基于SOFM的调度员培训仿真系统智能评估

An Intelligent Evaluation of DTS Based Self-Organizing Feature Map

自组织特征映射(SOFM)网络利用神经元权值向量表示输入数据的结构、具有较好的分类能力。本文针对电力调度员培训仿真系统(DTS)的培训评估,根据模式分类的思想,提出了基于自组织神经网络的智能评估方法,使仿真培训结果的评价客观化、自动化,在南阳电力调度员培训仿真系统中的应用验证了该方法的可行性。

The self-organizing feature map （SOFM） uses weight of network to present structure of the input data and has preferable ability of classification. Aiming at evaluation of dispatcher training simulation （DTS） and according to the idea of pattern classification, an intelligent method based on the SOFM network is put forward in this paper. It makes the process of evaluation automation. The practical verifications of the DTS in Nanyang Power System have demonstrated the feasibility and effectiveness of the proposed methcd.