Research of system controller based on cascade analog field calculation model

Unconventional Extended Analog Computer (uEAC) is a new type of computer. It contains natural substances as its computing carrier, obeys the physical laws, and conducts the information processing by analog field calculation. Aiming at uEAC´s excellent characteristic of self-study, this paper analyzed the uEAC´s system unit model, and proposed an optimization method of analog field calculation controller, which is similar to neural networks. A controller of cascade analog field calculation model is built up, and a derivation of parameter optimization using steepest descent method is given. By the simulation on time-invariant and slow time-varying controlled object, this paper illustrated that the given adaptive controller has a high control accuracy, and converges fast. Therefore, the validity and feasibility of this model is proved.