Numerical simulation of distributed dynamic systems using hybrid intelligent computing combined with generalized similarity analysis

摘要

The hybrid use of generalized similarity analysis (GSA) with intelligent computing tools such as neural networks and fuzzy logic, provide accurate and fast numerical simulation for distributed dynamic systems. The GSA combines dimensional, inspectional, and order-of-magnitude methods to derive the complete set of a minimum number of high-level designable variables. Thus, the input variable space is reduced, and this in turn reduces the number of input vectors needed for model development. The generated concise fuzzy logic neural network leads to a shorter running time, and a greater accuracy. This approach is multidisciplinary and is demonstrated for three different applications: injectivity in an oil reservoir, oil reservoir characterization, and a driven distributed transmission line with its load.