In this chapter we examine the fundamental design trade-offs that apply to linear scalar filtering problems based on bounded error estimators. We will see that, due to the condition of bounded error estimation, the filtering sensitivity functions introduced in the last chapter are necessarily constrained at points in the complex plane determined by ORHP poles and zeros of the plant. These interpolation constraints, in turn, translate into Poisson and Bode-type integral relations, which show essential limitations in the achievable performance, and induce clear trade-offs in filter design.
Unable to display preview. Download preview PDF.
Notes and References
- This chapter was mainly based on Goodwin et al. (1995), Seron & Goodwin (1995) and Seron (1995).Google Scholar