基于最小二乘改进算法的时变系统参数辨识-福州大学学报.PDF

33 2 ( ) Vol. 33 No. 2 2005 4 Journal of Fuzhou Univ rsity( Natural Sci nc ) Apr. 2005 : 1000- 2243( 2005) 02- 0 163- 04 , ( , 350002) : , . , , , . , , , . , , . : ; ; ; : TP273 : A Parameter identification of time- varying systems based on the least- squares modified algorithm CHEN Xian- zhi, CHEN Chong ( Coll g of El ctrical Engin ring and Automation, Fuzhou Univ rsity, Fuzhou, Fijian 350002, China) Abstract: In th syst m id ntification domain, th l ast- squar s ar on of th most fundam ntal m th ods. How v r, as th obs rv d data ar incr as d, data saturation com s about, which l ads to a ph nom non that th stimat d valu s can! t b modifi d by n w obs rv d data. B caus th n w obs rv d data hav a mor r markabl ff ct on param t r stimation than th old obs rv d data, this pap r intro duc s th l ast- squar s modifi d algorithm to id ntify th param t r of tim varying proc ss. Finally, th simulat d xp rim nt is propos d. Th r sult indicat s that both r gr ssion m mory and r striction m mory m thod can ov rcom th data saturation ph nom non, and improv th pr cision of param t r i d ntification. Keywords: l ast- squar s; algorithm; param t r id ntification; tim - varying syst m 1 , . , ( CAR) , : [ 1] - 1 - 1 A(z )z(k) = B(z ) u(k) + e(k) (1) - 1 -1 - n - 1 - 1 - n