Automatic Design of Spiking Neural P Systems for an Adder with Multiple Natural Numbers
Guoliang Li, Jianping Dong, Gexiang Zhang, Haina Rong, Ao Gu, Qiyu Liu and Sergey Verlan
The automatic design of spiking neural P (SN P) systems is an important research area within the field of membrane computing. In this paper, we present an automated design method that introduces a novel approach for generating the initial population in evolutionary processes. Our method specifically addresses the challenge of multi-natural-number addition combinations by dynamically adjusting the rules of neurons and their interconnection. The proposed approach consists of three key steps. First, for each addition combination in the problem, evolutionary processes are applied to search for successful individuals, which are then stored in a designated repository. This process constructs a candidate repository of SN P systems. Second, the initial population for further evolution is formed by extracting successful individuals from the repository. Finally, an evolutionary search is conducted providing individuals that allow the simultaneous computation of multiple addition combinations. Experimental results demonstrate that the proposed method is both feasible and effective in the automatic design of SN P systems for solving multi-natural-number addition combination problems.
Keywords: Spiking neural P systems, multi-natural-number addition combination, automatic design, evolutionary search
DOI: 10.32908/ijuc.v21.zhang02