Part II presents some of the mathematical and numerical methods which are used to solve the device modelling equations. Short sections of code are given in order to illustrate the implementation of some of the methods. The discretisation of the Poisson and Schrödinger equations are introduced in terms of finite differences in Chapter 7. This chapter also includes descriptions of the solution of linear equations, time discretisation, and function fitting and updating. Since sections of code written in C++ are given in some later chapters, this chapter also includes a brief description of some of the simpler elements of the language. Chapter 8 introduces the Fermi in- tegrals and associated integrals which arise in the calculation of electron densities and energies, and describes how they may be approximated. Chapter 9 introduces the upwinding method. Chapter 10 presents a description of the Newton method for solving nonlinear equations, with a description of how this method can be directed towards device modelling by the judicious grouping of the physical variables. The phaseplane method is introduced in Chapter 11, and the multigrid method is intro- duced in Chapter 12. The approximate and numerical solutions of the Schrödinger equation are discussed in Chapter 13. Genetic algorithms and simulated annealing are introduced in Chapter 14. These methods have their limitations, and they cer- tainly cannot be applied for the full solution of the device equations. But I have found them to be useful in subsidiary optimisation problems for which the standard, and more rigorous, optimisation methods are difficult to implement. The process of imposing, and refining/de-refining, a non-uniform rectangular grid over a device is discussed in Chapter 15.
đang được dịch, vui lòng đợi..
