Based on the behavior pattern of users in both office and residential buildings in Abu Dhabi, we discovered a huge amount of energy waste. In order to achieve the sustainable lifestyle, we proposed an implementation which combines the idea of IoT(Internet of Thing) and Smart building. Moreover, we will conduct a survey to discover the user behavior pattern in Masdar Villas residential area as a case study to estimate how much energy and money we can save by applying our implementation.
Buildings are a main source of global energy consumption and CO2 emission; accounting for about 40% of earth's energy consumed yearly [1]. Advancements in photovoltaic (PV) technology have birth an innovative strategy leading to more energyefficient buildings. This paper aims to investigate the possibilities and potentials of this strategy; along with its unique benefits and challenges; highlighting also possible research opportunities.
A ray-tracing model of a beam down solar concentrator available at the Masdar Institute Solar Platform is validated. The solar reflection of operative heliostats on a lambertian target at the top of the tower is compared to the solar flux density distribution predicted by the ray-tracing model. Each image is captured by a CCD camera installed at the base of the tower, and processed to derive the solar flux density concentration distribution on the target. The ray-tracing model is implemented in TracePro 6.0.2, following the geometrical specifications of the beam down facility. The results indicate that the model is able to capture the shape of the reflection, although significant deviations are noticed between the experiments and the predictions.
Heat recovery using organic Rankine cycle is a widely used technique. Rankine cycle usually uses water as working fluid, but for heat recovery purposes especially low grade heat source, organic fluids are used. The heat source in this paper is the aluminium melting furnaces' flue gases. This paper presents the performance of organic Rankine cycle using different working fluids and its impact on the amount of fuel recovered and CO2 emission reduction. It also shows how changing heat source parameter such as flue gases' mass flow rate and temperature would impact the output power and the performance of the cycle.
Owing to their high porosity, low density, and low electrical resistance, carbon paper materials embody important characteristics for use as high performance electrodes in Vanadium Redox Flow Batteries (VRFBs), yet only little has been done to investigate the methods and effects of their activation and enhancement. In this work, thermal and electrochemical treatment methods are investigated in an attempt to enhance and unlock the potential of carbon paper electrodes. The respective electrochemical performance is characterized by employing cyclic voltammetry and tafel plot electrochemical techniques to understand the effects induced on reversibility and performance. Significant improvements of performance have been attained by the different treatment methods employed, of which the most significant is the thermal treatment method.
This paper describes a new synthesis of hybrid Petri net sliding mode control (PNSMC) applied to reach the maximum power point tracking (MPPT) of a variable speed wind energy conversion system. To solve the main and major undesired phenomenon faced by conventional sliding mode control, the high frequency oscillations known as chattering, the design of a hybrid controller based on Petri network sliding mode control (PNSMC) is proposed, in which a Petri network controller replaces the discontinuous part of the classical sliding mode control law. The new hybrid controller law has been tested in Simulink/Matlab environment. Simulations results of the proposed control theme present good dynamic and steady-state performances compared to the classical SMC from aspects of the reduction chattering phenomenon.
Fuel cell systems are considered to be very complex systems that are very difficult to model. This is because their models consists of thermal, electrical, mechanical, chemical, and fluidic phenomena that are inter-acting with one another. The use of Evolutionary algorithms (EA) proved to be successful in approximating solutions to various types of problems in vast areas such as since, art, mathematics, biology, chemistry, physics and engineering. In this paper, the ability of two famous EAs to correctly identify the parameters of two PEMFC systems will be explored and compared to one another.
In this work, the lipid content and the count of microalgae cells suspended in a cultivation medium are quantified electrically without the need for any preprocessing. The proposed technique is based on finding the cell effective dielectric constant which is directly related to the cell composition and can be used as figure of merit to be correlated with the lipid content. The electrical measurements of the capacitance voltage concept is employed to determine microalgae cells counts, suspended in a cultivation medium without any sample treatment or pre-processing steps. In the proposed technique, the microalgae cells are considered as dopants embedded inside a relevant medium. The cells count is then estimated by subtracting the intrinsic impurities of the medium from the effective ensemble impurities of the suspension inside a defined volume.
Autoscaling of virtual machines based on real-time dynamic workload variation on cloud clusters are well known. These techniques can be applied on dynamic scaling of cores in mega-core servers. Frequently known performance metrics such as CPU utilization and latency are used to estimate the scaling of the number of cores irrespective of the type of workloads. It results into excess power consumption and chip temperature resulting in performance degradation. In this work, the hardware performance counters (PCs) of the processor are used to predict the characteristic of workload executing on multicore processors. It includes only those PCs that are most correlated with the workload power or thermal behavior. Such PC subsets are considered as workload signatures and it is used to trigger core's scaling in real-time. Experiments are performed using workloads from the SPEC CPU 2006 and PARSEC benchmark suites.
In this work, we are concerned by non-linear Kalman filtering scheme where the extended Kalman filter (EKF) is used to train multi-layer perceptron (MLP) to be applied to noisy speech. However, when only noise-corrupted speech is available, our enhancement experiments use a NOIZEUS corpus where the proposed method achieves higher Perceptual Evaluation of Speech Quality (PESQ) score and better subjective tests than the basic scheme of Kalman filter as well as other enhancement methods.
