No abstract

High-level analog circuit design is a complex problem domain in which evolutionary search has recently produced encouraging results. However, little is known about how to best structure evolution for these tasks. The choices of circuit representation, tness evaluation technique, and genetic operators clearly have a profound e ect on the search process. In this paper, we examine tness evaluation by comparing the e ectiveness of four tness schedules. Three tness schedules are dynamic the evaluation function changes over the course of the run, and one is static. Coevolutionary search is included, and we present a method of evaluating the problem population that is conducive to multiobjective optimization. Twenty-ve runs of an analog ampli er design task using each tness schedule are presented. The results indicate that solution quality is highest with static and coevolving tness schedules as compared to the other two dynamic schedules. We discuss these results and o er two possible explanations for the observed b ehavior: retention of useful information, and alignment of problem di culty with circuit pro ciency.

No abstract

The relatively new field of Evolvable Harding for rapid solution testing (a common bottleneck in many evolutionary algorithms). Some of the pioneering work in this area was done by Higuchi [5] and Thompson The other circuit design domain and the focus of this paper is analog circuitry. Analog circuits are of great importance in electronic system design since the world is fundamentally analog in nature. While the amount of digital design activity far outpaces that of analog de-https://ntrs.nasa.gov/search.jsp?R=20000116210 2020-07-14T08:35:43+00:00Z sign, uu)stdigitalsyst(_rns reqttir(_ ;taah)g m(),ht[es for int('rfa('ing t() the ext(,rmd world. Te('hniques for analog (:ir('iiit iI(_sigii ;uitoination t)egan appi:ariug al)out two (l(',_a(h,._ .'_g() ((:.g., [_01). Etf,,rts .si._ te('hniqu(_ fr,)in evoillti()n;try (:()UlliUti).tion have appeared over the last few y(;ars. Tlii,se inchi(le the I1_ of gem:tic algorithms (GAs) {7i t,) seh;ct filter c()mp()t,,at sizes [_], t() select filter top(Jhigi(;s [3J, and to design operational amplifiers using a sniall set of topologies [12 i. Various analog filter design prol)lems have been solved using genetic programruing (e.g., It li) , and an overview of these techniques, including eight analog circuit synthesis problems, is found in llOl.

No abstract