Optimization of Multi-Micro Extended Analog Computer Array and its Applications to Data Mining

摘要

The micro-extended analog computer (uEAC) is a prototype implementation of extended analog computer (EAC). In this paper, a dynamic multi-uEAC array is proposed based on the analysis of the uEAC unit mathematical model. By constructing a multi-uEAC array, the limitations of single uEAC unit on nonlinear mapping can be overcome. In order to obtain an optimal multi-uEAC array structure, a comprehensive optimization strategy based on particle swarm optimizer (PSO) is developed to optimize multi-uEAC array structure matrix and connection matrix simultaneously. Two simulation examples are provided to verify the effectiveness of the proposed model.