复杂过程的智能集成故障诊断模型研究

The Intelligent Integrated Model Research for Complex Process Fault Diagnosis

由于复杂过程与一般工业过程的本质区别,使得传统的依赖对象精确数学模型的故障诊断方法难以取得令人满意的结果。而智能技术具有无需建立对象精确模型的优势,并且可以充分利用人类专家的经验知识,因此利用智能故障诊断模型研究适合复杂过程实现的故障诊断技术既是必要的也是可行的。通过比较目前研究较多的智能故障诊断模型,提出基于模糊逻辑、神经网络与专家系统的智能集成故障诊断模型,以便有效地解决复杂过程的故障诊断问题,并具体分析了模型组成机理、结构和网络学习算法,从而为复杂过程的故障诊断技术研究提供了新的途径。

Owing to the essential differences of complex process with usual industrial process,it is difficult to get the satisfied results by the traditional fault diagnosis methods which depend on the exact mathematic model of the objects.However,the intelligent technology does not need to construct the exact mathematic model of the objects and can fully take advantage of human experts' experience,so it is not only necessary but also feasible to study fault diagnosis techniques which adapt to complex pro cess by using the intelligent fault diagnosis model.The intelligent integrated fault diagnosis model based on fuzzy logic,neural network and expert system is suggested by comparing the present general intelligent fault diagnosis models,which can effectively resolve the fault diagnosis problems of complex process.The compositive mechanism of model,construction and learning algorithm of network are analyzed in detail.Thereby a new approach of fault diagnosis technology research for complex process is proposed.