LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Actuator Fault Detection in Nonlinear Systems Using Neural Networks Rastko Selmic, Ph.D. Department of Electrical Engineering and Institute for Micromanufacturing Louisiana Tech University Ruston, LA 71272, USA Email: rselmic@latech.edu Web: http://www2.latech.edu/~rselmic/ 1 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Contents Introduction Problem Formulation Actuator Fault Detection, Fault Dynamics, and Fault Detectability Two cases considered: - State feedback - Output feedback Simulation Results Conclusion Other projects, ideas, etc. 2 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Introduction Collaborative work with Marios Polycarpou and Thomas Parisini An actuator fault identification in unknown, input-affine, nonlinear systems using neural networks is presented Two cases are considered: state feedback and output feedback case Neural net tuning algorithms and identifier have been developed using the Lyapunov approach A rigorous detectability condition is given for actuator faults relating the actuator desired input signal and neural netbased observer sensitivity Simulation results are presented to illustrate the detectability criteria and fault detection in nonlinear systems. 3 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Questions to be Answered What kind of actuator faults can be detected? Under what conditions faults are detectable using NN identifiers? If faults are not presently detectable, how identifier parameters need to be adjusted in order to detect the faults? 4 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Problem Formulation Consider a nonlinear, input-affine system given by x f ( x) g ( x)u d (t ) y h( x ) where x [ x1 , x2 ,, xn ]T is a system state and f, g: n n , h: n are unknown, smooth functions, u is the input control signal, y is the output, and d represents a system disturbance. It is assumed that the disturbance is bounded such that d (t ) DB It is assumed first that the full state measurement is available. 5 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Problem Formulation The actuator of the nonlinear system generates the control signal u based on desired actuator signal v. We consider the Gang Tao’s actuator fault model given by u (t ) v(t ) (u v(t )) , where the fault value of the actuator input is u . 0 models the actuator without a fault. 1 models the actuator with a fault. The actuator fault value can be considered as a time-varying function u (t ) v(t ) (u (t ) v(t )) Combined, the nonlinear system and the actuator fault model is x f ( x) g ( x)v(t ) (u (t ) v(t )) d (t ) 6 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Case I: State Feedback A NN system observer is given by xˆ Axˆ Ax Wˆ fT f ( x) Wˆ gT g ( x)v(t ) , where the matrix A is Hurwitz and two NNs are used to approximate nonlinear functions f(x) and g(x) f ( x) W fT f ( x) f ( x) , g ( x) WgT g ( x) g ( x) . W f , Wg are some ideal target NN weights, and f (x ) , g (x) are NN approximation errors bounded by fM and gM . An observer error given by e x xˆ 7 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS A NN System Observer u v Nonlinear System with Actuator Failure Model 1 u x f ( x) g ( x)u d (t ) 0 x e NN OBSERVER x̂ Figure 1. NN system observer – fault identifier. 8 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS NN Tuning Law A nonlinear system and observer dynamics are given by ~ ~ e Ae W fT f ( x) f ( x) WgT g ( x)v(t ) g ( x)v d (t ) where W~f W f Wˆ f , W~g Wg Wˆ g . Theorem: Stable NN Observer Tuning Law (Lewis’ NN Tuning) Given the nonlinear system and the NN observer, let the estimated NN weights be provided by the NN tuning algorithm Wˆ f C f f ( x)eT kC f e Wˆ f Wˆ C ( x)v(t )eT kC e Wˆ g g g g g with any constant matrices C f C f 0 , C g C g 0 , and a small design parameter k. Then the state observer error e and the NN weight estimation ~ ~ errors W f , Wg are uniformly ultimately bounded. The convergence region for the state observer error e can be reduced by increasing the gain A. T T 9 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Stability Analysis Stability proof is carried out using the Lyapunov function 1 1 ~ 1 ~ 1 ~ 1 ~ L eT e tr W fT C f W f tr WgT C g Wg 2 2 2 The bound on the tracking error is given by 1 2 1 2 W fM WgM fM gMVB DB 4 e 4 B min where min is a minimum eigenvalue of A, and VB is a bound on a control signal. We assume that the control signal is bounded (the system has been stabilized). 10 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS The State Observer Error Modified closed-loop error dynamics due to actuator fault is given by ~ ~ e1 Ae1 W fT1 f ( x ) f ( x ) WgT1 g ( x )v(t ) g ( x )v(t ) d (t ) g ( x )(u (t ) v(t )) where x denotes the state x when there is a fault in the actuator. It is important to consider the difference in behavior of the state observer error when the actuator is healthy and when there is an actuator fault. This difference has its own dynamics - fault dynamics. 11 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Dynamics of a Fault Fault, as a process, has its own dynamics. The fault dynamics is given by ~ ~ ~ e~ Ae~ W fT f ( x) W fT1 f ( x ) f ( x) f ( x ) WgT g ( x)v(t ) ~ WgT1 g ( x )v(t ) g ( x)v(t ) g ( x )v(t ) g ( x )(u (t ) v(t )) where e~ e e1 Equivalently, the dynamics is given by ~ ~ e~ h(e~,W fT ,W fT1 , x, x , t ) g ( x )(u (t ) v(t )) , where the function h( ) is given by ~ ~ ~ ~ h(e~,W fT ,W fT1 , x, x , t ) Ae~ W fT f ( x) W fT1 f ( x ) f ( x) ~ ~ f ( x ) WgT g ( x)v(t ) WgT1 g ( x )v(t ) g ( x)v(t ) g ( x )v(t ) 12 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Detectability of Actuator Faults Theorem: Detectability of Actuator Faults Given a NN observer from Theorem 1, a fault in the system actuator can be detected after time interval T if the following condition is satisfied t 0 T g ( x )(u ( ) v( ))d 2eB T C1 t0 ~ ~ where C1 eB max 2W fB 2WgB 2 M VB 1. The above condition provides rigorous justification for an intuitive concept of fault detectability which says that for the fault to be detected there should be “enough” difference between the desired input signal and failed actuator values. The result also relates the time before fault has been detected and the NN estimator parameters. 13 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Case II: Output Feedback Consider a single-input single-output nonlinear system in a normal form given by x Ax b f ( x) g ( x)u d (t ) y hT x where x [ x1 , x2 ,, xn ]T denotes the state vector, f, g: n are unknown smooth functions, u is the input control signal, y is the output, and d represents a system disturbance. We assume that the disturbance is bounded, that is d (t ) DB . 0 0 A 0 0 1 0 0 1 0 0 1 0 , h . 0 0 0 1 0 0 0 0 14 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS A NN Observer A NN system observer is given by xˆ Axˆ b Wˆ fT f ( xˆ ) Wˆ gT g ( xˆ )v(t ) r0 (t ) L( y hT xˆ ) yˆ hT xˆ where x̂ is an estimate of the state vector x and ŷ is an estimate of the output y. The observer gain matrix L is chosen so that As A Lh T is stable. Two NNs are used to approximate the nonlinear functions f(x) and g(x): f ( x) W fT f ( x) f ( x) g ( x) WgT g ( x) g ( x) where W f , Wg are some ideal target NN weights, and f (x) , g (x) are NN approximation errors. 15 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS A NN Observer Nonlinear System with Actuator Failure u v 1 0 u NONLINEAR SYSTEM y ~ y ŷ NN OBSERVER Figure 2. NN system observer – fault identifier. 16 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS NN Observer Tuning Law Theorem: Stable NN Observer Tuning Law (Lewis’ NN Tuning). Given the nonlinear system and the NN observer, choose the robustifying term as ~ y ro kr ~ y with a scalar k r . Let the estimated NN weights be provided by the NN tuning algorithm Wˆ f C f f ( xˆ ) ~ y kC f ~ y Wˆ f Wˆ C ( xˆ )v(t ) ~ y kC ~ y Wˆ g g g g g with any constant, symmetric matrices C f C f T 0 , Cg Cg T 0 , and a x and the suitably small design parameter k. Then the state observer error ~ ~ ~ NN weight estimation errors W f , Wg are uniformly ultimately bounded. Proof: Select the Lyapunov function as 1 T ~ 1 ~ T 1 ~ 1 ~ ~ L ~ x Px tr W f C f W f tr WgT C g 1Wg 2 2 2 17 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Dynamics of a Fault The fault dynamics is given by ~ ~ ex As ex b{W fT f ( xˆ ) WgT g ( xˆ )v(t ) c1 (t ) r0 (t )} ~T ~T b{W fF fF ( xˆ F ) WgF gF ( xˆ F )v(t ) c1F (t ) r0 F (t ) g ( xF )(u (t ) v(t ))} e y hT e x where ex ~ y~ yF . The system can then be represented in x ~ xF and e y ~ the following form: ~ ~T ~T ~T ex As ex (W fT ,W fF ,Wg ,WgF , x, xˆ , t ) bg ( xF )(u (t ) v(t )) e y hT e x 18 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Detectability of Actuator Faults Lemma. Given the NN observer and its NN parameters, the fault detectability condition is given by the following inequality t 0 T t 0 T t0 t0 T bWˆ g g (u ( ) v( ))d (WgM Ng ) (u ( ) v( )) d exp( As T )(2 ~ xB 2C2T ) 19 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Detectability of Actuator Faults The result relates observer parameters, i.e. NN weights, with fault detectability and the actuator control signal It also shows when actuator faults can not be detected or what needs to be done with NN observer to improve sensitivity. 20 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Simulation Example Consider the second order nonlinear system given by x1 x2 x2 5 x13 2 x2 ( x1 1)u (t ) y x1 Fault model is given by u (t ) v(t ) (u (t ) v(t )) Two NNs are used for the system observer. Both NNs have 2, 30, and 2 neurons at the input, hidden, and output layers, respectively. Standard sigmoid activation function is used. For both NNs, the first-layer weights are uniformly randomly distributed between -1 and 1. 21 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Simulation Example The threshold weights for the first layer are uniformly randomly distributed between -20 and 20. The second layer weights W are initialized to zero for both neural networks. Neural network tuning parameters are chosen as k=0.0001, Cf=20, Cg=20. The state observer matrix is A=diag{30,30}. An ideal control signal v(t) is given by v(t ) 10 sin(t ) 22 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Simulation Example State observer errors e1(t), e2(t) Norm of state observer error 0.4 1.5 0.2 0 1 -0.2 -0.4 0.5 -0.6 -0.8 -1 0 5 10 15 time (sec) System state observer errors e1(t) (full line) and e2(t) (dotted line). 0 0 5 10 15 time (sec) Norm of the error e(t). 23 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Simulation Example NN weights bounds and approximation error bounds are estimated as W fM WgM 3 , fM gM 0.1. Actual simulation or experimental results can also be used to estimate the above parameters. The error bound is estimated as eB 0.2 . Assume that there is a fault at t=5sec where u 18 . 24 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Simulation Example State observer errors e1(t), e2(t) Norm of state observer error 0.5 1.5 0 1 -0.5 0.5 -1 0 5 10 15 time (sec) Actuator fault at t=5sec; system state observer errors e1(t) (full line) and e2(t) (dotted line). 0 0 5 10 15 time (sec) Actuator fault at t=5sec; norm of the error e(t). 25 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Conclusions It is shown how neural net-based system can be used for actuator fault detection in unknown, nonlinear, input-affine systems. Stable neural net tuning laws are given and estimate on the state observer error is provided using Lyapunov approach. Sufficient conditions for actuator fault detectability are presented. An open research problem is to combine active fault detection methods in case detectability conditions are not satisfied. 26 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Intelligent Sensors and Actuators Group Research interests: Wireless sensor networks for chemical agents monitoring Suboptimal coverage control missions in mobile sensor networks. Intelligent actuator control using neural networks Actuators and sensors failure detection and compensation Intelligent wireless sensor networks Group members: Dr. Rastko Selmic, 3 Ph.D. students, 4 M.S. students, and 2 undergraduate students. The group has two laboratories with several control system setups, sensors, wireless sensor nodes, two mobile robots, 11 PC computers. 27 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Intelligent Sensors and Actuators Laboratory The newest lab in EE – 11 PC computers, 8 control system experimental setups, sensors, wireless sensor nodes, two mobile robots, 2 oscilloscopes. 28 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Smart Actuator Control Using IEEE 1451 Standard Develop a smart actuator control that is compatible with IEEE 1451 standard for smart transducers. The concept allows for intelligent control based on data and metadata gathered by the network of smart sensors. Control action depends on sensor data and information stored in TEDS and HEDS. Smart Actuator STIM Network (TCP/IP) NCAP Memory TEDS, HEDS Actuator Power Module 29 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Testbed Development – Chemical Agent Monitoring Developed a chemical sensor board for WSN applications based on Xbow motes. Sensor nodes monitor for carbon monoxide (CO), nitrogen dioxide (NO2), and methane (CH4). Research problem: a suboptimal sensor network coverage of the area of interest while providing quasi real-time tracking and monitoring of the focus area observation space. Link Local Computer Remote Sensor Nodes RS-232 Remote Computer Base Station Radio Link 30 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Simulation Tool for Coverage Control in Mobile Sensor Networks Simulation tool is needed to experiment with variety of algorithms for sensor node deployment under localization and network connectivity conditions. Development based on C (optimization, network conditions) and C++ (GUI). C language chosen so simulation can be ported to High Performance Computing machines in case it is needed for very large networks. Examples of different scenarios in sensor network coverage control: uniform coverage, focused coverage, balanced coverage control of sensor nodes. 31 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS 32 LOUISIANA TECH UNIVERSITY Department of Electrical Engineering and Institute for Micromanufacturing INTRODUCTION PROBLEM FORMULATION STATE FEEDBACK OUTPUT FEEDBACK OTHER PROJECTS Thank you! Any questions? 33