L1 Penalized Sequential Convex Programming for Fast Trajectory Optimization: With Application to Optimal Missile Guidance

노희건,오영재,탁민제,권기정,권혁훈 한국항공우주학회 2020 International Journal of Aeronautical and Space Sc Vol.21 No.2

In this paper, an L1 penalized sequential convex programming (LPSCP) method for trajectory optimization is proposed. Sequential convex methods dramatically reduce computation time for nonlinear trajectory optimizations, based on the desirable properties of convex optimization algorithms. Due to its fast convergence speed, the sequential convex method is considered as a near-future candidate for real-time optimal online guidance. However, the generic sequential convex method seldom suffers from poor robustness to crude the initial trajectory estimates and constraint perturbations, which limits its application to real-world systems. Suggested LPSCP method resolves the robustness issue of generic sequential convex methods, by addressing the convex subproblem infeasibility. While improving its robustness, proposed LPSCP method maintains its fast convergence property. Consequently, LPSCP method enhances overall versatility of sequential convex method for trajectory optimization, which includes ameliorated robustness toward initial trajectory estimates, and improved stability to state and constraint perturbations. Throughout the paper, details of LPSCP method are outlined along with an application example to missile trajectory optimization problems. Desirable properties of LPSCP method are demonstrated through several simulated examples. Optimized trajectories are compared with results from the pseudospectral method. Numerical simulation and Monte Carlo analysis show that LPSCP method accelerates optimization process dramatically (−98.2% on average) while producing the same optimal trajectory.

Convex 최적화를 이용한 화성착륙 유도제어 알고리즘 연구

강상욱,변수영,김호영,방효충 한국항공우주학회 2013 한국항공우주학회 학술발표회 논문집 Vol.2013 No.11

화성탐사의 성공적인 임무수행을 위해서는 화성 착륙선을 정해진 목표지점에 정확히 착륙시키는 것이 매우 중요하다. 본 연구에서는 Convex optimization을 이용하여 목표지점까지 착륙선을 유도제어할 수 있는 추력하강단계의 착륙알고리즘에 관해 연구를 수행하였다. 미리 정해진 목표지점과 최소의 착륙 오차를 갖는 궤적을 생성한 후 연료소모를 최소화하는 궤적을 설계하였다. 착륙오차를 최소화하는 궤적은 추력벡터의 크기가 0이 아닌 하한값을 갖기 때문에 nonconvex 문제이다. 따라서 이 문제를 convex 최적화 문제로 변형하여 전역영역에서의 최적화 궤적을 찾는 알고리즘을 소개하였다. 이 연구는 실제 화성탐사시 실시간으로 적용할 수 있는 추력단계에서의 착륙알고리즘 개발에 많은 도움이 될 것이다. To perform the Mars mission successfully, it is important to land the lander precisely on the target of Mars surface which is determined in advance. In this study, powered descent guidance algorithm to guide lander on target was investigated using convex optimization. Minimum landing error trajectory is generated before designing trajectory of minimum fuel consumption for this study. the problem of minimum landing error trajectory is noconvex optimal control problem because of nonzero lower bound on the magnitude of thrust vector. So it is introduced to solve the optimal trajectory in global domain after converting this problem to convex optimal control problem. This study will be helped on development of powered descent guidance algorithm to apply on real time implementation for Mars mission.

컨벡스 최적화 기반의 가변유동형 덕티드 로켓 유도탄의 중기궤적 최적화 연구

김보석,정철구,이용건,성홍계,이창훈 제어·로봇·시스템학회 2023 제어·로봇·시스템학회 논문지 Vol.29 No.8

This paper presents a convex optimization-based trajectory optimization method for variable-flow ducted rocket missiles during midcourse engagement to reach the predicted intercept point (PIP). The minimum-intercept time problem is established by reflecting the dynamics and some flight constraints for ducted rocket missiles. An artificial neural network (ANN) is utilized to approximate the nonlinear relationship between several flight conditions and the performance of ducted rockets. In addition, this study, unlike previous studies, proposes selecting the air-to-fuel ratio instead of the fuel mass-flow rate as a control input. This can alter the nonlinear constraint as an equivalent linear constraint by combining it with the pseudospectral method, which is demonstrated in this paper. A convex sub-problem is established by applying successive linearization to the nonlinear dynamic constraint. An improved trust-region algorithm is utilized with the convex sub-problem to solve the original non-convex problem. Numerical optimization results are provided to demonstrate the performance of the proposed method and investigate the optimal trajectory for variable-flow ducted rocket missiles.

Consistent dynamic model identification of the Stäubli RX-160 industrial robot using convex optimization method

Omer Faruk Argin,Zeki Yagiz Bayraktaroglu 대한기계학회 2021 JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY Vol.35 No.5

Dynamic models of robot manipulators with standard dynamic parameters are required for simulations, model-based controller design and external force estimation. The aim of this work is to identify the complete dynamic model of the 6-axis Stäubli RX-160 industrial robot. A convex optimization-based method is used for parameter identification. Consistent model parameters are obtained as the result of the optimization procedure subject to physical constraints. Low-speed behavior of the robot being dominated by joint friction, the dynamic model includes an algebraic friction model consisting of the Coulomb and viscous friction components along with the Stribeck effect. The coupled mechanical structure of the 5th and 6th joints, and elasticity due to the presence of balancing springs are also represented in the proposed dynamic model. The ordinary least square error method is used for the performance evaluation of the convex optimization-based method. Estimated parameters from both methods are experimentally verified over identification and test trajectories. The identified model is finally used as a basis in the estimation of external forces acting on the robot’s end-effector. The proposed sensor-less model-based approach for the estimation of external forces constitutes an alternative mean of experimental validation. Comparison of computed external forces with measured ones by an F/T transducer shows that the dynamic model obtained with the proposed method provides an accurate estimation.

Convex Optimization-based Entry Guidance for Spaceplane

Juho Bae,Sang-Don Lee,Young-Won Kim,Chang-Hun Lee,Sung-Yug Kim 제어·로봇·시스템학회 2022 International Journal of Control, Automation, and Vol.20 No.5

This paper aims to propose convex-optimization-based entry guidance for a spaceplane, which has potential in online implementation with less sensitivity to initial guess accuracy while mitigating a high-frequency jittering issue in the entry trajectory optimization problem. To this end, a highly nonlinear, constrained, and nonconvex entry guidance problem is converted into sequential convex sub-problems in the second-order cone programming (SOCP) form by an appropriate combination of successive linearization and convexification techniques. From the investigation on the potential sub-problem infeasibility due to a rough initial guess for radial distance, a linear penalized term associated with a virtual control for an inequality constraint is used to relieve the sub-problem infeasibility while preserving the standardized SOCP form. An adjustable trust-region bound is also adopted in the proposed approach to improve the convergence property further. Additionally, a change of control variables and a relaxation technique are utilized to relieve the high-frequency jittering issue. It is proven that the Lossless convexification property is preserved for the relaxed problem even in the presence of the penalty terms. The feasibility of the proposed method is investigated through numerical simulations.

통합유도조종에서 실시간 모델 예측 제어를 위한 FPGA 기반 PD-IPM 가속화 기법

김대연(Daeyeon Kim),이헌철(Heonchoel Lee),최원석(Wonseok Choi),정보라(Bora Jeong),조영기(Youngki Cho) 한국정보기술학회 2024 한국정보기술학회논문지 Vol.22 No.3

This paper addresses the integrated guidance and control system as a key element to secure effective maneuverability, especially in the military domain. Recent research has shown a growing interest in Model Predictive Control(MPC) and convex optimization. However, despite the excellent performance of convex optimization, there is a challenge in ensuring real-time capability in embedded systems due to high computational requirements. Therefore, the paper proposes a method to accelerate convex optimization using Primal-Dual Interior-Point Method(PD-IPM) on FPGA-based platforms and emphasizes enhancing the real-time capability of the system. In the experimental results, the proposed algorithm successfully performs target tracking while reducing the overall execution time by approximately 33%.

Convex Optimization Algorithms for Multiple Source Localization Based on Received Signal Strength Measurements

Xiaodun Deng 보안공학연구지원센터 2016 International Journal of Future Generation Communi Vol.9 No.5

Decaying with the increasing of signal propagation distance, Received Signal Strength (RSS) is used in the wireless localization due to its low cost and easily implementation. When the transmit power is unavailable, two convex optimization algorithms including semi definite programming (SDP), second order cone and semi definite programming (SOC/SDP) are designed to estimate the source locations by relaxing the non-convex problem as convex optimization. The corresponding Cramér-Rao lower bound (CRLB) of the problem is derived. The simulations demonstrate that the SOC/SDP algorithm provides the similar accuracy performance compared with the SDP algorithm. However the computational complexity of SOC/SDP is lower than that of the SDP due to the less variables and equality constraints. When perfect knowledge of the path loss exponent is available, the simulations also show that the accuracy performance of the proposed convex optimization algorithms degrades as the path loss exponent increases.

Artificial Noise Assisted Secure Transmission for Distributed Antenna Systems

Wang, Hui-Ming,Wang, Chao,Ng, Derrick Wing Kwan,Lee, Moon Ho,Xiao, Jia Institute of Electrical and Electronics Engineers 2016 IEEE transactions on signal processing Vol.64 No.15

<P>This paper studies the artificial noise (AN) assisted secure transmission for a distributed antenna systems (DAS). To avoid a significant overhead caused by full legitimate channel state information (CSI) acquisition, tracking and collection in the central processor, we propose a distributed AN scheme utilizing the large-scale CSI of the legitimate receiver and eavesdropper. Our objective is to maximize the ergodic secrecy rate (ESR) via optimizing the power allocation between the confidential signal and AN for each remote antenna (RA) under the per-antenna power constraint. Specifically, exploiting random matrix theory, we first establish an analytical expression of the achievable ESR, which leads to a non-convex optimization problem with multiple non-convex constraints in the form of high-order fixed-point equations. To handle the intractable constraints, we recast it into a max-min optimization problem, and propose an iterative block coordinate descent (BCD) algorithm to provide a stationary solution. The BCD algorithm is composed of three subproblems, where the first two subproblems are convex with closed-form solutions, and the last one is a convex-concave game whose saddle-point is located by a tailored barrier algorithm. Simulation results validate the effectiveness of the proposed iterative algorithm and show that our scheme not only reduces the system overhead greatly but also maintains a good secrecy performance.</P>

Pragmatic Assessment of Optimizers in Deep Learning

Ajeet K. Jain,PVRD Prasad Rao,K. Venkatesh Sharma International Journal of Computer ScienceNetwork S 2023 International journal of computer science and netw Vol.23 No.10

Deep learning has been incorporating various optimization techniques motivated by new pragmatic optimizing algorithm advancements and their usage has a central role in Machine learning. In recent past, new avatars of various optimizers are being put into practice and their suitability and applicability has been reported on various domains. The resurgence of novelty starts from Stochastic Gradient Descent to convex and non-convex and derivative-free approaches. In the contemporary of these horizons of optimizers, choosing a best-fit or appropriate optimizer is an important consideration in deep learning theme as these working-horse engines determines the final performance predicted by the model. Moreover with increasing number of deep layers tantamount higher complexity with hyper-parameter tuning and consequently need to delve for a befitting optimizer. We empirically examine most popular and widely used optimizers on various data sets and networks-like MNIST and GAN plus others. The pragmatic comparison focuses on their similarities, differences and possibilities of their suitability for a given application. Additionally, the recent optimizer variants are highlighted with their subtlety. The article emphasizes on their critical role and pinpoints buttress options while choosing among them.

Active and Passive Beamforming for IRS-Aided Vehicle Communication

( Xiangping Kong ),( Yu Wang ),( Lei Zhang ),( Yulong Shang ),( Ziyan Jia ) 한국인터넷정보학회 2023 KSII Transactions on Internet and Information Syst Vol.17 No.5

This paper considers the jointly active and passive beamforming design in the IRS-aided MISO downlink vehicle communication system where both V2I and V2V communication paradigms coexist. We formulate the problem as an optimization problem aiming to minimize the total transmit power of the base station subject to SINR requirements of both V2I and V2V users, total transmit power of base station and IRS’s phase shift constraints. To deal with this non-convex problem, we propose a method which can alternately optimize the active beamforming at the base station and the passive beamforming at the IRS. By using first-order Taylor expansion, matrix analysis theory and penalized convex-concave process method, the non-convex optimization problem with coupled variables is converted into two decoupled convex sub-problems. The simulation results show that the proposed alternate optimization algorithm can significantly decrease the total transmit power of the vehicle base station.