免疫克隆算法在船舶故障诊断中的应用

The application of immune clone algorithm in ship fault diagnosis

分析免疫克隆算法的缺点,提出利用聚类竞争机制进行算法改进。详细阐述算法的实现流程,包括聚类竞争、克隆扩增等。最后通过与遗传算法、BP神经网络算法、免疫克隆和本文改进的算法,从诊断精度、收敛速度等方面进行对比。实验结果表明,本文算法识别率高、收敛速度快。

This paper fully analyzed the disadvantages of immune clone algorithm,and put forward the improved algorithm based on cluster competition mechanism. The realization process of the algorithm was described in detail,including cluster competition,clone and amplification,etc.. Finally through with genetic algorithm,BP neural network algorithm,immune clone and the improved algorithm,from the diagnosis accuracy and convergence speed were compared. Experimental results showed that the algorithm recognition rate is high and the speed of convergence was fast.