摘要
油气管道腐蚀失效检测具有多因素性、复杂性、非线性和随机性等多个特点,利用精确的数学模型描述有一定的难度。论文提出了一种基于混合蚁群聚类算法的模糊神经网络的管道腐蚀失效检测方法。聚类采用与K-均值方法混合的蚁群聚类方法,将该聚类方法用于模糊神经网络构建中,建立了基于模糊神经网络的管道腐蚀失效检测模型。通过利用实际的管道腐蚀失效检测数据进行诊断应用,取得了较好的识别效果,验证了该模型及算法的有效性及可行性。
The failure detection of the oil and gas pipeline corrosion involve many characteristics, such as multiple factors, complex, non-linear, randomness and so on, it is difficult to describe by using the precise mathematical model. This paper proposes a failure detection method on pipeline corrosion, which is the Fuzzy neural network based on Hybrid ant colony clustering algorithm. The clustering method used the Ant colony clustering algorithm that mixed with the K-Means algorithm, which is used in the building of the Fuzzy neural network, adoption a hybrid of ant colony with K-means clustering method, the clustering method for fuzzy neural network construction, and in this way, it sets up a pipeline corrosion failure detection model. It obtains a good recognition effect, by using the actual pipeline corrosion failure detection data to diagnostic applications, and the result verify the validity and feasibility of the model and algorithm.
出处
《计算机与数字工程》
2013年第6期887-890,950,共5页
Computer & Digital Engineering
关键词
蚁群聚类方法
K-均值算法
模糊神经网络
管道腐蚀诊断
ant colony clustering algorithm
K-Means algorithm
fuzzy neural network
pipeline corrosion diagnosis
