Improving Machine-Learning Diagnostics with Model-Based Data Augmentation Showcased for a Transformer Fault

Machine-learning diagnostic systems are widely used to detect abnormal conditions in electrical equipment. Training robust and accurate diagnostic systems is challenging because only small databases of abnormal-condition data are available. However, the performance of the diagnostic systems depends on the quantity and quality of the data. The training database can be augmented utilizing data augmentation techniques that generate synthetic data to improve diagnostic performance. However, existing data augmentation techniques are generic methods that do not include additional information in the synthetic data. In this paper, we develop a model-based data augmentation technique integrating computer-implementable electromechanical models. Synthetic normal- and abnormal-condition data are generated with an electromechanical model and a stochastic parameter value sampling method. The model-based data augmentation is showcased to detect an abnormal condition of a distribution transformer. First, the synthetic data are compared with the measurements to verify the synthetic data. Then, ML-based diagnostic systems are created using model-based data augmentation and are compared with state-of-the-art diagnostic systems. It is shown that using the model-based data augmentation results in an improved accuracy compared to state-of-the-art diagnostic systems. This holds especially true when only a small abnormal-condition database is available.

Simulation for Predicting Condition of Plant Equipment

Condition-based maintenance (CBM) is a maintenance program that recommends maintenance decisions based on maintenance information by condition monitoring. To apply CBM to industrial systems, it is important to diagnose condition of equipment appropriately based on the knowledge of abnormal condition of the equipment. In nuclear power plant, however, it is quite difficult to obtain the knowledge of abnormal condition. For this reason, simulation is expected to be an effective approach to estimate conditions of equipment instead of data acquisition. In this study, we focused on and built models for a motor operated valve and a pump, since the inspection cost is generally expensive due to the number of them in nuclear power plant. We considered a pump system and analyzed the measurement values in normal and abnormal conditions. Our analytical result reproduces the behavior of the measurement values and shows the characteristic deviation from normal condition in abnormal conditions.

Smart Farming System by Creating Artificial Environment using IoT for Efficient Crop Growth

This paper presents the growth of plants effectively in less duration of time compared to traditional farming. In this the farming is done inside the packed electronic environment by using the artificial environment through Led’s for growth of plants. The photosynthesis process is carried out by the plants is dependent on the led’s. In this project we are using two sensors they are DHT11 which is used to monitor the temperature and humidity parameters and MQ135 which is an air quality sensor for monitoring the environment of the particular region and controlled using Node MCU. All the equipments are monitored using Iot. If any of these parameters is in abnormal condition then exhaust fan get turned on, so that we can reduce the humidity. Blynk application is used for displaying information. By using these parameters the rate of plant growth is doubled. Results shown that when all the factors of plant growth are stabilized, then it is possible to grow a plant in less time compared through normal plant time because photosynthesis is carries throughout the day.

Heart Rate Monitor using Average of Two Fingertip Sensors with Graphical ECG Record

This project proposes the measurement of heart rate accurately using an average of two readings of two different fingertip sensors and feeding them to a microcontroller and record simultaneously the photoplethysmographic data, using which an ECG waveform is plotted. The earlier projects on this topic used only one fingertip sensor in calculating the heart rate. This lacked accuracy and thus, even if the projects were economical, they could not be used at small hospitals in rural areas and village clinics. Also, they lacked the feature of graphical ECG plot that may be extremely useful for diagnosis by the medical experts. Our project is based on obtaining the heart rate by averaging the readings of two fingertip sensors. The average is then fed to the microcontroller after proper signal conditioning, and then displayed on a LCD screen. The graph of electrical activity of the heart is displayed on a graphical LCD. Also, it features sending of SMS to family members via GSM module in case of an abnormal condition. The novelty in this project is increased accuracy.

Modelling and Blood Flow Analysis of Internal Pudendal Artery

A common male sexual disorder is erectile dysfunction which has multidimensions. In this fast-moving world, it is prominently seen in lots of males. There are many causes for Erectile Dysfunction, one of the major causes is the improper supply of blood to the penile organ. That may be due to vasoconstriction or blockage in the internal pudendal artery which supplies oxygen to the penile organ. A simulated model of the internal iliac artery to the internal prudential artery is designed and a flow simulation is done using Solid works software. The Computed Tomography of a male subject is obtained and a three-dimensional model of the abdominal artery is extracted using MIMICS (Materialize Interactive Medical Image Control System) software. By making use of the measured dimensions from the three-dimensional image. The 3D models (Normal condition, Abnormal condition with blockage, and Abnormal condition with constrictions) are designed and the Flow analysis is done in Solid works software. By the end of the study, we came to a conclusion that at normal temperature and pressure, the simulated normal volumetric blood flow at the internal pudendal artery is 6.88701e-09 m3/s and for abnormal cases the simulated volumetric blood flow is 2.6107e-09m3/s.

A novel expulsion control strategy with abnormal condition adaptability for resistance spot welding

Welding expulsion is a common problem in Resistance Spot Welding (RSW) process, which severely impacts weld quality and surrounding facilities. Existing expulsion control strategies are ineffective for complex and changeable welding conditions. This article studied the growth relationship between weld nugget and corona bond under two abnormal conditions: edge proximity (EP) and initial sheet gaps (IG). It is testified that expulsion would occur when the nugget size exceeds the corona bond size under EP and IG conditions. Reducing the welding current before the expulsion time can increase the size difference between the corona bond and the weld nugget, thereby delay and even eliminate the occurrence of expulsion. In this way, a novel online expulsion control strategy, named short-time current regulation (STCR), is proposed through expulsion moment analysis of historical weld data. The effect of the new control strategy is verified with workpieces ranging from low carbon steel to ultra-high strength steel. Experimental results showed that STCR can effectively reduce the amount of expelled metal, decrease the indentation depth and increase the nugget diameter. The method not only works well under one specific abnormal condition, but also adapts to the transition between different welding conditions. This novel expulsion control strategy can help achieve expulsion-free RSW process in mass production without frequent manual offline optimization of welding parameters.

Hybrid Governance in Northeastern Mexico: Crime, Violence, and Legal-Illegal Energy Markets

Despite the presence of organized crime in northeastern Mexico, the region has a functioning economy that attracts new investment to energy projects, especially those involving fossil fuels. This may be because legal and illegal markets there tend to overlap and function under hybrid governance schemes. The hybridization of governance is an expression of the fact that legality and illegality are embedded in contemporary capitalist markets. This embeddedness is not an abnormal condition but the way in which societies deal locally with organized crime, and violence serves as the primary regulatory mechanism in disputed territories and markets. A pesar de la presencia del crimen organizado en el noreste de México, la región tiene una economía funcional que atrae nuevas inversiones a proyectos energéticos, especialmente aquellos relacionados con combustibles fósiles. Esto puede deberse a que, ahí, los mercados legales e ilegales tienden a solaparse y funcionar en el marco de los sistemas de gobernanza híbrida. La hibridación de la gobernanza es una expresión del hecho de que la legalidad y la ilegalidad están incrustadas en los mercados capitalistas contemporáneos. Dicha incrustación no es una condición anormal, sino la forma en que las sociedades se ocupan localmente del crimen organizado, y la violencia sirve como el principal mecanismo regulador en los territorios y mercados en disputa.

Simulation for Predicting Condition of Plant Equipment

Condition-based maintenance (CBM) is a maintenance program that recommends maintenance decisions based on maintenance information by condition monitoring. To apply CBM to industrial systems, it is important to diagnose condition of equipment appropriately based on the knowledge of abnormal condition of the equipment. In nuclear power plant, however, it is quite difficult to obtain the knowledge of abnormal condition. For this reason, simulation is expected to be an effective approach to estimate conditions of equipment instead of data acquisition. In this study, we focused on and built models for a motor operated valve and a pump, since the inspection cost is generally expensive due to the number of them in nuclear power plant. We considered a pump system and analyzed the measurement values in normal and abnormal conditions. Our analytical result reproduces the behavior of the measurement values and shows the characteristic deviation from normal condition in abnormal conditions.