DEVELOPMENT OF A MAN-MACHINE INTERFACE FOR MANAGING PHOTOVOLTAIC ENERGY: APPLIED STUDY

Objective: Develop and evaluate a human-machine interface (HMI) that integrates advanced monitoring, forecasting, and management functionalities for photovoltaic solar energy systems, aiming to optimize energy production and operational efficiency. Theoretical Framework: This study is based on concepts of modeling and simulation, solar energy management, and problem-solving methodologies such as Soft System Methodology (SSM). Method: An applied approach was adopted using modeling, simulation, and statistical analysis techniques. The research included a bibliographic review in scientific databases, a case study, and SSM to organize and solve complex problems. 121 digital solar energy platforms in Brazil were analyzed to define the interface requirements. The interface was developed with React JS, Axios, Bootstrap v5, Apache Echarts, HTML, CSS, JavaScript, and Python libraries for forecasting models. Results and Discussion: The interface, named "Solar Smart Manager," enables efficient monitoring and management of energy production using critical data such as temperature, time of day, and solar irradiation. Tests in a real operational environment demonstrated improvements in energy management, incident response, and preventive maintenance. The functionality of validating solar radiation incidence data represents a significant contribution to the energy sector, promoting sustainability and innovation. Research Implications: The practical and theoretical implications of this research provide insights into the efficient and optimized management of photovoltaic solar energy systems, contributing to a better understanding and optimization of available solar resources. Originality/Value: This study contributes to the literature by developing an innovative interface that improves operational efficiency and solar energy management. The relevance and value of this research are evidenced by its positive impact on the energy sector, promoting sustainability and innovation.


INTRODUCTION
As the global demand for energy continues to grow, it becomes essential to search for more sustainable energy sources, but also to implement robust policies that promote environmentally responsible practices ( UN, 2015) .In this context, the transition to renewable energy is not only a strategic priority, but an urgent need to mitigate the adverse environmental impacts associated with conventional energy sources.This change is driven both by the need to reduce greenhouse gas emissions and by the potential of renewable energies to meet growing energy demand in a sustainable way ( Assunção;Deus, 2022;EPE, 2023) .
Among alternative sources, photovoltaic solar energy emerges as one of the most promising alternatives due to its availability being an abundant source ( Sarkar ;Odyuo, 2019) .Brazil exhibits a high average incidence of global irradiation, which extends across the entire national territory ( Pereira et al., 2017) .This characteristic translates into great potential for solar energy generation, a resource that is still underutilized in the country ( Alves et al., 2018) .According to the Electric Energy Statistical Yearbook (2021), the installed capacity for electrical generation from solar sources expanded by 32.9% in 2020, continuing a growth trend observed since 2018.Currently, the country has around 4,982.18 megawatts of installed capacity in photovoltaic (PV) generation, with forecasts for 2023 indicating an additional increase of more than 2,270.53MW, reaching the expectation of 16,799 MW by 2029( ANEEL, 2023;Pereira, 2022;EPE, 2022) .
The process of converting solar radiation into electricity is carried out through photovoltaic solar modules that include inverters, circuit breakers and meters ( Mariano;Urbanetz Jr, 2022) .These modules consist of solar cells made from semiconductor materials, such as silicon, that directly convert solar energy into electricity.With the technological evolution of these components, especially cells, which can reach efficiencies of up to 47.1% in laboratory conditions NREL (2021) , the solar energy generation capacity has increased significantly.
Furthermore, the variability in the incidence of solar radiation due to climatic and atmospheric factors, such as humidity, cloudiness, and the presence of dust, requires accurate radiation prediction to optimize energy generation.Studies have addressed these variabilities and developed predictive models that improve the forecasting and management of solar energy generation ( Rahimi et al., 2023) .The implementation of advanced technologies, including neural networks, plays a crucial role in reducing uncertainty in predictions and maximizing the potential of installed photovoltaic systems ( Singla et al., 2022) .
However, to maximize the use of this technology, it is essential to have advanced management systems that optimize both the capture and use of the energy generated ( AlDahoul et al., 2022;Pinheiro;Müller, 2023) .The increasing complexity and diversity of photovoltaic systems, combined with dynamic environmental variables, impose significant challenges in the effective monitoring and management of these systems ( Lima et al., 2021;Hasan et al., 2022) Faced with these challenges posed by the complexity and diversity of photovoltaic systems, in addition to dynamic environmental variables, it becomes crucial not only to continuously monitor energy production, but also to employ advanced strategies to effectively predict and respond to significant variations ( Huang et al., 2018;Das et al., 2018) .The efficient management of these systems requires the development and implementation of forecasting technologies that use a complex set of data to anticipate changes in solar irradiation conditions and energy demand ( Mellit ;Kalogirou, 2021) .Such technologies allow operators to proactively adjust system parameters to maximize operational efficiency and energy production, thus ensuring more robust and adaptive management that can meet the challenges brought by an ever-changing environment.
Thus, efficient management of photovoltaic systems requires not only continuous monitoring of energy production, but also the ability to predict and respond to variations based on a complex set of data.Therefore, the research question is: How can a human-machine interface be developed to significantly improve the management of solar photovoltaic energy systems by integrating predictions and optimizing system operation?
Therefore, the objective of this study is to develop and evaluate a human-machine interface that integrates advanced monitoring, forecasting and management functionalities for photovoltaic solar energy systems, aiming to optimize energy production and operational efficiency.In order to improve photovoltaic energy management, this study proposes the development of an intelligent interface for web systems.This article details the creation of a low-cost solution designed to monitor, report and analyze the energy production and operational condition of photovoltaic panels.The interface aims to facilitate informed decision-making about the operation and maintenance of these systems, thus contributing to energy efficiency and sustainability ( Lins Chung et al., 2021) .
As a contribution, this study aims to create an interface to manage energy generation from six different technologies, a project with great potential to better manage photovoltaic solar energy in the country.
The article is organized into four sections.Section 1 presents the introduction, problem, objective and justification for the development of this article.Section 2 addresses the methodology used and how its application was conducted, the tools used and specificities of the interface, as well as the databases used.Section 3 presents the development of the interface and the results obtained.Section 4 presents the conclusions obtained in this study.And finally, the references used.

SCIENTIFIC RESEARCH METHODS
To develop this research, we adopted robust methodological methods and procedures, aiming to efficiently address the study problem related to the management of grid-connected photovoltaic systems for six different technologies.

APPLIED METHODOLOGY
This study is classified as applied, according to Marconi;Lakatos (2021) and GIL (2020)

6
, as it aims to develop practical solutions for specific challenges in the area of photovoltaic energy.We use a qualitative and quantitative approach , with modeling, simulation and statistical analysis techniques, following the guidelines of ( Andrade, 2017) .The study is descriptive and explanatory in nature, focusing on the development of interfaces that facilitate efficient energy management of photovoltaic systems, with a focus on identifying and describing models and algorithms that not only optimize the simulation of energy generation, but also improve the interactivity and usability of user interfaces ( Leão, 2017) .Regarding technical procedures, we conducted a literature search, case study and Soft System Methodology .The bibliographical research was based on previously published materials, such as scientific articles, books, theses and conference proceedings, Scopus and Web of Science databases ( Andrade, 2017;Marconi;Lakatos, 2021 Son;Santos, 2017) .The case study, recommended by Fachin (2017) , allowed a detailed investigation into particular aspects of the energy generation of PV systems and also the particularities of their different technologies, providing solid information to be used at the interface.And the Soft Systems Methodology (SSM) methodology helped in the process of organizing and making flexible which is ideal for approaching and solving complex problems, as described by ( Checkland ;Poulter, 2020) .It allowed us to have a deeper understanding of user needs and technical complexities, allowing a holistic approach to interface design.
For sensitivity simulations of the photovoltaic system, we chose to use a neural network with 20 neurons in the hidden layer.This method was chosen as the most suitable for this type of analysis after an extensive training process, which involved seven different types of networks and three different network algorithms, as documented by (Pinheiro, Lovato, et al., 2017 ;Pinheiro, Ruther, et al., 2017) .The results obtained through this method were used to develop the interface proposed in this study.

METHODOLOGY APPLIED TO THE INTERFACE
The objective of this research is to develop a low-cost web system to optimize energy generation systems.The proposed interface will be adaptable to six different types of photovoltaic technologies, with the function of analyzing the sensitivity of the PV system as well as predicting the power generated in the system with a delay of 30 minutes, including those built by users who have implemented their own photovoltaic energy systems .
For its construction, some steps were analyzed: initially defining the specific objectives of the management interface, subsequently designing the user interface, followed by the choice

DEVELOPMENT AND RESULTS
The results presented in this document follow the typical steps of a software development project, including: requirements analysis, architecture design, code implementation, unit testing, integration and system testing, and production deployment.It is worth noting that, in this article, the conduct of the first three stages of the methodology were described: requirements elicitation , prototyping and prototype evaluation.
Table 1 highlights the main digital platforms and applications used by integrators, highlighting a selection of 121 platforms, of which a representative from each segment is exemplified.This table not only catalogs the available applications, but also maps the effective use of these tools by integrating companies in the sector, as documented by ( Greener , 2021) .Notably, among all the solutions analyzed, none offers predictive energy analysis functionalities, thus highlighting the importance and innovation proposed by this new interface.As the main objective, at this point real-time data will not be used, the main focus is to work with historical data analysis: with access to the historical energy generation database to identify trends, patterns and anomalies; with generation forecasting: using a network trained with predictive models to estimate future energy generation, assisting in planning and optimizing the operation and with generation optimization: implementing optimization algorithms to maximize energy generation and minimize costs, considering variables such as weather conditions and energy demand.Firstly, the data was collected using the Scada system , after collection, the data was processed, considering one year, totaling approximately 774,526 data points by variables from the six technologies.(See letter A).Subsequently, as input data for the network, global irradiance (X1), ambient temperature (X2), collection time (X3) and module temperature (X4) were used.As output data, or Target, power data generated by the photovoltaic system of each of the six technologies were used (See Letter B).Algorithms trained for the six different technologies and which presented satisfactory prediction results were used to build the interface.The first sketch of the interface can be seen in Figure 2 Letter C.

Figure 2
Sequence of steps for developing the Interface prototype.
Source : Prepared by the authors.
Based on document analysis, the essential requirements for the web system were defined.To implement these requirements, the React JS framework was used , which provides an efficient structure for developing dynamic interfaces.HTTP requests were managed with the Axios library , while the Bootstrap v5 library was used to build the visual elements, ensuring a responsive and modern interface.For data visualization, Apache Echarts was used , which offers interactive and highly configurable graphs.Basic web development technologies such as HTML, CSS and JavaScript formed the basis of the system.Furthermore, for the solar energy forecast models, Python libraries were used, due to their robustness and wide range of tools for data analysis and modeling.This combination of technologies allowed the creation of an integrated, efficient and easy-to-use platform, meeting previously established requirements.

PROTOTYPE DEVELOPMENT
The development of the first prototype was conducted in an academic manner to evaluate the integration of trained models into an interface, resulting in the first sketches visible in Figure 3. Initially, an interface was designed that used different climatic conditions, such as ambient temperature, relative humidity , wind speed and direction, atmospheric pressure, rainfall and the five solar irradiance variables.The objective was to instantly compare the estimated energy generation for the six photovoltaic technologies.
For systems that have only one technology, the six samples can be maintained or reset at the interface.Figures 3 and 4 present the first ideas of how information would be demonstrated in the proposed interface.

Figure 3
Demonstrative interface for energy estimation with sensitivity analysis of climate variables Source : Authors.
The second suggestion for the interface includes a field for choosing the photovoltaic technology.In addition, temperature information and the incidence of solar irradiance were added.Finally, the interface generates a graph with an estimate of how much energy was produced throughout the day every 30 minutes.

Figure 4
Interface for estimating energy generation every half hour.
With the aim of professional refinement, a collaboration was carried out with experts in Human-Machine Interaction, who have extensive experience in developing interactive systems 12 in business and academic contexts.These professionals provided critical evaluations and valuable suggestions, essential for the prototyping and evaluation stages of the project.As a result, the prototype's interface design was strategically developed to facilitate visual and intuitive access to energy production details, as shown in Figure 5. Named " Solar Smart Manager ", this prototype uses temperature, time of day and irradiation data .The home page includes login and password fields, as well as a description of the system's purpose.In addition to comparing different technologies, the platform also allows individual analysis of each system.This way, the system using a single technology can monitor its performance and simultaneously obtain information about how another technology would behave under the same specific weather conditions.Figure 7 illustrates these features.

Figure 7
Interface window design with comparisons by technology Source : Authors.
Thus, it is observed that the intuitive and visually accessible interface allows users to monitor and manage energy production efficiently, using critical data such as temperature, time of day and irradiation.This approach not only makes systems easier to understand and use, but also promotes more informed decision-making.Thus, the project highlights the relevance of integrating technical expertise with usability principles to create solutions that meet the real

FINAL CONSIDERATIONS
This study addressed the construction of a human-machine interface (HMI) aimed at significantly improving the management of photovoltaic solar energy systems, integrating predictions and optimizing system operation.The main objective of the research was to develop and evaluate an HMI that consolidated advanced monitoring, forecasting and management functionalities, aiming to optimize energy production and operational efficiency.The development of an effective and intuitive human-machine interface such as the " Solar Smart Manager " demonstrates the importance of user-centered design and an interdisciplinary approach.The specific objectives were corroborated as follows.Regarding the development of the interface, it was created in an intuitive way that consolidates critical information from different photovoltaic systems in an accessible and understandable way, facilitating decisionmaking by users.Regarding the implementation of Forecasting Algorithms, solar energy forecasting algorithms were tested and implemented in the interface, providing estimates of future production and allowing proactive preparation for environmental variations.The effectiveness of the interface was tested in a real operational environment, demonstrating improvements in energy management, incident response and preventative maintenance.
Furthermore, the developed interface includes a crucial functionality for validating solar radiation incidence data in the country, which represents a significant contribution to the energy sector.This validation allows for a better understanding and optimization of available solar resources, promoting sustainability and innovation in the energy field.Therefore, it is concluded that the human-machine interface developed in this study provides a powerful tool for the efficient and optimized management of photovoltaic solar energy systems, meeting the established objectives and offering concrete benefits for the country's energy sector.
photovoltaic systems and presents real-time data in an accessible and understandable way.II.Implement solar power forecasting algorithms within the interface to provide estimates of future production, helping with proactive decision making and preparing for environmental variations.III.Test the effectiveness of the interface in a real operational environment, measuring improvements in energy management, incident response and preventative maintenance.
Development of A Man-Machine Interface for Managing Photovoltaic Energy: Applied Study ___________________________________________________________________________ Rev. Gest.Soc.Ambient.| Miami | v.18.n.7 | p.1-18 | e08363 | 2024.7 of development technologies and implementation of the interface ( Bassani et al., 2010) .Figure 1 presents a summary diagram of the development phases of an interface ( Passos; Silva, 2012) Therefore, the methodology adopted in this interface development project in the area of solar energy included the following steps: (1) Requirements elicitation , to clearly understand needs and expectations; (2) Construction of the prototype, allowing an initial visualization of the interface; (3) Evaluation of the prototype, where functionality and usability were tested; (4) System development, phase in which the interface was finalized and prepared for implementation; and (5) Case study, used to validate the interface in a real usage scenario ( Lins Chung et al., 2021) .

Figure 1
Figure 1Main steps for development and an interface

Figure 2
Figure 2 Letters A, B and C present a schematic of the sequence considered for the development of the interface.

Figure 5
Figure 5Design of the proposed interface home window

Figure 6
Figure 6Letter A -Design of the second window of the proposed interface.

Figure 6 Figure 6
Figure 6Letter B -Continuation of the second window of the proposed interface Development of A Man-Machine Interface for Managing Photovoltaic Energy: Applied Study ___________________________________________________________________________ Rev. Gest.Soc.Ambient.| Miami | v.18.n.7 | p.1-18 | e08363 | 2024.15 needs of users and promote sustainability in the energy sector.