control algorithm is respectively shown as in Fig.3. Compared with PID control algorithm, it is short to transient process time for the HSIC of control algorithm based on control strategy of intelligence fusion, and it reveals very strong robustness of control algorithm. By means of process deviation signal, the strategy strengthens the control function for the process. According to the control rule, it must make the adjustment for control parameter in each control period in terms of direction, being big or small of process deviation and its deviation change trend, and it forced the vulcanizing process to produce the response so as to make the process deviation press on forwards to zero. The control of intelligence fusion control is the nonlinearity control. For shortening the time of transient process and enhancing the speed of process response, the Bang-Bang control could be fused into the control strategy, and it permits control amplitude to be more. But it also applies the other control rule to make its overshoot obtain the constraint, and therefore it could avoid the originated oscillation. So the vulcanizing process can rapidly reach the steady status.

Fig.3 Response curve at pulse interference

The analysis of curve in vulcanizing shaping shows that the control effect is satisfactory for strategy of intelligence fusion, and it makes the robustness and accuracy of vulcanizing process control all reach the expected result.

5.Conclusions

The control of vulcanizing process with uncertainty is very complex. All the above theory analysis based on control strategy of intelligence fusion and simulation experiment show that adopting proposed control strategy can ensure better in dynamic and steady control quality of vulcanizing process response, high in control precision, and short in transient process response. Under the condition of outer interference, compared with conventional PID control, the intelligence fusion control strategy has very strong robustness, and it is a better control strategy for control of complex vulcanizing process with uncertainty.

[1] AO Chaohua,ZHANG Shiping.Vulcanizing Process Control of I-Fuzzy-Smith Based Fusion Strategy[J].Journal of Sichuan Ordnance，2010，31(8)：82-84.

[2] LIU Yucheng，LIU Yubin，LI Taifu，et al.Human Simulated Intelligent Control of Vulcanization Process for Plastic & Rubber Product[J].Microcomputer Information,2008，24(7)：80-81.

[3] HUANG Wei,GUO Bing.Intelligence based control strategy of affusion system for oil field stratum[J].Journal of Liaoning Technical University：Natural Science,2010，29(5)：799-802.

[4] LI Zushu,TU Yaqing.Human Simulated Intelligent Controller[M].Beijing:National Industry defense Press,2003.

[5] Carmona P,Castro J L,Zurita J M.FRIwE：fuzzy rule identification with exceptions[J].Fuzzy Systems,IEEE Transactions on,2004,12:140 -151.

[6] Cruz C,Pelta D A,Royo A S,et al.Soft computing and cooperative strategies for optimization[C]//Soft Computing in Industrial Applications,SMCia/05,Proceedings of the 2005 IEEE Mid-Summer Workshop on.[S.l.]:IEEE,2005:75-78.

[7] Mazinan A H,Kazemi M F.Notes on intelligence based model predictive control scheme:A case study[C]//Intelligent Systems (IS),2010 5th IEEE International Conference.[S.l.]:IEEE,2010:474-478.

復(fù)雜硫化過程的智能融合控制策略

裴玉玲*

重慶工業(yè)職業(yè)技術(shù)學(xué)院 自動化系，重慶 401120

針對硫化過程中因模糊性與不確定性導(dǎo)致的塑膠產(chǎn)品成形過程難以控制、產(chǎn)品質(zhì)量穩(wěn)定性差的問題，提出了一種智能融合控制策略?？偨Y(jié)了復(fù)雜硫化過程中的控制難點(diǎn)，探討了硫化過程的控制論特性，研究了對不確定性復(fù)雜過程的控制策略，討論了控制模型，提出了智能融合控制策略，基于仿人智能構(gòu)造了基本控制算法。仿真結(jié)果顯示該策略的魯棒性強(qiáng)與控制精度高。研究結(jié)果表明：智能融合控制策略對不確定性復(fù)雜硫化過程實(shí)施控制是可行與有效的。

硫化過程；不確定性；等效硫化；仿人智能；融合控制策略

TP273

2012-10-21

*PEI Yuling, Associate Professor. E-mail：Beilei425@163.com

10.3969/j.issn.1001-3881.2013.06.018