Self-healing Control for Multi-input Multioutput Systems Based on a Fuzzy Neural Network Inverse Model

Abnormal working conditions such as valve wear, pipeline leakage and environmental change often occur in modern industrial processes. According to a specific situation, abnormal working conditions are divided into three categories. In this paper, a self-healing controller that can automatically recover the key performance indicators of the whole system is designed to be utilized in some abnormal conditions that do not affect the safety of the system. On this basis, this paper proposes a self-healing controller based on fuzzy neural network inverse model. The controller obtains the historical data under normal operation status, and establishes a fuzzy neural network inverse model of a nonlinear multivariable process. The value of the adjustable variable in the industrial process is calculated through the real-time data collected, and then the adjustable variable is adjusted to return the key performance indicators of the entire system to the normal or close-to-normal level.