Pattern Recognition and Data Classification
Pattern recognition is the research area that studies the operation and design of systems that recognise and classify patterns in data. The classification decision in such systems is made on the basis of observed attributes or features, and each datum is assigned to one class from a set of predefined classes. In the following, a few basic classification techniques, i.e., statistical or Bayesian classifiers, classification by decision trees and neural networks are reviewed. Statistical...
Info Qwt
Figure 8.4. The bond graph 0-junction a , and 1-junction b . Components, along flow paths in CFS, form bond graph junctions type 1 junction - that corresponds to a loop of interconnected components, type 0 junction - that corresponds to a node of interconnected components. Each junction's common variables are effort in 0-junction and flow in 1-junction the noncommon power variable is specific to each component and all enter a sum e.g., the flow in the 0-junction , as in Figure 8.4a,b. In the...
Info Eau
point-to-set similarity functions for the two categories are shown in Figures 4.5 and 4.6 left , respectively. The two plots are drawn for all the points in the Cartesian product 0,16 x 0,16 . If we decrease the value of p to 1.8 for the second category, the contour plot for this category matches more accurately the topology of the area occupied by points in the category Figure 4.6, right .
Fault Detectability Fault Isolability
In the presence of system faults, the residual vector will be r s Hy s Gf s f s Grf s f s r s Gf s i fi s G.rf s 2 f2 s lt Gf s g fg s 10 where Grj s Hy s Gj s represents the relation between residual and faults, Gfs is the i-th column of matrix Grf andf s is the i-th component off s . The fault fi is detectable in the residual r s if the corresponding column of Gf s is nonzero, Gr s i 0 this is called the fault detectability condition of the residual r s to the fault f Chen and Patton, 1999 ....
Benchmark Applications
The applications of computational intelligence techniques to fault diagnosis tasks presented in this book have been validated using five benchmarks. The book revolves around two main benchmarks aero-engines gas path faults Chapters 2 and 6 , and the control valve faults used in the European Commission's FP5 DAMADICS project Chapters 3, 4, 7 and 10 , respectively. Other three chapters are concerned with diagnosis of a power generation plant Chapter 9 , a rolling mill plant Chapter 8 , and...
sm h [vhm1 vhmvhm16
The reason the filtered segment is split again into three frames in Eq. 6 is very simple. The filter was not actually applied to all frames in sm but to its main frame, the central one. The lateral frames are only context signals that tell to the filter there are nonnull signal values before and after the main frame. Since filters are shift invariant linear systems Oppenheim and Schafer, 1985 Proakis and Manolakis, 1996 , the main frame in Eq. 6 is also the central one. Therefore, from the...
PointtoPoint Similarity Measure Based on Pearson Correlation
The Pearson correlation Weisstein, 1999 measures the similarity in the trends of two signals. Let us suppose that s and t represent the measurements of two signals over the same time window. The formula used to compute the correlation between the vectors s and t is given in Eq. 3. The terms zs and zt represent the z-scores of s and t, respectively. The z-score of a vector is obtained by first subtracting the mean value and then dividing by its standard deviation. The product between zs and zt...
References 1
1. Kandel A 1986 Fuzzy mathematical techniques with applications. Addison-Welsey, USA. 2. Kosko B 1997 Fuzzy engineering. Prentice Hall, New Jersey. 3. Marinai L, Ogaji S, Sampath S and Singh R 2003a Engine Diagnostics -Fuzzy Logic Approach. In Proceedings of the Seventh International Conference on Knowledge-Based Intelligent Information amp Engineering Systems - KES'03, Oxford, 3-5 September. 4. Marinai L, Singh R and Curnock B 2003b Fuzzy-logic-based diagnostic process for turbo fan engines....
Automated Procedure
The procedure to generate fuzzy rules was automated via the graphical user interface GUI shown in Figure 2.14. This GUI constitutes the first of two windows of the diagnostics module based on fuzzy logic described in Marinai, 2004 . This first GUI is aimed at setting up fuzzy logic diagnostics models for a given engine. A second interface is aimed at operating the diagnostics models created to estimate the possible faults - see section 2.7. Figure 2.14. Fuzzy diagnostic model setup GUI. Figure...
Fault Diagnosis Techniques and Approaches
Fault diagnosis research deals with real-world problems as plant efficiency, maintainability and reliability. For safety-critical systems, such as nuclear plants and aircrafts, the problem of detecting the occurrence of faults is of high importance. The consequences of faults in such systems could be disastrous in terms of human mortality and environmental impact. To a lesser extent, fault detection in process and manufacturing industries is also crucial in order to improve production...
Engine Reliability Availability and Diagnostic Techniques
Operation and maintenance costs of a gas turbine contribute a major portion of the annual maintenance budget of a company. In view of the changes in world economy towards globalisation and openness of the market, any efforts that can reduce the total cost of ownership and life cycle cost of the equipment will be added advantages. The primary objectives of all maintenance strategies are to reduce equipment downtime, increase reliability and availability of the equipment, which at the same time...
A
where kA and kB are the number of data points of groups A and B, respectively, yi and yiB are the output data of groups A and B, respectively, y AB is the model output for group A estimated using the data from group B, and y BA is the model output for group B estimated using the data from group A. Thus, using two groups of data, A and B, two fuzzy models are built for each group, starting with only one input. The RC is computed for each model, and the one that minimizes RC is selected as the...
What Is The Diagnosis Hoh
The outcome of this analysis highlighted two optimal combinations of functional parameters that show a minimum number of MS and HS cases and a minimum average value of RMS. These best layouts are for the cases 1 and 9 that correspond respectively to the following layout Best choice AND Product, Implication Product, Aggregation Summation, Defuzzification Centroid. Second best choice AND Product, Implication Product, Aggregation Summation, Defuzzification Centre of Maximum. Case 1 was selected as...
References
1. Ariton V and Palade V 2005 Human-like fault diagnosis using a neural network implementation of plausibility and relevance. Neural Computing amp Applications 14 2 149-165 2. Ayoubi M 1994 Fault diagnosis with dynamic neural structure and application to a turbocharger. In Proceedings of 1st IF AC Symposium SAFEPROCESS'94, Espoo, Finland, vol. 2, pp. 618-623 3. Babuska R 2002 Neuro-fuzzy methods for modeling and identification. In Abraham A, Jain LC and Kacprzyk J eds Recent Advances in...
NeuroFuzzy Systems for TakagiSugeno Fuzzy Model Implementation
The most general Takagi-Sugeno model has as consequence of the fuzzy rules ARMA AutoRegressive Moving Average models of higher order Palade et al., 2002 , as shown in Eq. 37. THEN yt t cl X pjx t - j X sjy t - j K ' where i 1, ,r, r is the number of rules, x x1, x2, , xk is the input vector, Pj Pj1, .,Pjk , sj sJ1i, ., sjk , and x t-j , y t-j , j 1, or n2, represent the past values for the inputs and output of the system. If the two sums in the consequent of the rule given in Eq. 37 are...
Info Gvc
where the parameter a e 0,1 controls the percentage of rectangular window centred inside. For the vibration segment, a good choice is a 1 3, since the central frame takes only one third of the whole segment. Figure 5.4. Nine of the most utilized signal processing windows. Figure 5.4. Nine of the most utilized signal processing windows. All windows above are symmetric, as shown in Figure 5.4, where, beside the window shape, the parameter values are also depicted for Chebyshev, Kaiser, Lanczos...
Fuzzy Algebra Basic Elements of a Fuzzy System Architecture
Engineering science typically deals with uncertain variables and approximations to a fixed number of decimal places that depend on the accuracy capability but also on the necessity and costs of being accurate. When a decision has to be made based on uncertain values of a set of variables, a binary logic based on either-or laws can become a limitation. A fuzzy system based on multivalue logic can help in modelling a process when a mathematical model of how the system's outputs depend on the...
Contributors
Danubius University of Galati Lunca Siretului no. 3, 800416 Galati, Romania Computer Science and Engineering Department Dunarea de Jos University of Galati Domneasca 47, Galati, Romania Email cosmin.bocaniala ugal.ro IDMEC ISEL, Polytechnic Institute of Lisbon Mechanical Engineering Studies Centre Rua Conselheiro Emidio Navarro, 1950-062 Lisbon, Portugal Email jcalado dem.isel.ipl.pt School of Electrical and Electronic Engineering University of Science Malaysia Engineering Campus, 14300 Nibong...