This study has proposed a method for detecting turbulence, a primary factor that influences safe aircraft operation. The number of observed turbulence events is limited, thereby indicating the requirement of an appropriate flow for detecting turbulence events from a small number of samples. In addition, the opinions and experiences of pilots must be reflected at the initial stage to address the high risk of turbulence occurrence, which can result in airline operations being cancelled. Thus, this study proposed a method for predicting turbulence occurrence based on the turbulence occurrence date information provided by airlines as well as meteorological data sets obtained from open data available in Japan as teacher data. However, because commonly used machine learning methods are unable to detect the turbulence occurrence date, the proposed method employed principal component analysis coupled with the K-Means method to generate risk clusters with a high likelihood of turbulence occurrence and consequently perform statistical checks. Subsequently, the risk clusters were utilized as supervisory data for turbulence occurrence, while the support vector machine was used for predicting turbulence occurrence. Furthermore, the results obtained with the proposed method were statistically checked as well as practically verified by a pilot to confirm the appropriateness of the turbulence occurrence date predicted.
In this study, a closed BCMP queueing network, which is one of the most flexible queueing models relative to customer class and multiple service types that can be selected, was used for optimal node placement. Although the closed BCMP queueing network offers flexibility, it is computationally expensive when calculating performance evaluation quantities, such as the average number of customers in a system. Thus, the application of the closed BCMP to large-scale problems is considered challenging. In this study, an optimal node placement model, which uses a genetic algorithm to select the nodes where customers are not concentrated and are appropriately distributed, was implemented using a computational engine that calculated the performance evaluation quantity of a closed BCMP queueing network in real time. Two types of objective functions were designed: a penalty type, which applied a penalty when customers were concentrated at a node, and a variance type, which minimized the standard deviation of the average number of customers to equalize the average in the system. The features of this study, including the large-scale application of the closed BCMP queueing network and optimal node placement model, find application as a general-purpose model for node allocation planning in numerous situations and have high academic significance in the application of queueing theory to the real world.
Various measures have been devised to reduce crowdedness and alleviate the transmission of COVID-19. In this study, we propose a method for reducing intra-facility crowdedness based on the usage of Wi-Fi networks. We analyze Wi-Fi logs generated continually in vast quantities in the ever-expanding wireless network environment to calculate the transition probabilities between the nodes and the mean stay time at each node. Subsequently, we model this data as a continuous-time Markov chain to determine the variance of the stationary distribution, which is used as a metric of intra-facility crowdedness. Therefore, we solved the optimization problem by using stay rate as a parameter and developed a numerical solution to minimize the intra-facility crowdedness. The optimization results demonstrate that the intra-facility crowding is reduced by approximately 30%. This solution can practically reduce intra-facility crowdedness as it adjusts people’s stay times without making any changes to their movements. We categorized Wi-Fi users into a set of classes using the k-means method and documented the behavioral characteristics of each class to help implement class-specific measures to reduce intra-facility crowdedness, thus enabling facility managers to implement effective countermeasures against crowdedness based on the circumstances. We present a detailed description of our computing environment and workflow used for the basic analysis of vast quantities of Wi-Fi logs. We believe this research will be useful for analysts and facility operators because we have used general-purpose data for analysis.
Recently, increased measures are being devised to reduce crowdedness as a countermeasure for the spread of COVID-19. In this study, we propose a solution to reduce intra-facility crowdedness based on the usage of Wifi networks. This study maximizes the Wi-Fi logs that are continually generated in vast quantities in the ever-expanding Wi-Fi network environment to calculate the transition probabilities between nodes and the mean stay time at each node. Then, we model this data as a continuous-time Markov chain to obtain the variance of the stationary distribution, which we use as a metric of intra-facility crowdedness. Therefore, to minimize intra-facility crowdedness, we solved the optimization problem using stay rate as a parameter and demonstrated a numerical solution. In the optimization results, we succeeded in reducing intra-facility crowding by approximately 30%. This solution is a realistic approach for reducing intra-facility crowdedness as it makes adjustments to people’s stay times without any changes in their movements. We used the k-means method to categorize Wi-Fi users into a set of classes and documented the behavioral characteristics of each class, which helped implement class-specific measures to reduce intra-facility crowdedness. This enables facility managers to implement fine-grained countermeasures against crowdedness according to circumstances. Herein, we describe our computing environment and workflow for the basic analysis of vast quantities of Wi-Fi logs. Because the data we used are general-purpose, we believe that this research will be useful for both analysts and facility operators.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.