A visual method of fault diagnosis for a complicated process was developed based on self-organizing map (SOM).Due to the high dimensionality of the complicated process

principal component analysis (PCA) was introduced to reduce the dimension of the process data.Then the self-organizing map was utilized to project the preprocessed data onto a 2D visualization space in which different process conditions were represented by different regions.Online monitoring could be achieved by the dynamic trajectory in the visualization space.The cause of certain fault could be deduced from the U-matrix of the derived SOM network and the loadings vector of the principal components.The application to the Tennessee Eastman process (TEP) demonstrated that fault detection and diagnosis could be carried out in a more intuitional and practical manner by using the proposed method.