1994 Volume 27 Pages 1-5
The automation of roof crane system is a important field of FA. Our purpose is to attain the automation of this system using a fuzzy neural network (FNN). Fuzzy variables for input and output layers are introduced for the method of FNN in this study. And learning using a neural network is carried out. The final output value of the FNN is obtained by applying fuzzy reasoning in output layer units. Then, by giving initial deflection and disturbance for the crane, the amount of hanging load deflection and control time to reach target position were analyzed by using computer simulation. In the case of the control to initial deflection, though residual deflection existed, it was observed that the trolley in the crane reached the target position accurately. In the case of the control to disturbance, it was found that the state of deflection was changed greatly by moving speed of the trolley.