NEUROPT: neurocomputing for multiobjective design optimization forprinted circuit board component placement

摘要

The traveling salesman problem (TSP) can be mapped to the problemnof optimal component placement for printed circuit boards (PCBs). Anninnovation that consists of adapting a neural network formulation of TSPnto multiobjective component placement on the basis of wire-lengthncriteria, as well as thermal reliability, is described. The author showsnthat the mathematical formulation of the Hopfield energy function fornTSP is identical to the energy for the placement problem except for thencost (distance) function. The Hopfield cost function can be modified bynintroducing terms to model the wire-length and thermal reliability fornalternative component placements. This approach was tested by coding anneural network simulator and comparing the quality of the resultingnplacement with standard methods. The positive results of that testing,nthe potential for a dramatic improvement in the time needed to calculatensuch optimal placements, and the natural way of extending thenformulation to include more design criteria lead to a confidence thatnsimilar approaches will have a significant impact on multiobjectivendesign