摘要：机器学习和数据科学技术的飞速发展,很大程度上满足了许多领域通过数据预测比较实际或复杂的动力系统的需求.列维噪声是一个比高斯白噪声更普适和复杂的涨落模型,它被广泛应用于许多非高斯情形来模拟爆炸或跳跃行为.在本文中,我们设计了一个系统的数据驱动方法来识别受高斯白噪声和非高斯列维噪声扰动的系统的最大似然转移路径.其中涉及的主要理论和数值概念包括经典动力系统理论中的非局部Kramers-Moyal公式、非局部Fokker-Planck方程以及一个通过交叉验证解决稀疏回归问题的机器学习框架.接下来我们给出两个例子从细节方面展现了该方法的可操作性,并且对最大似然转移路径做了简要的分岔分析.该方法将作为相关理论研究对应的数值方法以及数值角度的验证,并为该高斯扩散过程的推广复杂系统的其他诸如平均离出时间或离出概率等动力学指征的数值识别工作提供一些洞见.

摘要：具有随机孔隙的多孔泡沫在声音管理应用中得到越来越多的利用,其吸声系数通常作为性能评估和设计优化的关键声学参数.根据不同的加工方法,泡沫中的孔隙可以是全开或半开的,并且通常表现为分形分布特征.为了便于工程应用,有必要对这些泡沫的吸声系数进行分析预测.然而,分析预测泡沫的吸声系数是一个挑战.因此,本研究提出了半开孔和全开孔的简化代表结构来微观分析泡沫内部的流动特性,并应用分形理论来描述随机分布的孔隙.引入一个叫“开孔度”的参数来表征孔隙的开放程度,可以解析地获得代表结构的粘性和热特征长度.随后,在经典吸声系数Johnson-Champoux-Allard模型的基础上,建立了一种将代表结构与分形理论统一起来的解析模型,从而可以将吸声系数作为具有全开/半开孔泡沫的关键形态参数的函数进行预测.与已有的实验测量和数值模拟结果相比,本文提出的分析模型可以较好地预测了半开孔和全开孔泡沫的关键流动特性和吸声系数.在0到4500 Hz的频率范围内,半开孔泡沫相对于具有相同孔隙率的全开孔泡沫可以更好地衰减声波.在孔隙率为0.95的条件下,半开孔泡沫的总体吸声系数分别提高了21.2%、57.7%和75.8%,此时开孔度由0.75到0.50降低了0.25.

摘要：In this paper,design and optimization technique of slotted tube grain for solid rocket motors has been *** doing so,the design objectives and constraints have been set,geometric parameters identified,performance prediction parameters calculated,thereafter preliminary designs completed and finally optimal design *** model for slotted tube grain configuration has been *** thrust has been taken as the objective function with constraints of burning time,mass of propellant,fixed length and diameter of chamber *** parameter method has been used for calculating the performance prediction parameters.A set of preliminary designs has been completed and an analysis of these results *** all the preliminary results fulfill the design requirements in terms of objective function and constraints,however in order to attain the optimal design,Sequential quadratic programming optimization technique has been *** the slotted tube grain geometry is totally dependent upon various independent variables and each of these variables has a bearing on explicit characteristic of grain designing,hence affects of the independent variables on performance parameters have been examined,thus variation laws have been *** on the variation laws and the analysis of preliminary design results,upper and lower limits have been defined for the independent geometric variables and an initial guess provided for conducting *** attained exhibits that an optimal result has been attained and the value of objective function has been *** the design constraint limits have also been met while ensuring sound values of volumetric loading fraction,web fraction and *** methodology of design and optimization of slotted tube grain for solid rocket motors can be used by engineers as a reference guide for actual design and engineering purposes.

摘要：This paper presents studies of aeroelastic optimization on composite skins of large aircraft wings subject to aeroelastic constraints and strength/strain constraints. The design variable for optimization was the ply thickness of the wing skin panels, and the structural weight was the objective function to be minimised. The impacts of three strength/strain constraints and the ply proportion of the wing skin panels on the optimization results are discussed. The results indicate that the optimal composite wings that satisfy different constraints have remarkable weight advantages over metal wing. High levels of stiffness can be achieved while satisfying the constraints regarding allowable design strains and failure criteria. The optimization results with variable-proportions indicate that wing skins with higher proportions of 0° plies from the root to the middle segment and ±45° plies outboard have a more efficient and reasonable stiffness distribution.

摘要：The optimization design of the power system is essential for stratospheric airships with paradoxical requirements of high reliability and low weight. The methodology of orthogonal experiment is presented to deal with the problem of the optimization design of the airship's power system. Mathematical models of the solar array, regenerative fuel cell, and power management subsystem (PMS) are presented. The basic theory of the method of orthogonal experiment is discussed, and the selection of factors and levels of the experiment and the choice of the evaluation function are also revealed. The proposed methodology is validated in the optimization design of the power system of the ZhiYuan-2 stratospheric airship. Results show that the optimal configuration is easily obtained through this methodology. Furthermore, the optimal configuration and three sub-optimal configurations are in the Pareto frontier of the design space. Sensitivity analyses for the weight and reliability of the airship's power system are presented.

摘要：An ammonia self-managed vaporization propulsion (ASVP) system for micro-nano satellites is presented. Compared with a normal cold gas or liquefied gas propulsion system, a multiplex parallel sieve type vaporizer and related vaporization control methods are put forward to achieve self-managed vaporization of liquefied propellant. The problems of high vaporization latent heat and incomplete vaporization of liquefied ammonia are solved, so that the ASVP system takes great advantage of high theoretical specific impulse and high propellant storage density. Furthermore, the ASVP operation procedure and its physical chemistry theories and mathematical models are thoroughly analyzed. An optimal strategy of thrust control is proposed with consideration of thrust performance and energy efficiency. The ground tests indicate that the ASVP system weighs 1.8 kg (with 0.34-kg liquefied ammonia propellant) and reaches a specific impulse of more than 100 s, while the power consumption is less than 10 W. The ASVP system meets multiple requirements including high specific impulse, low power consumption, easy fabrication, and uniform adjustable thrust output, and thus is suitable for micro-nano satellites.

摘要：An aerodynamic design optimization platform(ADOP) has been *** numerical optimization method is based on genetic algorithm(GA),Pareto ranking and fitness sharing *** platform was used for design optimization of the stator of an advanced transonic stage to seek high adiabatic *** compressor stage efficiency is increased by 0.502% at optimal point and the stall margin is enlarged by nearly 1.0% at design rotating *** flow fields of the transonic stage were simulated with FINE/Turbo software *** optimization result indicates that the optimization platform is effective in 3D numerical design optimization problems.

摘要：The ant system algorithm （ASA） has proved to be a novel meta-heuristic algorithm to solve many multivariable problems. In this paper, the earth coverage of satellite constellation is analyzed and a n ＋ 1^ -fold coverage rate is put forward to evaluate the coverage performance of a satellite constellation. An optimization model of constellation parameters is established on the basis of the coverage performance. As a newly developed method, ASA can be applied to optimize the constellation parameters. In order to improve the ASA, a rule for adaptive number of ants is proposed, by which the search range is obviously enlarged and the convergence speed increased. Simulation results have shown that the ASA is more quick and efficient than other methodV211.71s.

摘要：Shape sensitivities of flutter characteristics can predict the moving of flutter boundary as wing shape varies. The nonlinear relationship between mass, stiffness and damping matrices of aeroelastic systems and shape variables makes the flutter characteristics vary nonlinearly as shape variables change. The computation cost of finite difference method is high and it cannot solve precisely shape sensitivities. An analytic method is developed to compute sensitivities of flutter characteristics of low aspect ratio wings to shape parameters, which include aspect ratio, taper ratio, sweep angle, and area. On the basis of the equivalent plate model and piston theory, analytic sensitivities of mass, stiffness and damping matrices with respect to various shape parameters are computed. The equivalent plate model is a continuous aeroelasticity analysis model oriented toward wing design. The flutter equation is solved by tracking the root locus of the system state space model. Lancaster's adjoint method is used to solve the eigenvalue derivatives and shape sensitivities of flutter characteristics. Linear Taylor approximation based on the analytic sensitivities is used to predict the variation of flutter speed with respect to shape variables. Comparison of these results with those from reanalysis indicates that Taylor approximation based on analytic sensitivities can precisely predict trends of flutter characteristics near the baseline configuration, but the applied neighborhood is small for sweep and area. The method can help designers make a judicious choice of wing shape parameters for preventing flutter in the preliminary design phase of aircraft.

摘要：This article seeks to outline an integrated and practical geometric optimization design system （GODS） incorporating hybrid graphical electromagnetic computing-wedge modeling （GRECO-WM） scheme and the genetic algorithm （GA） for calculating the radar cross section （RCS） and optimizing the geometric parameters of a large and complex target respectively. A new wedge modeling （WM） scheme is presented for calculating the high-frequency RCS of wedge with only one visible facet based on the method of equivalent currents （MEC）. The applications of GODS to 2D cross-section and 3D surface are respectively implemented by choosing an average of monostatic RCS values corresponding to a series of incident angles over a frequency band as the optimum objective function. And the results demonstrate that the RCS can be effectively and conveniently reduced by the GODS presented in this article.