解决约束违背问题的一种自适应调整方法

带约束的优化问题的目的是要找到满足等式或者不等式约束的最优点。在某些情况下,优化求解得到的"最优点"可能会使得某个或某几个约束条件超出目标约束限,或者在所有约束条件中的最大值远远小于目标约束限。针对这一类问题,本文提出一种在寻优过程的每一次迭代中自适应调整约束限的方法,通过动态调整迭代过程中迭代模型约束限的值,将约束条件中最大值的约束条件变为等式约束,使得迭代解始终在可行域范围内,且收敛后的最优解不违背任何约束条件。本文将该方法成功应用于位移约束下结构重量最小化拓扑优化模型,原来不满足约束条件的情况在使用该方法后都能使约束得到满足,解决了约束条件被违背的问题。     

Constrained receding horizon controls for nonlinear time-delay systems

This paper proposes a constrained receding horizon control (RHC) for a nonlinear time-delay system with input and state delays. The control law is obtained by minimizing a receding horizon cost function with weighting functions of inputs and states on the end portion of the horizon. For stability, a general condition on the weighting functions is presented and its feasibility is illustrated via a certain type of nonlinear time-delay systems. In order to deal with input and state constraints, an invariant set is obtained, where the trajectories of the inputs and the states satisfy given constraints and stay forever under some conditions. It is shown in a numerical example that the proposed RHC guarantees the closed-loop stability for nonlinear time-delay systems while meeting the constraints.     

Further study of the equivalent theorem of Hellinger-Reissner and Hu-Washizu variational principles

ＩｎｔｒｏｄｕｃｔｉｏｎＩｎ１９５４,ｂｙｔｈｅｓｏ＿ｃａｌｅｄｔｒｉａｌ＿ａｎｄ＿ｅｒｏｒｍｅｔｈｏｄＨｕ［１］ｄｅｄｕｃｅｄｔｈｅｗｅｌ＿ｋｎｏｗｎＨｕ＿Ｗａｓｈｉｚｕｐｒｉｎｃｉｐｌｅ,ｗｈｉｃｈｐｌａｙｓａｎｉｍｐｏｒｔａｎｔｒｏｌｅｉｎｔｈ...     

基于对偶变量变分原理的完整约束系统保辛算法

基于对偶变量变分原理,选择积分区间两端位移为独立变量,构造了求解完整约束哈密顿动力系统的高阶保辛算法。首先,利用拉格朗日多项式对作用量中的位移、动量及拉格朗日乘子进行近似;然后,对作用量中不包含约束的积分项采用Gauss积分近似,对作用量中包含约束的积分项采用Lobatto积分近似,从而得到近似作用量;最后,在此近似作用量的基础上,利用对偶变量变分原理,将求解完整约束哈密顿动力系统问题转化为一组非线性方程组的求解。算法具有保辛性和高阶收敛性,能够在位移的插值点处高精度地满足完整约束。算法的收敛阶数及数值性质通过数值算例验证。     

An eigenvector-based approach for improved parameter variation for multibody mechanical systems

Optimal design of complex multibody mechanical systems involving numerous design parameters and constraints is a challenging problem. System parameters are typically determined through a recursive process which leads to a set of nominal values. While these nominal values satisfy design constraints they generally result in suboptimal performance. In practice, nominal design values can undergo a final adjustment step to improve performance while ensuring that design constraints are not violated. In this article, we present a framework for design parameter variation to optimize the performance of a multibody system. As an example, we apply the proposed method to an exploration rover and show that the proposed approach provides superior results compared to the standard sensitivity derivative approach.     

基于阈值因子的结构可靠性解耦优化方法

基于结构可靠性理论,引入阈值因子的概念,其物理意义为当前最优解满足可靠性约束需将阈值增减的幅度。随着迭代的进行,对于有效约束,阈值因子收敛于1,能够使得最可能失效点快速向满足可靠性约束的极限状态曲面靠拢,优化效率得以提升。解耦模型中,优化变量可为随机变量也可为非随机变量,当优化变量为随机变量时,采取优化变量拆解方式进行计算。数值算例表明,本文方法对优化变量拆解方式不敏感,对有效约束和非有效约束均能够获得满意的优化结果,且计算效率明显高于经典方法。     

多位移等式约束拓扑优化及在光电精密结构设计中的应用

为了满足光电精密跟踪设备中光学系统对支撑结构变形位移相等的设计要求,基于变密度法,以刚度极大为目标,同时以体积约束和位移等式约束作为约束条件,构建结构拓扑优化模型。位移等式约束通过增广拉格朗日乘子法引入原目标函数,在拉格朗日乘子的求解中,采用考虑具有真实物理意义的近似替代法而非传统的纯数学迭代逼近方法。在利用伴随方法得到增广目标函数敏度基础上,采用MMA优化算法,在满足体积约束的同时进行迭代优化得到新结构。算例验证结果表明,本文方法能够有效解决具有多个位移等式约束的刚度极大结构轻量化设计问题。     

具有内部特征约束的四边形网格生成方法

提出一种具有内部特征约束的四边形网格自动生成方法,可以处理内部具有孔洞、约束线、约束点以及密度线、密度点等特征约束,满足数值分析中对网格生成的特殊要求。以区域分解法为基础,提出了确定最佳切割线的方法以及在切割线上生成过渡均匀网格节点的方法;将约束线、约束点以及密度线、密度点作为面积为零的孔洞,提出了对特征约束的处理方法;针对特征约束分隔区域的情况,提出了自动确定各子区域边界的方法。基于本文提出的方法已开发出具有特征约束的四边形网格自动生成程序,并成功应用于洪水分析系统和楼面设计分析系统。     

Constraints-induced emergence of functional novelty in complex neurobiological systems: a basis for creativity in sport

In this paper we present a model of creativity captured as exploration and production of novel and functionally efficient behaviors, based on the statistical mechanics of disordered systems. In support of the modelling, we highlight examples of creative behaviors from our research in sports like boxing and rugby union. Our experimental results show how manipulation of practice task constraints changes the exploratory breadth of the hierarchically soft-assembled action landscape. Because of action metastability and differing task constraints, the specificity of each assembled movement configuration is unique. Empirically, a movement pattern's degree of novelty may be assessed by the value of the order parameter describing action. We show that creative and adaptive movement behavior may be induced by at least two types of interven-tions, based on relaxing task constraints which we term direct and indirect. Direct relaxing is typically a function of changing task constraints so that the number of affordances that can satisfy goal constraints increases. Indirect relaxing of constraints occurs when a habitual action is suppressed by, for example, stringent instructional constraints during sports training. That suppression simultaneously relaxes other correlated constraints that enable larger exploratory capacity and new affordances to emerge for the athlete or team.     

On the redundancy of complex modal parameters

Generating the simulation transfer function ( TF ) is indispensable to modal analysis, such as examining modal parameters identification algorithm, and assessing modal analysis software. Comparing 3 feasible algorithms to simulate TF shows that, one of them is preperable, which is expressing the TF as the function of the complex modal parameters(CMPs), because the deliberate behaviors of CMPs can be implemented easily, such as, dense modals, large damping, and complex modal shape, etc. Nonetheless, even this preferable algorithms is elected, the complex modal shapes cannot be specified arbitrarily, because the number of CMPs far more exceeds that in physical coordinate. So for physical realizable system, there are redundant constraints in CMPs. By analyzing the eigenvalue problem of a complex modal system, and the inversion equations from CMPs to physical parameters, the explicit redundancy constraints were presented. For the special cases, such as the real modal, the damping free modal, and non-complete identification, the specific forms of the redundancy constraints were discussed, along with the number of independent parameters. It is worthy of noting that, redundancy constraints are automatically satisfied for the real modal case. Their equivalent forms on the transfer matrix and a column of transfer matrix were also provided. These results are applicable to generate TF, to implement identification by optimization and appreciate the identification results, to evaluate residual modal, and to verify the complementary of identified modal orders.     

基于Gauss伪谱法的3D刚体摆姿态最优控制

研究Gauss伪谱法求解3D刚体摆姿态最优控制问题．针对其最优姿态控制问题,既要满足由任意位置运动到平衡位置姿态运动规划问题,又要满足系统含有动力学约束的力学模型问题,提出基于四元数来描述3D刚体摆的数学模型,建立3D刚体摆姿态的动力学和运动学方程,为了解决3D刚体摆在平衡位置处的姿态最优控制问题,设计基于Gauss伪谱算法的最优姿态开环控制器,得到了3D刚体摆的姿态最优控制轨迹,得到满足的可行解,通过仿真实验验证了其开环解在平衡位置的控制姿态最优性．     

A direct approach to design of geometry and forces of tensegrity systems

In the process of designing a tensegrity system, some constraints are usually introduced for geometry and/or forces to ensure uniqueness of the solution, because the tensegrity systems are underdetermined in most cases. In this paper, a new approach is presented to enable designers to specify independent sets of axial forces and nodal coordinates consecutively, under the equilibrium conditions and the given constraints, to satisfy the distinctly different requirements of architects and structural engineers. The proposed method can be used very efficiently for practical applications because only linear algebraic equations are to be solved, and no equation of kinematics or material property is needed. Some numerical examples are given to show not only efficiency of the proposed method but also its ability of searching new configurations.     

Geometry of constraint manifolds and wave propagation in internally constrained elastic bodies

The implications of the non-Euclidean structure of constraint manifolds on differentiation of the stress in internally constrained elastic bodies are examined, and the equations governing propagation of acceleration waves in such bodies are deduced differentiating the reactive stress consistently with the assumption that it does no work in any admissible motion. This yields a treatment of the subject in which the presence of internal constraints imposes restrictions on the set of possible amplitudes of waves but the condition for local existence of waves, that amplitudes must satisfy, is of the same type as that for bodies free from internal constraints, in the sense that it depends on the properties of the response map of the material and is independent of reactive stress.     

A new approximate solution technique for randomly excited non-linear oscillators—II

The method of weighted residuals is applied to the reduced Fokker-Planck equation associated with a non-linear oscillator, which is subjected to both additive and multiplicative Gaussian white noise excitations. A set of constraints are deduced for obtaining an approximate stationary probability density for the system response. One of the constraints coincides with the previously proposed criterion of dissipation energy balancing, and the others are useful for calculating the equivalent conservative force. It is shown that these constraints imply certain relationships among certain statistical moments; their imposition guarantees that such moments computed from the approximate probability density satisfy the corresponding exact equations derived from the original equation of motion. Moreover, the well-known procedure of stochastic linearization and its improved version of partial linearization are shown to be special cases of this scheme, and they are less accurate since the approximations are not chosen from the entire set of the solution pool of generalized stationary potential. Applications of the scheme are illustrated by examples, and its accuracy is substantiated by Monte Carlo simulation results.     

Characterization of hole-diameter in thin metallic plates perforated by spherical projectiles using genetic algorithms

The empirical and semi-empirical models available in literature for the estimation of hole-diameter in thin metallic plates by the strike of spherical projectile are mostly valid for the data for which these have been developed. This may be partly attributed to the form of the model employed for their development. The behavioural constraints and the limiting conditions are not satisfied by these models. In the present paper, some of the non-dimensional models have been developed that satisfy the behavioural constraints and limiting conditions. The data used in the development of earlier statistical models has been reanalyzed for the development of new models for the characterization of hole-diameter with a view towards seeing whether better characterization is possible. The genetic algorithm coupled with the penalty function method has been used for the constrained optimization of model parameters that result in low errors and high correlation coefficients.     

复合材料导向叶片的结构与材料一体化优化设计

将材料细观结构优化和宏观结构优化结合起来,从结构与材料两个尺度出发,发展复合材料导向叶片的结构与材料一体化优化设计方法。在建立平纹编织复合材料细观结构分析模型的基础上,采用细观力学有限元法进行材料刚度性能预测,热-固耦合分析方法进行涡轮导向叶片结构分析。同时以细观结构参数和宏观结构参数为设计变量,以叶片质量最小为优化目标,并要求满足应力和位移方面的约束,进行编织复合材料导向叶片的结构与材料一体化优化设计。     

基于一种免疫算法的结构动态参数识别

受生物免疫系统的启发提出了一些基于免疫概念的优化算法。该方法有效地避免了传统遗传算法易于出现早熟、搜索效率低及不能很好保持个体多样性等问题。本文将免疫算法应用于结构动态参数识别领域,通过其在VC环境下调用ANSYS有限元程序中的APDL语言,并在后期引入加速循环操作,从而实现了对结构动态参数快速而精确的识别,仿真算例表明其正确性、有效性和较好的抗观测噪声能力。该方法特别适合于对大型复杂结构的参数识别计算。     

Geometrically-based Consequences of Internal Constraints

When a body is subject to simple internal constraints, the deformation gradient must belong to a certain manifold. This is in contrast to the situation in the unconstrained case, where the deformation gradient is an element of the open subset of second-order tensors with positive determinant. Commonly, following Truesdell and Noll [1], modern treatments of constrained theories start with an a priori additive decomposition of the stress into reactive and active components with the reactive component assumed to be powerless in all motions that satisfy the constraints and the active component given by a constitutive equation. Here, we obtain this same decomposition automatically by making a purely geometrical and general direct sum decomposition of the space of all second-order tensors in terms of the normal and tangent spaces of the constraint manifold. As an example, our approach is used to recover the familiar theory of constrained hyperelasticity.     

Truss optimization under stiffness, stability constraints and random loading

An actual design of light-weight structures must evaluate strength, stiffness and stability constraints as well as the nature of external loading. A designed structure must satisfy optimality and safety criterions per prescribed maintenance period. One faces the known difficulties when trying to implement several from the above mentioned requirements into optimization problem for further successful numerical realization. A method to formulate the optimization problem, incorporating all above described criterions, the mathematical model and algorithm to solve it numerically, taking into account the stochastic nature of external loading, are presented for elastic–plastic truss-type structure.