Visual method for process monitoring and its application to Tennessee Eastman challenge problem

Online process monitoring is extremely important for the successful operation of any process. A visual data-based method suitable for online monitoring of complex systems is proposed. The self-organizing map is used to project a high-dimensional vector of process data onto a 2D visualization space in which different process conditions are represented by different regions. The process state can be indicated by the trajectory in visualization space. The effectiveness of the proposed method is illustrated by the application on the Tennessee Eastman process. Online monitoring and fault detection can be carried in a more intuitionistic and practical manner by using this method.