Saint-Petersburg State Technological Institute FAULT DIAGNOSIS IN CHEMICAL PROCESSES AND EQUIPMENT WITH FEEDBACKS L.A.Rusinov N.V.Vorobjev V.V.Kurkina I.V.Rudakova FAULT DIAGNOSIS IN OBJECTS WITH FEEDBACKS The subject-matter of my report concerns the application of chemometrics approach to solving industrial problems relevant in particular to developing diagnostic systems for fault diagnosing in difficult cases. 08.04.2015 2 Why is the diagnostics necessary? • • • • • The majority of chemical technological processes refer to the class of potentially dangerous (PDTP) or hazardous. As a rule PDTP are characterized by: high level of uncertainty, large uncontrollable disturbances, essential internal nonlinearity, bad observability. lack of mathematical descriptions (often). 08.04.2015 3 Why is the diagnostics necessary? The operation of protection systems obligatory in the PDTP is usually accompanied • by emergency dumping of a reactionary mass, • irreversible repressing of a reaction and • other operations resulting in essential losses. The early diagnostics can define the faults at their incipient stages and thus allows undertaking necessary acts to avoid the protection systems actuating. 08.04.2015 4 What is necessary for diagnostics? Continious monitoring and fault diagnostics are carried out on the basis of a diagnostic models (DM) connecting faults in the process under control (abnormal situations) with their observable symptoms. For this reason, mathematical descriptions of PDP cannot often be used as DM because they are usually valid only in PDP working zones and unsuitable for abnormal situations. 08.04.2015 5 Classification of diagnostic models Diagnostic Models Quantitative Models Observers Parity Space EKF Qualitative Models Causal Models Digraphs Process History Data Models Abstraction Hierarchy Qualitative Quantitative Expert Fuzzy Systems expert systems Statistical Functional Structural Fault Trees PCA/PLS Neural Networks Statistical Classifiers V. Venkatasubramanian, R. Rengaswamy, K Yin, S.N. Kavuri, Computers and Chemical Engineering 27 (2003) 293-311) 08.04.2015 6 THE MASKING EFFECT FROM THE FEEDBACK Control action 100% 0% Regulator has exhausted its resource The object of diagnostics The sensor signal PROCESS Recycle or Regulator Regulator has exhausted its resource Emergency situation ABNORMAL SITUATION Fault development Normal regime 08.04.2015 7 THE WAY OF SOLVING THE PROBLEM For solving the problem, the diagnostic model (DM) should be built for the process section or equipment in the closed loop formed by feedback (further - object). Deviations from this model can be used in detecting the fact that an abnormal situation has arisen. However, it is impossible to identify a cause of the fault while using deviations from the model. For this purpose, the models describing all possible abnormal situations are required. As a result, it is necessary to have a bank of models 08.04.2015 8 THE STRUCTURE FLOW CHART OF THE OBJECT DIAGNOSTIC SYSTEM 08.04.2015 9 THE FLOW-CHART OF THE DIAGNOSTICS’ PROCEDURE 1. THE OBJECT MONITORING FAULT DETECTION THE FAULT DETECTION CRITERION: i>; - threshold 2. INITIATION OF THE MODELS OF THE BANK DESCRIBING OBJECT FAULTS 3. THE FAULT REASON (FR) IDENTIFICATION: Restrictions of the method: 1. The necessity of the presence of diagnostic models, easy-in-use in real time. 2. The necessity to obtain the knowledge of all possible faults of the object under control. 08.04.2015 10 THE FUZZY DIAGNOSTIC MODEL Each fault Fi is described by fuzzy rule Ri of Takagi-Sugeno type: Ri : IF x1 == Ai1 AND... AND xn = = Ain, THEN yi = ai1x1 +... + ain xn + bi, where Ri – i-th rule of fuzzy model, i [1,k]; k – the number of rules; X = {xkj} - a matrix of input variables samples [Nxn]; k [1, N]; N – number of samples; j[1,n]; n - number of inputs; Аi = {Aij} - the fuzzy terms-sets that enter into the conditional part of each i-th rule; yi - an output of i th fuzzy rule; ai = [ai1, …, ai n] and bi - parameters of fuzzy model: Ti=[aTi, bi]. 08.04.2015 11 FLOW-CHART OF COMPUTING THE RESULT OF FUZZY MODEL The result of fuzzy model computing is determined by combination of contributions of all rules in a common k inference: y i yi i 1 k i i 1 where βi - the degree of activation of i-th rule that is determined by max-product composition: n i П A (xj), j 1 ij A ( x j ) [0,1] - the contribution of each term-set Aij to the conditional part of the rule Ri. ij 08.04.2015 12 FLOW-CHART OF COMPUTING THE RESULT OF FUZZY MODEL Coefficients of a right part of rules are determined by solution of the system equations by means of weighed МLS: The solution is: y=Xch θ ; Xch=[X,1] 1 i [ X ch W i X ch ] X ch W i y T T A ( x j ) - the diagonal matrix with the degrees of activation β ij on its diagonal 08.04.2015 13 THE DETERMINE OF MEMBERSHIP FUNCTIONS. By means of fuzzy clustering of the data object array, for example, by the algorithm of Gustafson-Kessel, the number of clusters с and the matrix of their fuzzy separation U are defined. Membership functions of fuzzy sets in the conditional part of rules are extracted from the matrix U which (g, s)-th member mgs[0,1] characterizes the value of membership of an input-output combination in s-th column in the cluster g. 08.04.2015 14 THE DETERMINE OF MEMBERSHIP FUNCTIONS. To obtain one-dimensional fuzzy set Ggj, the multidimensional fuzzy sets, defined pointwise in g-th row of the separation matrix U, are projected into input variables space Xj. Resulting fuzzy sets Ggj are usually nonconvex. To obtain the convex (unimodal) fuzzy sets, approximating by appropriate forms of membership functions (for example, Gaussian) is needed. 08.04.2015 15 MEMBERSHIP FUNCTIONS OBTAINED BY CLUSTERISATION 6 clusters – by threes clusters for a forward and reverse valve strokes M E M B E R S H I P 1 2 3 4 5 6 F U N C TI O N S THE NORMALIZED VALUES OF INPUT VARIABLES 08.04.2015 16 THE DIAGNOSTIC MODEL BASED ON THE KALMAN FILTER In this case DM is developed in the space of object states. The Kalman filter is actually searching for an optimum estimate with the least-squares method. The linear object model for Kalman filter is of the form: x ( k 1) A x ( k ) B u ( k ) n y ( k 1) C x ( k ) w where x(k) is the state vector, y(k) - the vector of filter output variables at the kth step, x(k-+1) - the predicted (extrapolated) state vector value at the (k+1)th step; A, B, C - are known prediction, control and observation matrices; n,w – noises. 08.04.2015 17 FLOW-CHART OF COMPUTING THE RESULT OF KALMAN MODEL The matrix of filter gain factors is given as: K ( k ) P ( k ) H [ S ( k )] where: P ( k ) T 1 , S(k) – correlation matrixes P ( k ) А P ( k 1) А w k T S (k ) C P (k ) C T v k The specified estimation for the system state vector is: x (k ) x (k ) K Re sidual(k); Re sidual(k) y(k ) C x (k ) And finally, the specified covariance matrix of estimation of the system state vector is given in the form: P ( k ) I K ( k ) C P ( k ) 08.04.2015 18 THE DIAGNOSTIC MODEL BASED ON THE EXTENDED KALMAN FILTER For nonlinear objects the model is of the form: x ( k ) f ( x ( k 1), u ( k 1)) y ( k ) h ( x ( k 1), u ( k 1)) In this case, the filter does not use fixed matrices A(x) and C(x), but linearizes them recursively based on the previous state estimate with the use of matrices of first partial derivatives of the state equations: Аi , j C i, j f i ( x ( k 1), u ( k 1)) x j ( k 1) x ( k 1 ) x ( k ) h i ( x ( k 1), u ( k 1)) x j ( k 1) x ( k 1 ) x ( k ) These matrices are calculated at every step and then inserted into the standard Kalman filter formulas. 08.04.2015 19 CASE STUDY Case study was carried out on two types of objects: 1. The object in control loop electropneumatic valve with the positioner 2. The object in recycle circuit - Tennessee Eastman process 08.04.2015 20 THE STRUCTURE FLOW CHART OF THE ELECTROPNEUMATIC VALVE WITH THE POSITIONER 08.04.2015 21 THE STRUCTURE FLOW CHART OF THE TENNESSEE EASTMAN PROCESS *) Applicability of the method to processes with recycles is presented in the Vorobiev’s poster presentation. 08.04.2015 22 THE ELECTROPNEUMATIC VALVE WITH THE POSITIONER The POSITIONER has the mathematical model developed in the European Interuniversity Project DAMADICS. 19 various positioner faults have been considered by the model. But it does not fulfill to the first restriction for diagnostic models of objects of the class under study: it is difficult and not easy-to-use in real time. So, models on the basis of fuzzy logic and Kalman filtering are developed and the DAMADICS model was used for their training. 08.04.2015 23 MODELED POSITIONER FAULTS F2 ABRUPT FAULT «FLUID BOILING UP IN THE VALVE CAVITY AT THE EXTREME FLOW RATE» F1 INCIPIENT FAULT «SEDIMENTATION» Input variables CV1(k), CV1(k-1), CV1(k-2) – Control signal values entered from controller at kth, (k-1)th and (k-2)th steps ZT – position of the valve plunger at the kth step Output variable 08.04.2015 FT – The flow rate through the valve 24 OUTPUT RESIDUALS OF FUZZY MODELS (INCIPIENT FAULT F1) DETECTION RESIDUALS ,% (MODEL OF NORMAL REGIME) time RESIDUALS ,% (MODEL OF FAULT F1 time RESIDUALS ,% (MODEL OF FAULT F2 IDENTIFICATION time 08.04.2015 25 OUTPUT RESIDUALS OF FUZZY MODELS (ABRUPT FAULT F2) DETECTION Forward plunger stroke RESIDUALS ,% (MODEL OF NORMAL REGIME) time Reverse plunger stroke RESIDUALS ,% (MODEL OF FAULT F1 time RESIDUALS ,% (MODEL OF FAULT F2 IDENTIFICATION time 08.04.2015 26 OUTPUT RESIDUALS OF MODELS BASED ON THE KALMAN FILTERS (INCIPIENT FAULT F1) DETECTION RESIDUALS ,% (MODEL OF NORMAL REGIME) time RESIDUALS ,% (MODEL OF FAULT F1 time IDENTIFICATION RESIDUALS ,% (MODEL OF FAULT F2 time 08.04.2015 27 OUTPUT RESIDUALS OF MODELS BASED ON THE KALMAN FILTERS (ABRUPT FAULT F2) RESIDUALS ,% (MODEL OF NORMAL REGIME) DETECTION time RESIDUALS ,% (MODEL OF FAULT F1 time IDENTIFICATION RESIDUALS ,% (MODEL OF FAULT F2 time 08.04.2015 28 CONCLUSIONS Сhemometrics methods are very effective for execution the monitoring and diagnostics of technological processes in chemical and related industries, even in difficult cases at diagnostics of the objects in circuits with feedbacks because of feedback masking effects Statistical methods allow constructing diagnostic models on the base of the history process data, not demanding the knowledge of process chemism and the presence of its mathematical descriptions. 08.04.2015 29 CONCLUSIONS For diagnosing such faults, it is suggested to use the bank of diagnostic models describing normal operation of the objects under control and their operation when faults are available. Applicability of the method is illustrated by the example of system development with two types of diagnostic models: the model with fuzzy rules of Takagi-Sugeno type and on the basis of extended Kalman filters. Both models have demonstrated approximately equal results when diagnosing both incipient and abrupt faults. 08.04.2015 30