A comparison of the different CNN implementations in solving the problem of spatiotemporal dynamics in mechanical systems

The large computing power of cellular neural networks (CNNs) is used to compute the transient response of a mechanical vibrating system. A basic question is how an inhomogeneous mechanical system can be computed by using different CNNUM (cellular neural net Universal Machine ) implementations. How complexity to time transformations were used in different CNNUM implementations, namely in software, in emulated digital CNN chips and in analog chips is discussed.