Combinatorial optimization with higher order neural networks-cost oriented competing processes in flexible manufacturing systems

In this paper, higher order neural networks are applied to handle combinatorial optimization problems by using cost oriented competing processes (COCP). This method has a high adaptability to complicated problems. The COCP are adapted to flexible manufacturing systems (FMS) which are based on the concept of cellular robotic systems (CEBOT). In these systems a number of optimization problems have to be solved which cannot be easily handled by using known heuristics. In contrast to neural networks without higher order couplings the output of the COCP are only valid solutions of the optimization problem. The competing process of each neuron favours the selection of the lowest costs by considering the constraints of the problem. The neural network dynamics with higher order couplings used here can be described by a potential function and a gradient descent method with suitable initial conditions.