摘要
BP算法在神经网络中应用较为广泛,但有收敛速度慢、易于陷入局部极小的缺点,而蚁群算法是一种新型的模拟进化算法,有正反馈、分布式计算、全局收敛、启发式学习等特点。将蚁群算法和神经网络结合起来,应用于设备故障专家系统的知识获取和诊断推理中,可以提高运算效率,具有很好的应用前景。利用该方法,对测得的样本数据进行实验分析,证明此系统具有推理效率及准确性较高的特点。
BP Algorithm is widely used in neural networks. But it has some shortcomings, such as slow convergent speed and easy convergence to the local minimum points. Ant colony system is a novel simulated evolutionary algorithm. It has characteristics of positive feedback, distributed computation, and so on. The combination of ant colony algorithm with neural network is adopted in knowledge acquisition and reasoning of Device Failure diagnosis expert system and it can improve the efficiency of operation. It has wide application prospects. Using this method, a lot of sample data are analyzed on the experiment, which has proved this system has characteristics of higher inference efficiency and accuracy.
出处
《北京联合大学学报》
CAS
2008年第4期63-66,共4页
Journal of Beijing Union University
关键词
神经网络
蚁群算法
故障诊断
知识获取
BP算法
neural network
ant colony algorithm
failure diagnosis
knowledge acquisition
BP algorithm
