Khaled M. Abo-Al-Ez

A data mining based load forecasting strategy for smart electrical grids

Smart electrical grids, which involve the application of intelligent information and communication technologies, are becoming the core ingredient in the ongoing modernization of the electricity delivery infrastructure. Thanks to data mining and artificial intelligence techniques that allow the accurate forecasting of power, which alleviates many of the cost and operational challenges because, power predictions become more certain. Load forecasting (LF) is a vital process for the electrical system operation and planning as it provides intelligence to energy management. In this paper, a novel LF strategy is proposed by employing data mining techniques. In addition to a novel load estimation, the proposed LF strategy employs new outlier rejection and feature selection methodologies. Outliers are rejected through a Distance Based Outlier Rejection (DBOR) methodology. On the other hand, selecting the effective features is accomplished through a Hybrid technique that combines evidence from two proposed feature selectors. The first is a Genetic Based Feature Selector (GBFS), while the second is a Rough set Base Feature Selector (RBFS). Then, the filtered data is used to give fast and accurate load prediction through a hybrid KN3B predictor, which combines KNN and NB classifiers. Experimental results have proven the effectiveness of the new outlier rejection, feature selection, and load estimation methodologies. Moreover, the proposed LF strategy has been compared against recent LF strategies. It is shown that the proposed LF strategy has a good impact in maximizing system reliability, resilience and stability as it introduces accurate load predictions.

A proposed reactive power controller for DG grid connected systems

Most of currently grid connected Decentralized Generation (DG) systems are power electronics interfaced, and only manage their active power production. With industrial inductive loads, the utility grid has not only to supply their reactive power needs, but also to supply extra reactive power to compensate for the expected power factor deterioration caused by DG active power injection. Among various DG options, Fuel cell technology makes a clean, highly controllable and economically viable DG option. Local reactive power compensation is a particular feature of fuel cell systems, as they will always be very close to the point of usage of electricity. With smart control algorithm of the grid coupling inverter, it is possible to enable reactive power management of the fuel cell DG system. This paper presents a smart control algorithm of the fuel cell DG grid coupling inverter, which provides active/reactive power management capability for the DG unit. The controller modeling and performance validation is performed at a test system, for a proposed operation scenario using Matlab/Simulink and Simpower system blocks. This validation proves the capability of using the proposed controller for power factor improvement.

A proposed dynamic model of Photovoltaic-DG system

Dynamic modeling is important to predict the energy production of Photovoltaic (PV) systems. It is needed to make informed technical and economical decisions. The simulation models of PV systems in literature are good enough for steady state analysis, but they are not suitable for dynamic analysis of grid operating and control conditions. This paper proposes a dynamic PV model suitable for Decentralized Generation (DG) applications. The proposed model relates the electrical output of the PV system to various input environmental parameters. The model is developed in Matlab-Simulink environment, and it is validated comparing the developed PV performance characteristic curves with those of the manufacturers data sheet and those developed by a commercial software package for a Solarex-MSX 60W PV type.

A proposed framework for dynamic modelling of photovoltaic systems for DG applications

Dynamic modelling and simulation is essential to predict the overall electrical performance of photovoltaic (PV) systems. PV simulation models in the literature are not suitable for dynamic analysis with decentralised generation (DG) applications. This article proposes a framework for PV system dynamic modelling and simulation process. This framework presents the steps required to model the process of solar power generation, reflecting the environmental variables affecting the generation process. Based on the framework steps, a computer simulation model is developed in MATLAB-Simulink of the PV generator, and validated by comparing the developed PV electrical performance characteristic curves with those of the manufacturers data sheet and the ones developed by commercial software. The last step of the proposed framework is dedicated for testing the developed PV model for grid-connected operation. The proposed framework resulted in a simulation photovoltaic decentralised generation model which constitutes a computer-aided design tool that is helpful for real-world solar energy engineering.

A Multi-Aware Query Driven (MAQD) routing protocol for mobile wireless sensor networks based on neuro-fuzzy inference

A Mobile Wireless Sensor Network (MWSN) is composed of mobile sensor nodes, whereby the life time of a sensor is specified by the life time of its battery. MWSN is used to measure a mobile object parameters then transmit the sensed data to center station called Sink Node (SN). Routing of the sensed data is a challenging issue since several parameters and restrictions should be managed carefully in MWSN. Although the sensors power (e.g., sensors battery level) is a critical issue, managing data transmission time is also a considerable subject especially for real time applications. Several routing protocols had been proposed for MWSN, however, each protocol considers a single type of awareness (such as; long life, delay time, total energy). While some protocols consider the sensors energy to maximize the network lifetime, others aim to find the shortest path to reach the destination. In this paper, a multi-aware query driven (MAQD) routing protocol will be proposed for MWSN based on a neuro-fuzzy inference system. MAQD considers four types of awareness; the long life of the sensor, delay time of data transmission, total cost of the network, and the shortest transmission path. Hence, based on the selected type of awareness, a fuzzy inference system selects the proper path for routing data. MAQD is a query driven protocol, accordingly, SN can collect data from some/all sensors by employing a request (REQ) message in which the awareness type is specified. MAQD is simulated and tested using OPNET 14.5 and compared with the latest MWSN routing protocols. Simulation results have shown that MAQD outperforms the selected competitors routing protocols (LEACH, ERTLD, RACE, SPIN, EAR2, DCBM, and Rumor routing) as it introduces the best data delivery with the minimum routing overheads in terms of time penalties and power consumed.

A Reliable Routing Protocol for Vehicular Ad hoc Networks

Abstract This paper introduces a Reliable Routing Protocol (R2P) for Vehicular Ad-hoc Networks (VANETs), which divides the network into overlapping zones. For each zone, a special node is promoted to be the Master Node (MN), which maintains an up-to-date routing boards for inter/intra-zone communication. R2P depends on two types of boards, namely; Internal Routing Board (IRB) and External Routing Board (ERB). Two types of IRB are used, namely; Zone Routing Board (ZRB) that is maintained by MNs, and Private Routing Board (PRB) that is maintained by each network node. Both ZRB and PRB register routes among zone nodes, while ERB, which is maintained by MN, registers available gateways to neighboring zones. R2P employs a special route discovery mechanism to discover available routes to the destination, and then elects the most reliable route. It has been compared against the recent VANETs routing protocols. Experimental results have shown that R2P outperforms the others.