• Energy Research

The growing interest in renewable energy and the falling prices of solar panels place solar electricity in a favourable position for adoption. However, the high-rate adoption of intermittent renewable energy introduces challenges and the potential to create power instability between the available power generation and the load demand. Hence, accurate solar Photovoltaic (PV) power forecasting is essential to maintain system reliability and maximize renewable energy integration. The current solar PV power forecasting approaches are an essential tool to maintain system reliability and maximize renewable energy integration. This paper presents a comprehensive and comparative review of existing Machine Learning (ML) based approaches used in PV power forecasting, focusing on short-term horizons. We provide an overview of factors affecting solar PV power forecasting and an overview of existing PV power forecasting methods in the literature, with a specific focus on ML-based models. To further enhance the comparison and provide more insights into the advancement in the area, we simulate the performance of different ML methods used in solar PV power forecasting and, finally, a discussion on the results of the work.

<span lang="EN-US">In a photovoltaic system, a stable voltage and of tolerable power equilibrium is needed. Hence, a dedicated analog charge controller for a storage system which controls energy flow to impose power equilibrium, and therefore, voltage stability on the load is required. We demonstrate here our successful design considerations employing supercapacitors as main energy storage as well as a buffer in a standalone photovoltaic system, incorporating a dedicated supercapacitor charge controller for the first time. Firstly, we demonstrated a photovoltaic system employing supercapacitors as main energy storage as well as a buffer in a standalone photovoltaic system. Secondly, we design a constant voltage maximum power point tracker (MPPT) for peak power extraction from the photovoltaic generator. Thirdly, we incorporated a supercapacitor charge controller for power equilibrium and voltage stability through a dedicated analog charge controller in our design, the first of its kind. Fourthly, we analyzed the use of supercapacitor storage to mitigate disequilibrium between power supply and demands, which, in turn, causes overvoltage or under voltage across the load. Lastly, we then went ahead to demonstrate the control of the energy flow in the system so as to maintain rated voltage across a variant demand load.</span>

Environmental pollution and global warming due to the emission of gases from internal combustion engines has become a global issue that has made automobile companies develop emission-free or pure electric vehicles. The objective of this work is to convert the Bajaj three-wheeler (Indian-made auto-rickshaw) into a pure electric three-wheeler with an onboard battery charging system with a solar panel. Conversion methods were used and the engine and fuel-related components of the existing Bajaj were removed and replaced with the electric motor, battery, and associated components. Furthermore, the performance of the prototype was also analyzed for its maximum vehicle speed, grading capacity, acceleration performance, and driving range on flat and sloped roads. The proposed electric vehicle prototype is equipped with four solar panels installed on top of the system, while four batteries of 12 V / 26 Ah each were placed under the driver’s seat. The DC motor was arranged to deliver the required torque and power through the chain-sprocket mechanism of the differential. The vehicle has a mechanical reverse system to drive the car rearward without changing the motor’s rotational direction. For evaluation, the vehicle’s dynamic performance was determined, and the results show that the maximum vehicle speed, grading capacity, and driving range are 30 km/hr, 1 degree, and 34 km, respectively. The acceleration and gradability performance was computed as 0.106 m/s2 and 2.2% or 1.26 degrees respectively. The time required to accelerate from 0 to 30 km/hr is 78 s. In addition, it was observed that the onboard solar charging system increased the driving range by 17 km. The results were auspicious, and the proposed system performed satisfactorily for the purpose it was developed.

The main problem for event gathering in wireless sensor networks (WSNs) is the restricted communication range for each node. Due to the restricted communication range and high network density, event forwarding in WSNs is very challenging and requires multihop data forwarding. Currently, the energy-efficient ant based routing (EEABR) algorithm, based on the ant colony optimization (ACO) metaheuristic, is one of the state-of-the-art energy-aware routing protocols. In this paper, we propose three improvements to the EEABR algorithm to further improve its energy efficiency. The improvements to the original EEABR are based on the following: (1) a new scheme to intelligently initialize the routing tables giving priority to neighboring nodes that simultaneously could be the destination, (2) intelligent update of routing tables in case of a node or link failure, and (3) reducing the flooding ability of ants for congestion control. The energy efficiency improvements are significant particularly for dynamic routing environments. Experimental results using the RMASE simulation environment show that the proposed method increases the energy efficiency by up to 9% and 64% in converge-cast and target-tracking scenarios, respectively, over the original EEABR without incurring a significant increase in complexity. The method is also compared and found to also outperform other swarm-based routing protocols such as sensor-driven and cost-aware ant routing (SC) and Beesensor.

High efficient routing is an important issue for the design of wireless sensor network (WSN) protocols to meet the severe hardware and resource constraints. This paper presents a comprehensive survey and comparison of routing protocols in WSNs. The first part of the paper surveys state-of-the-art routing protocols in WSNs from classical routing protocols to swarm intelligence based protocols. The routing protocols are categorized based on their computational complexity, network structure, energy efficiency and path establishment. The second part of the paper presents a comparison of a representative number of classical and swarm based protocols. Comparing routing protocols in WSNs is currently a very challenging task for protocol designers. Often, much time is required to re-create and re-simulate algorithms from descriptions in published papers to perform the comparison. Compounding the difficulty is that some simulation parameters and performance metrics may not be mentioned. We see a need in the research community to have standard simulation and performance metrics for comparing different protocols. To this end, the final part of the paper re-simulates different protocols using a Matlab based simulator: routing modeling application simulation environment (RMASE), and gives simulation results for standard simulation and performance metrics which we hope will serve as a benchmark for future comparisons for the research community.

Big sensor-based data environment and the emergence of large-scale wireless sensor networks (LS-WSNs), which are spread over wide geographic areas and contain thousands of sensor nodes, require new techniques for energy-efficient data collection. Recent approaches for data collection in WSNs have focused on techniques using mobile data collectors (MDCs) or sinks. Compared to traditional methods using static sinks, the MDC techniques give two advantages for data collection in LS-WSNs. These techniques can handle data collection over spatially separated geographical regions, and have been shown to require lower node energy consumption. Two common models for data collection using MDCs have been proposed: data collection using data mule (MULE), and sensor network with mobile access point (SENMA). The MULE and SENMA approaches can be characterized as representative of the multihop and the single-hop approaches for mobile data collection in WSNs. Although the basic architectures for MULE and SENMA have been well studied, the emergence of LS-WSNs which require partitioning the network into multiple groups and clusters prior to data collection has not been particularly addressed. This paper presents analytical approaches to determine the node energy consumption for LS-WSN MDC schemes and gives models for determining the optimal number of clusters for minimizing the energy consumption. The paper also addresses the tradeoffs when the LS-WSN MULE and SENMA models perform well.

High efficient and energy-aware routing is an important issue for the design of resource constrained environments like Wireless Sensor Networks (WSNs). Many protocols have been developed for WSN that try to overcome the constraints that characterized this type of networks. Termite based routing protocols can add a significant contribution to assist in the maximization of the network lifetime without performance degradation. But this is only possible by means of an adaptable and balanced algorithm that takes into account the main constraints of WSN. This paper presents a biological inspired self-organized routing protocol for WSN which is based on termite colony optimization metaheuristic termed Termite-hill. The main objective of the proposed algorithm is to efficiently relay all the traffic destined for the sink, and also balance the network energy. The results of our extensive experiments on Routing Modeling Application Simulation Environment (RMASE) demonstrated that with sink mobility, our proposed routing algorithm was able to balance the network traffic load and prolong the network lifetime without performance degradation.

Sorting of products is a very difficult industrial process. Continuous manual sorting is tedious, time-consuming, and creates consistency issues. It is, therefore, admirable for companies to diversify into automation. Automation in the industrial workplace provides the advantages of improving productivity and quality while reducing errors and waste, increasing safety, and adding flexibility and efficiency to the manufacturing process. This paper presents the design and simulation of an automatic colour sensor and sorting machine (ACSSM) that automates grading and separating bell pepper according to colour. The system was designed for the food and packaging industry in Botswana, who experienced problems of spending enormous time and effort in sorting red, green, and yellow bell pepper according to colour after harvesting. Bell pepper is of high demand in all its ripening stages, and colour plays a crucial role in their pricing. Thus sorting is vital to meet the varying needs of customers and to obtain the most profits. The focal point was to design and test a circuit that would decrease the time, workspace, and effort while providing the essential function which was sorting of bell pepper. The implemented system is based on colour detection, object detection, and ejection to achieve sorting. The system performed well according to the design requirements when simulated on Proteus software.