Optimization simplex

As with programmed temperature GC, the application of the Simplex optimization procedure to programmed solvent LC is relatively straightforward. The same procedure can be used both for isocratic and for gradient optimization, as long as an appropriate criterion is selected for each case.  

An indication of this latter effect can be found in figure 6.11, which shows the result of the Simplex optimization procedure applied to the programmed solvent LC separation of three antioxidants [621]. 

The sum of peak-valley ratios was used as the resolution term in a composite optimization criterion, which otherwise corresponds to eqn.(4.30). Berridge also added a term to describe the contribution of the number of peaks (n). With this, the complete optimization criterion became 

The desired analysis time (fmax) was set equal to 4 min., whereas the value of the minimum time (tmin, which is irrelevant for the optimization process see section 4.4.2) was taken to be 1.5 min. 

It can be seen in the chromatogram of figure 6.11 that four peaks (the three antioxidants plus an unknown impurity) are amply resolved to the baseline. This implies that all values for the peak-valley ratio P are equal to 1 and that the criterion has become very insensitive to (minor) variations in the resolution between the different peak pairs. In the area of the parameter space in which four well-resolved peaks are observed, the only remaining aim of the optimization procedure is to approach the desired analysis time of 4 minutes. The irrelevance of the minimum time tmin is illustrated by the occurrence of the first peak in figure 4.9 well within the value of 1.5 min chosen for this parameter. 

The simplex methods, as the very name implies, are based on very simple algorithms that can be very easily implemented on analjrtic instruments, transforming the optimization of their performance into an automatic procedure. On the other hand, simplex optimization is always sequential since we can only go to the next step after we know the result of the immediately preceding step. Whereas when we are determining a response siuface we can perform several experiments at the same time to complete a factorial design, the simplex methods only permit us to do one experiment at a time (that is why they are called sequential). This characteristic makes simplex use most convenient for rapid response instruments that are ofl en encountered in anal3d ical chemistry laboratories. 

The simplex methods have other limitations that we should take into account when we choose an optimization method. In the first place, they 

The 10th edition of Webster s Collegiate Dictionary defines simplex as a spatial configuration of n-dimensions determined by n+1 points in a space of dimension equal to or greater than n. In the simplex optimization methods this configuration is a polygon (or its multidimensional equivalent) of p+1 vertices, where p is the number of independent variables that we wish to adjust. With two variables, therefore, the simplex is a triangle. With three, a tetrahedron. With four or more, a hyperpolyhedron. The number of factors defines the number of dimensions in which the simplex moves. 

It is evident that factorial design and the method of steepest ascent will be very complicated when several factors are involved. The next section describes a method of optimization which is conceptually much simpler. 

The position of the new vertex of a simplex is in practice found by calculation rather than drawing this is essential when there are more than two factors. The calculation (using constant step sizes) is most easily set out as shown in Table 7.7, the calculation lines being labelled (i)-(v). in this example there are five factors and hence the simplex has six vertices. (Note that it is not essential for each factor to have a different level for each of the vertices for example factor A takes the value 2.5 for each of the vertices 3-6.) in the initial simplex the response for vertex 4 is the lowest and so this vertex is to be replaced. The coordinates of the centroid of the vertices which are to be retained are found by summing the coordinates for the retained vertices and dividing by the number of factors, k. The displacement of the new point from the centroid is given by (iv) = (ii) - (iii), and the coordinates of the new vertex, vertex 7, by (v) = (ii) -i- (iv). 

The effect of these variable step sizes is that (when two factors are being studied) the triangles making up each simplex are not necessarily equilateral ones. The benefit of the variable step sizes is that initially the simplex is large, and gives rapid 

It can be seen that in contrast to factorial designs used in the method of steepest ascent, the number of experiments required in the simplex method does not increase rapidly with the number of factors. For this reason all the factors which might reasonably be thought to have a bearing on the response should be included in the optimization. 

Once an optimum has been found, the effect on the response when one factor is varied while the others are held at their optimum levels can be investigated for each factor in turn. This procedure can be used to check the optimization. It also indicates how important deviations from the optimum level are for each factor the sharper the response peak in the region of the optimum the more critical any variation in factor level. 

---

Muller K and Brown L D 1979 Location of saddle points and minimum energy paths by a constrained simplex optimization procedure Theor. Chim. Acta 53 75... 

Simplex optimizations have been tried in the past. These do not assume a quadratic surface, but require far more computer time and thus are seldom incorporated in commercial software. Due to the unavailability of this method to most researchers, it will not be discussed further here. 

Because the size of the simplex remains constant during the search, this algorithm is called a fixed-sized simplex optimization. Example 14.1 illustrates the application of these rules. 

Progress of Fixed-Sized Simplex Optimization for Response Surface in Figure 14.10... 

Progress of a fixed-sized simplex optimization for the response surface of Example 14.1. The optimum response at (3, 7) corresponds to vertex 25. 

The following set of experiments provides practical examples of the optimization of experimental conditions. Examples include simplex optimization, factorial designs used to develop empirical models of response surfaces, and the fitting of experimental data to theoretical models of the response surface. 

Leggett, D. L. Instrumental Simplex Optimization, /. Chem. Educ. 1983, 60, 707-710. 

A variable-size simplex optimization of a gas chromatographic separation using oven temperature and carrier gas flow rate as factors is described in this experiment. 

This experiment describes a fixed-size simplex optimization of a system involving four factors. The goal of the optimization is to maximize the absorbance of As by hydride generation atomic absorption spectroscopy using the concentration of HCl, the N2 flow rate, the mass of NaBH4, and reaction time as factors. 

This experiment describes a variable-size simplex optimization of the quantitative analysis of vanadium as... 

In this experiment the goal is to mix solutions of 1 M HCl and 20-ppm methyl violet to give the maximum absorbance at a wavelength of 425 nm (corresponding to a maximum concentration for the acid form of methyl violet). A variable-size simplex optimization is used to find the optimum mixture. 

R. J. Senorans, J. Villen, J. Tabera and M. Heiraiz, Simplex optimization of the direct analysis of free sterols in sunflower oil by on-line coupled reversed phase liquid chromatography-gas clnomatography , 7. Agric. Food Chem. 46 1022-1026 (1998). 

A simplex-optimization program that incorporates this scheme is used in the example nonlinear fitting (Section 4.2). 

Three paths can be advanced (1) expansion, e.g., Taylor series (2) trial and error, e.g., generating curves on the plotter and (3) simplex optimization algorithm. (See Section 3.1.)... 

Deming, S. N., and Morgan, S. L., Simplex Optimization of Variables in Analytical Chemistry, Anal. Chem. 45, 1973, 278A-283A. 

Multivariate chemometric techniques have subsequently broadened the arsenal of tools that can be applied in QSAR. These include, among others. Multivariate ANOVA [9], Simplex optimization (Section 26.2.2), cluster analysis (Chapter 30) and various factor analytic methods such as principal components analysis (Chapter 31), discriminant analysis (Section 33.2.2) and canonical correlation analysis (Section 35.3). An advantage of multivariate methods is that they can be applied in... 

Identification of sources of analytical bias in method development and method validation is another very important application of reference materials in geochemical laboratories. USGS applied simplex optimization in establishing the best measurement conditions when the ICP-AES method was introduced as a substitute for AAS in the rapid rock procedure for major oxide determinations (Leary et al. 1982). The optimized measurement parameters were then validated by analyzing a number of USGS rock reference samples for which reference values had been established first by classical analyses. Similar optimization of an ICP-AES procedure for a number of trace elements was validated by the analysis of U S G S manganese nodule P-i (Montaser et al. 1984). 

Sllylation, preparation of bonded phases (LC) 324 Simplex optimization selectivity (LC) 478 temperature program (GC) 56 Single column ion chroBatography 434, 438 Sintered glass layer (TLC) 671 Size-exclusion chroBatography 439... 

The Simplex optimization method can also be used in the search for optimal experimental conditions (Walters et al. 1991). A starting simplex is usually formed from existing experimental information. Subsequently, the response that plays the... 

Walters, F.H., Parker, J Llyod, R., Morgan, S.L., and S.N. Deming, Sequential Simplex Optimization A Technique for Improving Quality and Productivity in Research, Development, and Manufacturing, CRC Press Inc., Boca Raton, Florida, 1991. 

Because this proceeding is relatively expensive, an effective semi-quantitative method is widely used in optimization, the sequential simplex optimization. Simplex optimization is done without estimation of gradients and setting step widths. Instead of this, the progress of the optimization... 

As an example, in Fig. 5.7 a simplex optimization is shown in a simplified way, i.e., only by reflexions and with simplexes of invariable size. The approach to the optimum is indicated by rotation or oscillation of the simplex. Then contractions should be included into the operations. 

Fig. 5.7. Simplex optimization within a response surface of two factors X and x2 with simplexes of invariable size... 

Kinetics Analysis of Consecutive Reactions Using Nelder-Mead Simplex Optimization... 

A three-dimensional simplex optimal search for a minimum provides the following intermediate results ... 

---