Solution by Matrix Inversion

In the case of two identical equations, the determinant of the matrix A vanishes, from property 2 of determinants, deseribed in Chapter 9. The determinant of A will also vanish in more eomplieated types of linear dependence. If det(A) vanishes, either the equations are linearly dependent or they are inconsistent. 

It is possible for a set of equations to appear to be overdetermined and not actually be overdetermined if some of the set of equations are linearly dependent. 

We write a set of linear inhomogeneous equations in matrix form  

The solution is represented by a column vector that is equal to the matrix product A C. In order for a matrix to possess an inverse, it must be nonsingular, which means that its determinant does not vanish. If the matrix is singular, the system of equations cannot be solved because it is either linearly independent or inconsistent. We have already discussed the inversion of a matrix in Chapter 9. The difficulty with carrying out this procedure by hand is that it is probably more work to invert an by matrix than to solve the set of equations by other means. However, with access to Mathematica, BASIC, or another computer language that automatically inverts matrices, you can solve such a set of equations very quickly. 

The inverse of the relevant matrix has already been obtained in Example 9.10. 

---

Regardless of the type of model used, a method must be chosen for the self-consistent solution of the polarizable degrees of freedom. Direct solution via matrix inversion is nearly always avoided by most researchers in the field, because of the prohibitive O(N ) scaling with system size, N. Both iterative and predictive methods reduce the scaling to match that of the potential evaluation [O(N ) for direct summation 0(N In N) for Ewald-based meth-ods ° " 0 N) if interactions are neglected beyond some distance cutoff], but the cost of the iterations means that the predictive methods are always more efficient. Extended Lagrangian methods have been implemented for all four types of polarizable... 

When evaluated, the summation in Eq. (11) is a rank-one polytensor that represents the potential experienced at molecule A in terms of field components, field gradient components, and so on. This can be used with response properties such as shielding polarizabilities to find property changes dues to electrical influence. The evaluation is analogous to Eq. (11). The incorporation of mutual or back polarization/hyperpolarization requires a self-consistent solution for the induced moments, and this can be done iteratively [170] or if there are no hyperpolarizabilities, it can be done by matrix inversion. 

This is the primary means of obtaining information about the canopy source distribution of a scalar from atmospheric concentration measurements. A formal discrete solution is found by matrix inversion of Et]. (17), choosing the number of source layers (m) to be ecjual to the number of concentration measurements (n) so that D j is a scjuare matrix. However, this solution provides no redundancy in concentration information, and therefore no possibility for smoothing measurement errors in the concentration profile, which can cause large errors in the inferred source profile. A simple means of overcoming this problem is to include redundant concentration information, and then find the sources , which produce the best fit to the measured concentrations c, by maximum-likelihood estimation. By minimizing the squared error between measured values and concentrations predicted by Eq. (17), 4>j is found (Raupach, 1989b) to be the solution of m linear ec[uations... 

The MM/PCM simulation in Fig. 11.9 consists of 124,000 surface discretization points (those for which F, > 10 ). As such, solution of Eq. (11.22) by matrix inversion or other 0[N ) methods is clearly infeasible, and a linear-scaling approach (in both memory and CPU... 

The linear equation (Eq. 12.272) can be readily solved by matrix inversion to find the solution for the vector Y. Knowing this vector Y, the quantity of interest / is calculated in Eq. 12.246. It is written in terms of the quadrature as... 

Method of Solution We solve this problem by matrix inversion because matrix operations are much faster than element-by-element operations in M ATLAB, especially when a large number of equations are to be solved. In order to solve the set of equations in matrix format, u j values have to be rearranged as a column vector. Therefore, we put in order all the dependent variables in a vector and renumber them from to (p+ ) q + 1), as illustrated in Fig. E6.1<2. Using this single numbering system, Eq. (6.50) can be written as... 

Next in the function are several sections dealing with building the matrix of coefficients and the vector of constants according to what is discussed in the method of solution. Finally, the function calculates all the u values by matrix inversion method using the built-in MATLAB function inv. The outputs of elUptic.m are the vectors of jc and y and the matrix of values. 

Looking at the matrix equation Ax = b, one would be tempted to divide both sides by matrix A to obtain the solution set x = b/A. Unfortunately, division by a matrix is not defined, but for some matrices, including nonsingular coefficient matrices, the inverse of A is defined. 

Phase Inversion (Solution Precipitation). Phase inversion, also known as solution precipitation or polymer precipitation, is the most important asymmetric membrane preparation method. In this process, a clear polymer solution is precipitated into two phases a soHd polymer-rich phase that forms the matrix of the membrane, and a Hquid polymer-poor phase that forms the membrane pores. If precipitation is rapid, the pore-forming Hquid droplets tend to be small and the membranes formed are markedly asymmetric. If precipitation is slow, the pore-forming Hquid droplets tend to agglomerate while the casting solution is stiU fluid, so that the final pores are relatively large and the membrane stmcture is more symmetrical. Polymer precipitation from a solution can be achieved in several ways, such as cooling, solvent evaporation, precipitation by immersion in water, or imbibition of... 

There are various ways to obtain the solutions to this problem. The most straightforward method is to solve the full problem by first computing the Lagrange multipliers from the time-differentiated constraint equations and then using the values obtained to solve the equations of motion [7,8,37]. This method, however, is not computationally cheap because it requires a matrix inversion at every iteration. In practice, therefore, the problem is solved by a simple iterative scheme to satisfy the constraints. This scheme is called SHAKE [6,14] (see Section V.B). Note that the computational advantage has to be balanced against the additional work required to solve the constraint equations. This approach allows a modest increase in speed by a factor of 2 or 3 if all bonds are constrained. 

The above model was solved numerically by writing finite difference approximations for each term. The equations were decoupled by writing the reaction terms on the previous time steps where the concentrations are known. Similarly the equations were linearized by writing the diffusivities on the previous time step also. The model was solved numerically using a linear matrix inversion routine, updating the solution matrix between iterations to include the proper concentration dependent diffusivities and reactions. 

There is no defined operation of division for matrices. However, a comparable result can be obtained by multiplying both sides of an equation (such as equation 4-2 by the inverse of matrix [A], The inverse (of matrix [A], for example) is conventionally written as [A]-1. Thus, the symbolic solution to equation 4-2 is generated by multiplying both sides of equation 4-2 by [A]-1 ... 

And again, since any matrix multiplied by its inverse is a unit matrix, this provides us with the explicit solution for b, which was to be determined. 

Buccal dosage forms can be of the reservoir or the matrix type. Formulations of the reservoir type are surrounded by a polymeric membrane, which controls the release rate. Reservoir systems present a constant release profile provided (1) that the polymeric membrane is rate limiting, and (2) that an excess amoimt of drug is present in the reservoir. Condition (1) may be achieved with a thicker membrane (i.e., rate controlling) and lower diffusivity in which case the rate of drug release is directly proportional to the polymer solubility and membrane diffusivity, and inversely proportional to membrane thickness. Condition (2) may be achieved, if the intrinsic thermodynamic activity of the drug is very low and the device has a thick hydrodynamic diffusion layer. In this case the release rate of the drug is directly proportional to solution solubility and solution diffusivity, and inversely proportional to the thickness of the hydrodynamic diffusion layer. 

Consistent with the notion of approximating polynomials as in the preceding paragraph, finite-element approaches attempt to simplify the solution process by carefully choosing polynomials so as to minimize the number of coefficients required to determine and simplify the matrix inversion problem. One choice is the Bezier polynomials, defined by... 

In Example 1.1.4 we could replace e by the vector ai in the basis for all i. Matrix inversion (or solution of a matrix equation) is, however, not always as simple. Indeed, we run into trouble if we want to replace by... 

At convergence, the errors associated with the solution of the inversion procedure can be characterized by the variance-covariance matrix x of x given by ... 

This allows us to represent partial differential equations as found in the balance equations using the collocation method. Equation (11.47) is a solution to a partial differential equation represented by a system of linear algebraic equations, formed by the interpolation coefficients, oij, and the operated radial functions. The interpolation coefficients are solved for using matrix inversion techniques to approximately satisfy the partial differential equation... 

If there are n0 open channels at energy E, there are n linearly independent degenerate solutions of the Schrodinger equation. Each solution is characterized by a vector of coefficients aips, for i = 0,1, defined by the asymptotic form of the multichannel wave function in Eq. (8.1). The rectangular column matrix a consists of the two n0 x n0 coefficient matrices ao, < i Any nonsingular linear combination of the column vectors of a produces a physically equivalent set of solutions. When multiplied on the right by the inverse of the original matrix a0, the transformed a-matrix takes the canonical form... 

---