Australian Journal of Basic and Applied Sciences

March 2020


A novel frontal collision avoidance system using wireless network for railways

Chellaswamy C., Kannan G., Vanathi A., Chinnammal V.

Abstract Railways are an efficient means of transportation as they transport a large number of people and goods simultaneously. More than 1050 million tones of fright and more than 10 billion people travel by train annually. The modern rail network becomes busier with an increased number of trains, carrying more passengers, and heavy axle load than ever before. Frequent frontal collision is happening for the train running on the same track. Hence, monitoring of collision of a rail system is necessary. The efficiency of the rail monitoring system suffers as a result of frequent accidents that occur due to the carelessness of the drivers, signal problems, and track irregularities. To overcome these problems an automatic monitoring system is required. In this work, the frontal collision problem is considered and it can be avoided by introducing a novel Frontal Collision Avoidance System (FCAS) for detecting the trains on the same track using a wireless network. Wireless Monitoring Unit (WMU) called as nodes are deployed at each end of the branch track to enable detection of the arrival and departure of the trains for the specific branch. WMU sends the information to the central monitoring station (CMS) to keep monitoring the status of the train. A test route has been taken for this study and the performance of the proposed system is verified. The arrival and departure of the nodes at different speed scenarios have been carried out. Finally, the power consumption of WMU and CMS, and the distance of the data transmission between WMU and CMS are discussed. An experimental set up has been developed for the proposed FCAS. Alert information is sent immediately to the drivers of both the trains on the same track. The result obtained shows that the proposed FCAS exactly detects the trains on the same track. So the authors believe that it is an efficient method for avoiding frontal collision avoidance.

[ FULL TEXT PDF 1-12 ] DOI: 10.22587/ajbas.2020.14.3.1


A Survey on Data mining Techniques for Water Flow Forecasting

Atika Hussein, Johnson Agbinya, Iatimad Satti

Abstract Knowledge is inconsistent and metamorphic. The levels of knowledge are most challenging to mental faculty, and whenever you feel, you adopted knowledge, a new level of ignorance crops out which makes us still striving to achieve a higher level of knowledge. Forecasting of river flow is vital in hydrology and hydraulic engineering because of its application in managing and designing projects in the water field’s events. So, models for river flow forecasting are required to be built accurately and precisely to control water levels and to run water resources. If we don’t stick to reality, hydrology will surrender us disaster such as panicking situation visa-vee Renusunce dam. So, it is essential in hydrology to allow accurate evaluation in water budget, floods erosion, and even for a local river. This paper aims to review some machine learning techniques and some ensemble methods for water flow forecasting in different areas. Some of the data mining methods found in the review of the literature are discussed. After reviewing many algorithms for water flow forecasting, some models were built (ANN, SVR, and Markov chain) for forecasting water flow and compared. After that, an ensemble model using the same algorithms (ANN, SVR, Markov chain) is built. Two technique of ensemble modeling is implemented, which are bagging and voting. The real data set is applied to our models; it is composed of eight parameters at Eldeim station near the Blue Nile Sudan. It is concluded that the bagging technique gives better accuracy than voting. The ensemble results were compared with the single models, and it was found that the bagging technique gives the highest accuracy of all models, which is 97%. As the main conclusion from this paper, single models give predictions that do not consider all situations and phenomena compared with ensemble models.

[ FULL TEXT PDF 13-27 ] DOI: 10.22587/ajbas.2020.14.3.2


Determinants of Willingness To Pay for sustainable products: a study using Healthy Lifestyle and Environmental Awareness

Victor Henrique Medronha da Silva, Robson Andreazza, Everton Anger Cavalheiro, Marcela Afonso, Ázlan Lakus Pretto

Abstract Consumer buying behavior, in addition to the factors that precede it, has been highlighted as a key element for the adoption of more sustainable production processes. Willingness To Pay (WTP) for sustainable products, as well as a strong indication of consumers' perception of value, can be seen as an important strategic tool in search for competitive advantages. This way the main of this research sought to identify which constructs related to this buying behavior and how the interactions between them relates in the formation of consumer’s WTP for sustainable products. For that, the quantitative research method was used, through the survey technique, where 216 consumers were interviewed in different commercial locations in Brazil. Also, with the intuition to test the instrument, a pre-test was carried out with 12 master degree students, with research areas adhering to the area of this research. Using the chosen methodological procedure, it was started by the preliminary treatment of the data, checking the existence of outliers through the Hadi test. As a result, 3 multivariate outliers observations were found, which was disregard in this research. After analyzing the data, it was possible to identify the constructs recognized as predictors of Willingness To Pay for products with sustainable attributes . In the end, as a result, it was possible to confirm that a healthy lifestyle is correlated with consumer’s Environmental Awareness and Environmental Awareness is correlated with WTP. It was also verified that these connections have a moderating effect on variables such as income and schooling.

[ FULL TEXT PDF 28-34 ] DOI: 10.22587/ajbas.2020.14.3.3