دانلود مقاله ترجمه شده طراحی خلبان اتوماتیک موشک LPV با ترکیب بهره L2 بهبود یافته – مجله IEEE

• فهرست مطالب:

چکیده
۱ مقدمه
۲ تعریف مسئله
۳ سنتز بهره L2
۴ طراحی اتوپایلوت
۴ ۱ مدل موشک و هدف عملکرد
۴ ۲ مدل سازی PALPV
۴ ۳ طراحی و شبیه سازی اتوپایلوت
۴ ۴ مقایسه روش های کنترل LPV
۵ نتیجه گیری

• بخشی از ترجمه:

این مقاله، یک فرمول  پیشرفته LMI را برای سنتز بهره L2 برای سیستم لیاپنوف و کاربرد آن در طراحی اتوپایلوت موشک ارایه می کند.  فرایند طراحی نشان می دهد که روش  ما  قابلیت اطمینان بیشتر  در طراحی  اتوپایلوت موشک نسبت به روش های منتشر شده دارد. این بهبود ها را می توان به استفاده از مدل دقیق تر و کلاس کلی PDLF نسبت داد.

• بخشی از مقاله انگلیسی:

۱ Introduction A missile autopilot design problem has attracted the attention of many researchers because it involves a large variation in the system dynamics and stringent performance requirements [13,15]. While the gain-scheduling approach has been successfully applied to many interesting problems [17], there are several potential difficulties when this approach is applied to the advanced missile autopilot design. For example, some parameters, such as the angle-of-attack, are arbitrarily fast varying, which violates the heuristic guideline that the scheduling parameters should be slowly varying [17]. A complicated scheduling scheme is also required to account for nonlinearities in the system that have been ignored in the design process [13]. These potential difficulties have recently been overcome using a linear parameter-varying (LPV) control approach [2,9,16,18]. An LPV system is characterized as a linear system that depends on time-varying parameters that are assumed to be exogenous signals. It is further assumed that the trajectory of the parameter is not known in advance but is constrained a priori to lie in some known, bounded set, and its value can be measured in real time. Associated with the LPV system have been various synthesis tools such as the scaled smallgain framework [3,9,14,16] and the dissipative systems framework using smooth parameter-dependent Lyapunov functions (PDLFs) [1,4,10,18]. The authors have also proposed a dissipative systems framework using a nonsmooth quasi-piecewise afine parameter-dependent Lyapunov function (QPAL) for a PALPV system. The results in [12] showed that while the ‘Engr. specialist at Space Systems/Loral, limsyOssd.lora1. com 2Assistant Professor, howjoOsun-valley.stanford.edu 0-7803-4990-6/99 $10.00 0 1999 AACC 3733 QPAL approach can be computationally intensive, it leads to less conservative, guaranteed synthesis results than previously published techniques based on QALs or LurBPostnikov Lyapunov functions. This paper extends our previous results [12] in an attempt to reduce the computational complexity by using the multiconvexity approach [7] in the convexifying step of the algorithm development rather than the S-procedure [6]. Furthermore, this paper generalizes the previous results by eliminating the restriction on the number of parameters. This new approach is then applied to the missile autopilot design problem. For a comparison, we also consider other controller design techniques, such as a “naive” gain-scheduling [13], complex-p controller [15], and the LPV control design based on the gridding technique (181. For this application, we show that the PALPV system can yield a more accurate model than the typical LFT because it provides a better approximation of the nonlinear missile dynamics. The QPAL approach reduces conservatism of the synthesis result because of the richness of the PDLFs used in the synthesis. Furthermore, we address the potential difficulty in selecting a PDLF for the gridding technique in the LPV control I181 and show that our new technique can reduce this specific difficulty.