An overview of the self-organizing map algorithm, on which the papers in this issue are based, is presented in this article.

The Self-Organizing Map

A novel wavelet sequence based on deep bidirectional LSTM network model for ECG signal classification

This paper presents results obtained from monitoring a 1.72 kWp photovoltaic system installed on a flat roof of a 12 m high building in Dublin, Ireland (latitude 53.4°N and longitude 6.3°E). The system was monitored between November 2008 and October 2009 and all the electricity generated was fed into the low voltage supply to the building. Monthly average daily and annual performance parameters of the PV system evaluated include: final yield, reference yield, array yield, system losses, array capture losses, cell temperature losses, PV module efficiency, system efficiency, inverter efficiency, performance ratio and capacity factor. The maximum solar radiation, ambient temperature and PV module temperature recorded were 1241 W/m2 in March, 29.5 °C and 46.9 °C in June respectively. The annual total energy generated was 885.1 kW h/kWp while the annual average daily final yield, reference yield and array yield were 2.41 kW h/kWp/day, 2.85 kW h/kWp/day and 2.62 kW h/kWp/day respectively. The annual average daily PV module efficiency, system efficiency and inverter efficiency were 14.9%, 12.6% and 89.2% respectively while the annual average daily performance ratio and capacity factor were 81.5% and 10.1% respectively. The annual average daily system losses, capture losses and cell temperature losses were 0.23 h/day, 0.22 h/day and 0.00 h/day respectively. Comparison of this system with other systems in different locations showed that the system had the highest annual average daily PV module efficiency, system efficiency and performance ratio of 14.9%, 12.6% and 81.5% respectively. The PV system’s annual average daily final yield of 2.4 kW h/kWp/day is higher than those reported in Germany, Poland and Northern Ireland. It is comparable to results from some parts of Spain but it is lower than the reported yields in most parts of Italy and Spain. Despite low insolation levels, high average wind speeds and low ambient temperature improve Ireland’s suitability.

Measured Performance of a 1.72 kW Rooftop Grid Connected Photovoltaic System in Ireland

Signal based condition monitoring techniques for fault detection and diagnosis of induction motors: A state-of-the-art review

With the increase in the amount of data captured during the manufacturing process, monitoring systems are becoming important factors in decision making for management Current technologies such as Internet of Things (IoT)-based sensors can be considered a solution to provide efficient monitoring of the manufacturing process In this study, a real-time monitoring system that utilizes IoT-based sensors, big data processing, and a hybrid prediction model is proposed Firstly, an IoT-based sensor that collects temperature, humidity, accelerometer, and gyroscope data was developed The characteristics of IoT-generated sensor data from the manufacturing process are: real-time, large amounts, and unstructured type The proposed big data processing platform utilizes Apache Kafka as a message queue, Apache Storm as a real-time processing engine and MongoDB to store the sensor data from the manufacturing process Secondly, for the proposed hybrid prediction model, Density-Based Spatial Clustering of Applications with Noise (DBSCAN)-based outlier detection and Random Forest classification were used to remove outlier sensor data and provide fault detection during the manufacturing process, respectively The proposed model was evaluated and tested at an automotive manufacturing assembly line in Korea The results showed that IoT-based sensors and the proposed big data processing system are sufficiently efficient to monitor the manufacturing process Furthermore, the proposed hybrid prediction model has better fault prediction accuracy than other models given the sensor data as input The proposed system is expected to support management by improving decision-making and will help prevent unexpected losses caused by faults during the manufacturing process

Performance Analysis of IoT-Based Sensor, Big Data Processing, and Machine Learning Model for Real-Time Monitoring System in Automotive Manufacturing.

Solar radiation forecasting using MARS, CART, M5, and random forest model: A case study for India.

Summary Fault detection and diagnosis is the most important technology in condition-based maintenance (CBM) system for rotating machinery. This paper experimentally explores the development of a random forest (RF) classifier, a recently emerged machine learning technique, for multi-class mechanical fault diagnosis in bearing of an induction motor. Firstly, the vibration signals are collected from the bearing using accelerometer sensor. Parameters from the vibration signal are extracted in the form of statistical features and used as input feature for the classification problem. These features are classified through RF classifiers for four class problems. The prime objective of this paper is to evaluate effectiveness of random forest classifier on bearing fault diagnosis. The obtained results compared with the existing artificial intelligence techniques, neural network. The analysis of results shows the better performance and higher accuracy than the well existing techniques.

Feature selection and classification of mechanical fault of an induction motor using random forest classifier

Technical communication: Improved modified AZTEC technique for ECG data compression: Effect of length of parabolic filter on reconstructed signal

An overview on performance of PV plants commissioned at different places in the world

The Fast Fourier Transform (FFT) method is successfully used for the broken rotor bar fault detection purpose in the induction machines. It is based on the common-steady state analysis of the motor. This method is successfully used with Motor Current Signature Analysis (MCSA) technique for last three decades. However, this method is suffered from some serious drawbacks such as; it is applicable only in the constant load condition not for the variable load. The frequency-domain methods which are commonly used need accurate slip estimation for frequency components localization in any spectrum. It is also not suitable at the no-load or light load condition of the motor. At light load condition, it is quite difficult to distinguish between healthy and faulty rotors because the characteristic of broken rotor bar fault frequencies are very close to fundamental component and their amplitude are small in comparison. As a result, detection of the fault and classification of the fault severity under light load is almost impossible. In order to overcome the above problems of the FFT based technique, the Short Time Fourier Transform (STFT) Method was proposed. The excellent feature of this method is that it is capable to diagnose broken rotor bar fault in the transient condition. The STFT method also suffered from the drawback that it shows the constant window for all the frequencies. Therefore, it shows poor frequency resolution. In order to overcome all the problems stated so far, the most recent powerful mathematical tool i.e. Wavelet Transform (WT) has been used in the rotor broken bar fault detection purpose at all loading conditions. It shows variable window size for all the frequencies. Therefore, the WT method does not have resolution problem due to its multiresolution feature. This paper investigates the detection of rotor faults in induction machines by analyzing the starting current using a newly developed quantification technique based on the wavelet transform. The technique applies the wavelet transform to the envelope of the starting current. The envelope extraction is used to remove the strong fundamental component, which overshadows the characteristic differences between a healthy motor and a faulty motor with broken rotor bars.

Vinod Kumar Giri

Papers

Solar radiation forecasting using MARS, CART, M5, and random forest model: A case study for India.

Feature selection and classification of mechanical fault of an induction motor using random forest classifier

Technical communication: Improved modified AZTEC technique for ECG data compression: Effect of length of parabolic filter on reconstructed signal

An overview on performance of PV plants commissioned at different places in the world

Broken rotor bar fault detection in induction motors using Wavelet Transform