Search

Abstract The advancement of Photovoltaic (PV) systems hinges on the precise optimization of their parameters. Among the numerous optimization techniques, the effectiveness of each often rests on their inherent parameters. This research introduces a new methodology, the Reinforcement Learning-based Golden Jackal Optimizer (RL-GJO). This approach uniquely combines reinforcement learning with the Golden Jackal Optimizer to enhance its efficiency and adaptability in handling various optimization problems. Furthermore, the research incorporates an advanced non-linear hunting strategy to optimize the algorithm’s performance. The proposed algorithm is first validated using 29 CEC2017 benchmark test functions and five engineering-constrained design problems. Secondly, rigorous testing on PV parameter estimation benchmark datasets, including the single-diode model, double-diode model, three-diode model, and a representative PV module, was carried out to highlight the superiority of RL-GJO. The results were compelling: the root mean square error values achieved by RL-GJO were markedly lower than those of the original algorithm and other prevalent optimization methods. The synergy between reinforcement learning and GJO in this approach facilitates faster convergence and improved solution quality. This integration not only improves the performance metrics but also ensures a more efficient optimization process, especially in complex PV scenarios. With an average Freidman’s rank test values of 1.564 for numerical and engineering design problems and 1.742 for parameter estimation problems, the proposed RL-GJO is performing better than the original GJO and other peers. The proposed RL-GJO stands out as a reliable tool for PV parameter estimation. By seamlessly combining reinforcement learning with the golden jackal optimizer, it sets a new benchmark in PV optimization, indicating a promising avenue for future research and applications.

Abstract The accurate classification of power quality disturbances (PQDs) is crucial for advancing real‐time monitoring and classification systems within the modern power grid. The proposed system must ensure dependable, safeguarded, and stable operating conditions amidst diverse power quality issues. This paper presents an approach to classifying power quality disturbances using a deep learning model that synergizes deep convolutional neural networks (DCNN) and Bidirectional Long Short‐Term Memory (BiLSTM). This amalgamation effectively extracts and classifies disturbance signals in real time, grounded on noise levels. The initial feature extraction from the signal is accomplished through a time‐frequency matrix. Subsequently, secondary extraction employs the BiLSTM layer to intricately and significantly classify disturbances in the power signal. This aids in transforming high‐dimensional matrices into a reduced set for enhanced performance. The detailed classification is facilitated by the softmax layer. The simulation results support the power quality evaluations under varied constraints and underscore the substantial classification of power quality disturbances through the DCNN‐BiLSTM algorithm, in comparison to alternative classification algorithms in terms of computational speed and accuracy.

Abstract The Optimal Power Flow (OPF) is a primary tool in planning and installing power systems. It attempts to minimize the operating costs associated with generating and transmitting electrical power by modifying control parameters to satisfy environmental, economic, and operational constraints. Implementing an efficient and robust optimization algorithm for the above‐said problem is critical to achieving such a typical objective. Therefore, this paper introduces and evaluates new variants of the Successive History‐based Adaptive Differential Evolutionary (SHADE) algorithm called ESHADE, SHADE‐SFS, and SHADE‐SAP to solve the OPF problems with equality and inequality constraints. Generally, the static penalty approach is widely used for eliminating infeasible solutions discovered during the search phase when searching for feasible solutions. This approach requires the accurate selection of penalty coefficients, accomplished through the trial‐and‐error method. The proposed ESHADE algorithm is formulated using Self‐Adaptive Penalty (SAP) and Superiority of Feasible Solution (SFS) mechanisms to obtain feasible solutions for OPF problems. Two IEEE bus systems are used to demonstrate the effectiveness of the proposed algorithm in handling OPF problems. The fuel cost and active power loss obtained by the proposed algorithm are better than other state‐of‐the‐art algorithms. The results reveal that the proposed framework offers significant advantages over other algorithms.

Abstract The reliability of the photovoltaic models is strongly reliant on their parameters, which are primarily determined by the optimization algorithm and the objective function. As a result, obtaining the parameters under different environmental conditions is critical for increasing their performance, reliability and significantly lowering cost. Many optimization techniques are reported to address this problem based on the complexity. As a result, an enhanced version of the recently reported Hunger Games Search Optimizer (HGSO) method called Gaussian and Cauchy Mutation‐based HGSO (GCMHGSO) algorithm for defining the requirements of the Three‐Diode equivalent Model (TDeM) by utilizing multiple representations in the algorithm along with an efficient objective function. The Cauchy mutation increases the exploration ability, and Gaussian mutation increases the exploitation ability of the basic HGSO. Furthermore, an Enhanced Newton–Raphson Method (ENRM) is presented to effectively solve the behaviour of the current–voltage relation of the TDeM. The robust optimization is also considered to demonstrate the impact of the measurement error. Comparing the GCMHGSO‐ENRM to other competitors reveals that the proposed GCMHGSO‐ENRM can accurately find the best solution, and its effectiveness is verified in many statistical parameters. It is found that the GCMHGSO‐ENRM algorithm is stable and robust compared to other competitors.

The recent trend in hybrid electric vehicles (HEV) has increased the need for vehicle charging stations (VCS) in the distribution system. In this condition, the additional load in the system leads to an increase in power loss, reduction in voltage and reliability of the system. The drawbacks of introducing this additional load can be rectified by integrating distributed generation (DG) into the distribution system. In this paper, the ideal location for placing DG is identified through the voltage stability index. The power loss minimization objective function is formulated with all the required constraints to estimate the size of DG required for the distribution system. Moreover, loss of load probability is used as a reliability assessment technique, through which the system reliability is analyzed after assessing the impact of integrating VCS and DG. Simulations are carried out to compare the following cases: a system without VCS and DG, a system that has only VCS and a system that has both VCS and DG. The IEEE 12-bus and 33-bus test systems are considered. In the 12-bus system with both VCS and DG, the power loss is reduced by 56% when compared with the system with only VCS, while the net reliability is also improved. The reliability of the system is evaluated for a 24 h load variation. The proposed work provides an efficient tool to improve the reliability of the system with support from DG.

Abstract Chitosan nanoparticles (CSNPs) have been recently used for various applications in aquaculture, especially as drug carriers. The aim of this study was to synthesise and investigate a superlative method of CSNP synthesis for application in aquaculture through aquaculture‐based toxicology screening methods. Two different methods were analysed: the first a direct ionic gelation method (A) and the other involving a low‐molecular‐weight chitosan microparticle intermediate method (B). Dynamic light scattering characterisation revealed that the CSNP particle sizes were 192.7 ± 11.8 and 22.9 nm from methods A and B, respectively. The LC50 values for brine shrimp toxicity were found to be 1.51 and 0.02 ppt in 24 h for methods A and B, respectively. Acute toxicity studies in Litopenaeus vannamei rendered LC50 values of 3235.94 and 2884.03 ppt in 24 h for methods A and B, respectively. Zebrafish toxicity studies revealed mortality rates of 21.67% and 55% at 20 mg/L concentration for methods A and B, respectively, with an increased expression of intracellular reactive oxygen species in method B. From these findings, it can be concluded that a comparatively reduced toxicity of CSNPs derived from ionic gelation method makes it more appropriate for application in aquaculture.

This paper discusses a modified V-I relationship for the solar photovoltaic (PV) single diode based equivalent model. The model is derived from an equivalent circuit of the PV cell. A PV cell is used to convert the solar incident light to electrical energy. The PV module is derived from the group of series connected PV cells and PV array, or PV string is formed by connecting the group of series and parallel connected PV panels. The model proposed in this paper is applicable for both series and parallel connected PV string/array systems. Initially, the V-I characteristics are derived for a single PV cell, and finally, it is extended to the PV panel and, to string/array. The solar PV cell model is derived based on five parameters model which requires the data’s from the manufacturer’s data sheet. The derived PV model is precisely forecasting the P-V characteristics, V-I characteristics, open circuit voltage, short circuit current and maximum power point (MPP) for the various temperature and solar irradiation conditions. The model in this paper forecasts the required data for both polycrystalline silicon and monocrystalline silicon panels. This PV model is suitable for the PV system of any capacity. The proposed model is simulated using Matlab/Simulink for various PV array configurations, and finally, the derived model is examined in partial shading condition under the various environmental conditions to find the optimal configuration. The PV model proposed in this paper can achieve 99.5% accuracy in producing maximum output power as similar to manufacturers datasheet.©2020. CBIORE-IJRED. All rights reserved

Nanoparticles are potential candidates for wastewater treatment especially for the removal of heavy metals due to their strong affinity. Many biopolymers are used as adsorbents and encapsulation of nanoparticle onto them can increase their efficiency. In this study, SPIONs, alginate, and SPIONs incorporated on alginate beads have been synthesized and characterized both microscopically and spectroscopically. These were then used for the removal of chromium metal and the percentage of removal was evaluated using a batch adsorption study. The percent removal of chromium using SPIONs, alginate and alginate–SPIONs beads were recorded to be 93%, 91% and 94%, respectively. The adsorption of chromium using SPIONs and alginate–SPIONs beads followed the Tempkin isotherm, whereas adsorption of chromium metal by alginate beads was found to be homogeneous in nature and followed the Langmuir isotherm with an R2 value of 0.9784. An in-vivo study using Danio rerio as a model organism was done to examine the toxicity and the removal efficiency of the samples. It was observed that chromium water treated with alginate–SPIONs beads, which were removed after water treatment showed less damage to the fishes when compared to SPIONs and alginate beads treated with chromium water where the SPIONs and alginate beads were not removed after the treatment period.

A modified single-stage Quasi-Z-Source (QZS) boost DC-DC converter using a single MOSFET switch with a single-stage switched-inductor (SI) network is proposed in this paper. The DC-DC converter proposed in this study provides an extra voltage gain compared to the traditional QZS DC-DC converter. An additional capacitor and diode circuit are added to the existing QZS converter to decrease the voltage stress on the MOSFET switch. Therefore, compared to the traditional QZS boost converter, the proposed Modified QZS (MQZS) converter provides large voltage gain under a low duty ratio, less voltage stress, and continuous input current. Moreover, the reliability and the conversion efficiency can be increased. The derivation of the proposed MQZS converter and its operation, selection of parameters, and comparison with another similar converter are discussed in this paper. Lastly, the simulation and experimental results are illustrated to prove the notional deliberations of the proposed MQZS converter.

Summary: One of the nondeterministic polynomial (NP) hard problem in the area of wireless networks is the quality of service (QoS) routing problem. This has always been a key area wherein research is required to enhance the QoS metrics considered with respect to the considered network model. In this paper, mobile ad hoc network (MANET) model is identified for which an effective QoS routing protocol which meets the set QoS constraints is designed employing the novel variant of differential evolution (DE) model. The ultimate aim of the proposed work is to identify an optimal feasible path for the MANET model with the set objective function criterion being met. A variant of DE capsule net model is built by hybridizing it with the nature‐inspired firefly (FF) algorithmic approach. The effectiveness of the proposed technique is achieved by minimizing the set objective function on satisfying the QoS constraints in the MANETs. The developed technique chose better QoS path with respect to the best fitness value than that of the route replay the shortest path algorithm of ad hoc on‐demand distance vector (AODV) protocol. The modeled variant DE‐FF model achieves better QoS metrics with all the set constrained being met. Simulated results depict the superiority and better performance of the developed DE‐FF model than that of the other methods considered for comparison from the existing literature works. [ABSTRACT FROM AUTHOR]

This article presents the design of an advanced smart grid model with an improved optimal variation strategy by implementing renewable energy sources. The novelty involved in this article is designing a renewable energy-based smart grid system incorporating with Artificial Neural Network for maintaining an improved voltage profile with balanced reactive power level across the grid network. The objective is attained by implementing renewable energy sources, namely solar and wind energy modules, with artificial intelligence techniques for effective reactive power control using DSTATCOM. The power requirement across the power grid network is satisfied by providing solar and wind energy sources managing the smart-grid system's var level. The optimized power is trained on the Feed Forward Neural Network-based for calculating the amount of voltage profile and power loss based on system load. The DSTATCOM is an optimal var compensating device for maintaining the reactive power within the permissible limit across the smart grid system. The proposed system improves the voltage profile, reduces the power loss, and satisfies the grid by managing the system's reactive power balance. The proposed approach is executed in MATLAB/Simulink work site, and the performance is analyzed with three different cases. The smart grid system powered with hybrid energy sources strives to attain the objective functions effectively among the three different cases. Integration of hybrid renewable energy source with Artificial Neural Network achieves an improved voltage profile to about 98.45%, which states that the profile rating has improved from 1.056 to 1.083 p.u minimizations of real power loss has done from to 3.23 MW from 3.80 MW. [ABSTRACT FROM AUTHOR]