n m K NSGA NSGA-Ⅱ MOP 18 miny = F X = f 1 X f 2 X f m X X Ω g i X 0 i = 1 s.t 2 p 1 { h j X = 0 j = 1 2 q X = x 1 x 2 y 1 y 2 y m R m m g i X 0

Size: px

Start display at page:

Download "n m K NSGA NSGA-Ⅱ MOP 18 miny = F X = f 1 X f 2 X f m X X Ω g i X 0 i = 1 s.t 2 p 1 { h j X = 0 j = 1 2 q X = x 1 x 2 y 1 y 2 y m R m m g i X 0"

吹吉
7 years ago
Views:

1 Journal of Northwestern Polytechnical University Apr. Vol No.2 姚从潮 1,2, 王新民 2, 陈晓 2 2, 尹海韬高超声速飞行器再入飞行段需要反推力器进行辅助姿态控制, 控制力矩向气动舵面和反推力器的合理分配是再入飞行控制的一个关键问题文章考虑燃料消耗舵面偏转状态和误差三要素, 将控制分配视为一个多目标优化的问题, 提出了一种改进的多目标遗传算法引入一种改进的模拟退火算法提高局部搜索能力 ; 为保证种群的多样性, 提出了一种改进的小生境技术, 将距离参数设置为动态的函数, 并引入种群的繁殖代数分别设计燃料消耗舵面偏转和误差的代价系数约束条件等, 获得多目标 Pareto 解集, 并基于模糊逻辑, 在最优解集中寻获最优解通过仿真验证了方法的有效性高超声速飞行器, 再入, 遗传算法, 多目标优化, 控制分配,Pareto 最优解, 模糊控制 V249 A Pareto Pareto multi-objective optimization problem MOP 17 MOP

2 n m K NSGA NSGA-Ⅱ MOP 18 miny = F X = f 1 X f 2 X f m X X Ω g i X 0 i = 1 s.t 2 p 1 { h j X = 0 j = 1 2 q X = x 1 x 2 y 1 y 2 y m R m m g i X 0 i = 1 2 p p h j X = 0 j = 1 2 q q 1. 2 Pareto K T 0 Pareto 2 E = f x Vilfredo Pareto Ω Δx X n x + Δx minf X E n = f X n h ΔE T X X n E E n ΔE = E n - E Ω = X R n g i X 0 h j X = 0 i = 1 2 p j = 1 2 q Ω. 2 ΔE > minf X Ω exp - ΔE /T > ε ε 0 1 X 1 X 2 Ω F X 1 F X 2 P * 22 NSGA-Ⅱ 2. 1 x n R n n Y = = X * = X Ω X' Ω f j X' f j X j = 1 2 m Pareto P * ΔE = Δ j = 槡 Δ Δ Δ 2 n 4 Pareto n Pareto Pareto exp - ΔE /T > ε Rosenberg 20 Multi-objective Genetic Algorithm MOGA Srinivas Deb Non-dominated Sorting Genetic Algorithm NSGA 2002 Deb 1 T 3 g Δx T 4 ΔE 0 X 1 X 2 F X 1 < F X 2 X 1 X 2 5 K K 5. 3 minf X Ω 2 X * Ω X * X Ω X * Pareto X * X 3 Ω X * Pareto X * X Ω X * Pareto 5. 4 ΔE i i = 1 2 N N 4 ΔE i 0 ΔE j > 0 j = 1 2 n L

3 2 317 L L L L 3. 1 F X 1 L f 1 f m F X L X = x 1 x 2 T x 1 x 2 L 2 槡 d ij = M i - M j = q n = 1 2 f in x - f jn x i = 1 2 M + N - 1 j = i + 1 M + N 5 M i M j i j c ij i = j = 1 2 L' M L' = +N-1 i = 1 M +N j = i +1 M d ij / +N-1 i = 1 M +N j = i Pareto f 1 = c 11 x n 1 + c 12x n 2 f 2 = c 21 x n 1 + c 22x n 2 { 8 f 3 = c 31 x n + c 1 32x n 2 n n 1. 4 t L = L' /t 7 c d ij = M i - M j < L 2 c 11 c 11 2 L c 11 c c 11 N P t c 11 = k 11 δ e - δ trim + k 12 e k 13δ e 9 E i i = 1 2 N k 11 k 13 Q t P t Q t 2N k 11 k 12 R t c 11 R' t R' t c 12 R t k 14 c 11 N c 12 Pareto 2

4 f 1 = k 11 δ e - δ trim + k 12 e k 13δ e x n + k 1 14x n 2 f 2 = k 21 x n + k 22 1 h xn 2 f 3 = k 31 x n + k 32 1 x n 2 δ α δ 0 10 δ 0 40 /s c 22 = k 22 /h km k 22 c % 3 Pareto 3. 5 Pareto k 11 k 12 k 13 k 14 k 21 k 22 k 31 k 32 c 32 δ δ e δ trim δ e h M c δ n 5 3 c 32 = k 32 /δ k x 1 x 2 x 1 + x 2 = M c 12 M c x 1 M m 13 1 M m δ m M m = 1 2 ρv2 ScC Mδm C Mδm = δ m - α ~ 11 2

5 2 319 E con E h E δ 1E con δ con 1 2 δ con = m 1 δ e - δ trim + m 2δe δ e - δ trim Pareto δ e 2 E h h 3E δ 3 w 1 w 2 w 3 E con E h E δ w 1 w 2 w 3 4 vs rs mo rb vb 2 Pareto Pareto 1 E con minf = w 1 f 2 + w 1 2f 2 + w 2 3f w 1 w 2 w 3 w 1 w 2 w 3 w 3 f δ e δ trim M c n k 11 k 12 k 13 k 14 k 21 k 22 k 31 k % 25 /s 4 45% 25 /s 5 85% 40 /s

6 km 5% 7 75 km 45% km 85% km 25 /s km 25 /s km 55 /s Pareto 5

7 J Zhu Yunji Shi Zhongke. Several Problems of Flight Characteristics and Flight Control for Hypersonic Vehicles J. Flight Dynamics in Chinese 2 Wu S F Rodrigo R. Nonlinear Dynamic Modeling and Simulation of an Atmospheric Reentry Spacecraft J. Aerospace Science and Technology M Zhou Jun. Control Principle of Spacecraft M. Xi'an Northwestern Polytechnical University Press 2001 in Chinese 4 Clack R N Franklin G F. Limit Cycle Oscillations in Pulse-Modulated Systems J. Spacecraft Mendel J M. On-Off Limit-Cycle Controllers for Reaction Jet Controlled Systems J. IEEE Trans on Automatic Control Song G Agrawal B N. Vibration Suppression of the Flexible Spacecraft during Attitude Control J. Acta Astronautica Song G Buck N Agrawal B N. Spacecraft Vibration Reduction Using Pulse-Width Pulse-Frequency Modulated Input Shaper J. Journal of Guidance Control and Dynamics J Fang Yuanpeng. Research on Composite Control Method for Reusable Launch Vehicle J. Flight Dynamics in Chinese 9 Tomatis C Bouaziz L Franck T Kauffmann J. RLV Candidates for European Future Launchers Preparatory Programme J. Acta Astronautica J Yang Enquan Gao Jinyuan. Research and Development on Advance Fighter Control Allocation Methods J. Flight Dynamics in Chinese 11 Ning Guodong Zhang Shuguang Fang Zhenping. Entry Control Allocation Using Sliding Modes and State Observer Synthesis for Reusable Launch Vehicle J. Journal of Astronautics Durham W C Bordgon K A. Multiple Control Effectors Rate Limiting J. Journal of Guidance Control and Dynamics Buffington J Chandler P Pachter M. On-line System Identification for Aircraft with Distributed Control Effectors J. International Journal of Robust and Nonlinear Control Fletcher R. Practical Methods of Optimization Second Edition M. New York John Wiley & Sons M Xie Kexin Han Jian Lin Youlian. Optimization Methods M. Tianjin Tianjin University Press 2004 in Chinese 16 Peterson J A M Bodson M. Constrained Quadratic Programming Techniques for Control Allocation J. IEEE Trans on Control Systems Technology Deb K Pratap A Argrawal S et al. A Fast and Elitist Multi-Objective Genetic Algorithm NSGAⅡ J. IEEE Trans on Evolutionary Computation M Cui Xunxue. Multi-Objective Evolutionary Algorithm and Its Application M. Beijing National Defense Industry Press 2006 in Chinese 19. M Zheng Jinhua. Multi-Objective Evolutionary Algorithm and Its Application M. Beijing Science Press 2007 in Chinese 20 Rosenberg R S. Simulation of Genetie PoPulations with Bioehemieal ProPerties D. Ann Harbor Miehigan University of Miehigan 1967

8 Srinivas N Deb K. Multi Objective Optimization Using Nondominated Sorting in Genetic Algorithms J. Evolutionary Computation Deb K Agarwal S. A Fast and Elitist Multi-Objective Genetic Algorithm NSGA-II J. IEEE Trans on Evolutionary Computation Reentry Flight Control Allocation Research Based on Improved Multi-Objective Genetic Algorithm Yao Congchao 1 2 Wang Xinmin 2 Chen Xiao 2 Yin Haitao 2 1.No.38 Research Institute of China Electronics Technology Group Corporation Hefei China ( 2.Department of Automatic Control Northwestern Polytechnical University Xi'an China ) Abstract The reentry of the hypersonic vehicle requires reaction thruster to assist attitude control. How to allocate the control moment to aerodynamic surfaces and reaction thruster properly is a key problem. The fuel consumption the surface deflection and the error are considered in this paper. An improved multi-objective optimization genetic algorithm is proposed when regarding control allocation as a multi-objective optimization problem. A modified simulated annealing algorithm is introduced. An improved niche selection technology is put forward to maintain the population diversity. The distance parameter is set as a dynamic function and generation quantity of the population is introduced at the same time. The cost coefficient and constraint are designed respectively to obtain Pareto optimal solutions set. On the basis of fuzzy logic the optimal solution is selected. Finally effectiveness is verified by simulation. Key words hypersonic vehicles reentry genetic algorithms multiobjective optimization control allocation pareto-optimal solutions fuzzy logic

698 39,., [6].,,,, : 1) ; 2) ,, 14,, [7].,,,,, : 1) :,. 2) :,,, 3) :,,,., [8].,. 1.,,,, ,,,. : 1) :,, 2) :,, 200, s, ) :,.

698 39,., [6].,,,, : 1) ; 2) ,, 14,, [7].,,,,, : 1) :,. 2) :,,, 3) :,,,., [8].,. 1.,,,, ,,,. : 1) :,, 2) :,, 200, s, ) :,. 39 6 Vol. 39, No. 6 2013 6 ACTA AUTOMATICA SINICA June, 2013 1, 2,,,. DOI,,,., 2013, 39(6): 697 702 10.3724/SP.J.1004.2013.00697 Present Situation and Development Tendency of Aerospace Control Techniques