ABSTRACT
Neural network models inspired by biological nervous systems are providing a new approach to problem solving. Optimization is an area in which the neural network approach has not been used widely. Two different types of optimization problems can be solved using neural networks. The first type is control of mobile robots, where a neural network is used to learn the relationship between sensory input and behavior (Miyamoto et al., 1988). The second type is classical optimization, where a neural network is used to find a node configuration (an equilibrium point) which minimizes an energy or objective function (Tagliarmi et al., 1991; Hui and Zák, 1992).
