Parameter optimization-based TE process fault diagnosis method for deep belief network model

Abstract

The invention discloses a parameter optimization-based TE process fault diagnosis method for a deep belief network model. The method comprises the following steps: carrying out test simulation by taking a TE process as a research object to obtain simulation data, and dividing the simulation data into a training set data sample and a test set data sample; based on a quantum particle swarm optimization algorithm, optimizing a deep belief network to obtain an optimized deep belief network; substituting the training set data sample into the optimized deep belief network for training to obtain a TEprocess deep belief network fault diagnosis model; substituting the test set data sample into the TE process deep belief network fault diagnosis model, and diagnosing a fault of the TE process to obtain a test set fault data sample; and evaluating a fault diagnosis result according to the test set fault data sample. The method has the beneficial effects that the convergence speed is higher, the global convergence capability is stronger, and the phenomena that a DBN algorithm easily falls into local minimum, is insufficiently trained and is premature can be avoided.

Description

technical field [0001] The invention relates to the technical field of TE process fault diagnosis, in particular to a TE process fault diagnosis method based on a parameter optimized deep belief network model. Background technique [0002] The modern chemical production process is increasingly large-scale, integrated and refined. When a failure occurs in a certain part of the system, if it is not dealt with in time, it may cause the failure to expand and lead to major accidents. Therefore, it has become the key to the whole production process to establish an efficient and accurate real-time fault detection and diagnosis system, eliminate hidden troubles, troubleshoot in time, and ensure safe, stable and high-quality production. [0003] For example: TE process is a simulation of an actual chemical process proposed by J.J.Downs and E.F.Vogel of the process control group of Tennessee Eastman Chemical Company in the United States, and is widely used in the research of process ...

AI Technical Summary

Problems solved by technology

Although the above method has improved the accuracy rate, due to the large network size, difficulty in convergence, and longer training time, the generalization ability of the network is reduced, and the diagnosis result cannot be obtained quickly.

Images