The design for the fuzzy PID control of the intelligent following vehicle with gas floating

第六届全国通信新理论与新技术学术大会(CC2012)优秀论文TTelecom marketThe design for the fuzzy PId control of the intelligentfollowing vehicle with gas floatingHe Yi Song Xiaodong Chen Ming(Beijing Institute of Technology, Beijing, China, 10081)Abstract: The intelligent following vehicle with gas floating has the characteristics ofplicatedstructure and large quality. In this paper the author first establish the mathematicalodeof themotion system land and then design a controller using the fuzzy PID control method which couldrealize auto-tuning PID parameters. By the MATLAB simulation analysis, the results show thatfuzzy self-tuning PID control can enhance the response speed of system and has a betteradaptability.Index Terms: fuzzy control, fuzzy PID, DC motor, mathematical model, PID parametersI. IntroductionThe spatial intelligent vehicle with gas floating is a kind of mechanism suitable for largeexpansion. It could move in the 3d space unfolding its intelligently following mechanism. Thisspatial intelligent vehicle includes the intelligent following vehicle, the small gas floating platformand gravity balance device. The comprehensive function of the intelligent follow car and gravitybalance system realized the car small gas floating platform for large expansion of mechanismmotion in space. It accomplishes laboratory simulation of outer space mechanics environment inground with no expansion mechanism which could be additional force and additional moment. Sot could make sure good precision and accuracy during the ground testis to simulate the usualappearance of papers in a Journal of the Engineering and Technology Publishing. We arerequesting that you follow these guidelines as closely as possibleI. The mathematical ModelThe motion control system of the vehicle with brainpower is composed of controller, DC motorretarder and the vehicle etc4. The Mathematical Model Of The DC MotorThe servo motor is the actuator of the servo system. The actuator is supposed to turn electricroller(armature)is mostly used in driving the exterior burden The figure is as Fge erated by theenergy into mechanical energy by the retarder driving the roller. The torque gerFigure 1. The model of the DC motorThe transfer function of the DC motor is an approximately describe to concrete motor as alinear model. Some high-order influence such as the voltage of the brush could be ignored. Theinput voltage could be used both in the magnetic field and the armatureThe mathematic model of the motor could be educed as Fig. 2中国煤化工CNMHG通信市场·2012年1-12月第140页第六届全国通信新理论与新技术学术大会(CTC2012)优秀论文Telecom marketT(s)Js+BFigure 2. The block diagram of the transfer function of the DC servo-motorU,-the voltage of armature,V;L, -the inductance of the armature, HR,the resistance of the armature,n;K-the electromagnetism torque coefficient of the DC motor, N.m/AT-the electromagnetism torque of the DC motor, N.m;T-the hinder torque effected on the roller of the DC motor, including the hinder torque ofthe DC motor and the hinder torque of the burden, N.m;J-the equivalent rotary inertia of the hinder torque of dC motor and the hinder torque ofthe burden,Kg·mB-the equivalent viscous damping coefficient of the roller of the DC motor;The equivalent system diagram of the burden model of the DC motor is as Fig 3Figure 3. The model of the DC motorIn this diagram, K is the coefficient of the system. The quality of the vehicle body is M. Inconsider of the burden and the stiffness, the analysis of the system is as follows.The mechanical deformation of the system will generate an elastic force, which is named FF=Ky=k VdtAfter Laplace tansform, it is turmed to beF(s)=KY(s)/sIn this formular is the radius of the driving wheel, 2 is the anglerAmong which, V, is the velocity of the vehicle body. After Laplace tansform, it is turmed toThe moment of resistance, After Laplace tansform, it is turned to beT(s)=F(S)RThe structure diagram of the system is as followsHakFigure 4. The model of the DC motor中国煤化工Generally speaking, the inductance and the viscosity coefficientCNMHGsmall,which can be approximated to zero. According to the structure of tneen uausiuilllation,the system transfer function is finally obtained for通信市场2012年11-12月第141页第六届全国通信新理论与新技术学术大会(CTC2012)优秀论文Telecom marketKRKG(s)=JR。MS3+KKM32+RKRM(5)Substituting the related data aG(s)2.726s3+1212275s2+24192sII. The Parameters of The Pld ControllerThe conventional PID controller has the algorithm such as simple, good stability, highreliability. It could be easily designed and widely used. So it is the mostly popular controller inthe process control. It could get satisfactory control effect in all kinds of linear fixed length systemcontrol, especially suitable for the system whose controlled object parameters fixed and thenonlinear is not very serious systemHowever, in the real industrial production process, the load of the controlled object ischangeable and the interference facts are complicated. We need PID parameters constantlyadjusted online to obtain get satisfactory control effect. Sometimes because of the chop andchange of these parameters, there is no mathematical model and rules to follow. It would be a kindof practical, simple and feasible way to use fuzzy controller to control the parameters. The fuzzycontroller could make use of the Real-time nonlinear control experience of the operator. It couldalso give full play to the Pid controller of good control effect to make sure the system could getthe best control effectSuppose the input of the PID controller is u(r), the output is e(n), the relations between themare as followsu(i Kpe()+k, Ce(ride+k de(r)In which, Kp is proportional gain, K is integral gain, Kp is differential gain. For the bettercontrol effect, these three parameters should be true time control according to the system state. Ifthe control object is known advanced, we usually accomplish this task by the way of the onlineidentification. But it dose not work in the conditions that interference and the load are changeableIn this way, we should use the fuzzy controller to adjust it. And it turned out to be a very good anduseful methodBy the accumulation of a large number of operation experience, we knparameters are related to the deviation input of the controller e(t) and the deviationchange de(t)dt. And this relation is non-linear, which is not easy to describe by clearmathematical expressions. We could use the fuzzy language insteadB. The Fuzzy Rules OfAdjust The Parameters Of PID ControllerThrough the multiple summing-up of experience or multiple data processing, we could get atheory with theoretical analysis. The theory is that the deviation e and the deviation changehave a relationship with the parameters of the PId controller, which is expounded as followsWhen e(t is a little bigger, to get a faster respond speed, we should give K a bigger valueThen K could make the time constant and damping coefficient of the system decrease. Of coursecan not be too large otherwise, it will lead to the ever-changing stability. To avoid system in thebeginning may cause beyond control function, we should give Ko a smaller value to speed up thesystem response. To avoid system appearing larger overshoot, we can take out the integral action,take,=0.When ZO is not so big and not so small we should give kp a small to make sure the overshootof the system response smaller. At this time, it is very imp中国煤化工 value. Inorder to ensure the response speed of system, the value of Kappropriately increased a little but not too muchYHe may beCNMHG通信市场·2012年11-12月第142页六届全国通信新理论与新技术学术大会(CTC2012)优秀论文Telecom marketWhen e(o is a little smallIn order toake surethe system has good steady stateperformance, we probably give a bigger value to Kp and K To avoid system in balance volatilewe should give K, an appropriate valueBased on the above sums up of the input variables and the qualitative relationship of the threearameters, combined with engineering technical personnel's analysis and actual operationexperience, in consideration of the deviation rate of change, we finally get figure. These are therules that adjust PID controller correction three parametersIn the table, e and ec respectively stand for the absolute value of the deviation and thevariation of the deviation.NB、NM、NS、Z、PS、 PM and PB respectively stand for subsetcovering variable fuzzy, which is separated in large, middle, small and zero. It could also beexpressed by fuzzy number. In the figure AKp. AK, and AKp respectively stand for thecorrections of the design parameters of the system PID controller. The system real-time parameterselection should be respectively K+△KpK1+△ K. and K+△KDTABLthe rules of accommodation of thector△KPS PMI PB3 PBPB PM PMPSTABLE IL. the rules of accommodation of the adjustment ofPMPBTABLE IIL. the rules of accommodation of the adjustment of AKB NM NS ZPMPBNS NB NM NMNM NMPMPBPBPB PS PMPM PSPS PBThese fuzzy subsets of domain and its subordinate function could be obtained according to thelarge amounts of data analysis to the system f subset membership function often take simpletriangle or trapezoidal, depending on the specific situationC. The Fuzzy Controller For The Adjustment Of The Three Parameters Of The PID controllerAccording to the rules above, we could design a fuzzy controller. It can be connected with thePID controller as Fig 4中国煤化工Figure 5. The schematic diagram of the adjustment to PID paramCNMHG通信市场2012年11-12月第143页中国煤化工CNMHG