• <tr id="yyy80"></tr>
  • <sup id="yyy80"></sup>
  • <tfoot id="yyy80"><noscript id="yyy80"></noscript></tfoot>
  • 99热精品在线国产_美女午夜性视频免费_国产精品国产高清国产av_av欧美777_自拍偷自拍亚洲精品老妇_亚洲熟女精品中文字幕_www日本黄色视频网_国产精品野战在线观看 ?

    An adaptive neuro-fuzzy sliding mode controller for MIMO systems with disturbance

    2017-05-28 08:50:18MahmoudSaafanMohamedAbdelsalamMohamedElksasSabrySarayaFayezAreed

    Mahmoud M.Saafan*,Mohamed M.Abdelsalam,Mohamed S.Elksas,Sabry F.Saraya,Fayez F.G.Areed

    Computers and Control Systems Engineering Department,Faculty of Engineering,Mansoura University,Egypt

    1.Introduction

    Ammonia(NH3)is one of the most widely used chemicals in industry.NH3is produced from the reaction between nitrogen gas(N2)and hydrogen gas(H2).The reaction for ammonia production is described as[1–7]:

    There are six steps for ammonia synthesis process:hydrogen production stage,nitrogen addition stage,removal of carbon monoxide stage,water removal stage,removal of carbon oxides,and finally,synthesis of ammonia stage.The flow chart of the ammonia process is shown in Fig.1.An ammonia reactor with quenching cooling and beds is shown in Fig.2.[6].

    The functions of CO conversion unit are reducing the COslip,providing the highest yield of hydrogen,and maximizing the CO2production to be used in urea plant[4–6].

    The reaction is reversible and exothermic,pressure has no effect on it,and an increase of steam increases the products.Because the reaction is exothermic,as the temperature increases the extent of conversion will decrease.To overcome that,we use a two stage system with suitable inter-cooling stages.This is the reason to use a high temperature reactor and low temperature reactor.Low temperature reactor allows water shift reaction to proceed at low inlet temperature limited by the dew point temperature.Usually,the inlet temperature is 15°C greater than the dew point temperature.According to the water shift reaction,as the steam increases the conversion of CO will increase too.But as the steam decreases the following bad effects will happen:

    –The conversion of CO will decrease.

    –Steam is used to moderate the reduction effect,and without steam the catalyst will be over-reduced to metallic iron.If this happens the structure of the catalyst will be changed and be weaker also metallic iron reacts with CO forming iron carbide which promotes other side reactions such as:

    Carbon will make a layer on the catalyst bed and increase the pressure drop on the reactor.The gas enters through the inlet of the high temperature reactor at a temperature of nearly 345°C.For protection of the condensation of water droplets the catalyst bed is covered by a layer of ceramic balls,where,the droplets of water vaporize on it.The reactor contains two layers,where,the reaction is proceeded on reaching 2.65 vol%.As the temperature of the gas inlet of the low temperature reactor must be around 198°C,so the gas that exits the high temperature reactor must be cooled.For this purpose,the gas is passed through the exchanger to exchange with the gas inlet methanator.Sometimes this exchanger is not used to increase the temperature of the inlet of the methanator,it is only used during start up or to increase the methanator inlet temperature to about 300°C.Then it is sent through the gas cooler,where the gas temperature is decreased from 410 °C to 198 °C by exchanging with boiler feed water[4–6].

    Urea is a chemical compound.The formula for urea is NH2CONH2.Urea is used mainly as a nitrogen fertilizer to grow crops,to produce urea formaldehyde resins,in wood industries,pharmaceutical industries and as an additive to animal food.Urea is produced by two chemical reactions, first,between ammonia and carbon dioxide.This reaction produces ammonium carbamate,which then reacts to produce urea.The first reaction is exothermic and fast,while the second reaction is endothermic and slow,the ammonium carbamate dehydrates to produce urea(NH2CONH2)and water(H2O).The two reactions are described as[8–9]:

    –The reaction of condensation

    Fig.1.Flow chart of ammonia synthesis process.

    Fig.2.Ammonia reactor with quenching and bed sections.

    The condensation is dependent on condensation temperature,pressure of reactor,the molar ratio between ammonia and carbon dioxide(NH3/CO2ratio),and molar ratio between water and carbon dioxide(H2O/CO2ratio).The molar ratio between ammonia and carbon dioxide must be equal to 3[10]in order to achieve the maximum urea conversion and reduce the pollution effect.There are five steps for urea plant:synthesis stage,recirculation stage,desorption and hydrolyzation stage,evaporation stage,and finally,granulation stage.The block diagramofurea process is shown in Fig.3.The schematic representation of the urea process is shown in Fig.4.The schematic representation of the ammonia and urea processes is shown in Fig.5.

    The NH3/CO2measurement in the outlet of the reactor indicates the molar ratio in the liquid phase.The normal operating range is between molar NH3/CO2ratios of about3.0 and 3.1.If the ratio is lower than 3.0 or higher than 3.1 the composition of the liquid in the reactor will deviate from optimum composition.In this case reactor outlet temperature will not be maximum,thus conversion of ammonia and carbon dioxide to urea in the reactor will be less.But all non-converted ammonia and carbon dioxide have to be recycledviaeither a high pressure heat exchanger or recirculation stage.Consequently a wrong NH3/CO2ratio will lead to higher steam consumption of the High Pressure Stripper as well as possibly more recycle of carbamate from the recirculation stage,which in its turn again will decrease ammonia and carbon dioxide conversion due to more water introduced into the synthesis section.The normal operating range is between 180 and 183°C.At an operating pressure of 136×105Pa–143×105Pa and with a normal quantity of water in the reactor,the latter temperature will never be exceeded.If the temperature in the reactor top is too low,this may be due to:The operating pressure being below the requirement,to a wrong NH3/CO2ratio,and the degree of condensation in the high pressure carbamate condenser being too high.Whenever the NH3/CO2ratio is wrong,the composition of the liquid in the reactor top will differ from the optimum,so that the temperature will not reach its maximum value either[9–12].

    2.Mathematical Model of Ammonia Synthesis Reactor

    The conversion of nitrogen and the feed ratio of nitrogen and hydrogen are represented as[3,8]:

    The flow rates of nitrogen,hydrogen,methane,and ammonia are represented as:

    The mole fractions of nitrogen,hydrogen,methane,air,and ammonia can be evaluated from flow rates which are represented as:

    Fig.3.Block diagram of urea synthesis process.

    Fig.4.Schematic representation of the urea process.

    Fig.5.Schematic representation of the ammonia and urea processes.

    2.1.Bed modeling

    The mathematical model of the ammonia reactor consists of one model for each bed.Each bed is discretized into ten segments,and is represented by a mass and energy balance[8].

    The subscriptqrefers to the quench steam,brefers to the reactor flow before the quench,and mix refers to the reactor flow after the quench.T,C,andMrefer to temperature,concentration of ammonia and mass flow,respectively.

    2.2.Heat exchanger modeling

    The relation between in-and outlet temperatures

    whereTi,Toand ε are reactor in temperature,outlet temperature,and heat exchanger efficiency,respectively.

    Table 1 illustrates previous variable's definition.

    Table 1Variable's definition of NH3 reactor model

    3.Mathematical Model of Urea Synthesis Reactor[9,12–13]

    Eqs.(5)and(6)will be converted as ε1and ε2while the overall conversion as ε1ε2=ε:

    where,FU,FCandFDare the flow rates of urea,carbamate and carbon dioxide respectively.FCi,FDiandFUi,are the initial flow rate of urea,carbamate and carbon dioxide respectively.

    The flow rates of ureaFU,carbamateFC,carbon dioxideFD,ammoniaFN,waterFHand total rateFTwhich is the sum of all individual flow rate are represented as:

    whereaandbare ammonia and water feed ratios which are represented as:

    where,k1Fandk2Fare kinetic for the forward two urea reaction equations,Cd,Cc,andCuare molar flow rates for carbon dioxide,carbamate and urea,andk1andk2are equilibrium constants.

    Eq.(58)is true when the temperature is in the range of 293–405 K but in the range of216–304 Kfor Eq.(59).When the reactor parameters are changed or disturbed,the resultant effect can be described as:slower rate of reactions,reversible reactions,reduction in the concentration of the reactor outlet,change in the NH3/CO2ratio,reduction in the CO conversion,and reduction the quantity of the ammonia and urea productions.So,it is needed to design adaptive and intelligent controllers to overcome the parameter variables and external disturbances that affect the ammonia–urea reactors.In designing a control system for ammonia and urea the system is subjected to:

    –Parameter variations.

    –Parameter uncertainty.

    –Due to the complexity,the system is divided into subsystems.Some subsystems have influences to other subsystems as disturbances.

    –Dynamical system characteristics vary considerably over the operating regime.

    –The system is exposed to a fast and wide range of parameter variations.

    –Some processes are complicated to be modeled.

    So we want to design a smart robust controller as a neural network,fuzzy controller and sliding mode control that can control a process with the previous challenges.In this paper,three design methods are suggested to achieve a stable outlet concentration of ammonia and urea,stable reaction rate,an increase in the conversion of CO to reduce the pollution effect,and an increase in the ammonia and urea productions,keeping the NH3/CO2ratio equal to 3 to reduce the unreacted CO2and NH3,and the two reactors'temperature in operating ranges.The first method is based on AMPC controller techniques.This paper is arranged as follows.Section 2 represents the mathematical model of the ammonia synthesis reactor.Section 3 represents the mathematical model of the urea synthesis reactor.Section 4 describes the suggested method Adaptive Model Predictive Control.Section 5 describes the suggested method Adaptive Neural Network Model Predictive Control.Section 6 describes the suggested method Adaptive neuro-fuzzy sliding mode controller.Section 7 presents the simulation results obtained from the ammonia and urea synthesis reactors.Then,a comparison analysis between three methods and then,a comparison analysis with previously related works were performed.

    4.Adaptive Model Predictive Control(AMPC)

    AMPC is an optimal control method based on optimization techniques.The future output of the plant and the inputs of controller are predicted using a model of system and optimized at regular intervals.A dynamic system is used to predict the future output of the plant.AMPC is shown in Fig.6.The process can be described by a discretetime and state space model[13–16]:

    where,A,B,and C are state matrices,X represents the process state vector,u is the input vector,y is the output vector,Wis the state disturbance,and ξ is the measurement noise.

    The noise ξ and state disturbanceWare independent Gaussian noise with zero mean.

    Φ is a function used to minimize values of squared state and input and includes Q and R,which are separate state and input weight matrices.

    The plant output is used to obtain an optimal state estimate^Xk k.

    where,the optimal inputuk=?Kc^Xkk.

    AMPC uses a Kalman filter(KF)to update the states of the controller.It includes states of the plant,states of the disturbance,and measurement noise states.KF requires two gain matrices L and M.The AMPC controller depends on the variation of noise parameters and disturbance to calculate the L and M gains.

    where,R,N,and Q are constant matrices defined as in the state estimation of AMPC.Pkk?1is the state estimate error matrix at timeKbased onk?1 time.

    5.Adaptive Neural Network Model Predictive Control(ANNMPC)

    Predictive future plant is based on the prediction of the potential control signals to be controlled.The optimization block is used to optimize the plant performance[17].The ANN is used to represent the nonlinear model of the plant.ANNs are adaptable dynamical systems that are less sensitive to variations of parameters,and estimate the functions of input–output.The proposed ANN structure of this adaptive NN controller is shown in Fig.7.The proposed ANNMPC contains two networks,the first network is neural network identification(NNI).The error of NNI between the output of the neuralyNN(k)and the plant outputy(k)is computed by the back propagation algorithm.The error function to be minimized is where,mandnare delayed elements of the vectors U and Y starting from the time instant(k?1).

    The weights of NNI are adjusting the gradient descent in order to minimize the cost functionJ(k).The adapted weights are represented by

    Fig.6.Adaptive Model Predictive Control.

    where α is a learning rate.

    The second network is the neural network predictor(NNP).The new weights of NNP and the optimization block are used to compute the predictive control signal to be applied to the plant.

    6.Adaptive Neuro-fuzzy Sliding Mode Controller(ANFSMC)

    The sliding mode controller(SMC)is robust control to uncertainty plants and insensitive to any disturbances.Unfortunately,the chattering can affect the system behavior significantly[6].Another control method is based on adaptive neuro-fuzzy inference system(ANFIS)with SMC.The flow chart of ANFSMC is shown in Fig.8.The block diagram of ANFSMC is shown in Fig.9.

    ANFIS is divided into artificial neural network(ANN)and fuzzy inference system.The structure proposed of ANFIS consists of five network layers.The structure has two inputs(x,y)and produces one output(f).The ANFIS structure is shown in Fig.10.All nodes in the first layer are adaptive and every node contains a triangle membership function[18–21].

    Fig.7.Adaptive Neural Network Model Predictive Control.

    Fig.8.Flow chart of ANFSMC.

    where,iandlare the numbers of node and layer respectively andOliis the output of the node.

    All nodes in the second layer are fixed and they multiply the signal before outputting as shown in the equation:

    Fig.9.Block diagram of ANFSMC.

    Fig.10.Structure of ANFIS.

    Also,all nodes in the third layer are fixed and perform a normalization of the strength from the second layer.The output of each node is represented as:

    All nodes in the fourth layer are adaptive and the output is calculated from the product of the normalized strength and a first order polynomial and represented as:

    where,p,q,x,y,andr,are parameters which are referred to as consequent parameters in fuzzy rules.Finally,a single node in the fifth layer which is computing the overall output described by:

    The ANFIS controller combines the capability of neural network learning and the ability of fuzzy reasoning.The ANFIS is used to eliminate the chattering introduced by SMC.While,the combination of the fuzzy control and the SMC causes the reduction of the fuzzy rules significantly.ANFSMC is divided into two control signals.The first signal is equivalent controllerueq.Usingsands˙ to ANFIS structure to calculate the equivalent control signal.The second signal is hitting controluh.It is used to eliminate the chattering by using fuzzy logic.Steps of the ANFSMC technique.

    ?Set the input and output of the plant.

    ?Update the parameters of the ANFSMC technique.

    ?Using defuzzification process to calculate the output of the ANFSMC technique.

    ?Output of the controller added with disturbances is given to system.

    ?Repeat the steps until converge the optimal solution.

    7.Simulation Results

    The feed temperature of ammonia reactor has been disturbed for about 30%from the regular feed temperature from 100 min to 230 min.This is shown in Fig.10.The bed3 outlet temperature,ammonia concentration,and CO conversion due to this disturbance are shown in Fig.11.

    When the feed temperature is reduced to 210°C,high bed3 outlet temperature oscillations occurred as shown in Fig.12a.The temperature drop was about 36.5%from the steady state value at 207 min.As in Fig.12b,high oscillations in the ammonia outlet concentration,are reduced to about 63.8%from its steady state value at 207 min.As in Fig.12c,high oscillations in the conversion of CO,are reduced to about 32.4%from the normal value.The system has oscillations for about 200 min.Also,this disturbance that has occurred in the ammonia reactor has a strong influence and impact significantly on the urea reactor.The ratio of ammonia to carbon dioxide is affected by this disorder,as shown in Fig.12d.The NH3/CO2ratio is reduced to 0.497.The urea reactor temperature is also shown in Fig.12e,with a high oscillation,which is reduced to about 35.75%from its steady state stable value.The urea concentration is shown in Fig.12f,with high oscillation and drop from 29.6%to 9.2%at 207 min.All oscillations that occurred in the ammonia urea reactor gradually increased at the beginning of the disturbance that has happened to the feed temperature,then gradually decreased when removing this disturbance.

    Fig.11.Disturbed feed temperature.

    The originality of this paper is dealing with the total production line of urea starting from nitrogen and hydrogen then ammonia production till urea outcome with external disturbances and parameter variations to achieve the following objectives:

    1.Stable ammonia concentration.

    2.Reduce pollution effect by increase the CO2conversion.

    3.Increase the ammonia and urea production.

    The most diffuse model for ammonia and urea is used,with compatibility with several reactors in our country.

    Fig.12.Disturbed ammonia urea reactors.a.Bed3 outlet temperature.b.NH3 concentration ofbed3.c.COto CO2 conversion(%).d.NH3/CO2 ratio.e.Urea reactor temperature(°C).f.Urea concentration.For the proposed Adaptive Model Predictive Control,disturbed ammonia–urea reactors using AMPC for temperature disturbance are shown in Fig.13.Bed3 outlet temperature curve is shown in Fig.13a.It has oscillations from 100 to 130 min and from 210 to 227 min,which reach their minimum value at 431.5 °C then stabilize at 500 °C.The NH3 concentration of the bed3 curve is shown in Fig.13b.It has very small oscillation for about 15.3 min then the system stabilizes with 31.38%,at 150 min the system stabilizes with 30.11%NH3 concentration.The conversion of CO is shown in Fig.13c.It has small oscillation and it has dropped for about 3.9%from its steady state value then the system stabilizes with 96.5%.The NH3/CO2 ratio curve is shown in Fig.13d.It has very small oscillations at 100 min,which reach their minimum value at 2.9.The urea reactor temperature and urea concentration are shown in Fig.13e–f.There are small oscillations.The urea reactor temperature is stabilized at 181.6 °C.The urea concentration is about 30.23%.The system is stabilized approximately after 15 min.

    The control algorithm is better due to online learning for the neural network and fuzzy logic rather than using off-line learning as several issues.Table 2 presents the comparison analysis between different related works and the three suggested methods.It's observed that the proposed Adaptive neuro-fuzzy sliding mode controller succeeded in avoiding the external disturbances and parameter variations with the adaptive neural network model predictive controller proposed and Adaptive Model Predictive Control.Every control system can overcome the disturbances till certain limits.Due to the differences operation status,resources availability,technology in several ammonia and urea manufacturers,ANFSMC has achieved the fast recovery time without any oscillations.It takes 2.7 min to overcome the disturbance.While AMPC controller takes 15 min,and ANNMPC takes 11 min.The NH3concentration is 31.3%for ANFSMC,30.11%for AMPC,and 30.4%for ANNMPC.The bed3 outlet temperature is 510 °C for ANFSMC,500 °C for AMPC,and 502°C for ANNMPC.The CO conversion is 98.7%for ANFSMC,96.5%for AMPC,and 97.2%for ANNMPC.The urea concentration is 31.34%for ANFSMC,30.23%for AMPC,and 30.4%for ANNMPC.The urea reactor temperature is 182 °C for ANFSMC,181.6 °C for AMPC,and 182 °C for ANNMPC.Moreover,this result is compared to other publications for more validation.F.G.Areedet al.,presented a PID and Decoupled Sliding Mode Controller(DSMC)to overcome the disturbances of the ammonia reactor[6].For the PID controller the system suffers from high temperature oscillations.The controller takes a longtime 105 min which is a maximum time compared with the three suggested controllers to overcome the system.The NH3concentration was 29.3%which is still lower than the proposed controllers.For the DSMC the system suffers from low temperature oscillations.The controller takes a longtime 26.5 min which is a maximum time compared with the three suggested controllers to overcome the system.The NH3concentration was 28.94%which is still lower than the proposed controllers.E.Holter and M.Hovd proposed a feed forward controller to overcome the disturbances of the ammonia reactor[7].The system suffers from temperature oscillations with ripples±1%.The controller takes a longtime 50 min which is a maximum time compared with the three suggested controllers to overcome the system.Then the oscillations are reduced to±0.5%.

    Fig.13.Disturbed ammonia–urea reactors using AMPC for temperature disturbance.a.Disturbed NH3outlet temperature using AMPC for temperature disturbance.b.Disturbed NH3 concentration using AMPC for temperature disturbance.c.Disturbed CO conversion using AMPC for temperature disturbance.d.Disturbed NH3/CO2 ratio using AMPC for temperature disturbance.e.Disturbed urea reactor temperature using AMPC for temperature disturbance.f.Disturbed urea concentration using AMPC for temperature disturbance.For the proposed adaptive neural network model predictive controller,disturbed ammonia–urea reactors using ANNMPC for temperature disturbance are shown in Fig.14.There are lightly oscillations at the start of the disturbances.The bed3 outlet temperature is stabilized at 502°C as shown in Fig.14a.The NH3 concentration of bed3 is about 30.4%as shown in Fig.14b.The conversion of CO is stabilized at 97.2%as shown in Fig.14c.The NH3/CO2 ratio curve is shown in Fig.14d.It has very small oscillations at 100 min,which reach their minimum value at 2.976.The urea reactor temperature and urea concentration are shown in Fig.13e–f.They have small oscillations.The urea reactor temperature is stabilized at 182 °C.The urea concentration is about 30.4%.The system is stabilized approximately after 11 min.

    Fig.14.Disturbed ammonia–urea reactors using ANNMPC for temperature disturbance.a.Disturbed NH3 outlet temperature using ANNMPCfor temperature disturbance.b.Disturbed NH3 concentration using ANNMPC for temperature disturbance.c.Disturbed CO conversion using ANNMPC for temperature disturbance.d.Disturbed NH3/CO2 ratio using ANNMPC for temperature disturbance.e.Disturbed urea reactor temperature using ANNMPC for temperature disturbance.f.Disturbed urea concentration using ANNMPC for temperature disturbance.For the proposed Adaptive neuro-fuzzy sliding mode controller,disturbed ammonia–urea reactors using ANFSMC for temperature disturbance are shown in Fig.15.There are not any oscillations.The bed3 outlet temperature is stabilized at 510°C as shown in Fig.15a.The NH3 concentration of bed3 is about 31.3%as shown in Fig.15b.The conversion of CO is stabilized at 98.7%as shown in Fig.15c.The NH3/CO2 ratio curve is shown in Fig.15d.The urea reactor temperature and urea concentration are shown in Fig.15e–f.The urea reactor temperature is stabilized at 182°C.The urea concentration is about 31.34%.The system is stabilized approximately after 2.7 min.

    8.Conclusions

    The kinetic model for ammonia synthesis and urea synthesis reactors in industrial scale was developed and simulated by MATLAB in the present paper.AMPC,ANNMPC,and ANFSMC are suggested to overcome the external disturbances and parameter variations in order to increase the CO conversion to reduce the pollution effect in such ammonia reactor,increase the CO2which is used in urea production and reduce the unreacted NH3and CO2,and an AMPC controller was tested to compare with the ANNMPC and ANFSMC.ANFSMC is a robust controller that can overcome several challenges as mentioned before,and due to its smartness and learning capability,ANFSMC is preferred in several control problems.

    Fig.15.Disturbed ammonia–urea reactors using ANFSMC for temperature disturbance.a.Disturbed NH3 outlet temperature using ANFSMC for temperature disturbance.b.Disturbed NH3 concentration using ANFSMC for temperature disturbance.c.Disturbed CO conversion using ANFSMC for temperature disturbance.d.Disturbed NH3/CO2 ratio using ANFSMC for temperature disturbance.e.Disturbed urea reactor temperature using ANFSMC for temperature disturbance.f.Disturbed urea concentration using ANFSMC for temperature disturbance.

    Table 2The comparison analysis between different related works and the three suggested methods

    [1]J.Morud,S.Skogestad,Analysis of instability in an industrial ammonia reactor,AICHE J.44(4)(1998)888–895.

    [2]R.Angira,Simulation and Optimization of an Auto-thermal Ammonia Synthesis Reactor,International Journal of Chemical Reactor Engineering9(1)(2014)42–43.

    [3]A.Murase,H.Roberts,A.Converse,Optimal thermal design of an autothermal ammonia synthesis reactor,Ind.Eng.Chem.Process Des.Dev.9(4)(1970)503–513.

    [4]R.Baddour,P.Brian,B.Logeais,J.Eymery,Steady-state simulation of an ammonia synthesis converter,Chem.Eng.Sci.20(4)(1965)281–292.

    [5]T.Mohammad,K.Azam,The Optimization of an Ammonia Synthesis Reactor Using Genetic Algorithm,International Journal of Chemical Reactor Engineering,6(A113),2008.

    [6]F.G.Areed,M.A.Badr,S.F.Saraya,M.S.Elksasy,M.M.Abdelsalam,Decoupled sliding mode control for a multivariable nonlinear system,Int.J.Comput.Appl.55(6)(2012)25–32.

    [7]E.Holter,M.Hovd,Feedforward for stabilization of an ammonia synthesis reactor,MSc.Thesis,Norwegian University of Science and Technology,Norwegian,2010.

    [8]Z.Umair,et al.,Kinetic model for ammonia and urea production processes,International Conference on Process Systems Engineering,Elsevier 2013,pp.25–27.

    [9]M.A.Satyro,et al.,Modelling urea processes:A new thermodynamic model and software integration paradigm,Chem.Eng.(2003).

    [10]M.M.Saafan,M.M.Abdelsalam,M.S.Elksasy,et al.,A sliding mode controller for urea plant,IJCSIS J.14(3)(2016)115–126.

    [11]M.Frejacques,Theoretical basis of the industrial synthesis of urea,Chimie et Industrie.60(1)(1948)22–35.

    [12]S.Zendehboudi,G.Zahedi,A.Bahadori,et al.,A dual approach for modelling and optimization of industrial urea reactor,Can.J.Chem.Eng.92(3)(2014)469–485.

    [13]I.M.Fahmy,et al.,Real-time control of industrial urea evaporation process using model predictive control,Chem.Eng.Process Technol.(2015),http://dx.doi.org/10.4172/2157-7048.1000227(2015).

    [14]O.Mauricio,A.Manozca,J.J.Espinosa,J.Vandewalle,Control of the synthesis section of a urea plant by means of an MPC controller,Computer Aided Chemical Engineering21(06)(2006)1305–1310.

    [15]A.Alanqar,M.Ellis,P.D.Christo fides,Economic model predictive control of nonlinear process systems using empirical models,AIChE J61(3)(2015)4953–4958.

    [16]M.Shyamalagowri,et al.,Model predictive control design for nonlinear process control reactor case study,IOSR J.7(1)(2013)88–94.

    [17]Z.Yu,Y.Liang,Design and realization of optimization system of urea production process based on BP neural network,ICCASM,10,2010,pp.368–371.

    [18]M.Khaki,I.Yusoff,N.Islami,Application of the artificial neural network and neurofuzzy system for assessment of groundwater quality,Clean-Soil,Air,Water43(4)(2015)551–560.

    [19]A.Al-Hmouz,J.Shen,R.Al-Hmouz,J.Yan,Modeling and simulation of an adaptive neuro-fuzzy inference system(ANFIS)for mobile learning,IEEE Trans.Learn.Technol.5(3)(2012)226–237.

    [20]A.Swetapadma,A.Yadav,High-speed directional relaying using adaptive neurofuzzy inference system and fundamental component of currents,IEEJ Transactions on Electrical and Electronic Engineering10(6)(2015)653–663.

    [21]L.Hung,H.Chung,Decoupled sliding-mode with fuzzy-neural network controller for nonlinear systems,Int.J.Approx.Reason.46(2007)74–97.

    成人亚洲欧美一区二区av| 亚洲精品乱久久久久久| av天堂中文字幕网| 色尼玛亚洲综合影院| 99久久精品国产国产毛片| 免费看美女性在线毛片视频| 97人妻精品一区二区三区麻豆| 大陆偷拍与自拍| 一级爰片在线观看| 午夜爱爱视频在线播放| 色综合站精品国产| 91av网一区二区| 夜夜看夜夜爽夜夜摸| 国产色爽女视频免费观看| 人妻少妇偷人精品九色| 国产亚洲午夜精品一区二区久久 | 国产伦精品一区二区三区四那| 国产单亲对白刺激| 丝袜美腿在线中文| 免费播放大片免费观看视频在线观看| 男女边摸边吃奶| 日韩欧美 国产精品| 久久久久久久久久人人人人人人| 十八禁国产超污无遮挡网站| 亚洲欧美一区二区三区国产| 国产在视频线精品| 少妇裸体淫交视频免费看高清| 日韩欧美三级三区| 男女边摸边吃奶| 波野结衣二区三区在线| 欧美潮喷喷水| 精品久久久久久久久久久久久| av女优亚洲男人天堂| 国产av国产精品国产| 国产黄a三级三级三级人| 久久精品夜夜夜夜夜久久蜜豆| 国产午夜精品久久久久久一区二区三区| 亚洲经典国产精华液单| 久久久久免费精品人妻一区二区| 最近中文字幕2019免费版| 伊人久久精品亚洲午夜| 十八禁国产超污无遮挡网站| 韩国高清视频一区二区三区| 久久国内精品自在自线图片| 嫩草影院入口| 青春草亚洲视频在线观看| 久久久久久久久大av| av免费观看日本| 亚洲av日韩在线播放| 精品国产一区二区三区久久久樱花 | 日韩成人av中文字幕在线观看| 美女国产视频在线观看| 国产极品天堂在线| 国产永久视频网站| 网址你懂的国产日韩在线| 日韩欧美一区视频在线观看 | 九九爱精品视频在线观看| 国产精品一区www在线观看| 在线观看美女被高潮喷水网站| 五月玫瑰六月丁香| 国产永久视频网站| 精品熟女少妇av免费看| 中国美白少妇内射xxxbb| 极品少妇高潮喷水抽搐| 久久草成人影院| 秋霞在线观看毛片| 高清欧美精品videossex| 国产高清有码在线观看视频| av又黄又爽大尺度在线免费看| 亚洲欧美日韩无卡精品| 国产精品国产三级专区第一集| 亚洲第一区二区三区不卡| 一个人看视频在线观看www免费| 男女那种视频在线观看| 成人二区视频| 简卡轻食公司| 好男人在线观看高清免费视频| 国产伦理片在线播放av一区| 欧美3d第一页| 男人爽女人下面视频在线观看| 嫩草影院新地址| 一夜夜www| 色哟哟·www| av又黄又爽大尺度在线免费看| 亚洲国产精品专区欧美| 精华霜和精华液先用哪个| 国产极品天堂在线| 亚洲成人精品中文字幕电影| 国产女主播在线喷水免费视频网站 | 免费观看在线日韩| 亚洲经典国产精华液单| av在线亚洲专区| 亚洲精华国产精华液的使用体验| 国产大屁股一区二区在线视频| 天堂俺去俺来也www色官网 | 99热全是精品| 午夜激情欧美在线| 插阴视频在线观看视频| 国产黄色小视频在线观看| 中文精品一卡2卡3卡4更新| 日本午夜av视频| 午夜福利在线在线| 2021天堂中文幕一二区在线观| 免费观看在线日韩| 国产毛片a区久久久久| 麻豆成人av视频| 国产伦一二天堂av在线观看| 成人欧美大片| 免费少妇av软件| 亚洲美女搞黄在线观看| 久久久久性生活片| 日韩亚洲欧美综合| 精品久久久久久久末码| 美女内射精品一级片tv| 国产av国产精品国产| 最近视频中文字幕2019在线8| 九色成人免费人妻av| 久久久成人免费电影| 久久久久精品久久久久真实原创| 婷婷色综合大香蕉| 91午夜精品亚洲一区二区三区| 在线天堂最新版资源| 国产精品一区二区三区四区免费观看| 99re6热这里在线精品视频| 亚洲最大成人中文| 亚洲精品一二三| 国产伦精品一区二区三区四那| 一个人看的www免费观看视频| 成年女人看的毛片在线观看| 久久精品国产自在天天线| 国产三级在线视频| 亚洲精品aⅴ在线观看| 亚洲精品aⅴ在线观看| freevideosex欧美| 欧美日韩精品成人综合77777| www.色视频.com| 亚洲国产欧美在线一区| 日韩欧美精品免费久久| 男人狂女人下面高潮的视频| 成人午夜精彩视频在线观看| .国产精品久久| 亚洲精品自拍成人| 亚洲国产av新网站| 在线观看美女被高潮喷水网站| 国产亚洲av嫩草精品影院| 男人狂女人下面高潮的视频| 国产欧美日韩精品一区二区| 日本-黄色视频高清免费观看| 搡女人真爽免费视频火全软件| 日本爱情动作片www.在线观看| 2021少妇久久久久久久久久久| 久久精品夜色国产| 极品教师在线视频| 偷拍熟女少妇极品色| 精品一区二区三区视频在线| 国产色婷婷99| 99视频精品全部免费 在线| 久久久久精品久久久久真实原创| 在线观看免费高清a一片| 女的被弄到高潮叫床怎么办| 国产激情偷乱视频一区二区| 国产成人精品福利久久| 国产亚洲5aaaaa淫片| 亚洲欧洲日产国产| 欧美激情在线99| 午夜精品在线福利| videos熟女内射| 波多野结衣巨乳人妻| 国产精品1区2区在线观看.| 成人毛片60女人毛片免费| 欧美日韩精品成人综合77777| 午夜亚洲福利在线播放| 欧美3d第一页| 天天一区二区日本电影三级| 中文欧美无线码| 一级二级三级毛片免费看| 成人无遮挡网站| 最近2019中文字幕mv第一页| 精品人妻一区二区三区麻豆| 亚洲无线观看免费| 中文字幕av成人在线电影| 久久久午夜欧美精品| 亚洲怡红院男人天堂| 午夜亚洲福利在线播放| 国产精品一二三区在线看| 亚洲不卡免费看| 国产又色又爽无遮挡免| 国产av码专区亚洲av| 菩萨蛮人人尽说江南好唐韦庄| 日韩av在线大香蕉| 精品99又大又爽又粗少妇毛片| 伦理电影大哥的女人| 街头女战士在线观看网站| 亚洲最大成人av| 久久精品国产亚洲网站| 国产欧美日韩精品一区二区| 精品人妻偷拍中文字幕| 亚洲av.av天堂| 97在线视频观看| 女人十人毛片免费观看3o分钟| 久久久久精品性色| 女的被弄到高潮叫床怎么办| 亚洲国产精品国产精品| 国产精品1区2区在线观看.| av天堂中文字幕网| 国产极品天堂在线| 有码 亚洲区| 高清欧美精品videossex| videos熟女内射| 欧美区成人在线视频| 又大又黄又爽视频免费| 国产高清国产精品国产三级 | 99久久九九国产精品国产免费| 亚洲欧美精品专区久久| 国产精品日韩av在线免费观看| 综合色av麻豆| 亚洲精品日本国产第一区| 十八禁国产超污无遮挡网站| 波野结衣二区三区在线| 婷婷色综合大香蕉| 亚洲精品久久午夜乱码| 在线免费观看不下载黄p国产| 午夜精品一区二区三区免费看| 网址你懂的国产日韩在线| 午夜日本视频在线| 成人国产麻豆网| 国产综合精华液| 一本一本综合久久| 18禁动态无遮挡网站| 精品国内亚洲2022精品成人| 日韩,欧美,国产一区二区三区| 国产乱人视频| 热99在线观看视频| 久久久国产一区二区| 亚洲aⅴ乱码一区二区在线播放| 久久久久精品性色| av线在线观看网站| 亚洲,欧美,日韩| 国产精品久久久久久久电影| 国产一区亚洲一区在线观看| 午夜精品在线福利| 日日啪夜夜爽| 人妻一区二区av| 亚洲经典国产精华液单| 亚洲精品日韩av片在线观看| 亚洲av电影在线观看一区二区三区 | 亚洲第一区二区三区不卡| 国产伦精品一区二区三区视频9| 好男人视频免费观看在线| 观看免费一级毛片| 精品国产一区二区三区久久久樱花 | 日韩中字成人| 最后的刺客免费高清国语| 黄色欧美视频在线观看| 亚洲精品日韩在线中文字幕| 视频中文字幕在线观看| 色尼玛亚洲综合影院| 麻豆久久精品国产亚洲av| 丰满乱子伦码专区| 亚洲国产成人一精品久久久| 高清午夜精品一区二区三区| 久久精品综合一区二区三区| 亚洲精品第二区| 亚洲国产日韩欧美精品在线观看| 亚洲精品久久久久久婷婷小说| 有码 亚洲区| 大片免费播放器 马上看| 久99久视频精品免费| 亚洲国产色片| 又爽又黄无遮挡网站| 一个人看视频在线观看www免费| 亚洲av成人精品一二三区| 免费观看无遮挡的男女| 真实男女啪啪啪动态图| av福利片在线观看| 亚洲伊人久久精品综合| 亚洲成人久久爱视频| 久久精品国产亚洲av涩爱| 九九久久精品国产亚洲av麻豆| 国语对白做爰xxxⅹ性视频网站| 精品国内亚洲2022精品成人| 精品人妻偷拍中文字幕| 亚洲av一区综合| 午夜福利成人在线免费观看| 寂寞人妻少妇视频99o| 天美传媒精品一区二区| 免费黄频网站在线观看国产| 久久久午夜欧美精品| 精品久久久久久久末码| 最近最新中文字幕大全电影3| 成人毛片60女人毛片免费| 听说在线观看完整版免费高清| 成人二区视频| av天堂中文字幕网| 日日撸夜夜添| av女优亚洲男人天堂| 一个人看视频在线观看www免费| 又大又黄又爽视频免费| 国产淫片久久久久久久久| 免费av观看视频| 亚洲真实伦在线观看| 偷拍熟女少妇极品色| 国产高清三级在线| 亚洲乱码一区二区免费版| 国模一区二区三区四区视频| 日本av手机在线免费观看| 久久草成人影院| 国产av码专区亚洲av| 在线免费观看的www视频| 国产亚洲av嫩草精品影院| 可以在线观看毛片的网站| 国产亚洲5aaaaa淫片| av在线播放精品| 亚洲不卡免费看| 国产午夜福利久久久久久| 激情五月婷婷亚洲| 午夜精品国产一区二区电影 | 观看免费一级毛片| freevideosex欧美| 在线观看免费高清a一片| 天堂网av新在线| 亚洲精品日韩av片在线观看| 春色校园在线视频观看| 18+在线观看网站| 国产成人91sexporn| 成年人午夜在线观看视频 | 欧美区成人在线视频| 日韩制服骚丝袜av| 51国产日韩欧美| 国产精品99久久久久久久久| 久久久久精品久久久久真实原创| 国内精品宾馆在线| 尾随美女入室| 日韩不卡一区二区三区视频在线| 午夜精品在线福利| 日韩成人伦理影院| 亚洲伊人久久精品综合| 97精品久久久久久久久久精品| 亚洲在久久综合| 亚洲国产欧美在线一区| 干丝袜人妻中文字幕| 国产精品1区2区在线观看.| 色哟哟·www| 晚上一个人看的免费电影| 女人久久www免费人成看片| 一边亲一边摸免费视频| 免费在线观看成人毛片| 亚洲成色77777| 夫妻午夜视频| 如何舔出高潮| 国产免费一级a男人的天堂| 美女xxoo啪啪120秒动态图| 亚洲精品乱码久久久v下载方式| 边亲边吃奶的免费视频| 两个人视频免费观看高清| 中文在线观看免费www的网站| 久久热精品热| 国产高清有码在线观看视频| 日韩电影二区| 在线观看一区二区三区| 国产片特级美女逼逼视频| 亚洲最大成人中文| 久久久久久久久久黄片| 又爽又黄无遮挡网站| 欧美日韩一区二区视频在线观看视频在线 | 人妻一区二区av| 日日撸夜夜添| 超碰97精品在线观看| 波野结衣二区三区在线| 国产高清国产精品国产三级 | 免费看美女性在线毛片视频| 日本三级黄在线观看| 精品人妻偷拍中文字幕| 18禁在线无遮挡免费观看视频| av专区在线播放| 边亲边吃奶的免费视频| 日日摸夜夜添夜夜爱| 夜夜爽夜夜爽视频| 26uuu在线亚洲综合色| 中文乱码字字幕精品一区二区三区 | eeuss影院久久| 18禁在线无遮挡免费观看视频| 国产午夜精品论理片| 精品久久久久久电影网| 国产成人aa在线观看| 免费看光身美女| 麻豆国产97在线/欧美| 欧美日韩亚洲高清精品| 永久网站在线| 女人十人毛片免费观看3o分钟| 秋霞伦理黄片| 精华霜和精华液先用哪个| 久久久久精品性色| www.av在线官网国产| 国产男人的电影天堂91| 欧美 日韩 精品 国产| 男女啪啪激烈高潮av片| 一个人观看的视频www高清免费观看| 国产激情偷乱视频一区二区| 99热网站在线观看| 色视频www国产| 精品熟女少妇av免费看| 一区二区三区四区激情视频| 又黄又爽又刺激的免费视频.| 欧美日韩在线观看h| 国产在视频线精品| 99re6热这里在线精品视频| 少妇的逼好多水| 熟妇人妻不卡中文字幕| 午夜日本视频在线| 亚洲综合精品二区| 在线播放无遮挡| 日韩大片免费观看网站| 国产视频内射| 国产一级毛片七仙女欲春2| 69人妻影院| 99久久精品一区二区三区| 精华霜和精华液先用哪个| 日韩亚洲欧美综合| 黄色一级大片看看| 丰满乱子伦码专区| 国产 一区 欧美 日韩| 3wmmmm亚洲av在线观看| 97精品久久久久久久久久精品| 老司机影院毛片| 亚洲最大成人av| 尤物成人国产欧美一区二区三区| av网站免费在线观看视频 | 麻豆av噜噜一区二区三区| 老司机影院毛片| 日韩成人av中文字幕在线观看| 久久久久久久久久久丰满| 欧美精品一区二区大全| 日日干狠狠操夜夜爽| 小蜜桃在线观看免费完整版高清| 久久精品国产自在天天线| 最近2019中文字幕mv第一页| 日韩国内少妇激情av| 日韩制服骚丝袜av| 非洲黑人性xxxx精品又粗又长| 亚洲经典国产精华液单| 亚洲国产欧美人成| 欧美高清性xxxxhd video| 91aial.com中文字幕在线观看| 成人av在线播放网站| 国产男女超爽视频在线观看| 国产精品女同一区二区软件| 国产亚洲av片在线观看秒播厂 | 精品久久国产蜜桃| 中文字幕人妻熟人妻熟丝袜美| 亚洲av免费在线观看| 亚洲av中文字字幕乱码综合| 国产综合精华液| 成年版毛片免费区| 久久久久九九精品影院| 日韩精品青青久久久久久| 欧美丝袜亚洲另类| 美女主播在线视频| 99热这里只有精品一区| 黑人高潮一二区| 中文字幕av在线有码专区| 超碰av人人做人人爽久久| 亚洲av二区三区四区| 国产亚洲精品久久久com| 中文天堂在线官网| 国产精品久久久久久久久免| 一本—道久久a久久精品蜜桃钙片 精品乱码久久久久久99久播 | 秋霞在线观看毛片| 中文字幕av成人在线电影| 国产乱来视频区| 精品国产三级普通话版| 精品久久久噜噜| 听说在线观看完整版免费高清| 日本熟妇午夜| 国产亚洲最大av| 赤兔流量卡办理| xxx大片免费视频| 狠狠精品人妻久久久久久综合| 日韩欧美精品v在线| 能在线免费看毛片的网站| 亚洲欧美成人精品一区二区| 日韩精品有码人妻一区| 午夜精品在线福利| av天堂中文字幕网| 亚洲国产最新在线播放| 欧美激情久久久久久爽电影| 99九九线精品视频在线观看视频| 赤兔流量卡办理| 一级片'在线观看视频| 在线观看人妻少妇| 国产精品综合久久久久久久免费| 一区二区三区乱码不卡18| 成人亚洲欧美一区二区av| 免费观看无遮挡的男女| 99re6热这里在线精品视频| 免费大片18禁| 国产精品不卡视频一区二区| 国产视频首页在线观看| 成人美女网站在线观看视频| 免费电影在线观看免费观看| 国产在视频线在精品| 日韩伦理黄色片| 亚洲人成网站在线播| 欧美日韩在线观看h| 99热这里只有是精品在线观看| 亚洲精品自拍成人| 亚洲精品成人av观看孕妇| 日韩av不卡免费在线播放| 最近的中文字幕免费完整| 边亲边吃奶的免费视频| 久99久视频精品免费| 一区二区三区高清视频在线| 老司机影院成人| 日本免费在线观看一区| 水蜜桃什么品种好| 最新中文字幕久久久久| 免费人成在线观看视频色| 偷拍熟女少妇极品色| 色网站视频免费| 成人性生交大片免费视频hd| 国产 一区精品| 成人美女网站在线观看视频| 超碰av人人做人人爽久久| 你懂的网址亚洲精品在线观看| 在线免费观看不下载黄p国产| 亚洲美女视频黄频| a级一级毛片免费在线观看| 一区二区三区四区激情视频| 大话2 男鬼变身卡| 女的被弄到高潮叫床怎么办| 九色成人免费人妻av| 亚洲欧美日韩东京热| 99久久精品国产国产毛片| 只有这里有精品99| 美女高潮的动态| eeuss影院久久| 中文字幕制服av| 男人舔女人下体高潮全视频| 亚洲精品456在线播放app| 亚洲av中文av极速乱| 久久精品国产亚洲网站| 亚洲精品久久久久久婷婷小说| 99热这里只有是精品在线观看| av在线观看视频网站免费| 天堂俺去俺来也www色官网 | 综合色av麻豆| 中文资源天堂在线| 青青草视频在线视频观看| 熟妇人妻久久中文字幕3abv| 成人午夜高清在线视频| 伦理电影大哥的女人| 亚洲婷婷狠狠爱综合网| 国产精品久久久久久久电影| 日本欧美国产在线视频| 午夜福利高清视频| 少妇人妻精品综合一区二区| 国产午夜精品久久久久久一区二区三区| 国产精品一二三区在线看| 亚洲精品成人久久久久久| 亚洲精品国产av蜜桃| 精品国产三级普通话版| 伊人久久国产一区二区| 久久久久久伊人网av| 国产一区有黄有色的免费视频 | 男人和女人高潮做爰伦理| 22中文网久久字幕| 人妻系列 视频| 久久精品久久久久久噜噜老黄| 亚洲av电影不卡..在线观看| 久久精品夜夜夜夜夜久久蜜豆| 日本黄色片子视频| 国产高清不卡午夜福利| 99久国产av精品国产电影| 男的添女的下面高潮视频| 国产中年淑女户外野战色| 街头女战士在线观看网站| 午夜精品国产一区二区电影 | 看免费成人av毛片| 97在线视频观看| 国产伦在线观看视频一区| 一级毛片 在线播放| 精品亚洲乱码少妇综合久久| 精华霜和精华液先用哪个| 国产老妇伦熟女老妇高清| 天天躁夜夜躁狠狠久久av| 国产精品av视频在线免费观看| 国产 一区 欧美 日韩| 在线观看一区二区三区| 极品少妇高潮喷水抽搐| 男女啪啪激烈高潮av片| 色视频www国产| 国内揄拍国产精品人妻在线| 日本三级黄在线观看| 免费高清在线观看视频在线观看| 啦啦啦啦在线视频资源| 国产伦精品一区二区三区视频9| 精品一区二区三区视频在线| 中文欧美无线码| 久热久热在线精品观看| 夜夜看夜夜爽夜夜摸| 麻豆乱淫一区二区| 少妇被粗大猛烈的视频| 欧美 日韩 精品 国产| 亚洲怡红院男人天堂| 特大巨黑吊av在线直播| 久久这里只有精品中国| 极品教师在线视频| 亚州av有码| 日日摸夜夜添夜夜添av毛片| 男女国产视频网站| 国产一区二区三区综合在线观看 | 久久久久久国产a免费观看| 亚洲丝袜综合中文字幕| 少妇裸体淫交视频免费看高清| 精品少妇黑人巨大在线播放|