pdf

PID控制器的量子粒子群多目标优化设计

  • 1星
  • 日期: 2018-07-22
  • 大小: 488.99KB
  • 所需积分:1分
  • 下载次数:1
  • favicon收藏
  • rep举报
  • free评论
标签: 量子粒子

量子粒子

PID控制器

工业生产过程中,对于生产装置的温度、压力、流量、液位等工艺变量常常要求维持在一定的数值上,或按一定的规律变化,以满足生产工艺的要求。PID控制器是根据PID控制原理对整个控制系统进行偏差调节,从而使被控变量的实际值与工艺要求的预定值一致。

摘要:智能算法如粒子群算法已被应用于PID控制器的参数优化,以弥补传统优化方法容易产生振荡和较大超调量的不足,但是粒子群算法存在易于早熟的缺点,在分析量子粒子群算法的基础上,提出了使用量子粒子群算法优化PID控制器的参数。为了兼顾控制系统的各项性能指标,根据控制器的实际要求对各项指标进行加权作为算法的目标函数,对PID控制器进行多目标寻优。通过2个传递函数实例,分别使用Z-N、粒子群算法和量子粒子群算法进行了PID控制器参数优化设计,并对结果进行了分析。

关键词:量子粒子群算法;PID控制器;多目标优化

2008 4 Instrument Technique and Sensor PID , , P ID , , ( 2 , : TP229 ; P ID ; : A : , : P ID , , , 214121) Z - N P ID _ _ , _ _ _ _ _ 2008 No 4 P ID , _ _ : 1002 - 1841 (2008) 04 - 0084 - 04 M ulti Ba sed on Quan tum object Param eter O ptim iza tion of P ID Con troller behaved Particle Swarm A lgor ithm ( School of M echan ica l and Electr ica l, W ux i In stitute of Technology, W ux i 214121, Ch ina) LU Rong, WU Yong, ZHAO Ao dong Abstract: Heuristics such as particle swarm op tim ization is emp loyed to enhance the capability of traditional techniques, which is easy to p roduce surge and big overshoot, but PSO may be trapped in the local op tima of the objective and lead to poor per formance. This paper p roposed the quantum behaved particle swarm op tim ization for the parameter op tim ization of P ID controller. A fitness function containing performance indexes was defined and the algorithm was used in multi object op tim ization of P ID con trollers. Two examp les were given to illustrate the design p rocedure and exhibit the effectiveness of the p roposed method via com parison study with the existing Z N and PSO app roaches. Key words:QPSO; P ID controller; multi object op tim ization : G ( s) ; C ( s) 0 P ID [ 1 ] , , [ 2 ] , ( PSO) [ 5 ] P ID , [ 6 ] , P ID 1 P ID P ID , , r e y : P ID [ 3 ] , , , , Z - N , P ID ( GA ) [ 4 ] , , , 1 P ID 2 , [ 7 ] , ) , 1 P ID Kp Ti Td , , ( QPSO ) Kp Ti Td 1995 Kennedy Eberhart , ( ) , , , , , D ( , ( 2) ( 3) C ( s) = Kp (1 + 1 Ti s + Td s) ( 1) , i : 2007 - 02 - 11 : 2008 - 01 - 12 X i ( t + 1) = X i ( t) + v i ( t + 1) v i ( t + 1) = v i ( t) + 1 P i - v i ( t) + 2 P g - v i ( t) 1 i ( t) _ _ _ , ( 5) : P ID i ; c1 t , [ - vmax , + vmax ] i ) ; gbest. g = ( Pg1 , Pg2 , P 4 : c2 1 = c1 rand1 ; 2 = c2 ; X i ( t) , rand2 ; v i t [ 0, 1 ] ; pbest( , PgD ) Shi ; rand1 rand2 , P i = ( Pi1 , Pi2 , , PiD ) Clerc ( inertia weight) w w , w [9 ] , v i ( t + 1) =w v i ( t) + 1 ( P i - X g - X i ( t) ) ( 4) [ 8 ] , i ( t) ) + 2 ( P w , v i ( t + 1) = K [ v i ( t) + 1 P i - X i ( t) + 2 ( P g - X i ( t) ) ] : K = 2 | 2 - - 2 - 4 | ; = 1 max + 2 max , > 4; 1max 2 max 1 2 Jun Sun [ 10 ] (2) , (3) , ) , , PSO , , (QPSO ) , [ 10 ] : QPSO PSO QPSO , , : mbest = M i = 1 pi /M = ( M i = 1 pil /M , M i = 1 pi2 /M , p = ( 1 p1 + 2 pg ) / ( pid /M ) M , i = 1 1 + 2 ) x i ( t + 1) = p |mbest - x i ( t) | ln ( l/ u) : 1 2 ( 0, 1 ) pbest; pg ( t) gbest; mbest i t ; u ( 0, 1) ; ; pi ; M , ( , , 1 ( 6) ( 7) ( 8) i ; x i + , [ 12 ] , ( 0, 1) 5 0 [ 10 ] , - ; [ 11 ] [ 13 ] H [ 14 ] 1 P ID 1 85 , 1 1 , 1 n Kp Ti Td 1 1 , P ID : Step1 1 , : , xn ( 0) ], = [ x1 ( 0) , x2 (0) , Tj d (0) ] j 3 , Z - N (1 - ( 1 - ( 1 - ) Kp ) Ti ) Td Z - N : Kp Ti Td ; Kp Ti Td Step2 : ; w1 w2 w3 J ( pj ) =w1 ; ts , , ( 12) f ( pj ( 0) ) = J ( pj ( 0) ) : pj lbest ( 0) = pj ( 0) , j = 1, 2, , n X i ( 0) , ( 0 ) , xj ( 0) = [ Kj p ( 0) , Tj P ID , Z - N , , ( 9) ( 10) 3 ( 11) ( 9) (10) (11) (12) ; ( 1 + ) Kp ( 1 + ) Ti (1 + ) Td 3 P ID P ID J ( pj ) P ID +w2 ts +w3 e ; e P ID w1 w2 w3 Step3 Step4 : : : pgbest (0) (6) (7) ( 8) , , , f ( pj ( t) ) , t , f ( pj ( t) ) < f ( pj ( t - 1) ) , pj lbest ( t) = pj ( t) , ; f ( pj ( t) ) < f ( pgbest ( t - 1) ) , : f ( pgbest ( t) ) = f ( pgbest ( t - 1) ) pj lbest : f ( pgbest ( t) ) ( t) ) = f ( pj ( t) ) , Step4 t = t + 1, Step3 4 P ID QPSO ( 30 ) , , ( Z - N , PSO ) : G1 ( s) = 1 1) ( s + 0 ( s + 0 2) ( s + 0 3) 3 P ID P ID : G2 ( s) = 2 067s 75e - 8 0 ( 1 + 61 45s) 2 , 2 1 10 P ID 10 G1 ( s) Kp T i Td 0 0 0 0 0 1581 6 3004 12 2968 12 4965 17 5 3123 31 5 0 8211 1 2964 3 9597 3 ts 5828 49 6 1039 1208 6 7053 15 2401 15 24011 4 J 4054 0 0 1053 0941 0 Kp 5483 3304 2914 5447 3318 2851 Z - N PSO QPSO Z - N PSO QPSO 2 G1 ( s) 10 G2 ( s) T i 52 98 96 61 31 34 0028 8053 5193 ts 7460 8345 4631 Td 13 24 24 0007 7013 7013 J 1471 194 171 2362 1259 4913 ts J ts J 0 Z - N 0 PSO QPSO 0 4965 17 34974 6 30979 5 5828 49 08369 6 11937 6 4054 88485 0 10578 0 079195 2 000152 0 064079 000517 1 0 641828 00413 3 G2 ( s) 10 ts J ts J Z - N 0 PSO 0 QPSO 0 5447 61 35312 40 30427 33 7460 1471 03359 219 64153 171 2362 258 0 6548 0 078004 9 033699 1 945797 37 936933 0 14557 462212 QPSO PSO , G1 ( s) 1 P ID , ( s) SO G2 ( s) , PSO PSO 10 , 2 5 P ID W 1 , G2 ( s) 4 G2 ( s) 86 30, SO 10 1 Instrument Technique and Sensor Ap r 2008 30, PSO 1 0 0 5 w1 w2 w3 , , , c1 = c2 = 2, QP 1 0 0 05 1 0 P ID 4 4 , , , G2 ( s) 10 G2 ( s) (QPSO , 10 ) (w 1 w 2 w 3) ts J ts J 1 (1, 0 2 (2, 0 3 (4, 0 05, 0 05, 0 05, 0 01) 01) 01) 0 0 0 4582 26 3847 28 2788 34 3003 105 9316 90 88 6550 5980 0 5061 0 0478 0 004136 0 003633 0 004461 0 873185 27 714489 23 992897 14 16839 82217 49996 , G2 QP 2 3 QPSO 3 2 , , , 4 , , QPSO , 10 1 , 2 2 3 PSO P ID , , 10 4 QPSO G2 ( s) 5 P ID , , P ID P ID , , , Z - N PSO QPSO 3 , QPSO P ID , , QPSO : P ID [ 1 ] ZIEGLER J G, N ICHOLS N B. Op timum setting for automatic control lers, Trans. ASME 64, 1942: 759 - 768. 1 1 4 : P ID [ 2 ] L i Shan, L i J inhua L. IMC P ID Control of U ltra Sonic Motor Servo Sys tem Based on Neural Network, Intelligent Control and Automatin,WC I CA. 2006: 8275 - 8279. 2 2 2 2 2 2 2 87 [ 9 ] CLERC M. The swarm and the queen: towards a determ inistic and a dap tive particle swarm op tim ization. Proceedings of the Congress on Evolutionary Computation. Piscataway, NJ: IEEE Service Center, 1999: [ 3 ] Yang Yu, W ang wenge, Yu Deiid, et al. A fuzzy parameters adap tive P ID controller design of digital positional servo system,Machine Learn [ 10 ] ing and Cybernetics, ICMLC, 2002: 310 - 314. [ 4 ] Hu Yanan, Qu B in. App lication of Iterative Learning Genetic A lgo rithm s for P ID Parameters Auto Op tim isations of M issile controller, In [ 11 ] 1951 - 1957. SUN J, FENG B, XU W. Particle Swarm Op tim ization with Particles Having Quantum Behavior, IEEE Proc. of Congress on Evolutionary Computation, 2004: 325 - 331. FangW ei, Sun Jun, Xu W enbo, Design R D igital Filters U sing Quan telligent Control and Automation,WC ICA, 2006: 3435 - 3439. tum Behaved Particle Swarm Op tim ization. ICNC (2) 2006: 637 - 64. [ 5 ] L iu Yijian, zhang J ianm ing, wang shuqing. Op tim ization design based [ 12 ] Sun Jun, Xa W enbo, Fang wei. Solving Multi period Financial Plan on PSO algorithm for P ID controller, Intelligent Control and Automa ning Problem V ia Quantum Behaved Particle Swarm A lgorithm. IC IC tion,WC ICA, 2004: 2419 - 2422. (2) 2006: 1158 - 1169. [ 6 ] Van DEN BERGH F. An Analysis of Particle Swarm Op tim izers, PhD [ 13 ] Sun Jun, Xu W enbo, L iu J ing. Training RBF Neural Network V ia Thesis, University of Pretoria, 2001 Quantum Behaved Particle Swarm Op tim ization. ICON IP ( 2 ) 2006: [ 7 ] KENNEDY J, EBERHART R. Particle Swarm Op tim ization. Proceed 1156 - 1163. ings of IEEE international Conference On Neural Network, 1995: 1942 [ 14 ] X IM L, Sun J, Xu W B. Quantum Behaved Particle Swarm Op tim iza - 1948. tion for designing H infinity structured specified controllers DCABES [ 8 ] SH I Y, EBERHART R C. A modified particle swarm ontim izer, IEEE 2006 Proceedings, 2006, 2. International Conference of Evolutionary computation, Anchorage, : (1970 ) , , A laska,May 1998. ( 20 ) , ; , 5 , E - mail: wywx@163. com 5 , , , , 1 057 38% 0 3 6 , 146 124 , - 5, 1 053 , 6 [ 2 ] , . FBG . , 2006 (4) : 9 - 11. [ 3 ] Yu Youlong, Zhao Hongxia. A Novel Demodulation Scheme for Fiber B ragg Grating Sensor System. 2005, 17 (1) : 166 - 168. IEEE Photonics Technology Letters, [ 4 ] Yu Youlong. Fiber Laser Based W avelength D ivision Multip lexed Fi ber B ragg Grating Sensor System. IEEE Photonics Technology Letters. 2001, 13 (7) : 702 - 704. LabV IEW FBG , FBG , , : , [ 1 ] . , , 2007 (4) : 3 - 7. [ 5 ] [ 6 ] [ 7 ] , , . : , . LabVE IW8. 0. : , 2007: 1 - 8. , , . , 2001, 21 (8) : 987 - 989. , , . : , 2005: 194 - 237. . [ 8 ] N I Lock In Amp lifier Start Up Kit U ser Manual. National Instru ments, 2002: 3 - 4. (1983 : ) , , , , .
更多简介内容

推荐帖子

评论

登录/注册

意见反馈

求资源

回顶部

About Us 关于我们 客户服务 联系方式 器件索引 网站地图 最新更新 手机版 版权声明

北京市海淀区知春路23号集成电路设计园量子银座1305 电话:(010)82350740 邮编:100191

电子工程世界版权所有 京ICP证060456号 京ICP备10001474号-1 电信业务审批[2006]字第258号函 京公网安备 11010802033920号 Copyright © 2005-2021 EEWORLD.com.cn, Inc. All rights reserved
$(function(){ var appid = $(".select li a").data("channel"); $(".select li a").click(function(){ var appid = $(this).data("channel"); $('.select dt').html($(this).html()); $('#channel').val(appid); }) })
×