Dynamic Multiple-Fault Diagnosis With Imperfect Tests

Imperfect test outcomes, due to factors such as unreliable sensors, electromagnetic interference, and environmental conditions, manifest themselves as missed detections and false alarms. This paper develops near-optimal algorithms for dynamic multiple fault diagnosis (DMFD) problems in the presence of imperfect test outcomes. The DMFD problem is to determine the most likely evolution of component states, the one that best explains the observed test outcomes. Here, we discuss four formulations of the DMFD problem. These include the deterministic situation corresponding to perfectly observed coupled Markov decision processes to several partially observed factorial hidden Markov models ranging from the case where the imperfect test outcomes are functions of tests only to the case where the test outcomes are functions of faults and tests, as well as the case where the false alarms are associated with the nominal (fault free) case only. All these formulations are intractable NP-hard combinatorial optimization problems. Our solution scheme can be viewed as a two-level coordinated solution framework for the DMFD problem. At the top (coordination) level, we update the Lagrange multipliers (coordination variables, dual variables) using the subgradient method. At the bottom level, we use a dynamic programming technique (specifically, the Viterbi decoding or Max-sum algorithm) to solve each of the subproblems, one for each component state sequence. The key advantage of our approach is that it provides an approximate duality gap, which is a measure of the suboptimality of the DMFD solution. Computational results on real-world problems are presented. A detailed performance analysis of the proposed algorithm is also discussed.

Abstract: Inversion of self-potential anomaly for 2-D inclined sheets of infinite horizontal extent has been studied. Least-square inversion and very fast simulated annealing global optimization has been used to model the five parameters of self potential anomaly. The method of least square and very fast simulated annealing global optimization method is compared and analyzed. Very fast simulated annealing can model the noisy and field data of self potential anomaly very precisely than linear inversion technique. However, time taken by very fast simulated annealing inversion is larger than linearized inversion. The comparative analysis has been done on synthetic data (noise free and noisy) and two field data from Bavarian woods anomaly, Germany and Surda anomaly, India to show the efficacy of both the methods. The estimated parameters were compared with those from previous studies using various global optimization algorithms, mainly neural network, genetic algorithm and particle swarm optimization on the same field data sets. It can be concluded that the global optimization algorithms considered in this study were able to yield compatible solutions with those from least-square methods. The present global optimization method is in good agreement with the other global optimization methods in terms of results and computation time.

The present paper proposes a fault diagnosis methodology of three phase inverter circuit base on radial basis function (RBF) artificial neural network trained by particle swarm optimization (PSO) algorithm. Using the appropriate stimulus signal, fault features are extracted from efficient points in frequency response of the circuit directly and then a fault dictionary is created by collecting signatures of different fault conditions. Trained by the examples contained in the fault dictionary, the RBF neural network optimized by PSO has been demonstrated to provide robust diagnosis to the difficult problem of soft faults in three phase inverter circuits. The experimental result shows that the proposed technique is succeeded in diagnosing and locating faults effectively. Keyword: three phase inverter circuit, RBF neural network, fault diagnosis, particle swarm optimization.

Summary The Simulated Annealing (SA) algorithm has been widely used in geophysical data processing. However, the seri al SA algorithm is limited by low computing efficiency. In this paper, the annealing schedule has been improved in the SA algorithm, and a constrained inversion of magnetotelluric (MT) data using the parallel SA algorithm has been applied. The test of synthetic data shows that the improved Very Fast Simulated Annealing (VFSA) method is significantly better than the traditional method. Additionally, the computing efficiency has been improved using the parallel SA inversion algorithm. This method has also been applied in MT data processing for geothermal exploration and good geological results have been achieved.

Many complex networks display community structure—group of nodes within which connections are dense but between which they are sparser. A quantitative measure called modularity (Q) has been proposed to effectively assess the quality of community structure. Many community detection methods based on Q have been proposed. However, they have low accuracy or time consuming. In this paper, we suggest a fast simulated annealing method (FSA) to detect communities. An initial community partition is first obtained accord to similarity metric and then the FSA method is used to optimize the Q. Experimental results on real and synthetic networks show that compared with the existing simulated annealing method (SA), FSA method can not only maintain the quality of community but also improve the efficiency greatly.

Data-Driven Fault Diagnosis under Sparseness Assumption for LTI Systems

Contingency screening and ranking is a particularly important part of power system security analysis. In order to improve the speed and accuracy of screening and sorting of fault set in distribution network, this paper proposes a novel method of contingency screening and ranking in distribution network based on Phasor Measurement Unit (PMU). Based on PMU measurement data and historical fault data, the hierarchical cloud theory is used to calculate the weight relationship between voltage, short-circuit current, active power and reactive power loss under fault conditions, and pre-screening the expected fault set by constructing the branch parameter screening index to narrow the fault screening range. Due to the PMU-π branch model, the calculation method of active load margin and voltage threshold in view of PMU is proposed. The active load margin fault screening index and voltage stability ranking index are constructed to realize the rapid screening and sequencing of severe faults. Finally, this method proved to be fast, comprehensive and effective when applied to IEEE33 system.

An AC generator is an important component in producing an electric power and so it requires highly reliable protection relays to minimize the possibility of demage occurring under fault conditions. It is a need for research of digital generator protection system(DGPS) for the next-generation ECMS and an efficient operation of protection control system in power station. However, most of protection and control system used in power plants have been still imported as turn-key and operated in domestic. This may cause the lack of the correct understanding on the protection systems and methods, and thus have difficulties in optimal operation. In this paper, presented ratio differential relaying(RDR) is main protective element in generator protection IED. The fault detection technique, operation zone and setting value of the RDR were studied and, compared with two of the fault detection algorithm. For evaluation performance of the RDR, the data obtained from ATPDraw5.7p4 modeling was used. The proposed methods are shown to be able to rapidly identify internal fault and did not operate a miss-operation for all the external fault.<BR>

This paper deals with a time-scale method for stator and rotor incipient faults detection in wound-rotor induction machines. The initial study on the rotor incipient fault in induction machines consisted of the energy evaluation of a known low frequency bandwidth through time-scale analysis. It was shown that the stator current space vector magnitude (SCSVM) and the instantaneous magnitude of stator current (IMSC) are both good candidates to be used as an indicator for incipient rotor fault monitoring in the steady-state working condition of the induction machine. In this paper, it will be demonstrated that in the case of load fluctuation, there exists a time-varying dc component that may affect the energy criterion, resulting in a false alarm. Thus, a new technique based on a numerical phase lock loop (NPLL) is utilized to estimate and then remove this time-varying component before the SCSVM and IMSC energy calculations through time-scale analysis to handle this problem. Moreover, by definition of new frequency bandwidths associated with incipient stator fault in the rotor current signature, this technique can be extended to a more general slip independent incipient stator/rotor fault detection technique in wound-rotor induction machines. This proposal is validated through experimental tests based on a 10hp wound-rotor induction machine test-rig.

A novel diagnostic method for multiple open-circuit faults of voltage-source inverters (VSIs) is proposed based on output line voltage residuals (OLVRs) analysis. Possible OLVRs are characterized for single-phase and double-phase faulty operations. By analyzing the OLVRs, it is found that the OLVRs show different characteristics under different open-circuit (OC) fault conditions, and the positive and negative polarities of the line current residuals are deduced to be consistent with the corresponding OLVRs. Therefore, the fault can be located by line current residuals to avoid the use of extra sensors and meanwhile improve the diagnostic speed. In addition, the presented method has the features of simple structure, robustness against false alarms and being easily inserted to the existing control algorithms. Simulation and experimental results confirm the effectiveness of the proposed algorithm, and additional tests show that the diagnostic speed can be as fast as 5% and as slow as 83% of the current period.

At the initial stage of the mechanism, the collected samples are always in actual state, and the signals in fault conditions are gathered after a certain running time, so the general fault diagnosis model cannot be trained effectively. In this paper, a hybrid fault diagnosis scheme for pump in truck crane was proposed based on particle swarm optimization (PSO) SVDD and DBI K-Cluster method. Firstly, the SVDD procedure was constructed with the data in actual state, and the model parameters were optimized with PSO algorithm. Secondly, when the total number of novelty samples reached a given threshold, the K-Cluster method was utilized to classify the collected samples and the labels were allocated. In this procedure, the number of the class was determined with the Davies Bouldin index (DBI). Finally, each class data was trained with SVDD, and a whole diagnosis model was constructed with all the two-class classifiers. For the multi-fault mode samples of the pump in truck crane, experiments show that a promising classification performance is achieved.

This paper presents a fault detection algorithm based on SBT (Simulation Before Test) for testing linear analog circuits. In the proposed algorithm the transfer function of the Circuit Under Test (CUT) is derived for nominal value of the circuit components. The frequency response of the CUT is simulated under various fault conditions and the corresponding bandwidth for each fault is obtained. It can be observed that the bandwidth is different for different fault conditions. Therefore using fault dictionary all single and multiple faults in the CUT can be detected. The fuzzy inference system is used to classify the faults effectively. The feasibility of this method is validated through second order Sallen-Key band pass filter and fourth order Leapfrog low pass filter circuits. For faulty conditions the filter parameters are varied to 50 percent of their nominal value. This algorithm is robust to detect all single and multiple faults with encouraging results.

In safety-critical applications, the set of possible faults have to be divided into two groupes : the first groupe represents the subset of faults which lead to a system failure and the second one - the subset of faults which do not lead (at least immediately) to a failure. The detection/exclusion of a fault from the second group is considered as a false alarm. The problem of GNSS integrity monitoring is considered in the paper. It is proposed to design the statistical fault detection/exclusion algorithm which directly analyzes the impact of each satellite channel fault on the positioning accuracy and detects and isolates only faults which lead to a positioning failure.

Focusing on the character of nonlinear and the non-stationary characteristics of the fault signal of inverter, this paper proposes a method of open-circuit fault diagnosis for PWM inverter which based on local mean decomposition(LMD). the simulation model of inverter driving control system is built, getting three-phase output current of different fault conditions. LMD method is used to decompose three-phase output current respectively. the feature vector of fault is constructed with energy value of the principal component and proportional coefficient of the forward current in each phase. Finally, the model of fault diagnosis is constructed with neural network to locate the leg of bridge with open-circuit fault of the inverter. According to the simulation and experiment, the results show that the method by using local mean decomposition for fault signal are good, it provides a new approach of fault diagnosis for PWM inverter with open-circuit .

The bogie traction seat is the main part of urban rail vehicles and its fault status will affect the safe and smooth operation of the vehicles. For the low accuracy of the traditional detection methods, an intelligent fault diagnosis model of the traction seat based on principal component analysis with one versus one (PCA-OVO) and support vector machine (SVM) optimized by modified arithmetic optimization algorithm is proposed. Firstly, for the difficulty of high-frequency data collection under real working conditions, the simulation platform of the bogie of an urban rail vehicle is built, and the vibration signals of the traction seat are collected and processed in different domains, and then the feature extraction and fusion method based on PCA-OVO is used to obtain the sensitive feature set of the traction seat. Finally, the SVM intelligence recognition model is constructed based on the sensitive feature set data, and its parameters are optimally combined and selected by the modified arithmetic optimization algorithm after introducing the cosine factor. The experiments prove the effectiveness of the model. Experimental results show that the model is effective and provides a new model for fault diagnosis of traction seat of urban rail vehicles.