Cellular neural network to model and solve direct non-linear problems of steady-state heat transfer

Different non-electric problems can be efficiently modeled and solved by using their electric counterparts. The paper describes an original method to model and solve the direct nonlinear problem of heat transfer in solids by using the cellular neural network (CNN). By modifying constitutive equations of the hear transfer which are nonlinear functions of temperature, it is possible to rearrange them to become linear functions of some new variable. Similarly, by slightly modifying an original Chua CNN we can establish direct analogy between rearranged hear transfer equations and those of the modified CNN. Consequently, different heat transfer process in solids, having an arbitrary shape, can be solved by using CNN.