First-order optimization method

In this chapter, principle introductions are conducted in terms of seismic response control, the analytic hierarchy process (AHP) and first-order optimization method. Through the parameters sensitivity analyses of 3-dimensional model for the Runyang Suspension Bridge (RSB), damper optimization for seismic control of the RSB under certain seismic input are realized, on the basis of the combination of the AHP and first-order optimization method. 

The first-order optimization method (Wang, Li, Mao, 2005 Jaishi Ren, 2005), the zero order... 

The zero-order optimization method is similar to the first-order optimization method with the main difference being in the disposition of variables. Derivatives of the variables are used in the first-order optimization method whereas variables themselves are used in the zero-order optimization method. 

Determining Assessment Functions Based on First-Order Optimization Method... 

Four assessment functions are proposed as the optimization criteria, by first-order optimization method to deal with quantitive assessment. Different functions can lead to different optimization results and their optimization results are compared with each other to indentify the most effective one. 

Applications of continuum solvation approaches to MCSCF wavefunctions have required a more developed formulation with respect to the HF or DFT level. Even for an isolated molecule, the optimization of MCSFCF wavefunctions represents a difficult computational problem, owing to the marked nonlinearity of the MCSCF energy with respect to the orbital and configurational variational parameters. Only with the introduction of second-order optimization methods and of the variational parameters expressed in an exponential form, has the calculation of MCSCF wavefunction became routine. Thus, the requirements of the development of a second-order optimization method has been mandatory for any successful extension of the MCSCF approach to continuum solvation methods. In 1988 Mikkelsen el ol. [10] pioneered the second-order MCSCF within a multipole continuum model approach in a spherical cavity. Aguilar et al. [11] proposed the first implementation of the MCSCF method for the DPCM solvation model in 1991, and their PCM-MCSCF method has been the basis of many extensions to more robust second-order MCSCF optimization algorithms [12],... 

Most of the feature selection methods neglect second-order effects. If two features are highly correlated both features will be selected if they are valuable for the classification - although the second feature contains the same information as the first one. Optimal methods for the elimination of correlations generate new orthogonal features this can either be effected by the Karhunen-Loeve method (Chapter 8.2) or by the SELECT method (Chapter 9.5 C1583). 

Although the calculation of the Hessian is time consuming, the effort is quickly compensated by the excellent convergence properties of the Newton-Raphson approach [293-295]. This optimization technique solved the convergence problems of first-order MCSCF methods, which optimized orbitals and Cl coefficients in an alternating manner (recall chapter 9). Even perturbative improvements of the four-component CASSCF wave function are feasible and have been implemented and investigated [527]. 

Indirect or variational approaches are based on Pontryagin s maximum principle [8], in which the first-order optimality conditions are derived by applying calculus of variations. For problems without inequality constraints, the optimality conditions can be written as a set of DAEs and solved as a two-point boundary value problem. If there are inequality path constraints, additional optimality conditions are required, and the determination of entry and exit points for active constraints along the integration horizon renders a combinatorial problem, which is generally hard to solve. There are several developments and implementations of indirect methods, including [9] and [10]. 

Th c Newton-Raph son block dingotial method is a second order optim izer. It calculates both the first and second derivatives of potential energy with respect to Cartesian coordinates. I hese derivatives provide information ahont both the slope and curvature of lh e poten tial en ergy surface, Un like a full Newton -Raph son method, the block diagonal algorilh m calculates the second derivative matrix for one atom at a lime, avoiding the second derivatives with respect to two atoms. 

In order to obtain the best accuracy results as quickly as possible, it is often advantageous to do two geometry optimizations. The first geometry optimization should be done with a faster level of theory, such as molecular mechanics or a semiempirical method. Once a geometry close to the correct geometry has been obtained with this lower level of theory, it is used as the starting geometry for a second optimization at the final, more accurate level of theory. 

On the basis of data obtained the possibility of substrates distribution and their D-values prediction using the regressions which consider the hydrophobicity and stmcture of amines was investigated. The hydrophobicity of amines was estimated by the distribution coefficient value in the water-octanole system (Ig P). The molecular structure of aromatic amines was characterized by the first-order molecular connectivity indexes ( x)- H was shown the independent and cooperative influence of the Ig P and parameters of amines on their distribution. Evidently, this fact demonstrates the host-guest phenomenon which is inherent to the organized media. The obtained in the research data were used for optimization of the conditions of micellar-extraction preconcentrating of metal ions with amines into the NS-rich phase with the following determination by atomic-absorption method. 

In this chapter we described Euler s method for solving sets of ordinary differential equations. The method is extremely simple from a conceptual and programming viewpoint. It is computationally inefficient in the sense that a great many arithmetic operations are necessary to produce accurate solutions. More efficient techniques should be used when the same set of equations is to be solved many times, as in optimization studies. One such technique, fourth-order Runge-Kutta, has proved very popular and can be generally recommended for all but very stiff sets of first-order ordinary differential equations. The set of equations to be solved is... 

The method of steepest descent uses only first-order derivatives to determine the search direction. Alternatively, Newton s method for single-variable optimization can be adapted to carry out multivariable optimization, taking advantage of both first- and second-order derivatives to obtain better search directions1. However, second-order derivatives must be evaluated, either analytically or numerically, and multimodal functions can make the method unstable. Therefore, while this method is potentially very powerful, it also has some practical difficulties. 

As explained above, the QM/MM-FE method requires the calculation of the MEP. The MEP for a potential energy surface is the steepest descent path that connects a first order saddle point (transition state) with two minima (reactant and product). Several methods have been recently adapted by our lab to calculate MEPs in enzymes. These methods include coordinate driving (CD) [13,19], nudged elastic band (NEB) [20-25], a second order parallel path optimizer method [25, 26], a procedure that combines these last two methods in order to improve computational efficiency [27],... 

The well-known Box-Wilson optimization method (Box and Wilson [1951] Box [1954, 1957] Box and Draper [1969]) is based on a linear model (Fig. 5.6). For a selected start hyperplane, in the given case an area A0(xi,x2), described by a polynomial of first order, with the starting point yb, the gradient grad[y0] is estimated. Then one moves to the next area in direction of the steepest ascent (the gradient) by a step width of h, in general... 

In contrast, the NBO and NRT methods make no use of molecular geometry information (experimental or theoretical), but instead provide optimal descriptions of orbital composition or electron-density distributions based directly on the first-order density operator. For this reason the NBO/NRT indices have predictive utility for a broad range of chemical phenomena, without bias toward geometry or other particular empirical properties. 

From numerous tests involving optimization of nonlinear functions, methods that use derivatives have been demonstrated to be more efficient than those that do not. By replacing analytical derivatives with their finite difference substitutes, you can avoid having to code formulas for derivatives. Procedures that use second-order information are more accurate and require fewer iterations than those that use only first-order information(gradients), but keep in mind that usually the second-order information may be only approximate as it is based not on second derivatives themselves but their finite difference approximations. 

Liquid-liquid extraction is carried out either (1) in a series of well-mixed vessels or stages (well-mixed tanks or in plate column), or (2) in a continuous process, such as a spray column, packed column, or rotating disk column. If the process model is to be represented with integer variables, as in a staged process, MILNP (Glanz and Stichlmair, 1997) or one of the methods described in Chapters 9 and 10 can be employed. This example focuses on optimization in which the model is composed of two first-order, steady-state differential equations (a plug flow model). A similar treatment can be applied to an axial dispersion model. 

Wolbert et al. in 1991 proposed a method of obtaining accurate analytical first-order partial derivatives for use in modular-based optimization. Wolbert (1994) showed how to implement the method. They represented a module by a set of algebraic equations comprising the mass balances, energy balance, and phase relations ... 

The basic GC-model of the Constantinou and Gani method (Eq. 1) as presented above provides the basis for the formulation of the solvent replacement problem as a MILP-optimization problem. For purposes of simplicity, in this chapter, only the first-order approximation is taken into consideration (that is, W is equal to zero). In this way, the functions of the target properties of the generated molecules (solvent replacements) are written as monotonic functions of the property values, thereby, leading to a linear right hand side of the property constraints (property model equation), as follows,... 

---