In view of the problem of low diagnosis accuracy and poor adaptivity of diagnosis algorithm, an engine state detection algorithm based on danger theory and cloud decision making is proposed. A danger recogni-tion model is built based on the danger theory of immune system with the definitions of danger signal and safe signal given. A cloud decision model is created based on cloud model, and the danger recognition model and cloud deci-sion model are fused up to form a state diagnosis strategy suitable for complex mechanical system. The engine state detection algorithm is verified by tests on vehicle engine state detection test bench. The results show that the algo-rithm proposed can achieve a diagnosis accuracy rate up to 97. 5%.%针对发动机故障诊断中的诊断精度差和诊断算法自适应性差的问题，提出了一种基于危险理论云决策的发动机状态检测算法。根据免疫系统的危险理论建立了危险识别模型，给出了危险信号和安全信号的定义。基于云模型构建了云决策模型。通过融合危险识别模型和云决策模型，构建了适用于复杂机械系统的状态诊断策略。在汽车发动机状态检测试验台上，对发动机状态检测算法进行了试验验证，结果表明该算法的诊断准确率达到97.5%以上。
展开▼
