Adaptive Neural Network Control for Generating Random Shape

The adaptive neural network is trained to incorporate random deformation process into the given regular determined shape according to certain rules. Thus, the regular shape can be changed to irregular one with the adaptive learning approach, and both the global and local shape controllability can be enhanced. With improvement on the traditional adaptive neural network algorithm, the certainty and randomness can be fully combined, and the fuzzy controllability and adjustability can be achieved easily and concisely.