Through the characteristic analysis of linear engineering,this paper,above all,brings forward the conception of linear assessment(or band assessment) aimed at the risk analysis of large-scale linear engineering,and at the same time analyzes its meaning in detail.

Thus extracting solution of these models lead to solving the large linear systems.

This thesis mainly studies the generalized minimal error method (GMERR)from two aspects, which can be used to solve large linear systems.

Research on Symbiosis and Interaction between Large-scale Engineering and Ecosystem;

Large-scale engineering project has the characteristics of high investment,large scale,involving many factors and heavy effect.

Facing the flexible factors in the large project investing decision, the idea is put up that the large project investment should be decided flexibly in the direction of flexibility theory.

A lot of plans of investment on large projects aborted because of ignoring the risk.

In allusion to dynamic nonlinear large-scale industrial processes, to bring up gained the method of decentralized identification for the strong consistency estimates of the divisible steady-state models, it is used that property of polynomial can infinitely approach to the nonlinear function.

In this paper, the iterative learning control is studied for dynamics in steady-state hierarchical optimization of nonlinear large-scale industrial processes.