摘要
通过对CRBF的分析研究,提出一种基于CRBF神经网络的分类算法。采用数学模型和几何模型构造其应用模型,通过分类算法的训练过程修改应用模型中的相关参数,使得分类结果更趋合理。通过CENTUM3000和Visual Basic6.0平台开发了化工厂爆炸监控系统。实践表明,分类结果与监控设备运行结果吻合得很好,满足了工厂监控系统的实际需求,证明该分类算法和应用模型具有较高的理论和实用价值。
Based on the analysis and study of cosine radial basis function（CRBF）, a classification algorithm of neural network based on cosine radial basis function is proposed. Application model is constructed by mathematical model and geometric model, and the parameters of application model are modified through the training process of classification algorithm to make the classification result more reasonable. Classification result and supervision equipment are consistent after application of blast supervision system based on CENTUM CS3000 and Visual Basic 6.0, the factual demand is satisfied. The classification and application model are proved to be highly effective and valuable in practice and academic study.
出处
《计算机工程》
CAS
CSCD
北大核心
2007年第17期225-227,共3页
Computer Engineering
基金
国家"863"计划基金(2001AA113182)
陕西省科技攻关计划基金(2002K06-G5)
