Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.
Advertisement
Scientific Reports volume 16, Article number: 22528 (2026)
706
1
Metrics details
Ensuring reliable power delivery in standalone photovoltaic (PV) systems is challenging due to the intermittent nature of solar irradiance and dynamic load variations. Conventional switching control schemes often fail to maintain optimal performance under such varying conditions, leading to reduced efficiency and delayed source transition. Although several studies have explored PV converter control and MPPT algorithms, the integration of machine learning techniques for real-time switching control in standalone PV systems with grid backup remains an emerging area. This research work aims to address the identified gap by developing an ML- Decision Tree based switching strategy that enhances power management and system efficiency. This paper presents applying switching rules to choose the correct power source and train a machine learning model to predict future switching for a standalone PV system that operates autonomously and sources power from the utility grid only when solar energy becomes insufficient. The proposed control framework employs real-time monitoring of PV array voltage, current, irradiance, and load demand as input features to the trained DT model, which predicts the optimal switching decision between PV and grid supply. The Simulation work carried out in this paper using MATLAB/Simulink demonstrates application of machine learning – Decision Tree based switching to automate switching operation with reduced switching losses and improved voltage regulation compared to rule based conventional switching in a PV system.
The growing global demand for clean and sustainable energy has accelerated the adoption of photovoltaic (PV) systems as a viable renewable power source. Standalone PV systems, in particular, are increasingly deployed in remote or off-grid locations where conventional grid infrastructure is limited or unreliable. However, their performance is inherently dependent on environmental factors such as solar irradiance and temperature, which fluctuate significantly throughout the day. These variations cause intermittent power generation, making it difficult to ensure uninterrupted power supply to the connected load. To overcome this limitation, hybrid PV systems incorporating grid support are often employed, wherein the load is powered by the PV source under normal conditions and automatically switches to grid supply when solar energy becomes insufficient.
Machine learning algorithms are implemented for dynamic behavior of a power converter for energy harvesting applications. Data driven modeling of Power Electronic converters is a black box approach which results in high accuracy results1. For efficient operation of an energy system, balance between power demand and power supply is very important2. Smart home technology allows for managing and controlling of electrical energy sources with flexible and intelligently responding system providing comfort and needs for its occupants3. ML has the ability to revolutionize energy management practices providing accurate and effective solutions for integrating renewable energy into existing grid infrastructure4. Powerful predictive algorithms can improve system performance and can be used to solve control problems. Machine learning is the application of data driven algorithm to enhance the operation and performance of electrical systems5,6,7.
Powerful ML techniques use classification and regression-based learning algorithms and can be designed to make intelligent decisions which helps to optimize design parameters of a converter system8,9,10. Data driven modeling is used in multiple domains to model and predict system behavior under various operating conditions. ML models are emerged in the area of control and estimation of power electronic converters and in motor drives. ML algorithms can provide solutions to improve performance of PV system through integration of different techniques and modeling complex dynamics11,12,13.
Well trained ML based nonlinear controllers can improve efficiency, minimizing oscillations. ML model learns through information directly from data without assuming predetermined Eqs.14,15. ML technology helps to generate and utilize renewable power effectively with minimum losses and good efficiency, maintaining system reliability16,17. Controllers using ML technology can reduce settling time error and overshoot percentage optimizing control parameters design18,19,20. In renewable energy system ML algorithms are modelled for energy management in smart grid to ensure power availability, load predictions and information gathering21,22,23. Power electronic converters are designed using ML technology to enhance and optimize power system stability and performance. Success rate of ML models is increasing towards better performance of the system24,25. To analyze data, it is important to know what type of data source is used so that it helps in choosing techniques to be used in analyzing them. Different machine learning tasks have different performance evaluation metrics26,27,28.
Although extensive research has been carried out on photovoltaic (PV) converter control and maximum power point tracking (MPPT) techniques, the application of machine learning for real-time switching control in standalone PV systems with grid support remains relatively underexplored. Existing studies largely emphasize MPPT optimization and conventional control strategies, with limited focus on intelligent switching mechanisms for overall power management. Recent advancements in machine learning have demonstrated significant potential in enhancing the adaptability and decision-making capabilities of renewable energy systems. These techniques enable systems to learn from operational data, identify patterns, and respond effectively to dynamic environmental conditions. However, their application in controlling switching actions of power electronic converters particularly in coordinating DC–DC converters and inverter operations has not been sufficiently investigated.
In particular, the use of machine learning to determine optimal switching states for efficient power flow management in standalone PV systems has received minimal attention. This highlights a critical research gap in developing data-driven control strategies that go beyond traditional rule-based methods. To address this gap, the present work proposes an intelligent machine learning-based switching control framework aimed at improving power management, enhancing system efficiency, and ensuring reliable operation under varying irradiance and load conditions. The proposed approach contributes towards advancing smart and adaptive control in modern PV energy systems.
In this work, a machine learning switching control strategy using a Decision Tree algorithm is proposed for a standalone PV system with auxiliary grid support. The proposed controller continuously monitors system parameters such as PV voltage, current, irradiance, and load demand, and predicts the optimal switching action between PV and grid sources to ensure continuous and efficient power supply. Simulation studies performed in MATLAB/Simulink demonstrate that the proposed DT-based controller achieves faster transition response, minimized switching losses, and improved power quality compared to conventional rule-based switching methods. The results validate the potential of data-driven switching control to improve the reliability and operational efficiency of standalone PV systems with grid backup.
The remainder of this paper is organized as follows: Sect. 2 presents the system configuration and control architecture. Section 3 describes the proposed Decision Tree–based switching control algorithm and training methodology. Section 4 discusses the simulation setup and performance evaluation results. Finally, Sect. 5 concludes the paper with key findings and future research directions.
The proposed standalone photovoltaic (PV) system with grid support is designed to ensure uninterrupted power delivery to the connected load, even under varying solar irradiance conditions. The system operates primarily using solar energy generated by the PV array, while the utility grid acts as an auxiliary power source during periods of insufficient solar power. The control objective is to intelligently switch the load supply between the PV and grid sources, maintaining stable voltage and frequency with minimal power interruption and switching loss. The overall system consists of the following main components as shown in Fig. 1.
The PV array is connected to the boost converter with MPPT control, whose output feeds the inverter.
The inverter output is connected to the load.
A grid connection path with a controllable switch runs parallel to the inverter–load path.
Sensors feed data (irradiance, load 1 and load 2 current) to a Decision Tree–based switching controller.
The controller decides whether to maintain PV operation or switch to grid supply based on real-time conditions.
Block diagram illustrating the principal components of a solar PV system.
To meet 1 kW capacity PV system, 5 modules each of 215 W power rating with specified I-V and P-V ratings as shown in Fig. 2 are connected in series. For the selected PV module maximum PV voltage is 28.72 V with maximum current capacity of 7.49 A.
P–V and I–V characteristics of 215 W and 1 kW photovoltaic modules.
A flyback DC–DC converter, with specifications listed in Table 1, is employed to boost the PV output voltage.
A single-phase inverter operating in DQ control mode with capacitor current sensing, with specifications as shown in Table 2, is implemented.
The PV system is designed for a Loading capacity of 1 kW with maximum 850 W of connected load. The analysis is carried out with connected loads of:
Load 1:500 Watts.
Load 2: 350Watts
The grid serves as a secondary power source, supplying energy to the load when solar power is insufficient. A static switch connects the grid to the load through an DT based intelligent control circuit.
Voltage, current, and irradiance, sensors continuously monitor system parameters. These sensed values are processed and used as input features to the DT controller.
The control architecture comprises of two functional loops, a power management loop and machine learning based switching control loop as shown conceptually in Fig. 1. The power controller architecture of single-phase inverter designed for DQ mode of operation is as shown in Fig. 3. The output voltage of the inverter is controlled by implementing rotating reference frame control technique using a PI controller.
Power control architecture depicting control loops and power flow pathways.
In this study, two machine learning algorithms namely Decision Tree and Random Forest are evaluated for their suitability in switching control of the PV system. The comparison is carried out to identify an appropriate model that balances accuracy, computational efficiency, and real-time implementation feasibility. Both models are trained and tested using the same dataset derived from PV system operating conditions, including irradiance, voltage, and load variations. Performance evaluation is conducted based on metrics such as classification accuracy, response time, and implementation complexity as presented in Tables 3 and 4.
Although Random Forest demonstrated slightly higher prediction accuracy, the Decision Tree model was selected for the proposed system due to its faster response, lower computational requirement, and ease of integration into real-time control hardware. This makes it more suitable for practical implementation in standalone PV systems where rapid switching decisions are critical.
The proposed intelligent controller replaces conventional threshold-based switching logic with a DT model trained on system behaviour under different operating conditions.
Input Features: irradiance (G), load power (P_load), time.
Output Decision: Binary switching command (PV mode = 1/0, Grid mode 1/0 ,central load 1/0).
The DT controller is implemented as part of a MATLAB/Simulink model, where training data are obtained from simulated PV system responses under variable irradiance and load profiles. Once trained, the DT model executes in real time with low computational overhead, making it suitable for embedded digital controllers such as DSPs or microcontrollers. Two DT models DT1 and DT2 are developed to use solar power generation to the best by predicting power generation capacity and predict switching between load, PV and grid to have faster and more efficient way of switching control respectively. The expected results from the developed DT models are.
Automation of managing solar energy.
Apply conditional logic for power source switching.
Convert real-world energy system data for machine learning.
Train a Decision Tree classifier to predict switching decisions.
Visualize energy source behavior over different days.
Mechanism to record switching results.
The Simulink function block designed to predict power generated from PV based on irradiance is shown in Fig. 4a. The Predicted output from the DT1 model will be to evaluate if the power output from PV panel is sufficient enough to turn on only load1 or it can turn on both load1 and load2 simultaneously. Conditions considered for developing DT1 model predictions is shown in Table 5, wherein, 0 and 1 corresponds to turn off and turn on conditions of the load.
(a) DT1 simulink block. (b) DT2 simulink block.
The Simulink function block designed to predict switching action and to take decision to manage power is shown in Fig. 4b. Conditions for predictions and switching control in DT2 model is shown in Table 6.
A machine learning DT algorithm is developed for predictions based on conditions specified in Tables 5 and 6 using Python and is presented in the following section.
The tools and llibraries used in developing DT algorithms in this work are shown in Table 7 The model is evaluated using metrics such as classification accuracy,precision, and recall to assess its performance under unseen operating conditions. Once trained, the Decision Tree is exported that can be directly implemented in MATLAB/Simulink in a microcontroller. In this work DT model is developed for switching predictions between PV, central load and grid as shown in flowchart in Fig. 5.
Flowchart of the machine learning-based decision-making process.
To develop an accurate and generalizable DT model, representative datasets are generated through detailed simulations of the standalone PV system under various environmental and load conditions. MATLAB/Simulink is used to simulate different operating scenarios which include variable solar irradiance (300–1000 W/m2), varying load demands. The Python function utilizes a Decision tree classifier from the scikit-learn library to predict the state of three switches. The process of calling Python function in DT models developed in the present work is shown in Fig. 6.
Data communication process between MATLAB and Python during simulation.
The overall control algorithm is implemented using MATLAB/Simulink blocks as shown conceptually in Fig. 7a and b. The DT logic is embedded as a MATLAB Function block that processes real-time data and outputs the control signals for switches.The circuit includes PV panel with MPPT controller, Fly back DC-DC converter and single phase H bridge inverter with DQ mode with capacitor current sensing. Three switches namely PV switch, Grid switch, and Central switch are monitored and controlled based on the irradiance level and power availability from the PV system. The PV voltage which is obtained after boosting using Fly back DC-DC converter is taken as reference to take the decision. Simulink model circuit is shown in Fig. 7a and b respectively.
(a) Simulink-based test bench for performance evaluation of the conventional switching model. (b) Simulink model illustrating the integration of machine learning for system control.
To evaluate the performance and effectiveness of the proposed DT based intelligent switching controller, comprehensive simulations were carried out using MATLAB/Simulink. The simulation model integrates the photovoltaic array, boost converter with MPPT control, single-phase inverter, load and grid interface by DT based switching algorithm. The simulation is carried out for various scenarios. Conventional and DT based switching methods are compared in terms of switching speed, total harmonic distortion and converter efficiency under dynamic load conditions. Unless otherwise stated, time is expressed in seconds. Voltage, current, and power are measured in volts (V), amperes (A), and watts (W), respectively.
Three representative test cases were designed to evaluate the dynamic performance of the system under varying solar irradiance and load conditions:
Case 1: Grid mode (< 400 W/m²).
Case 2: Constant Irradiance (1000 W/m²).
Case 3: Irradiance Drop (1000 → 400 W/m²).
Initially grid is powering load as PV is not able to meet load demand. In such case grid does not transfer power to central load. when PV gains generation capacity grid is disconnected.The simulation results of the inverter output voltage obtained for a conventional PI based controller are as shown in Fig. 8.
Switching actions between PV and grid in conventional switching.
From results as obtained in Fig. 8,the following cases are discussed:
Case 1: 0 to 0.5 s as irradiance is less than 400 W/m2, grid switch is on, PV switch is off, load 1 is on, load 2 is off and no power goes to central load.
Case 2: 0.5 s to 1 s as irradiance is greater than 400 W/m2, grid switch is off, PV switch is on, load 1 is on, and load 2 is off, extra power goes to central load.
1 s to 1.5 s, grid switch is off, PV switch is on, Load 1 is on, load 2 is on and in such case no power goes to central load.
Case 3: 1.5 s to 2sec, grid switch is on, PV switch is off, as irradiance is less than 400 W/m2 load 1 is on, Load 2 is on and in such case no power goes to central load and the load demand is taken care by grid alone.
Output voltage at various stages under maximum load and maximum irradiance using conventional switching is shown in Fig. 9.
Voltage profiles illustrating input and output behavior in conventional switching.
THD of inverter output voltage and power obtained at various conversion stages at full load condition at maximum irradiance 1000 W/m2 is shown in Fig. 10a and b.
(a) Total Harmonic Distortion (THD) of inverter output voltage under conventional switching. (b) Power characteristics of the system under conventional switching.
It is observed from results that for constant irradiance, overall converter efficiency is 76.49% with fundamental peak inverter voltage 0f 310.5Vp-p.
Output voltage at various stages for maximum load and maximum irradiance is shown in Fig. 11.
Voltage profiles illustrating system behavior under DT switching control.
THD and power output at full load condition at maximum irradiance 1000 W/m2 is shown in Fig. 13a and b.
(a) Total Harmonic Distortion(THD) of inverter output voltage with DT switching. (b). Power characteristics of the system under decision tree (DT)-based switching.
Current waveforms in conventional switching with varying irradiance.
It is observed from results that for constant irradiance overall converter efficiency is 77.04% with fundamental peak inverter voltage 0f 307.2Vp-p.
The results obtained with DT model are evaluated with results from the conventional model and are found to be comparable. Comparative results of conventional and DT based model with 1000 W/m2 irradiance when full load is connected is presented in Table 8.
For conventional switching, output voltage, THD of output voltage and current flow at various stages for varying irradiance and varying loading pattern are shown in Fig. 14.
THD of inverter output voltage for conventional switching with varying irradiance.
THD of output voltage, power conversion efficiency for varying irradiance and varying loading pattern with conventional switching is shown in Figs. 15 and 16.
Power conversion curves for conventional switching.
Current waveforms in conventional switching with varying irradiance.
For DT based model switching between PV and grid with varying irradiance and varying loading pattern, output voltage and current flow at various stages are shown in Fig. 17.
THD of output voltage with DT switching with varying irradiance.
THD of inverter output voltage and power conversion curves for varying irradiance and varying loading pattern with DT based switching is shown in Figs. 18 and 19.
Power conversion curves for DT switching.
Predictions of variation of irradiance W/m2 with respect to time.
Tables 9 and 10 summarize the comparative performance of the proposed DT-based controller versus a conventional threshold-based switching controller.
The comparative analysis highlights the effectiveness of the proposed Decision Tree-based control strategy in improving the dynamic performance of the PV-grid system.
Data used in this work is mainly irradiance, time and loading pattern. To train and test the DT model, three days’ time and irradiance data is used and the prediction of irradiance with respect to time for the fourth day is shown in Fig. 20.
Dataset and predicted values for the loading pattern.
To train and test the DT model, three days loading data used and the fourth day predictions of loading with respect to time is shown in Fig. 21.
Dataset and predictions of decision tree (DT)-based switching.
The switching between PV, grid and load is based on power availability and loading condition.DT1 model first predicts power generation capacity of PV based on irradiance and accordingly predicts if the power is sufficient to turn on both loads and only one load. Based on the predictions of DT1 model, loads can be turned on according. DT 2 model takes inputs from DT1 model and switches between central, PV and grid accordingly The data used and the predicted results of DT model is shown in Table 11. The dataset and the predictions plot is shown in Fig. 22.
This paper presents a Decision Tree-based intelligent switching control strategy for PV-grid systems, enabling efficient real-time power management under varying operating conditions. The proposed approach demonstrates enhanced overall system characteristics by achieving a 5.7% reduction in electrical stress and an 8.8% improvement in output voltage regulation, contributing to improved stability and reliability. Furthermore, the power conversion efficiency is increased from 76.49% to 77.04%, representing a modest improvement of approximately 0.72%, indicating better energy utilization. To ensure a fair comparison, a limited dataset has been used for training the Decision Tree model, consistent with the dataset employed in the conventional switching simulation framework. While the results are based on simulation studies, the observed improvements indicate the potential of the proposed approach for fast and efficient decision-making.
The proposed control strategy enables optimized grid utilization, reducing dependency on the grid by prioritizing the use of available PV power. This ensures that solar energy is effectively utilized to its maximum potential, thereby minimizing unnecessary grid consumption.
Future work will focus on real-time hardware implementation and the integration of additional features such as temperature and advanced learning models to further enhance system performance.
All data generated or analysed during this study are included in the published article.
PV output voltage(Volts)
DC link voltage(Volts)
Inverter output voltage(Volts)
Grid Current (Amps)
Total Load current(Amps)
Current in central load(Amps)
PV power output(Watts)
Dc power output(Watts)
AC power output(Watts)
Vergos, G. et al. Machine Learning based Power Converter Large Signal Simulation for Energy Harvesting Applications, 2022 IFIP/IEEE 30th International Conference on Very Large Scale Integration (VLSI-SoC), Patras, Greece, pp. 1–5, (2022). https://doi.org/10.1109/VLSI-SoC54400.2022.9939641
Goswami, K. & Kandali, A. B. Machine learning algorithms for predicting electrical load demand: an evaluation and comparison. Sādhanā 49, 40. https://doi.org/10.1007/s12046-023-02354-2 (2024).
Article Google Scholar
Widianto, M. H., Gunawan, A. A. S., Heryadi, Y. & Budiharto, W. Evaluation of Machine Learning on Smart Home Data for Prediction of Electrical Energy Consumption, International Conference on Computer Science, Information Technology and Engineering (ICCoSITE), Jakarta, Indonesia, 2023, pp. 434–439, Jakarta, Indonesia, 2023, pp. 434–439, (2023). https://doi.org/10.1109/ICCoSITE57641.2023.10127700
Singh, R., Kumar, A. & Bajaj, R. S. Machine learning-based energy management and power forecasting in grid-connected microgrids with multiple distributed energy sources. Sci. Rep. 14, 19207. https://doi.org/10.1038/s41598-024-70336-3 (2024).
Article CAS ADS Google Scholar
Khashroum, Z., Rahimighazvini, H. & Bahrami, M. Applications of Machine Learning in Power Electronics: A Specialization on Convolutional Neural Networks. ENG Transactions. 4. (2023). https://doi.org/10.61186/engt.4.1.2866
Fassi, Y., Heiries, V., Boutet, J. & Boisseau, S. Toward Physics-Informed Machine-Learning-Based Predictive Maintenance for Power Converters—A Review, in IEEE Transactions on Power Electronics, vol. 39, no. 2, pp. 2692–2720, Feb. (2024). https://doi.org/10.1109/TPEL.2023.3328438
Bensoussan, A., Li, Y., Nguyen, D. P. C. & Tran, M. B. Sheung Chi Phillip Yam, Xiang Zhou, Chap. 16 – Machine learning and control theory, Elsevier, Volume 23,2022,Pages 531–558,https://doi.org/10.1016/bs.hna.2021.12.016
Matania, R., Forêt, J. M., Camarillo, V. & Walker, M. Process Control Decision Inference, Monitoring, and Execution. Annual Conference of the PHM Society. 11. (2019). https://doi.org/10.36001/phmconf.2019.v11i1.850
Wang, S. et al. Jan. An Intelligent System for Automatic Selection of DC-DC Converter Topology with Optimal Design., Conference: Denver, CO, United States, OSTI ID:1860710. (2022).
Chen, Y. et al. Evaluation of Machine Learning Models for Smart Grid Parameters: Performance Analysis of ARIMA and Bi-LSTM. Sustainability 15, 8555. https://doi.org/10.3390/su15118555 (2023).
Article ADS Google Scholar
Krishnamoorthy, H. S. & Narayanan Aayer, T. Machine Learning based Modeling of Power Electronic Converters, 2019 IEEE Energy Conversion Congress and Exposition (ECCE), Baltimore, MD, USA, pp. 666–672, (2019). https://doi.org/10.1109/ECCE.2019.8912608
Zhang, S., Wallscheid, O. & Porrmann, M. Machine Learning for the Control and Monitoring of Electric Machine Drives: Advances and Trends. IEEE Open. J. Ind. Appl. 4, 188–214. https://doi.org/10.1109/OJIA.2023.3284717 (2023).
Article Google Scholar
Gaviria, J. F., Narváez, G., Guillen, C. & Giraldo, L. F. Michael Bressan,Machine learning in photovoltaic systems: A review,Renewable Energy,Volume 196,2022,Pages 298–318,ISSN 0960–1481,https://doi.org/10.1016/j.renene.2022.06.105
Anup Marahatta, Y., Rajbhandari, A., Shrestha, S., Phuyal, A. & Thapa Petr Korba,Model predictive control of DC/DC boost converter with reinforcement learning,Heliyon,8, Issue 11,2022,e11416,ISSN 2405–8440,https://doi.org/10.1016/j.heliyon.2022.e11416
Escobar, C. & Morales-Menendez, R. Machine learning techniques for quality control in high conformance manufacturing environment. Adv. Mech. Eng. 10, 168781401875551. https://doi.org/10.1177/1687814018755519 (2018).
Article Google Scholar
Aanand Kumbhar, P. G., Dhawale, S., Kumbhar, U. & Patil Pravin Magdum,A comprehensive review: Machine learning and its application in integrated power system,Energy Reports,Volume 7,2021,Pages 5467–5474,ISSN 2352–4847,https://doi.org/10.1016/j.egyr.2021.08.133
Gaboitaolelwe, J. et al. Machine Learning Based Solar Photovoltaic Power Forecasting: A Review and Comparison. IEEE Access. 11, 40820–40845. https://doi.org/10.1109/ACCESS.2023.3270041 (2023).
Article Google Scholar
Arikuşu, Y. S. & Bayhan, N. Design of a Novel PID Controller Based on Machine Learning Algorithm for a Micro-Thermoelectric Cooler of the Polymerase Chain Reaction Device, in IEEE Access, vol. 12, pp. 61959–61977, (2024). https://doi.org/10.1109/ACCESS.2024.3392734
Abegaz, B. W. Dynamic switching control of buck converters using unsupervised machine learning methods. J. Eng. 2020 (12), 1155–1164. https://doi.org/10.1049/joe.2020.0086 (2020).
Article Google Scholar
Saha, J., Hazarika, D., Gorla, N. B. Y. & Panda, S. K. Machine-Learning-Aided Optimization Framework for Design of Medium-Voltage Grid-Connected Solid-State Transformers, in IEEE Journal of Emerging and Selected Topics in Power Electronics, vol. 9, no. 6, pp. 6886–6900, Dec. (2021). https://doi.org/10.1109/JESTPE.2021.3074408
Mansoury, I., El Bourakadi, D., Yahyaouy, A., Boumhidi, J. & & & A novel decision-making approach based on a decision tree for micro-grid energy management. Indonesian J. Electr. Eng. Comput. Sci. 30, 1150. https://doi.org/10.11591/ijeecs.v30.i2.pp1150-1158 (2023).
Article Google Scholar
Gros, I-C., Lü, X., Oprea, C., Lu, T. & Pintilie, L. Artificial intelligence (AI)-based optimization of power electronic converters for improved power system stability and performance, IEEE 14th International Symposium on Diagnostics for Electrical Machines, Power Electronics and Drives (SDEMPED), Chania, Greece, 2023, pp. 204–210, Chania, Greece, 2023, pp. 204–210, (2023). https://doi.org/10.1109/SDEMPED54949.2023.10271490
Solanki, S. D. & Solanki, A. D. Implementation of Big Data and Machine Learning in Smart Grid with Correlated Safety Considerations: Review, IEEE International Symposium on Sustainable Energy, Signal Processing and Cyber Security (iSSSC), Gunupur Odisha, India, 2020, pp. 1–6, Gunupur Odisha, India, 2020, pp. 1–6, (2020). https://doi.org/10.1109/iSSSC50941.2020.9358879
Kaledio, P. Machine Learning Applications in Electric Power Systems: Enhancing Efficiency, Reliability, and Sustainability (january 25, 2024). https://doi.org/10.2139/ssrn.4716389
Zhang, K. et al. Feb., Machine Learning-Based Temperature Prediction for Runtime Thermal Management Across System Components, in IEEE Transactions on Parallel and Distributed Systems, vol. 29, no. 2, pp. 405–419, 1 (2018). https://doi.org/10.1109/TPDS.2017.2732951
Andreas, C. M. Sarah Guido Introduction to Machine Learning with Python O’Reilly media,2016 ISBN 978-1-449-36941-5.
Saiket, D. Subramanian Chandramouli,Amit Kumar Das,Machine Learning Pearson (2018).
Alice ZhengEvaluating Machine Learning Models O’Reilly,2015.
Download references
Open access funding provided by B.M.S. College of Engineering.
Department of Electrical and Electronics Engineering, B.M.S. College of Engineering (Affiliated to V.T.U, Belagavi), Bangalore, India
K R Pushpa
B.M.S. College of Engineering Department of Electrical and Electronics Engineering, BMSCE, Bangalore, India
R S Geetha
PubMed Google Scholar
PubMed Google Scholar
Both the authors contributed equally towards the completion of the manuscript. The manuscript was written by Pushpa K R under the guidance of. Dr. Geetha R S.
Correspondence to K R Pushpa or R S Geetha.
The authors declare no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Reprints and permissions
Pushpa, K.R., Geetha, R.S. Application of machine learning decision tree based power switching control for a standalone PV system with grid support. Sci Rep 16, 22528 (2026). https://doi.org/10.1038/s41598-026-50546-7
Download citation
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41598-026-50546-7
Anyone you share the following link with will be able to read this content:
Sorry, a shareable link is not currently available for this article.
Provided by the Springer Nature SharedIt content-sharing initiative
Advertisement
Scientific Reports (Sci Rep)
ISSN 2045-2322 (online)
© 2026 Springer Nature Limited
Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.
