Research on Computer System Architecture for Artificial Intelligence

Abstract

A network storage resource allocation method based on genetic simulated annealing algorithm is proposed in this article, and it is applied for data storage and backup among heterogeneous systems, as well as the optimal allocation of mobile Agent dispatching sequence in storage process. This method takes into account the timing, priority and storage availability of each mobile Agent and makes the allocation of resources more reasonable on the premise of ensuring the effective execution of each mobile Agent. At the same time, SA is introduced to perform local fine search in GA, maintaining the diversity of the group and avoiding precocious convergence. Simulation experiments show that the mobile Agent storage resource allocation method based on GASA algorithm can not only improve the storage efficiency, but also improve the network real-time and storage performance.