ANN is a promising technique to solve optimization problems because it can emulate the
operations of the brain and uses parallel processing to save computational time [17]. The method
dates back to six decades. McCulloch and Pyne [18,19] utilized logical calculus to emulate
nervous activities. Since then, various types of analogue neural networks have been proposed for
computation. In particular, Hopfield and Tank [1] made a momentous contribution for neural
computation by solving a traveling salesman problem. Many investigators followed their work.
Although Shirazi and Yih [20] present some weakness of the approach, many followers make
further improvements and verifications (see [21]). Since then ANN approaches have been widely
accepted as a competitive approach to traditional optimization techniques.