Mateusz Kalisch

Application of selected classification schemes for fault diagnosis of actuator systems

The paper presents the application of various classification schemes for actuator fault diagnosis in industrial systems. The main objective of this study is to compare either single or meta-classification strategies that can be successfully used as reasoning means in off-line as well as on-line diagnostic expert systems. The applied research was conducted on the assumption that only classic and well-practised classification methods would be adopted. The comparison study was carried out within the DAMADICS benchmark problem which provides a popular framework for confronting different approaches in the development of fault diagnosis systems.

Actuator Fault Diagnosis Using Single and Meta-Classification Strategies

The paper presents the application of various classification schemes for actuator fault diagnosis in industrial systems. The main objective of this study is to compare either single or meta-classification strategies that can be successfully used as reasoning means in the diagnostic expert system that is realized within the frame of the DISESOR project. The applied research was conducted on the assumption that classic as well as soft computing classification methods would be adopted. The comparison study was carried out within the DAMADICS benchmark problem which provides a popular framework for confronting different approaches in the development of fault diagnosis systems.

Outlier Detection and Elimination in Stream Data – An Experimental Approach

In the paper the issue of outlier detection and substitution (correction) in stream data is raised. The previous research showed that even a small number of outliers in the data influences the prediction model application quality in a significant way. In this paper we try to find a proper complex method of outliers proceeding for stream data. The procedure consists of a method of outlier detection, a statistic used for the outstanding values replacement, a historic horizon for the replacing value calculation. To find the best strategy, a wide grid of experiments were prepared. All experiments were performed on semi–artificial data: data coming from the underground coal mining environment with an artificially introduced dependent variable and randomly introduced outliers. In the paper a new approach for the local outlier correction is presented, that in several cases improved the classification quality.

Data Intensive vs Sliding Window Outlier Detection in the Stream Data — An Experimental Approach

In the paper a problem of outlier detection in the stream data is raised. The authors propose a new approach, using well known outlier detection algorithms, of outlier detection in the stream data. The method is based on the definition of a sliding window, which means a sequence of stream data observations from the past that are closest to the newly coming object. As it may be expected the outlier detection accuracy level of this model becomes worse than the accuracy of the model that uses all historical data, but from the statistical point of view the difference is not significant. In the paper several well known methods of outlier detection are used as the basis of the model.

Fault Detection Method Using Context-Based Approach

The paper describes the context based and model-free fault detection method. The main purpose of the research is to present that there is the possibility of development of diagnostic schemes using ensemble learning and context based approach to obtain the high efficiency of the fault detection system. The achieved results confirm the effectiveness of the proposed approach and also show its limitations.

Influence of Outliers Introduction on Predictive Models Quality

The paper presents results of the research related to influence of the level of outliers in the data (train and test data considered separately) on the quality of a model prediction in a classification task. The set of 100 semi–artificial time series was taken into consideration, which independent variables was close to real ones, observed in a underground coal mining environment and dependent variable was generated with the decision tree. For every considered method (decision trees, naive bayes, logistic regression and kNN) a reference model was built (no outliers in the data) which quality was compared with the quality of two models: Out–Out (outliers in train and test data) and Non-out–Out (outliers only in test data). 50 levels of outliers in the data were considered, from 1 % to 50 %. Statistical comparison of models was done on the basis of sign test.

Development of Expert System Shell with Context-Based Reasoning

The paper focuses on the expert system shell which is proposed as a tool that can be used for a wide spectrum of industrial applications. A new architecture of the system enables reasoning by means of multi-domain knowledge representations and multi-inference engines. Moreover, the extended functionality of the system is developed using a context based approach. The system is implemented applying a data mining software which makes possible to acquire domain-specific knowledge and its direct application in the expert system shell. In this study, the preliminary verification is presented using the data registered by the SCADA system of the water supply network. The case study results are useful to illustrate the merits and limitations of the proposed approach.

Development of Expert System Shell for Coal Mining Industry

The paper deals with the design of an expert system shell for the decision support system that is developed to be used in coal mining industry. A proposed architecture of the system allows reasoning by means of multi-domain knowledge representations and multi-inference engines. The implementation of the system is based on data mining software (RapidMiner) which makes possible to acquire domain-specific knowledge and its application in the expert system shell. In this study, the preliminary verification is presented using DAMADICS simulator that was proposed to compare different fault diagnosis methods. The obtained results show the merits and limitations of the proposed approach.

Application of Context-Based Meta-Learning Schemes for an Industrial Device

The paper presents the application of fault diagnosis schemes for an industrial device. Presented schemes were based on various machine learning algorithms using single classifiers and different types of ensemble-classification methods. Besides the well-known methods of ensemble-classification, the author also implemented and used context-based meta-learning method. Presented schemes were tested using artificial datasets generated by the benchmark of a longwall scraper conveyor. Model allows for testing various scenarios of operation modes of the conveyor, including the possibility of modelling operational faults. Signals generated by model were connected with velocity sensor of the conveyor and current sensors of motors. Collected data was divided into smaller parts dependent on the values of discrete contextual feature. The contextual feature cannot be used directly by a classifier, but can be useful when it is combined with other features. It does not have direct information about operational faults of the device. Speaker’s sex, nationality and age are examples of contextual features for speech recognition problem. Operational states of the conveyor and a longwall shearer was used to determine contextual feature for case study described in the article. In the next step each part of data was used to calculate single value of the feature like average value, median, skewness, etc. New dataset containing only features of measured signals was used during training and verification process of classifiers using a cross-validation method. The author compared results of reasoning processes based on the proposed schemes and basic types of the classifiers. The achieved results confirm the effectiveness of the new proposed approach based on context and also show its limitations.

Genetic Optimization of Meta-Learning Schemes for Context-Based Fault Detection

The paper is focused on the problem of performance optimization of meta-learning schemes for context-based fault detection. The context-based reasoning is developed to increase the effectiveness of well-known classification methods due to the need for designing high-efficient fault diagnosis systems. The most important problem to solve in this approach is to find optimal structures as well as optimal values of behavioural parameters of component classifiers. This problem has been formulated as a multi-objective optimization task, and in the next step, the global criterion method was adapted for expressing the meta-criterion function taking into account different statistical measures (objectives) obtained from a table of confusion. It was decided to make use of the generational genetic algorithm in order to search for the optimal solution. The subject of the case study was the scraper conveyor simulator. A few variants of the meta-learning scheme for context-based fault detection were elaborated using common machine learning methods such as decision tree, naive Bayes, Bayesian network and k-nearest neighbours. The obtained results prove that the proposed approach has both theoretical and practical relevance and thus it should find potential applications in real-world conditions.