Application of chance-constrained programming to design centering, tolerance assignment, and tuning

Due to various influencing factors such as tolerance, temperature, humidity, electrical circuit design problems always involve uncertainty and probability. In this paper, chance-constrained programming, a stochastic mathematical model, is improved to fit this problem, mostly concerned with design centering, tolerance assignment, and tuning. The mathematical formulation of this approach is formed in a stochastic manner. Thus the essence of our original purpose in dealing with consumable electronics production, to minimize the cost in a practical manner while keeping the required maximal qualification rate, is well represented under this model. Stochastic simulation, neural networks and genetic algorithms are integrated to produce a hybrid intelligent algorithm for solving this model. A realistic circuit design illustrates this approach.