Control of an Inverted Pendulum

Project Proposal for IRS9 –Master of Science in Intelligent Reliable Systems (IRS)Fall 2005, Control Laboratory, Aalborg University EsbjergFault Diagnosis and Fault Tolerant Control of a Three-tank SystemBackground and ProblemThree-tank system has been proposed as a benchmark system for system modeling, identification, control, faultdetection and diagnosis, as well as for fault-tolerant control. The system exhibits typical characteristics of aconstrained hybrid system and has been proven useful to serve as a test environment for algorithms concerningstate estimation, parameter identification, and control of hybrid systems. Two lab-scale experimental systemscan be viewed in Fig.1 as below. Such a system can be viewed as a prototype of many industrial applications inprocess industry, such as (petro)chemical and oil & gas plants.The three-tank liquid level control system is a multi-input-multi-output (MIMO) system, but with the valves(actuators) between the tanks closed, each tank can be treated as a single-input-single-output system (SISO).The typical control issue involved in the system is how to keep the desired liquid level in each tank in thepresence of disturbances and even actuator, sensor or leakage faults in the tank system.a) A setup from University of Western Ontario, Canada b) A setup built by an IRS8 project, AAUEFig. 1. Two experimental setup examples of three-tank systemObjectiveThe objective of this project is to investigate integrated fault diagnosis and fault-tolerant control strategies forthe experimental three-tank system as shown in the right figure in Fig. 1. The main task is to design andimplement an active (or passive) fault-tolerant controller that is able to maintain the desired liquid level in thepresence of actuator, sensor or leakage faults in the tank system.Following tasks need to be investigated in the project:• Hardware: The three-tank system is available.• Modelling and simulation: Set up the mathematical model under normal and fault conditions, analyzeand simulate system performance under the Matlab\Simulink environment.• Fault diagnosis and fault-tolerant controller design:o Develop fault detection and diagnosis (FDD) algorithms based on state and/or parameterestimation approacheso Develop fault-tolerant control (FTC) algorithms in conjunction with the designed FDD scheme.o Integration of FDD and FTC and interaction analysis between FDD and FTC• Implementation and testing: The designed FDD and FTC algorithms need to be implemented andtested in the hardware system. Evaluation and validation of the designed methods in hardware-in-loopsimulation need to be conducted.

Project Proposal for IRS9 –Master of Science in Intelligent Reliable Systems (IRS)Fall 2005, Control Laboratory, Aalborg University EsbjergReferences:1.2.3.4.5.DE8-F05-4 project report, FDD for the three-tank system, AAUE, Spring 2005.D. Mignone and N. Monachino, The Total Three Tank Tutorial Text, Technical Report, ETH Zurich,2001.B. Heiming and J. Lunze. Definition of the Three-Tank Benchmark Problem for ControllerReconfiguration, Proceedings of the European Control Conference, Karlsruhe, Germany, 1999.J-C. Ponsart, D. Theilliol and H. Noura, Fault-tolerant control method for actuator and componentfaults applied to a hydraulic system, Proc. of the European Control Conference 2001, pp. 1044-1049.J-C. Ponsart, C. Join, D. Theilliol and D. Sauter, Sensor fault diagnosis and accommodation innonlinear system, Proc. of the European Control Conference 2001, pp. 2346-2351.Key words: Fault detection and diagnosis (FDD), fault tolerant control (FTC); PID (Proportional-Integral-Derivative) control, state-space model-based reconfigurable control.RemarksThis project is an extension and continuation of an IRS8 project investigated in Fall 2004, as shown in reference[1]. The emphasis of this project is to investigate mainly for active fault tolerant control (FTC), i.e., integratedFDD and FTC.Proposer: Youmin Zhang, FUV 0.22; Phone: 7912 7741Email: ymzhang@cs.aaue.dk; URL: http://www.cs.aue.auc.dk/~ymzhang/