In this article, a cost-effective technique for the synthesis of gamma iron oxide nanoparticles has been proposed for intelligent maghemite electrode applications pitched in the context of smart and efficient energy storage solution. A facile process-optimized technique for synthesis of gamma iron oxide nanoparticles has been designed in order to investigate the optimum temperature, doping and pH of the sodium hydroxide. By dint of morphological investigation, it has been established that the samples have high surface area, crystalline structure, and size in the range of fifty to hundred angstrom. The linearity of the magnetization feature coupled with its doping sensitivity points towards its usage for state estimation technology of the energy storage device management. The nano-scaled samples witness an increase of 75%-110% in the direct bandgap in comparison to its bulk existence. This band gap modulation establishes that the conductivity can be improved for electrode application by doping. High surface area for the active material ingredient nano-particles has also been confirmed by BET surface area of up to 75 m(2)/g. Thermal analyses of the samples establish the fidelity of the samples' constitution over a desirably wide temperature range. The cost-effectiveness of gamma-iron oxide batteries will be a crucial factor for faster adoption of indigenous renewable energy storage solutions.

This paper introduces a robust and adaptive control framework that integrates a Proportional-Integral-Derivative (PID) controller with the bio-inspired Grey Wolf Optimization (GWO) algorithm for real-time tuning of controller parameters in grid-connected photovoltaic (PV) inverter systems. Conventional controllers such as P and PI are widely used in PV applications due to their simplicity, but they exhibit notable limitations in dynamic environments, including increased Total Harmonic Distortion (THD), slower transient response, and poor voltage regulation-particularly under variable irradiance conditions. The proposed GWO-PID method overcomes these limitations by leveraging the GWO algorithm's global search capability to dynamically optimize the PID gains (Kp, Ki, Kd) based on a composite fitness function that minimizes Mean Squared Error (MSE) and THD. The system architecture, simulated in MATLAB/Simulink, comprises a 50 kW PV array with a boost converter employing an Incremental Conductance (INC) Maximum Power Point Tracking (MPPT) algorithm, a three-phase voltage source inverter, RLC filters, and a dual-loop (voltage and current) control system synchronized with the utility grid through a Phase-Locked Loop (PLL). The GWO algorithm iteratively refines PID parameters to achieve real-time adaptation to environmental fluctuations. Under standard irradiance (1000 W/m²), the GWO-PID controller achieved a rise time of 0.025 s, settling time of 0.035 s, THD of 3.7%, and MSE of 0.25 kW², while maintaining a stable DC-link voltage of 500 V, thereby ensuring compliance with IEEE 519-2014 power quality standards. Across irradiance levels ranging from 400 W/m² to 1000 W/m², the GWO-PID controller consistently maintained DC-link voltage stability and minimized oscillations in PV voltage and current. Compared to traditional PI and P controllers, the proposed method reduced settling time by over 45%, improved power tracking accuracy, and significantly lowered harmonic distortion. Furthermore, it ensured a power factor close to unity and exhibited excellent frequency stability under transient disturbances. Simulation results also confirm the superior performance of the GWO-PID controller in managing active and reactive power exchange, minimizing overshoot, and maintaining synchronization with the grid, even during rapid environmental transitions. By embedding intelligent metaheuristic optimization into a classical PID framework, this work advances the state of inverter control strategies for PV systems. The proposed GWO-PID technique provides a scalable, efficient, and real-time solution that enhances grid compliance, energy quality, and system stability, marking a key advancement in adaptive control for smart grid and microgrid applications.

The increasing frequency of forest fires, particularly in tropical deciduous forests, is causing severe damage to ecosystems. Currently, most available fire data is limited to active fire-point locations, and there is a lack of geospatial information of forest burn area and fire frequency in India. Timely and accurate mapping of fire scars is essential for planning appropriate forest management activities including enabling the assessment of fire frequency, risk zones, and identification of suitable areas for watch towers and fire closure areas etc. Current post-fire field surveys in India are inefficient, labour-intensive and lack objectivity. To address these challenges, this study proposes an automated method for forest burn scar extraction using spectral indices and machine learning algorithms with medium resolution Landsat-7 and 8 data. The study focuses on the tropical deciduous forests of Vidarbha region, Maharashtra state. Proposed approach involves two steps. First, generating the best Spectral Indices combination for burn scar delineation, which creates precise training samples for the segmentation model. Second, applying Deep Learning models to automatically map burn scars using the optimal outputs from the first step. This study presents an Automatic Weighted Average Ensembled learning U-Net (AWAE U-Net) model, where the learning of three individual backbone U-Net models i.e., VGG16, ResNet34 and Inception V3 were ensembled by applying the best weight calculated automatically. Experimental results show the model achieves 91.12

Owing to the swift depletion of conventional energy sources, the future of energy production relies on renewable sources. Amid these green energy options, solar and wind power emerge as pivotal players in partially mitigating the energy challenge. Nonetheless, their volatility makes it difficult to rely entirely on each source separately. Combining these two resources into hybrid energy systems looks to be a more reliable strategy to increase dependability and cost-efficiency. In this paper, author introduce an efficient energy management system designed for a micro-grid that encompasses wind and solar energy conversion systems, battery storage and control unit. The primary purpose of this study is to evaluate the operation of a renewable energy-powered hybrid energy source-based micro-grid system in order to maintain a stable power balance despite inherent variability in renewable energy supply and load demand. A POASMC (PO adaptive sliding mode control) has been proposed and developed to improve the performance of hybrid energy generation systems. Furthermore, a standard energy management system based on grid current, solar/wind power, and battery SOC has been established to enable smooth power distribution among diverse sources, loads, and the battery.

Adapting deep learning models to new domains often requires computationally intensive retraining and risks catastrophic forgetting. While fine-tuning enables domain-specific adaptation, it can reduce robustness to distribution shifts, impacting out-of-distribution (OOD) performance. Pre-trained zero-shot models like CLIP offer strong generalization but may suffer degraded robustness after fine-tuning. Building on Task Adaptive Parameter Sharing (TAPS), we propose a simple yet effective extension as a parameter-efficient fine-tuning (PEFT) method, using an indicator function to selectively activate Low-Rank Adaptation (LoRA) blocks. Our approach minimizes knowledge loss, retains its generalization strengths under domain shifts, and significantly reduces computational costs compared to traditional fine-tuning. We demonstrate that effective fine-tuning can be achieved with as few as 5% of active blocks, substantially improving efficiency. Evaluations on pre-trained models such as CLIP and DINO-ViT demonstrate our method's broad applicability and effectiveness in maintaining performance and knowledge retention.

The global energy landscape is increasingly dominated by solar power installations, driven by the sun’s position as Earth’s most abundant and sustainable energy resource. However, the intermittent nature of solar radiation, influenced by both astronomical cycles and meteorological conditions, creates significant challenges for reliable power generation and grid integration. To address the issue of uncertainty, this study proposes a robust and improved capacity machine learning framework with enhanced hypothesis functional space. The proposed model improves the capacity of an individual model by combining the hypothesis functions of individual machine learning models, increasing the representational capacity and hence the model’s generalization. Moreover, a non-linear second stage is stacked to increase the depth of the proposed model, which utilizes meta-data of first stage to further improve the forecasting accuracy. Furthermore, the proposed model is validated on four different climatic zones of the world for solar power forecasting. The proposed model achieves an average improvement of 66.7% in mean absolute error across all locations compared to the next best performing algorithm, with particularly strong performance in arid zones. Statistical validation through Cook’s distance analysis also confirms the model’s reliability with an average of 8.64% influential points across all locations.

For visual recognition, knowledge distillation typically involves transferring knowledge from a large, well-trained teacher model to a smaller student model. In this paper, we introduce an effective method to distill knowledge from an off-the-shelf vision-language model (VLM), demonstrating that it provides novel supervision in addition to those from a conventional vision-only teacher model. Our key technical contribution is the development of a framework that generates novel text supervision and distills free-form text into a vision encoder. We showcase the effectiveness of our approach, termed VLM-KD, across various benchmark datasets, showing that it surpasses several state-of-the-art long-tail visual classifiers. To our knowledge, this work is the first to utilize knowledge distillation with text supervision generated by an off-the-shelf VLM and apply it to vanilla randomly initialized vision encoders.

The largest flux and a crucial player in the terrestrial carbon cycle is the gross primary productivity (GPP). This study aimed to evaluate the effectiveness of different satellite based models for estimating the GPP of cotton agroecosystems in the Nagpur district of Maharashtra, India. To assess the spatial distribution and the estimation of GPP of rainfed cotton crop of the study area, three year’s satellite data of Landsat-8/9 and Sentinel-2A/B along with eddy covariance flux data were used. Two light use efficiency (LUE) based algorithms, namely vegetation photosynthesis model (VPM) and moderate resolution imaging spectroradiometer (MODIS), as well as a vegetation index (VI) based greenness and radiation (GR) model, were employed for GPP estimation. The overall accuracy for year wise cotton crop mapping were 91.6, 95.7 and 96.8

We introduce InVi, an approach for inserting or replacing objects within videos (referred to as inpainting) using off-the-shelf, text-to-image latent diffusion models. InVi targets controlled manipulation of objects and blending them seamlessly into a background video unlike existing video editing methods that focus on comprehensive re-styling or entire scene alterations. To achieve this goal, we tackle two key challenges. Firstly, for high quality control and blending, we employ a two-step process involving inpainting and matching. This process begins with inserting the object into a single frame using a ControlNet-based inpainting diffusion model, and then generating subsequent frames conditioned on features from an inpainted frame as an anchor to minimize the domain gap between the background and the object. Secondly, to ensure temporal coherence, we replace the diffusion model's self-attention layers with extended-attention layers. The anchor frame features serve as the keys and values for these layers, enhancing consistency across frames. Our approach removes the need for video-specific fine-tuning, presenting an efficient and adaptable solution. Experimental results demonstrate that InVi achieves realistic object insertion with consistent blending and coherence across frames, outperforming existing methods.

Energy harvesting is possible through capable energy transfer materials, and one such impressive material is graphene, which has exhibited promising properties like unprecedentedly high theoretical surface area, enhanced electrical conductivity, thermal conductivity, mechanical stability, flexibility, recyclability, and so on. Herein, for the sake of everyone desirous of contributing to the field of graphene materials for high-speed energy storage devices, the fundamentals, analytics, synthesis, prospects, and challenges of energy storage cell design for fast charging of electric vehicles have been reviewed. To overcome the limitations of mechanistic models for energy-storage devices, empiricism conjointly with data-driven machine learning has been quite effective in multiple domains. The limitations in modeling of energy storage devices, in terms of swiftness and accuracy in their state prediction can be surmounted by the aid of machine learning. Conclusively, in the context of energy management, we underscore the significant challenges related to modeling accuracy, performing original computations, and relevant big data generation. This review, by dint of its futuristic insights, will help researchers to develop digital twin approach for sustainable energy management using energy storage technology toward dependable, economic, and scalable optimization processes at the multiple stages of energy conversion and energy management.

In this work, a new practical approach for the widespread use of grid-connected renewable energy generation can be successfully implemented as a microgrid. Implementing the most efficient, secure, dependable, and synchronized use of renewable energy sources requires a microgrid System for Energy Management (EMS). This study describes the management of an (ESS) connected to a solar array in a microgrid that regulates battery discharge and charge operations using a converter in accordance with demand. The challenge of daily EM is underlined. The main focus of this work is on the investigation of heuristics and algorithmic for lowering the total variable power costs on clear and overcast days.

Recently, there has been a notable penetration of solar photovoltaic (PV) power into the grid, accompanied by a rising power demand for electric vehicles (EVs) load the distribution system. This convergence of PV and EV in the grid presents formidable challenges, particularly concerning grid synchronization during PV integration and power quality when EVs draw power from the grid. Managing the combined impact of PV penetration and EV utilization necessitates a controller capable of simultaneously addressing synchronization and power quality control. This study investigates the utilization of proportional-integral (PI) control methodologies to create an integrated controller that can effectively manage both photovoltaic (PV) and grid synchronization and address power quality concerns associated with electric vehicle (EVs) load. For PV power generation grid synchronization, a dual-control loop approach is employed, with one loop dedicated to EV battery charging control and the other focused on grid synchronization. To suppress harmonics in the grid current, the paper utilizes multicarrier Space Vector Pulse Width Modulation (SVPWM) in conjunction with PI control on the grid connected converter. The study includes the development of a MATLAB Simulink model to simulate the PV-EV-Grid setup, offering a thorough evaluation of the effectiveness of the controller. This paper contributes to the development of solutions for the challenges posed by PV integration and EV usage in the grid, enhancing the reliability and performance of sustainable energy systems.

Aluminium and copper foils are commonly used as current collectors due to low contact resistance, low price and high conductivity, in lithium ion (Li-ion) battery. Localized electrolyte corrosion during long-term cycling, weak bonding of the electrode material with the current collector and limited contact area are major issues still need to address. Advance level research is going on the above discussed paraments. This review study concentrated on employing graphene covering on copper and aluminum foil in lithium-ion batteries to address corrosion issues that the current collector was experiencing. Therefore, it is critical to commend the graphene coating on current collectors, such as copper and aluminum, for preventing corrosion and improving cycle performance while also extending their lifespan. This review study demonstrates that the graphene coating strengthens the bond between the electrode material and current collectors, which is critical for increasing power density. In addition, the modified graphene sheets behave as a protective layer against metal corrosion because of its hardness and mechanical strength can be treated as thin layer and also it offers high transparency and stability.

Advancements in graphics technology has increased the use of simulated data for training machine learning models. However, the simulated data often differs from real-world data, creating a distribution gap that can decrease the efficacy of models trained on simulation data in real-world applications. To mitigate this gap, sim-to-real domain transfer modifies simulated images to better match real-world data, enabling the effective use of simulation data in model training. Sim-to-real transfer utilizes image translation methods, which are divided into two main categories: paired and unpaired image-to-image translation. Paired image translation requires a perfect pixel match, making it difficult to apply in practice due to the lack of pixel-wise correspondence between simulation and real-world data. Unpaired image translation, while more suitable for sim-to-real transfer, is still challenging to learn for complex natural scenes. To address these challenges, we propose a third category: approximately-paired sim-to-real translation, where the source and target images do not need to be exactly paired. Our approximately-paired method, AptSim2Real, exploits the fact that simulators can generate scenes loosely resembling real-world scenes in terms of lighting, environment, and composition. Our novel training strategy results in significant qualitative and quantitative improvements, with up to a 24% improvement in FID score compared to the state-of-the-art unpaired image-translation methods.

In the fabric of energy generation, solar power is the most promising clean energy solution as an alternative to non-renewable energy sources. However, solar power's dependency on environmental factors adds uncertainty to energy production. In such a scenario, solar power forecasting provides an edge to mitigate this uncertainty and improves overall system stability. Recently, machine learning (ML) models have been extensively deployed for designing and forecasting solar power. However, data pre-processing, forecast horizon, and performance evaluation of ML algorithms have to be carefully evaluated to find an accurate model. This paper provides an empirical comparison of different generation ML models for solar power forecasting, which can help understand future research on which method to adopt, depending on the ML model's strengths and weaknesses. Therefore, an effective forecasting method is designated in aspects such as performance errors, convergence time, and computational complexity. So, this work rates different ML models on error performance metrics and convergence time. Moreover, cross-fold validations and hyperparameter are also examined for the top five performing models for a comprehensive evaluation and to give more intuitive and calibrated insight into various stakeholders working in the solar power plant modeling field.

Pearl millet ( Pennisetum glaucum ), commonly known as Bajra crop, is one of the most extensively cultivated cereals in the world, after rice, wheat, and sorghum, particularly in arid to semi-arid regions. India is the largest producer of this crop, both in terms of area and production. Temporal monitoring of crop area under cultivation is essential for the sustainable management of agricultural activities on both national and global level. The present study is envisaged to discriminate of bajra crop in the rainfed agroecosystem using multi-temporal Sentinel-1 Synthetic Aperture Radar (SAR) data with dual polarization (VH and VV) in Beed district of Maharashtra. The bajra area is extracted using Random Forest (RF) classification technique and validated using the ground observation collected extensively from the field. An area of 1401 square kilometers was found under rainfed bajra out of 10693 square kilometers area of entire Beed district which is 13.1% of the total geographical area. The user accuracy (omission error), producer accuracy (commission error) for bajra crop, overall accuracy and Kappa coefficients were 82.9, 80.1, 87 and 0.71%, respectively. The study demonstrated that SAR data can be successfully used for the discrimination and acreage estimation of rainfed bajra using RF classifier.

This work presents voltage-oriented MPPT approach that is in combination of classic P&O algorithm and an external voltage based adaptive sliding mode controller. The proposed algorithm has a novel sliding surface with derivative and integral terms, selected to reduce overshoot during sudden changes in solar irradiation and thus improve steady-state fluctuation. The proposed techniques also incorporate for changing the control parameters at each irradiance and temperature. In order to compare the outcome of solar MPPT the proposed technique compared with the most widely used MPPT techniques i.e P&O and Inc on MATLAB/Simulink simulations.

We propose a novel-view augmentation (NOVA) strategy to train NeRFs for photo-realistic 3D composition of dynamic objects in a static scene. Compared to prior work, our framework significantly reduces blending artifacts when inserting multiple dynamic objects into a 3D scene at novel views and times; achieves comparable PSNR without the need for additional ground truth modalities like optical flow; and overall provides ease, flexibility, and scalability in neural composition. Our codebase is on GitHub.

We present SHIFT3D, a differentiable pipeline for generating 3D shapes that are structurally plausible yet challenging to 3D object detectors. In safety-critical applications like autonomous driving, discovering such novel challenging objects can offer insight into unknown vulnerabilities of 3D detectors. By representing objects with a signed distanced function (SDF), we show that gradient error signals allow us to smoothly deform the shape or pose of a 3D object in order to confuse a downstream 3D detector. Importantly, the objects generated by SHIFT3D physically differ from the baseline object yet retain a semantically recognizable shape. Our approach provides interpretable failure modes for modern 3D object detectors, and can aid in preemptive discovery of potential safety risks within 3D perception systems before these risks become critical failures.

This paper explores the possibility of using visual object detection techniques for word localization in speech data. Object detection has been thoroughly studied in the contemporary literature for visual data. Noting that an audio can be interpreted as a 1-dimensional image, object localization techniques can be fundamentally useful for word localization. Building upon this idea, we propose a lightweight solution for word detection and localization. We use bounding box regression for word localization, which enables our model to detect the occurrence, offset, and duration of keywords in a given audio stream. We experiment with LibriSpeech and train a model to localize 1000 words. Compared to existing work [1], our method reduces model size by 94%, and improves the F1 score by 6.5%.