The prediction of wheel wear is still a great challenge for railway systems. This work examines the effect of radial basis function neural network (RBFNN) parameters such as spread, goal, maximum number of neurons, and number of neurons to add between displays on wheel wear prediction. VAMPIRE vehicle dynamic software was used to produce the vehicle performance data to train, validate, and test the neural network. The wheel wear was calculated using an energy dissipation approach and contact position on straight track. The neural network simulation results were implemented using the Matlab program. The percentage error for wheel wear prediction was calculated. Also, the accuracy of wheel wear prediction using the neural network was investigated and assessed in terms of mean absolute percentage error (MAPE). The results reveal that the railway wheel wear prediction using neural network is dependent on the correct selection of the neural network parameters.
Steeply falling installation costs during the last decade makes the Photovoltaic technology highly competitive particularly, of course, in a sunny region like the Libyan Desert. Photovoltaic power plants are easy to maintain and, once installed, consume practically no additional natural resources. New and innovative jobs will be created and local companies will take part in certain ways and benefit. The aim of this paper is to demonstrate how efficient and cost-saving solar energy can be exploited in the deserts of Libya. A feasibility analysis to the large-scale grid connected PV project with total capacity of 14 MW is dedicated. The power plant delivers approximately 27 GWh of electrical power in one year. It is shown that the cost for one kWh is in the range of 0.08 USD, which is approximately four times less than the cost for electricity produced with Diesel-generators in the same region. This leads to savings of Diesel (light oil) of approximately 6,400 tons per year. The value of the saved diesel is approximately 5.7 Million US Dollars (at a world market price of 900 USD per ton). Within its minimum lifetime of 25 years, the PV power plant saves 150,000 tons of Diesel or 133 Millions of US Dollars. Compared to the saved oil the break-even is within 7 years and the Return of Investment (ROI) is in the range of 14 %.
Self-organization for Wireless Sensor Network (WSN) is critical issue because of each sensor node’s limited energy, limited bandwidth and WSN’s scalability. Therefore, how to manage wireless sensor networks effectively is a big challenge task. This paper presents a novel self- organized architecture which is capable to avoid these problems. In this architecture, we suppose that each node does not know its location (Random clustering). Random clustering is practical to implement on some applications which deploy nodes into inaccessible unknown environment. We propose two algorithms to divide sensor nodes into cells. The first algorithm Active-Tree uses tree topology assign different role and node ID to each sensor node. The second algorithm Drawn-Grid divides sensor nodes into cells according to the radio coverage and the roles get from the Active-Tree algorithm. Based on sensor nodes with different role play different tasks in WSN. The result of numerical simulations will show that our algorithm performs better.
Silicon Carbide (SiC) is an important indirect wide band gap semiconductor with outstanding electronic properties. This work focuses on an investigations of silicon carbide (SiC) based vertical Double Implanted Metal Oxide Semiconductor Field Effect Transistor (DIMOSFET). Silicon Carbide (4H SiC as well as 6H SiC) is known to be highly anisotropic material. Among others, the transport parameters like low field mobility and saturation velocity are considerably different in c direction compared to a and b directions. The aim of this paper is to investigate the influence of variation of mentioned parameters, as well as the variation of parameters describing specific model, on "drift region voltage drop" in vertical DIMOS structure.
This paper, emphasis different growth techniques of two-dimensional hole gas of strained germanium (sGe) heterostructure, molecular beam epitaxy (MBE) and chemical vapor deposition (CVD). sGe heterostructure has become an important material as a replacement material to Silicon in P-type devices because of its higher hole mobility and lower effective mass. Researchers study this material in terms of electrical and spintronic devices according to technology demands for devices with higher efficiency and low power consumption. High hole mobility up to 1 × 10cmଶ/Vs at temperature of 1.5 K has been reported for normal structure declaring high quality samples with low density dislocation and low interface roughness. These samples were grown using Reduced Pressure Chemical Vapour Deposition (RP-CVD) indicating high purity system and the affordability of this technique in terms of electronic and spintronic devices.
Wheel and rail reprofiling costs millions of dollars around the world. The wheel/rail roughness is one of the important parameter which be able to effect on wheel/rail wear. The use of an artificial neural network to predict the wheel/rail roughness parameters can help to improve the design of the wheel/rail profiles. Wheel/rail roughness can define as the shorter frequency of real wheel/rail surfaces relative to the troughs. There are several roughness parameters, but the arithmetical mean roughness (Ra) is the common parameter, it is indicating the average of the absolute value along the sampling length. In this paper, both rail and wheel roughness were measured experimentally using Alicona profilometer and replica material, then, the arithmetical mean height of the wheel and rail was predicted using artificial neural networks. The results showed that the neural network predicted the wheel and rail roughness parameter efficiently.
MANET is a collection of wireless nodes that can dynamically form a network to exchange information without using any pre-existing fixed network infrastructure. In an Ad-Hoc network nodes cooperate to maintain network connectivity and perform various functions including routing. This paper focuses on two flat routing protocols the reactive Ad- Hoc on Demand Distance Vector Routing protocol (AODV), and the proactive Destination-Sequenced Distance- Vector Routing (DSDV).Where a comparison between these two protocols was done using the well-known Network Simulator 2 (NS2).