Process control via artificial neural networks and learning automata

The application of learning automata for control is shown to give significantly improved performance by two modifications (a) using a set of proportionality constants of a proportional feed-back controller, as the action set of learning automata and (b) a scheme which uses artificial neural networks as memory. This work examines a class problems where the objective function is reduced to inequality constraints and also where qualitative information from artificial neural networks can be used with learning automata as an effective decision maker. Two nonlinear process control examples are tackled. It is shown that despite deterioration of net´s prediction error the learning automata-ANN control strategy works while the performance improves resulting a better accuracy