摘要
介绍了一种使用人工神经网络的典型模型—BP算法,构造了棉花病虫害学习和诊断系统。系统性能测试结果表明该网络不仅能100%正确诊断已学样本,而且对未学样本也能以60%以上的正确率进行诊断,同时系统还有较强的容错和抗干扰能力。该系统已投入实际使用,收到了明显的经济效益。
A typical model
based on artificial networkcotton pest learning and diagnostic system based on BP algorithm is
introduced. Test results of system performance demonstrates that the network can fully diagnose
not only the samples which have learned, but also the samples which have not learned in the rate
of above 60%,and the system has powerful capabilities of tolerance and antijamming. This
system has come into use and obtained a visible economic benefit.
出处
《电脑开发与应用》
1999年第1期25-27,共3页
Computer Development & Applications
基金
国家自然科学基金
关键词
棉花病虫害
诊断
神经网络
BP neural network,
pest diagnosis, diagnostic system, BP algorithm, system test, module