基于支持向量机回归的lyapunov指数计算方法研究-控制与决策.pdfVIP

绗?6 鍗 绗? 鏈 鎺 鍒 涓 鍐 绛 2011 骞 5 鏈 Vol. 26 No. 5 Control and Decision May 2011 鏂囩珷缂栧彿: 1001-0920 (2011) 05-0785-04 鍩轰簬鏀寔鍚戦噺鏈哄洖褰掔殑Lyapunov 鎸囨暟璁＄畻鏂规硶鐮旂┒ 寮 楣? 鍊笘瀹 (绌哄啗宸ョ▼澶у宸ョ▼瀛﹂櫌锛岃タ瀹?10038) 鎽 瑕? Lyapunov 鎸囨暟鏄弿杩板姩鍔涘绯荤粺娣锋矊鎬ц川鐨勯噸瑕佹寚鏍? 鍦ㄥ皬鏍锋湰鏉′欢涓嬪噯纭€佸揩閫熷湴璁＄畻Lyapunov 鎸囨暟鏄 涓€涓毦棰? 瀵规, 鎻愬嚭涓€绉嶅熀浜庢敮鎸佸悜閲忔満鍥炲綊鐨凩yapunov 鎸囨暟璁＄畻鏂规硶, 閫氳繃閲忓瓙閬椾紶绠楁硶瀵规敮鎸佸悜閲忔満妯″瀷鐨 鍙傛暟杩涜浼樺寲, 鎺ㄥ浜嗘敮鎸佸悜閲忔満鍥炲綊搴旂敤浜庤绠桳yapunov 鎸囨暟鐨勫叕寮? 閫氳繃瀵规贩娌屽簭鍒楄繘琛屼豢鐪熷疄楠? 浠跨湡缁撴灉 琛ㄦ槑, 鍦ㄥ皬鏍锋湰鏁版嵁鎯呭喌涓? 姝ゆ柟娉曞彲琛屾湁鏁? 鍏抽敭璇岤 鏀寔鍚戦噺鏈哄洖褰掞紱Lyapunov 鎸囨暟锛涢噺瀛愰仐浼犵畻娉曪紱灏忔牱鏈 涓浘鍒嗙被鍙耳 TP183 鏂囩尞鏍囪瘑鐮? A Algorithm for computing Lyapunov exponent based on support vector ZHANG Peng, NI Shi-hong (The Engineering Institute锛孉ir Force Engineering University 锛孹i鈥檃n 710038 锛孋hina. Correspondent 锛歓HANG Peng, E-mail 锛歾hangpeng25@) Abstract: Lyapunov exponent(LE) is an important indicator to describe the nature of chaotic dynamical systems. It鈥檚 dif铿乧ult to calculate LE accurately and fast with small sample size. In this paper, an algorithm for computing LE based on support vector regression(SVR) is proposed, the parameters of support vector machine(SVM) model are optimized by using quantum genetic algorithm(QGA), and the formula of the LE based on SVR is derived. Simulation results show that the proposed algorithm is feasible and effective with much less samples 锛 Key words: support vector regression锛汱yapunov exponent 锛泀uantum genetic algorithm 锛泂mall sample size 1 寮曞紩寮 瑷€瑷€瑷€ 鐗规€? 鎻愬嚭浜嗗熀浜庣缁忕綉缁滅殑Lyapunov 鎸囨暟璁＄畻鏂 [7] 鍦ㄥ姩鍔涘绯荤粺鍒嗘瀽涓? Lyapunov 鎸囨暟鏄弿杩扮郴 娉 , 浣嗚鏂规硶瀛樺湪灞€閮ㄦ渶浼樺拰鎵€闇€鏍锋湰杈冨绛夐棶 缁熸€ц川鐨勯噸瑕佸畾閲忔寚鏍? 涓€涓郴缁熸槸鍚﹀瓨鍦ㄦ贩娌岀壒鎬? 棰? Vapnik 绛変汉鍩轰簬缁熻瀛︿範鐞嗚鎻愬嚭浜嗘敮鎸佸悜閲