论文部分内容阅读
在任何系统的设计中,了解系统参数变化对系统性能的影响十分重要。这种影响的测度用灵敏度来表示,记为S_x~T在网络问题中,T是任何所要研究的网络函数或电路响应;x是任何与电路元件有关的参数,它可以是元件值本身,也可以是影响元件值的某些物理参数。网络的灵敏度计算主要用于电路的容差分析、最优化设计和调整技术等方面。目前已研究出若干种面向计算机的灵敏度计算方法,本章只介绍其中应用最普遍、最有效的伴随网络法。由于计算灵敏度的实质就
In any system design, understanding the impact of system parameter changes on system performance is important. The measure of this effect is expressed in terms of sensitivity, denoted as S_x ~ T. In a network problem, T is any network function or circuit response to be studied; x is any parameter related to a circuit element, which can be the element value itself, It can be some physical parameter that affects component values. Network sensitivity calculation is mainly used for circuit tolerance analysis, optimal design and adjustment techniques and so on. So far, a number of computer-oriented methods for calculating the sensitivity have been developed. In this chapter, only the most common and effective adjoint network method is introduced. Because of the nature of computational sensitivity