Earthquakes are natural disasters caused by an unexpected release of seismic energy from extreme levels of stress within the earth’s crust. Over the years, earthquake prediction has been a controversial research subject that has challenged even the smartest of minds. Because numerous seismic precursors and other factors exist that may indicate the potential of an earthquake occurring, it is extremely difficult to predict the exact time, location, and magnitude of an impending quake. Nevertheless, evaluating a combination of these precursors through advances in Artificial Intelligence (AI) can certainly increase the possibility of predicting an earthquake. The sole purpose for predicting a seismic event at a pre-determined locality is to provide substantial time for the citizens to take precautionary measures. With this in mind, Artificial Neural Networks (ANNs) have been promising techniques for the detection and prediction of locally impending earthquakes based on valid seismic information. To highlight the recent trends in earthquake abnormality detection, including various ideas and applications, in the field of Neural Networks, valid papers related to ANNs are reviewed and presented herein.
In recent years, due to liberalization, power systems are being operated closer and closer to their limits. At the same time, they have increased in size and complexity. Both factors increase the risk of major power outages and blackouts. In emergency and abnormal conditions, a power system operator has to deal with large amounts of data. However, due to emotional and psychological stress, an operator may not be able to respond to critical conditions adequately and make correct decisions promptly. Mistakes can damage very expensive power system equipment or worse lead to major emergencies and catastrophic situations. Intelligent systems can play an important role by alarming the operator and suggesting the necessary actions to be taken to deal with a given emergency. This paper outlines some experience obtained at the University of Tasmania, Australia, Energy Systems Institute, Russia and TU-Dortmund University, Germany in developing intelligent systems for preventing large-scale emergencies and blackouts in modern power systems.
Development of novel offshore wind turbine designs and technologies are necessary to reduce the cost of offshore wind energy since offshore wind turbines need to withstand ice and waves in addition to wind, a markedly different environment from their onshore counterparts. This paper focuses on major design challenges of offshore wind turbines and offers an advanced concept wind turbine that can significantly reduce the cost of offshore wind energy as an alternative to the current popular designs. The design consists of a two-blade, downwind rotor configuration fitted to a fixed bottom or floating foundation. Preliminary results indicate that cost savings of nearly 25% are possible compared with the conventional upwind wind turbine designs.
Measurement using a moving camera and a 6-axis sensor under the camera is proposed to determine the distance from the camera to the surface of a moving object and the object’s position movement in two continuous frame images. This makes it possible to measure the 3D position of a moving object at half of the computational cost while keeping the same accuracy as using a stereo camera. 3D measurement experiments with several original images show that the computational time using the proposal is about twice as fast as that of a stereo camera. The proposed method is planning to be used to vehicles or mobile robots avoid obstacles, and its use as a depth meter is also investigated.
This paper presents the fusion of swarm behavior in multi robotic system specifically the quadrotors unmanned aerial vehicle (QUAV) operations. This study directed on using robot swarms because of its key feature of decentralized processing amongst its member. This characteristic leads to advantages of robot operations because an individual robot failure will not affect the group performance. The algorithm emulating the animal or insect swarm behaviors is presented in this paper and implemented into an artificial robotic agent (QUAV) in computer simulations. The simulation results concluded that for increasing number of QUAV the aggregation accuracy increases with an accuracy of 90.62%. The experiment for foraging revealed that the number of QUAV does not affect the accuracy of the swarm instead the iterations needed are greatly improved with an average of 160.53 iterations from 50 to 500 QUAV. For swarm tracking, the average accuracy is 89.23%. The accuracy of the swarm formation is 84.65%. These results clearly defined that the swarm system is accurate enough to perform the tasks and robust in any QUAV number.
In this study, a three-way factorial design is used to optimize the friction stir welding (FSW) process of Al 6061 alloy. Control factors are shoulder diameter, travel speed, and rotation speed of tool, and each factor has three levels. Tensile strength tests are also carried out to measure the mechanical properties under various FSW conditions. In this work, travel speed, shoulder diameter, and tool speed are shown to individually be meaningful factors in the tensile strength of the alloy, but interactions among the weld factors are not detected. The result of the study is that the optimum process condition for maximum tensile strength is estimated to be A3B3C3. In addition, the presumed range of tensile strength under the optimal conditions is estimated to be 257±23 (Mpa) with 95% reliability.
We have designed, developed and deployed a robust system of sensors in a small and shallow – 1 km diameter and about 7 m deep – inland lake used for aquaculture. The sensor system currently measures and sends out telemetry data on dissolved oxygen, conductivity and temperature in the lake at two depths – 0.5 and 2.5 meters – every 30 minutes. The measurements are sent out via SMS to a central server 80 km away and displayed in near-real time on a website which can be accessed by local fishermen, researchers and other stakeholders. The system also features an aerator, which operates automatically during the evening and early morning 12 hours at a time, and can also be turned on outside this period via missed-call (dropped call) from a cellular phone. This system can aid in decision-making of local fishermen, making it a potentially powerful approach to lake management, and in conjunction with other measures and long term planning, for averting or mitigating fishkill events.
We deposited graphite on silicon (111) and copper foil substrate through femtosecond pulsed laser deposition (fs-PLD). A high purity graphite target was placed inside a vacuum chamber at a base pressure of 10-2 mbar. The deposition time was varied for 3 hours, 4 hours and 5 hours. XRD spectra showed a (110) peak indicating an oriented growth for samples deposited on silicon (111) and copper foil substrates. AFM topographical images of the samples deposited on silicon (111) showed flake-like structures. However, samples deposited on copper foil showed the presence of defects and lack of deposited particles.
This paper presents a swarm robot simulator for implementing underwater wireless communication network. Swarm intelligence is based on the collective behavior of social insects and animals such as ants, bees and others. In this paper, swarm was applied to overcome the challenges of transmitting data in a large underwater environment. A robot considered to be a member of the swarm acts as a simple “physical” carrier of the data, it moves until they converge and manage to form a link connecting the data transmitter and receiver. The system is developed, simulated and tested using a coded simulator.
This paper presents machine vision for locating and identifying 23 highly dynamic objects on 4.4 meters by 2.8 meters micro robot soccer playing field. The approach is based from the idea that the two camera vision subsystems should be synchronized and well informed in real time of the combined vision data and a selection of objects to track under each other’s camera view. A measure of effectiveness on using incremental tracking for two-camera operation is developed and is used to evaluate the introduced approach through experimentation. A real-time visualization of the whole playfield containing the 22 micro robots and a golf ball is also provided for the system operator to validate the objects’ actual poses with the vision system’s measurements. Results show that the proposed technique is very fast, accurate, reliable, and robust to external disturbances.
We propose a novel bag-of-words (BoW) framework for building and retrieving a compact database of view images for use in robotic localization, mapping, and SLAM applications. Unlike most previous methods, our method does not describe an image based on its many small local features (e.g., bag-of-SIFT-features). Instead, the proposed bag-of-bounding-boxes (BoBB) approach attempts to describe an image based on fewer larger object patterns, which leads to a semantic and compact image descriptor. To make the view retrieval system more practical and autonomous, the object pattern discovery process is unsupervised through a common pattern discovery (CPD) between the input and known reference images without requiring the use of a pre-trained object detector. Moreover, our CPD subtask does not rely on good image segmentation techniques and is able to handle scale variations by exploiting the recently developed CPD technique, i.e., a spatial random partition. Following a traditional bounding-box based object annotation and knowledge transfer, we compactly describe an image in a BoBB form. Using a slightly modified inverted file system, we efficiently index and/or search for the BoBB descriptors. Experiments using the publicly available “RobotCar” dataset show that the proposed method achieves accurate object-level view image retrieval using significantly compact image descriptors, e.g., 20 words per image.
Colon cancer is one type of cancer that has a high death rate, but early diagnosis can improve the chances of patient recovery. Computer-assisted diagnosis can aid in determining whether images are of healthy or cancerous tissues. This study aims to contribute to the automatic classification of microscopic colonic images by implementing a 2-D wavelet transform for feature extraction and neural networks for classification. The colonic histopathological images are assigned to either the normal, cancerous, or adenomatous polyp classes. The proposed algorithm is able to determine which of the three classes the images belong to at a 91.11% rate of accuracy.
The current results on logistic Web services selection are not optimal due to some key quality indexes of logistic Web services excluded, in order to resolve the above problem, an evaluation system on quality of service is established by use of principal component analysis based on quality of logistic service, quality of Web service, and satisfaction of customers. The values of quality of service with subjective uncertainty in the evaluation system are given with trapezoidal fuzzy number according to the definition of logistic business and evaluation from domain experts and customers, besides, the weight on each quality of service is given by pairwise comparison, and an algorithm based on analytic hierarchy process for logistic Web service selection is established. The optimal service is got by adopting the algorithm in the logistic scenario on automotive transportation, which proves that the way on service selection in this paper is feasible and effective.
A small scale real time tiger prawn aquaculture setup was built and tested in the laboratory using ordinary aquariums to test the controllability and control of the four most important parameters in culturing tiger prawns, the temperature, salinity, pH and dissolved oxygen. These parameters were monitored using Vernier sensors via Labview program. The water quality index of the artificial habitat was monitored and computed using fuzzy logic. New values for the safe parameter conditions of the tiger prawns were observed and used in the computation of the water quality index. Lastly, electronic valves and actuators are used to automatically control the four said water parameters and set them to their optimal values. The control needed by each parameter to force them to stay within their optimal values was done using neural network. This control system is used to activate the electronic valves that will dispense correction fluids for each of the four monitored water parameter.
This paper presents a scheme of weather forecasting using artificial neural network (ANN) and Bayesian network. The study focuses on the data representing central Cebu weather conditions. The parameters used in this study are as follows: mean dew point, minimum temperature, maximum temperature, mean temperature, mean relative humidity, rainfall, average wind speed, prevailing wind direction, and mean cloudiness. The weather data were collected from the PAG-ASA Mactan-Cebu Station located at latitude: 10°19’, longitude: 123°59’ starting from January 2011 to December 2011 and the values available represent daily averages. These data were used for training the multi-layered backpropagation ANN in predicting the weather conditions of the succeeding days. Some outputs from the ANN, such as the humidity, temperature, and amount of rainfall, are fed to the Bayesian network for statistical analysis to forecast the probability of rain. Experiments show that the system achieved 93%–100% accuracy in forecasting weather conditions.
The objective of this paper is to raise the accuracy of multiclass text classification through Graph-Based Semi-Supervised Learning (GBSSL). In GBSSL, it is essential to construct a proper graph which expresses the relation among nodes. We propose a method to construct a similarity graph by employing both surface information and latent information to express similarity between nodes. Experimenting on a Reuters-21578 corpus, we have confirmed that our proposal works well in raising the accuracy of GBSSL in a multiclass text classification task.
In the optimization of seaweed cultivation now being extensively researched, a problem arises in avoiding twisting seaweed. Twisting is a complex phenomenon and difficult to formulate. Producing the optimal water flow, requires calculating the risk of twisting occurring. In this paper, we propose a method to calculate and estimate the twist state based on the results of physical simulation. We devise a seaweed model using multiple rigid bodies that mutually and physically interfere. One result of physical interference, is that the model has two internal state variables – contact time and the number of contact points between individual pieces of seaweed. We introduce an evaluation function to quantify twisting using these state variables in each time step, and propose a way to calculate twist risk based on the von Neumann and Laplacian diffusion kernels, in a dynamic network.
The goal of this study was to propose a three-dimensional evaluation of the EndoButton displacement direction after anterior cruciate ligament reconstruction in the multidetector-row computed tomography (MDCT) image by using the tunnel axis. The proposed method was applied experimentally to six subjects. The result of the simulated experiment revealed that the proposed method could analyze EndoButton displacement direction satisfactorily because the error was less than that of the MDCT image resolution. The clinical experiment results revealed displacement relative to the tunnel between time-zero and the follow-up point. We conclude that the proposed method can quantitatively evaluate the EndoButton displacement direction from the raw MDCT image after anterior cruciate ligament reconstruction; further, our findings suggest that the EndoButton was displaced relative to the tunnel between time-zero and the follow-up point.
As vital transportation carriers in trade, ships have the advantage of stability, economy, and bulk capacity over airplanes, trucks, and trains. Even so, their loss and cost due to collisions and other accidents exceed those of any other mode of transportation. To prevent ship collisions many ways have been suggested, e.g., the 1972 COLREGs which is the regulation for preventing collision between ships. Technologically speaking, many related studies have been conducted. The term “Ship domain” involves that area surrounding a ship that the navigator wants to keep other ships clear of. Ship domain alone is not sufficient, however, for enabling one or more ships to simultaneously determine the collision risk for all of the ships concerned. Fuzzy theory is useful in helping ships avoid collision in that fuzzy theory may define whether collision risk is based on distance to closest point of approach, time to closest point of approach, or relative bearing – algorithms that are difficult to apply to more than one ships at one time. The main purpose of this study is thus to reduce collision risk among multiple ships using a distributed local search algorithm (DLSA). By exchanging information on, for example, next-intended courses within a certain area among ships, ships having the maximum reduction in collision risk change courses simultaneously until all ships approach a destination without collision. In this paper, we introduce distributed local search and explain how it works using examples. We conducted experiments to test distributed local search performance for certain instances of ship collision avoidance. Experiments results showed that in most cases, our proposal applies well in ship collision avoidance among multiple ships.
Automatic road obstacle detection is one of the significant problem in Intelligent Transport Systems (ITS). Many studies have been conducted for this interesting problem by using on-vehicle cameras. However, those methods still needs a dozens of milliseconds for image processing. To develop the quick obstacle avoidance devices for vehicles, further computational time reduction is expected. Furthermore, regarding the applications, compact hardware is also expected for implementation. Thus, we study on computational time reduction of the road obstacle detection by using a small-type parallel image processor. Here, computational time is reduced by developing an obstacle detection algorithm which is appropriated to parallel processing concept of that hardware. According to the experimental evaluation of the new proposal, we could limit computational time for eleven milliseconds with a good obstacle detection performance.
Systems capable of autonomous thinking are sometimes required to cope with unanticipated situations. An important issue in this context is knowledge – especially common sense – acquisition. In this paper, we propose novel quantitative common sense estimation methods and apply them to an automatic membership function generation system. Our proposed system estimates threshold values corresponding to large and small for various kinds of object-attribute sets to form membership functions, where it attempts to relate each object to its corresponding impression. Two methods are proposed in this paper. The first, Method-1, obtains data from the top 1,000 snippets through a web search and estimates the global and local tendencies by clustering them. The second, Method-2, uses the number of hits from a web search together with parts of the results obtained through Method-1. In addition, we devise several techniques to eliminate unnecessary information in the retrieved web pages. We also carried out experiments that verified the effectiveness of our proposed methods and the method combining those two.