An integrated framework is presented to represent and classify process data for on-line identifying abnormal operating conditions. It is based on pattern recognition principles and consists of a feature extraction ste...An integrated framework is presented to represent and classify process data for on-line identifying abnormal operating conditions. It is based on pattern recognition principles and consists of a feature extraction step, by which wavelet transform and principal component analysis are used to capture the inherent characteristics from process measurements, followed by a similarity assessment step using hidden Markov model (HMM) for pattern comparison. In most previous cases, a fixed-length moving window was employed to track dynamic data, and often failed to capture enough information for each fault and sometimes even deteriorated the diagnostic performance. A variable moving window, the length of which is modified with time, is introduced in this paper and case studies on the Tennessee Eastman process illustrate the potential of the proposed method.展开更多
The prediction of magnitude （M） of reservoir induced earthquake is an important task in earthquake engineering. In this article, we employ a Support Vector Machine （SVM） and Gaussian Process Regression （GPR） for...The prediction of magnitude （M） of reservoir induced earthquake is an important task in earthquake engineering. In this article, we employ a Support Vector Machine （SVM） and Gaussian Process Regression （GPR） for prediction of reservoir induced earthquake M based on reservoir parameters. Comprehensive parameter （E） and maximum reservoir depth] （H） are considered as inputs to the SVM and GPR. We give an equation for determination oil reservoir induced earthquake M. The developed SVM and GPR have been compared with] the Artificial Neural Network （ANN） method. The results show that the developed SVM and] GPR are efficient tools for prediction of reservoir induced earthquake M. /展开更多
Expert system plays an important role in port machine diagnosis, which aims at automatic equipment test for higher availability and efficiency of port operations. In this study, a port machine diagnosis expert system ...Expert system plays an important role in port machine diagnosis, which aims at automatic equipment test for higher availability and efficiency of port operations. In this study, a port machine diagnosis expert system is proposed based on multi-reasoning mechanism. Relying on the knowledge acquired from the experienced experts in the port machine engineering, the system builds a library of relative experience and a set of rules of reasoning and estimating. Multi-reasoning mechanism that simulates the decision-making process of domain experts is employed to achieve reliable diagnosis results. The reasoning machine integrates artificial neural network, uncertain decision making and decision tree, which complements each other by sustainable growing voting mechanism. The effect of this multi-reasoning mechanism is evaluated and validated by means of Matthew＇s Correlation Coefficient （MCC）. The system incorporating the mechanism is successfully designed, implemented and applied in Shanghai Port.展开更多
Nowadays, the check of the organoleptic characteristics for the evaluation of extra virgin olive oil （EVOO） quality is regulated by the European Union （EU） authorities, which indicate the use of the panel test （P...Nowadays, the check of the organoleptic characteristics for the evaluation of extra virgin olive oil （EVOO） quality is regulated by the European Union （EU） authorities, which indicate the use of the panel test （PT）. It is composed by a team of specialists that give a numerical value to many characteristics about flavours, synthesising a sensory analysis. Each expert answers questions about the aroma by assigning the adequate scores to each oil. The evaluation becomes objective by applying the statistical analysis of all the scores given by the participants： This is the definition of ＂measure＂ of Russell. The PT can be considered a true standard ＂metrological system＂ （considering the number of questions in the questionnaire）, while the perceptions of the testers are the solicitations of it. To allow access to an expensive evaluation process by small companies, this work proposes to ＂disseminate＂ the properties of the metrology represented by PT. The results of the PT are arranged in an unsupervised artificial neural network （ANN）, the Kohonen map, which represents the synthesis of self-organised output that has only the goal, in this paper, to make readable PT results. The dissemination process is obtained by the gas chromatographic （GC） analysis of each oil sample and through the identification of peaks corresponding to the perceptions. These signals are used for the training of the supervised Multi Layer Perceptron （MLP） ANN, with the back propagation algorithm, whose outputs are represented by the results of the PT. This procedure is exact a ＂metrological dissemination of a standard＂ and also the aim of the work： to classify EVOO without always resorting to PT.展开更多
基金Supported by National High-Tech Program of China (No. 2001AA413110).
文摘An integrated framework is presented to represent and classify process data for on-line identifying abnormal operating conditions. It is based on pattern recognition principles and consists of a feature extraction step, by which wavelet transform and principal component analysis are used to capture the inherent characteristics from process measurements, followed by a similarity assessment step using hidden Markov model (HMM) for pattern comparison. In most previous cases, a fixed-length moving window was employed to track dynamic data, and often failed to capture enough information for each fault and sometimes even deteriorated the diagnostic performance. A variable moving window, the length of which is modified with time, is introduced in this paper and case studies on the Tennessee Eastman process illustrate the potential of the proposed method.
文摘The prediction of magnitude （M） of reservoir induced earthquake is an important task in earthquake engineering. In this article, we employ a Support Vector Machine （SVM） and Gaussian Process Regression （GPR） for prediction of reservoir induced earthquake M based on reservoir parameters. Comprehensive parameter （E） and maximum reservoir depth] （H） are considered as inputs to the SVM and GPR. We give an equation for determination oil reservoir induced earthquake M. The developed SVM and GPR have been compared with] the Artificial Neural Network （ANN） method. The results show that the developed SVM and] GPR are efficient tools for prediction of reservoir induced earthquake M. /
文摘Expert system plays an important role in port machine diagnosis, which aims at automatic equipment test for higher availability and efficiency of port operations. In this study, a port machine diagnosis expert system is proposed based on multi-reasoning mechanism. Relying on the knowledge acquired from the experienced experts in the port machine engineering, the system builds a library of relative experience and a set of rules of reasoning and estimating. Multi-reasoning mechanism that simulates the decision-making process of domain experts is employed to achieve reliable diagnosis results. The reasoning machine integrates artificial neural network, uncertain decision making and decision tree, which complements each other by sustainable growing voting mechanism. The effect of this multi-reasoning mechanism is evaluated and validated by means of Matthew＇s Correlation Coefficient （MCC）. The system incorporating the mechanism is successfully designed, implemented and applied in Shanghai Port.
文摘Nowadays, the check of the organoleptic characteristics for the evaluation of extra virgin olive oil （EVOO） quality is regulated by the European Union （EU） authorities, which indicate the use of the panel test （PT）. It is composed by a team of specialists that give a numerical value to many characteristics about flavours, synthesising a sensory analysis. Each expert answers questions about the aroma by assigning the adequate scores to each oil. The evaluation becomes objective by applying the statistical analysis of all the scores given by the participants： This is the definition of ＂measure＂ of Russell. The PT can be considered a true standard ＂metrological system＂ （considering the number of questions in the questionnaire）, while the perceptions of the testers are the solicitations of it. To allow access to an expensive evaluation process by small companies, this work proposes to ＂disseminate＂ the properties of the metrology represented by PT. The results of the PT are arranged in an unsupervised artificial neural network （ANN）, the Kohonen map, which represents the synthesis of self-organised output that has only the goal, in this paper, to make readable PT results. The dissemination process is obtained by the gas chromatographic （GC） analysis of each oil sample and through the identification of peaks corresponding to the perceptions. These signals are used for the training of the supervised Multi Layer Perceptron （MLP） ANN, with the back propagation algorithm, whose outputs are represented by the results of the PT. This procedure is exact a ＂metrological dissemination of a standard＂ and also the aim of the work： to classify EVOO without always resorting to PT.
