Circuit Optimization Via Sequential Computer Experiments: A Case Study (1997)

Summary:

In electrical engineering, circuit designs are now often optimized via circuit simula­tion computer models. Typically, there are many response variables characterizing the circuit's performance. Each response is a function of many input variables, including factors that can be set in the engineering design and noise factors representing manu­facturing conditions. We describe a modelling approach appropriate for the simulator's deterministic input-output relationships. Nonlinearities and interactions are identified without explicit assumptions about the functional form. These models lead to pre­dictors to guide the reduction of the ranges of the designable factors in a sequence of experiments. Ultimately, the predictors are used to optimize the engineering design. We also show how visualization of the fitted relationships facilitates understanding of the engineering trade-offs between responses. The example used to demonstrate these methods, the design of a buffer circuit, has multiple targets for the responses, representing different trade-offs between the key performance measures. 

Keywords:

Circuit simulator; Computer code; Computer model; Engineering design; Parameter design; Stochastic process; Visualization.