Integrating Model Based Fault Diagnosis with Model Predictive Control

详细信息    查看全文

• 作者：J. Prakash ; Shankar Narasimhan ; and Sachin C. Patwardhan
• 刊名：Industrial & Engineering Chemistry Research
• 出版年：2005
• 出版时间：June 8, 2005
• 年：2005
• 卷：44
• 期：12
• 页码：4344 - 4360
• 全文大小：410K
• 年卷期：v.44,no.12(June 8, 2005)
• ISSN：1520-5045

文摘

Model predictive control (MPC) schemes are typically developed under the assumption that thesensors and actuators are free from faults. Attempts to develop fault-tolerant MPC schemeshave mainly focused on dealing with hard faults, such as sensor or actuator failures, processleaks, etc. However, soft faults such as biases or drifts in sensors or actuators are more frequentlyencountered in the process industry. Occurrences of such faults can lead to degradation in theclosed loop performance of the MPC controller. Since MPC controllers are typically used to controlkey operations in a chemical plant, this can have an impact on safety and productivity of theentire plant. The conventional approach to dealing with such soft faults in MPC formulations isthrough the introduction of additional artificial states to the model. The main limitation of thisapproach is that number of artificial extra states introduced cannot exceed the number ofmeasurements. This implies that it is necessary to have a priori knowledge of which subset offaults are most likely to occur. In this paper, an active on-line fault-tolerant model predictivecontrol (FTMPC) scheme is proposed by integrating state space formulation of MPC with thefault detection and identification (FDI) method based on generalized likelihood ratios. The factthat both these schemes use a Kalman filter as their basis facilitates tight integration of thesetwo components. The main difference between the conventional MPC formulation and FTMPCformulation is that the bias corrections to the model are made as and when necessary and atqualified locations identified by the FDI component. The FTMPC eliminates offset between thetrue values and set points of controlled variables in the presence of a variety of faults whileconventional MPC does not. Also, the true values of state variables, manipulated inputs, andmeasured variables are maintained within their imposed bounds in FTMPC, while inconventional MPC, these may be violated when soft faults occur. These advantages of theproposed scheme are demonstrated using simulation studies on a CSTR process and experimentalstudies conducted on the temperature control of a coupled two tank heater system.