ES2949156A1 - SYSTEM AND PROCEDURE FOR COMPREHENSIVE ENERGY MANAGEMENT (Machine-translation by Google Translate, not legally binding) - Google Patents

Abstract

System and procedure for comprehensive energy management. System for comprehensive energy management based on information provided from different devices, either in real time or through historical data, which are processed through the use of AI (Artificial Intelligence) providing actions on any energy management device that can act on an energy distribution network, an energy generation system (renewable or non-renewable), accumulation as well as the internal energy networks of any type of consumer, proceeding to its activation/deactivation based on the optimization strategy developed by the system. The system includes: - A data collection subsystem (A) - Data lake storage subsystem (B) - Artificial Intelligence subsystem (C) - Interaction agents subsystem (D) - Interface subsystem (Frontend) (E) to interact directly with the user - An external application programming subsystem (F) - A control signal generation subsystem (G). (Machine-translation by Google Translate, not legally binding)

Description

DESCRIPTION

SYSTEM AND PROCEDURE FOR ENERGY MANAGEMENT

COMPREHENSIVE

OBJECT OF THE INVENTION

The object of the present invention, as the title of the invention establishes, is a system for comprehensive energy management, as well as the procedure carried out through said energy management system.

The present invention characterizes each of the elements that form part of it, as well as their functionalities and the interaction between them so that optimal energy management is achieved at any user level. The system object of the invention collects energy information from different devices, as well as from other data sources, both in real time and historical, to process it through the use of AI and obtain personalized solutions for each type of user, including recommendations or predictions or actions on energy management through different agents.

Therefore, the present invention is limited within the field of energy, as well as systems implemented through processing means based on Artificial Intelligence.

BACKGROUND OF THE INVENTION

Energy consulting is a key tool in the development and efficiency of modern energy systems. From old to current models, focused on sustainability and resource economy, all depend on energy services to supply society with the energy it needs. Energy advisors traditionally offered advice technical, financial advice and strategic analysis to provide tailored solutions that optimize the use of energy, its generation and applications, mainly in the business world.

Thanks to technological advances, there has been a significant advance in energy services to gradually adapt to the increasing inclusion of renewable energies, both in companies and individuals. In fact, the market is evolving at high speed and local energy communities are already developing as an effective way to collectively share and manage renewable energy. To achieve this, it is essential to have professional advice, both in the design and management of these systems.

Energy advice has become a strategic necessity for both consumers and producers and managers of energy networks or demand aggregators, not only to meet GHG (Greenhouse Gas) reduction objectives, but also to achieve the greatest efficiency. of the systems developed and, therefore, to ensure the investments made in this area.

The incorporation of Artificial Intelligence into energy consulting has opened the door to new energy service models, with advanced capabilities for predictive analysis, simulation and automated learning, which allows optimizing the decision making of all market players. , both final consumers and producers, prosumers, demand aggregators or network managers. In this context, it is necessary to have a solution that addresses these new energy needs and provides an advanced hyper-personalized advisory system, which integrates the use of Big Data and AI (Artificial Intelligence) analysis techniques in its architecture.

Therefore, it is the object of the present invention to develop a system for comprehensive energy management based on AI techniques, including Machine Learning, where it also has a solid database architecture, and means to interact with the managed systems. .

Our patent, based on a cloud data architecture and the use of AI (Artificial Intelligence) techniques, arises in response to these needs, being a unique solution on the market that can be integrated into the entire value chain of the energy sector. , from end consumers to producers, prosumers or network managers, allowing not only to optimize energy management in a personalized way but also to significantly reduce associated costs, opening a new range of possibilities for any type of user. Thanks to this innovation, our solution represents an important milestone in the evolution of energy services, establishing a new standard in advanced energy consulting.

DESCRIPTION OF THE INVENTION

The object of the present invention is included in its essentiality in the independent claim and the different embodiments are included in the dependent claims.

The purpose of the present invention is a system for comprehensive energy management based on the information provided from different devices, either in real time or through historical data, which are processed through the use of AI (Artificial Intelligence) providing personalized actions on any energy management device, such as opening and closing switches for storage media, such as batteries or similar, or for supply from the associated renewable (or non-renewable) energy network, or from the general distribution network .

The system object of the invention comprises:

- A data collection subsystem for the creation of a massive data set or Big Data. The origin of the data can be:

o Structured data that refers to measurements and numerical values that are recorded at certain times and stored in organized tables or databases

o Unstructured data that is stored in the cloud and is treated with natural language processing (NLP) and artificial intelligence (AI) techniques.

- Data lake storage subsystem. Once data has been collected, it needs to be stored efficiently and securely. This Data lake storage Subsystem is a type of data repository that allows large amounts of information from various sources to be stored in their original form, without the need to previously structure them as in a traditional relational database. In our case, this subsystem can employ a distributed data architecture, which involves storing the data on different servers to improve the scalability and performance of the system. This technique allows large amounts of data to be processed more quickly and efficiently, and improves data availability in the event of server failures. Furthermore, by distributing the workload across different servers, queries and operations in the system can be processed in parallel, significantly improving system performance.

- Artificial Intelligence Subsystem. It is a set of machine learning algorithms that analyze the captured data and generate personalized actions for each user. The AI system can use natural language processing techniques, neural networks, decision trees, among others. The goal is for the AI system to be able to detect patterns in the data and offer precise and personalized actions for each type of user.

This Artificial Intelligence subsystem includes:

or a data analysis platform with AI and compatible with distributed databases. This platform implemented in least one server provided with processing means carries out the following actions:

• Data processing: Various data analysis techniques such as machine learning, deep learning, natural language processing and statistical analysis are used to extract useful insights from the data.

• Creating AI models: Machine learning and deep learning tools are used to create custom AI models that can be used to predict, classify, or cluster data.

• AI model evaluation: Created AI models are evaluated using cross-validation and other techniques to ensure that they are producing accurate and reliable results.

• Deployment of AI models: AI models are deployed in a production environment so that they can be used by end users.

• Monitoring and maintenance of AI models: AI models are constantly monitored to ensure they are producing accurate and up-to-date results, and adjustments and improvements are made as necessary.

Internal APIs, which are application programming interfaces that allow users to access and use the platform's analysis tools, as well as interact with AI algorithms and machine learning models to create customized solutions and automate processes. analysis of data. These APIs allow the integration of the data analysis platform with other applications and systems, increasing efficiency and data processing capacity.

A cloud-native data storage platform designed to store large amounts of historical and current data, with the objective of carrying out analysis and generating reports of all kinds. These SQL databases are organized to support efficient querying and analysis of data and provide tools for data visualization and report generation.

- Subsystem of interaction agents, are those that interact with users to offer the actions generated by the AI system, which may be any of the following or combinations thereof.

o Recommendation agents: these are systems that use machine learning techniques and Big Data analysis to analyze consumption patterns and offer personalized energy management actions for each type of user. These agents can be both rule-based and recommendation systems based on collaborative filtering.

o Natural language processing (NLP) agents: these are systems that use NLP techniques to analyze large volumes of unstructured data, such as customer opinions, surveys, technical reports, regulations, user manuals and any other data from which to extract meaningful information for different types of users. These agents can be used to offer personalized actions, problem resolution or sentiment analysis to improve customer satisfaction, among others. o Chatbots: are systems that use natural language processing and machine learning techniques to offer a conversational user interface for personalized energy management. Chatbots can be used to offer personalized actions, energy consumption analysis and customer support among other functions.

o Optimization agents: these are systems that use mathematical optimization techniques and Big Data analysis to optimize energy management in real time. These agents can be used to offer personalized management actions energy, such as the optimization of energy production, consumption and storage, either through optimizations of existing equipment or through the inclusion of new equipment specifically sized for each customer profile (end user, demand aggregator, energy community... .). o Monitoring agents: these are systems that use Big Data analysis techniques to monitor energy consumption and detect anomalies or problems. These agents can be used to offer personalized actions to improve energy efficiency and prevent energy problems. They can be integrated into different platforms for real-time consultation, reporting or incident management, among other functions.

- Interface Subsystem (Frontend) to interact directly with the user, providing a graphical interface and functionalities so that the user can also interact with the system. This is the presentation layer of the software, which includes visual elements such as buttons, menus, forms and other interactive elements that allow the user to perform tasks and consume the data provided by the backend. The frontend communicates with the backend to request information and process the user's actions, and then presents the results to the user.

- An external application programming subsystem that allows communication and the exchange of information between systems or services of different organizations or companies. That is, it is an API (application programming interface) that allows a company to offer its services or products through a common interface, which can be used by other external systems or applications.

- A control signal generation subsystem for sending control signals that act on different devices such as switches, disconnectors, relays, temperature controllers, control valves, sensors, speed controllers, PLCs (Programmable Logic Controllers), frequency converters, lighting control systems and any other device that can act on an energy distribution network, an energy generation system (renewable or non-renewable), as well as the internal energy networks of any type of consumer, proceeding to activate/deactivate it based on the optimization strategy developed by the system.

All the indicated subsystems are implemented in processing means associated with a computer, server, computer or any other technical means provided with processing means.

Structured data of some of the following or combinations thereof:

• Smart meters: they are electronic devices that connect to a facility's electrical network and measure energy consumption and/or production in real time.

• Energy monitoring systems: used to collect data on energy consumption/production in a building or system. They can include sensors for temperature, air flow, and lighting level, among many others.

• Telemanagement systems: these are remote energy management systems, which allow the consumption or production of energy to be controlled and monitored remotely. They are used in buildings and production plants and can control lighting, air conditioning, hot and cold water supply, among others.

• Submetering meters: they are devices that are installed in the energy circuits of the equipment and systems of a building or facility, to measure energy consumption in real time. They can serve as a complement of information and/or verification of the information obtained from energy meters and allow the information available for each type of user to be customized in greater detail.

• Energy bills in structured format: used to collect billing data from energy companies and analyze consumption patterns, detect errors and as a basis for optimizing new contracts.

• Monitoring systems for energy storage equipment or for solar/wind/biomass energy installations: used to collect data on the production and storage of energy from renewable sources and analyze the efficiency of the systems involved.

• Electrical distribution network monitoring systems: used to collect data on the status of electrical networks, such as voltage and current. The data is used to optimize network operation and to detect and prevent possible system failures, among others.

The sources of unstructured data can be any of the following or combinations thereof:

- Free text data, such as comments in customer satisfaction surveys or reports of failures in the electrical network.

- Image data, such as photographs or thermal maps that show the distribution of energy in a building.

- Audio data, such as recordings of telephone conversations between a customer and their energy supplier.

- Video data, such as recordings from surveillance cameras that monitor energy consumption in real time or thermographic images used for facility maintenance.

- Social media data, such as posts and comments related to energy and the environment on platforms such as Twitter or Instagram.

- Energy bills from different suppliers in unstructured format.

- Data on energy regulations, legislation, catalogs of services and products, user manuals, regulatory bases for sectoral public aid...etc.

The Interface Subsystem (Frontend) to interact directly with the user can be any of the following or combinations thereof:

- A web application: which allows users to access personalized actions, monitor their energy consumption, obtain personalized reports and view graphs, reports and statistics of their energy performance. The web application could include interactive tools so that users can experiment with different scenarios and see how they affect their energy consumption, for example with savings calculators or sizing of self-consumption and/or storage facilities.

- A mobile application: which allows users to access personalized actions and monitor their energy consumption in real time as well as any other utility present on the web pages. The mobile application could include push notifications to alert users of energy consumption problems and offer advice to improve energy efficiency, as well as allow action on different devices to optimize energy management.

- A control panel: which shows the energy performance of the installation in real time and provides an overview of personalized actions. The dashboard could include interactive graphs, reports and visualizations so that different types of users can see how they are using energy in real time.

- A chatbot tool: that allows users to ask questions and receive personalized performances through a conversation. The chatbot could use natural language processing to understand users' questions and provide useful answers in real time based on personalized information for each type of user. - A touch screen (or normal): which can be installed in a visible location to display real-time graphs, reports and statistics on the energy performance or carbon footprint of a facility. The touch screen could include interactive tools for users to Users can experiment with different scenarios and see how they affect their energy consumption or emissions.

- Any of the interfaces described, without prejudice to others that could be developed, would have the capacity to interact with different types of users, but also to act on energy control devices based on these interactions and the data accumulated in the system.

Thanks to the system described, among other functionalities, the following are achieved:

- Energy demand/production prediction and management: AI can analyze customer's historical energy consumption/production data as well as real-time production/consumption load curves to predict future energy demand/production, thereby allows a more efficient and sustainable management of energy resources, for example, acting on the different existing control devices based on this data.

- Energy bill analysis and fraud detection: AI can analyze customer energy bills to detect possible errors or improper charges, and can also analyze customer energy billing data to detect possible fraud, such as energy theft or the manipulation of meters, which allows for more efficient and fair management of energy resources, for example, for an energy distributor that detects "illegal connections" and decides to cut off the affected supply.

- Optimization of energy storage and generation, minimizing production costs and maximizing energy efficiency. This optimization can be applied to both electricity, gas or green hydrogen, and can act through control devices in loading or discharging from storage systems automatically.

- Electricity contracting and demand aggregation: AI can act as a demand aggregator, analyzing the energy consumption data of a group of companies or an energy community to negotiate electricity contracts on behalf of its members, maximizing efficiency and reducing electricity costs for all users. It can also analyze production and/or storage data to manage energy demand and maximize the use of renewable energy sources through interaction with the control devices of the different technologies.

- Analysis of environmental impact and profitability of renewable energy projects: AI can analyze client energy consumption data and sectoral consumption profiles to evaluate the environmental impact of energy consumption and offer actions to reduce the carbon footprint. Additionally, you can analyze customer energy production/storage data and market prices to evaluate the profitability of renewable energy projects, such as installing solar panels or wind turbines. - Analysis of biomass, green hydrogen or natural gas data: AI can analyze the client's energy consumption data to design energy storage systems that optimize collection costs and manage, through control devices, the operation of consumer systems of energy.

- Energy storage optimization: AI can analyze customer energy production and consumption data and market prices to optimize, through device control devices, energy storage, minimizing production costs and maximizing revenue.

- Energy efficiency actions in buildings and communities: AI can analyze the energy consumption data of the client's buildings and communities to optimize energy management, offering personalized actions to reduce energy consumption and improve energy efficiency, acting, by For example, on control devices that manage temperature, humidity or energy consumption. - Renewable energy production prediction systems: AI can analyze renewable energy production data and data meteorological data to predict short- and long-term renewable energy production, allowing for more efficient energy management and the integration of renewable energy sources into the electrical grid.

- EV Charging Optimization: AI can analyze customer EV charging data and market prices to optimize EV charging, minimizing energy costs and maximizing efficiency. This system can, for example, manage control devices that act on the instantaneous power during electricity loads, distributing the loads equally or activating complementary energy systems such as batteries.

- Energy microgrid management systems: AI can act as an energy microgrid management system, analyzing the network's energy production and consumption data and optimizing energy management in the network in real time at through control devices and sensors.

- Personalized energy actions: AI can analyze customer energy consumption data and offer personalized actions to reduce energy consumption and improve energy efficiency. Also to optimize the contracting of services or products related to energy consumption or production based on the possibilities offered by being able to act on consumption, production and/or storage control devices based on the particular data of an installation.

- Optimization of energy management in smart buildings: AI can analyze the energy consumption data of smart buildings, as well as temperatures, humidity percentages, luminosity... etc. and optimize the management of these parameters, offering personalized actions to reduce energy consumption and improve energy efficiency.

- Integration of climate and weather data to optimize energy management: AI can analyze climate and weather data to predict energy production and consumption and optimize energy management in real time, allowing the integration of renewable energy sources into the electrical grid.

- Detection and early warning of extreme weather events: AI can analyze climate and weather data to detect and warn early about extreme weather events that may affect energy production, storage or supply, allowing for more efficient and safer management of energy resources.

- Prediction and prevention of power grid failures: AI can analyze power grid data in real time to predict possible power failures or interruptions and take measures to prevent them before they occur, allowing for more efficient management and security of energy resources.

- Analysis of energy consumption data at the city or country level: AI can analyze energy consumption data at the city or country level to offer large-scale solutions to improve energy efficiency and sustainability, allowing for more efficient management and responsible for energy resources globally.

Unless otherwise indicated, all technical and scientific elements used herein have the meaning normally understood by a person skilled in the art to which this invention belongs. Procedures and materials similar or equivalent to those described herein may be used in the practice of the present invention.

Throughout the description and claims the word "comprises" and its variants are not intended to exclude other technical characteristics, additives, components or steps. For those skilled in the art, other objects, advantages and characteristics of the invention will emerge in part of the description and part of the practice of the invention.

EXPLANATION OF THE FIGURES

To complement the description that is being made and in order to help a better understanding of the characteristics of the invention, in accordance with a preferred example of its practical implementation, a set of drawings is attached as an integral part of said description. where, for illustrative and non-limiting purposes, the following has been represented.

In Figure 1, we can see a general scheme of a possible application of energy management with different user levels.

Figure 2 shows a global diagram of the elements that are part of the system and the interaction between them.

In figure 3, we can see a representation of a distributed database.

PREFERRED EMBODIMENT OF THE INVENTION

In view of the figures, a preferred embodiment of the proposed invention is described below.

In Figure 1 we can see a system for comprehensive energy management (100) like the one that is the object of the invention, applied to different types of users. That is: an energy distribution network (110), renewable energy generation means (120) associated with energy storage means (130), where both elements supply energy to an energy community or company (150). ) through switches (140) controlled from an energy management system (100).

Associated with each of the consumption centers, which may well be neighborhood communities, individual users or companies, there may also be Additionally, opening and closing switches (140) commanded by the system in order to achieve optimization of energy consumption. These switches can act on the existing generation, consumption and/or accumulation elements based on the information collected by the system.

In figure 2 we can see that the energy management system includes:

- A data collection subsystem (A) for the creation of a massive data set or Big Data. The origin of the data can be:

o Structured data (3) that refers to measurements and numerical values that are recorded at certain times and stored in organized tables or databases and transmitted through various communication technologies (4), such as the Internet of Things ( IoT), 5G or short-range communication systems such as Bluetooth or Zigbee.

o Unstructured data (1) that is stored in the cloud and is processed with natural language processing (NLP) and artificial intelligence (AI) techniques that are transmitted through communication technologies (2) such as FTP, SFTP, API, SDX, Microsoft One Drive, Amazon S3, etc.

- A Data lake storage subsystem (B). Once data has been collected, it needs to be stored efficiently and securely. This Data lake storage Subsystem is a type of data repository that allows large amounts of information from various sources to be stored in their original form, without the need to previously structure them as in a traditional relational database. In our case, this subsystem can employ a distributed data architecture (5), which involves storing the data on different servers to improve the scalability and performance of the system. This technique allows large amounts of data to be processed more quickly and efficiently, and improves data availability in the event of server failures. Additionally, by distributing the workload across different servers, queries and Operations in the system can be processed in parallel, significantly improving system performance.

- Artificial Intelligence Subsystem (C). It is a set of machine learning algorithms that analyze the captured data and generate personalized actions for each user. The AI system can use natural language processing techniques, neural networks, decision trees, among others. The goal is for the AI system to be able to detect patterns in the data and offer precise and personalized actions for each user.

This Artificial Intelligence subsystem includes:

or an AI data analytics platform (6) supporting distributed and non-distributed databases. This platform, implemented on at least one server provided with processing means, carries out the following actions based on its relationships with the different systems (5)(7)(8)(9)(11):

• Data processing: Various data analysis techniques such as machine learning, deep learning, natural language processing and statistical analysis are used to extract useful insights from the data.

• Creating AI models: Machine learning and deep learning tools are used to create custom AI models that can be used to predict, classify, or cluster data.

• AI model evaluation: Created AI models are evaluated using cross-validation and other techniques to ensure that they are producing accurate and reliable results.

• Deployment of AI models: AI models are deployed in a production environment so that they can be used by end users.

• Monitoring and maintenance of AI models: AI models are constantly monitored to ensure they are producing accurate and up-to-date results, and adjustments and improvements are made as necessary.

o Internal APIs (7), which are application programming interfaces that allow users to access and use the platform's analysis tools (6), as well as interact with AI algorithms and machine learning models to create customized solutions and automate data analysis processes. These APIs allow the integration of the data analysis platform with other applications and systems, such as interaction agents (D), which increases efficiency and data processing capacity.

o A cloud-native data storage platform (8) designed to store large amounts of historical and current data, with the aim of performing analyzes and generating reports of all types. SQL databases are organized to support querying and analysis of data efficiently and provide tools for data visualization and report generation through interaction agents (D).

- Interaction agent subsystem (D), is responsible for interacting with users and offering the actions generated by the AI system. It is composed of different types of agents that may be some of the following agents (9), or combinations of these, to offer personalized and efficient energy management:

o Recommendation agents: these are systems that use machine learning techniques and Big Data analysis to analyze consumption patterns and offer personalized energy management actions. These agents can be both rule-based and recommendation systems based on collaborative filtering.

o Natural language processing (NLP) agents: These are systems that use NLP techniques to analyze large volumes of unstructured data, such as customer reviews, and extract useful information from this data. These agents can be used to offer personalized performances and sentiment analysis to improve customer satisfaction.

o Chatbots: are systems that use natural language processing and machine learning techniques to offer a conversational user interface for personalized energy management. Chatbots can be used to offer personalized actions, energy consumption analysis and customer support.

o Optimization agents: these are systems that use mathematical optimization techniques and Big Data analysis to optimize energy management in real time. These agents can be used to offer personalized energy management actions, such as optimizing energy production and storage, and reducing energy costs. o Monitoring agents: these are systems that use Big Data analysis techniques to monitor energy consumption and detect anomalies or problems. These agents can be used to offer personalized actions to improve energy efficiency and prevent energy problems. Any of the agents described, without prejudice to others that could be developed, would have the capacity to interact with different types of users, but also to act on devices. energetic control (13) from these interactions. Furthermore, the use of different types of agents can be combined to offer more complete and efficient energy management.

In order to improve the efficiency of the subsystem and thanks to the interaction with the data analysis platform with AI (6), predictive and prescriptive analysis techniques can be incorporated to anticipate needs and offer solutions before problems arise. Reinforcement learning (Machine Learning) can also be used to improve the performance of interaction agents as they interact with users and receive feedback.

- Interface Subsystem (E) (Frontend) to interact directly with the user, providing a graphical interface and functionalities (10) so that the user can also interact with the system, as well as customize the appearance and behavior of the interface to adapt to your preferences and needs. This is the presentation layer of the software, which allows integration with the interaction agents described in subsystem (D) to provide a more complete and personalized user experience, which includes visual elements such as buttons, menus, forms and others. interactive elements that allow the user to perform tasks and consume the data provided by the Data lake storage Subsystem (B). This interface can communicate with the Data lake storage (B) to request information and process user actions, and then present the results to the user. Any of the interfaces described, without prejudice to others that could be developed, would have the capacity to interact with different types of users, but also to act on energy control devices (G) based on these interactions.

- An external application programming subsystem (F) that allows communication and exchange of information between systems or services of different organizations or companies. That is, it is an API (application programming interface) (11) that allows a company to offer its services or products through a common interface, which can be used by other systems or external applications (12) that, in turn, At the same time, they can provide feedback to the comprehensive energy management system by providing new information in the form of structured (3) or unstructured (1) data. The outsourcing API can receive information directly from the data analysis platform with AI (6) and the cloud-native data storage platform (8).

- A control signal generation subsystem (G) for sending control signals (13) that act on different devices such as switches, disconnectors, relays, temperature controllers, control valves, sensors, speed controllers, PLCs ( Programmable Logic Controllers), frequency converters, lighting control systems and any other device that can act on an energy distribution network, an energy generation system (renewable or non-renewable), as well as internal energy networks of any type. of consumer, proceeding to its activation/deactivation based on the optimization strategy developed by the system. This subsystem can receive information from different sources (2)(4)(D)(E), interacting with them and/or feeding back the entire system by providing new information in the form of structured (3) or unstructured data (1).

From unstructured data sources (1), transmitted through various communication technologies (2), a data extraction, loading and transformation process (ELT: Extraction, Load and Transformation) takes place, by which raw data from a source server to the database subsystem (B) or destination server, where the information is prepared for subsequent use.

On the other hand, from the sources of structured data (3), transmitted through various communication technologies (4), in addition to the ELT mentioned above, an ETL (Extract, Transform, Load) process also takes place that occurs when The data transformation takes place directly at the target, that is, in the AI data analysis platform (6) where it will be transformed according to the needs of the system.

Figure 2 shows a possible representation of the Data lake storage Subsystem (B) using a distributed database architecture, which involves storing data on different servers to improve the scalability and performance of the system. This technique allows large amounts of data to be processed more quickly and efficiently, and improves data availability in the event of server failures. Furthermore, by distributing the workload across different servers, queries and operations in the system can be processed in parallel, significantly improving system performance.

Specifically, Figure 3 shows a Spark cluster architecture, which is a distributed architecture that uses parallel processing nodes to process large amounts of data. The architecture is made up of a master node (14) that coordinates the work and several slave nodes (15) that process the data. The data is stored in a distributed memory called RDD, which is divided into small partitions and distributed across different nodes. Data processing in Spark is done through transformational operations and actions, which are executed in parallel on the worker nodes. Spark's cluster architecture is highly scalable and fault-tolerant, making it ideal for processing large data sets, which is a desirable element for the system to operate efficiently.

Having sufficiently described the nature of the present invention, as well as the way of putting it into practice, it is stated that, within its essentiality, it may be put into practice in other embodiments that differ in detail from that indicated by way of example. , and to which the protection sought will also reach, as long as it does not alter, change or modify its fundamental principle.

Claims (1)

1.- System for comprehensive energy management characterized because it includes: - A data collection subsystem (A) for the creation of a massive data set or Big Data, where the origin of the data is from: o Structured data (3) that refers to measurements and numerical values that are recorded at certain times and stored in tables or organized databases, o Unstructured data (1) that is stored in the cloud and is treated with techniques natural language processing (NLP) and artificial intelligence (AI), - Data lake storage subsystem (B), which can use a distributed and/or non-distributed data architecture (5) in connection with the data collection subsystem (A) - Artificial Intelligence Subsystem (C) that uses natural language processing techniques, neural networks, decision trees, among others, which receives data from the data collection subsystem (A) and the Data lake storage Subsystem (B), where this Artificial Intelligence subsystem includes: or a data analysis platform with AI (6), compatible with distributed databases that is implemented on at least one server provided with processing means, o Internal APIs (7), which are application programming interfaces that allow users to access and use the platform's analysis tools (6), as well as interact with AI algorithms and machine learning models to create customized solutions and automate data analysis processes, o A cloud-native data storage platform (8) designed to store large amounts of historical data and current, with the aim of carrying out analysis and generating reports of all kinds, - Interaction agent subsystem (D), are those that interact with users to offer the actions generated by the AI subsystem (C) with which it is in connection, which may be some of the following agents (9) or combinations of these: o Recommendation agents: these are systems that use machine learning techniques and Big Data analysis to analyze consumption patterns and offer personalized energy management actions where these agents are based both on rules and based on recommendation systems based on collaborative filtering. . o Natural language processing (NLP) agents: o Chatbots: are systems that use natural language processing and machine learning techniques to offer a conversational user interface for personalized energy management. o Optimization agents: these are systems that use mathematical optimization techniques and Big Data analysis to optimize energy management in real time. o Monitoring agents: these are systems that use Big Data analysis techniques to monitor energy consumption and detect anomalies or problems. Any of the agents described, without prejudice to others that could be developed, would have the capacity to interact with different types of users, but also to act on energy control devices (G) based on these interactions. - Interface Subsystem (E) or Frontend to interact directly with the user, providing a graphical interface and functionalities (10) so that the user can also interact with the system, where this interface communicates with the distributed database subsystem (B ) and with the interaction agent subsystem (D) to request information and process the user's actions, and then present the results to the user. Any of the interfaces described, without prejudice to others that could be developed, would have the capacity to interact with different types of users, but also to act on energy control devices (G) based on these interactions. - An external application programming subsystem (F) that allows communication and the exchange of information between systems or services of different organizations or companies and that is in connection with the interface subsystem (E) and the AI subsystem (C) . - A control signal generation subsystem (G) for sending control signals (13) that act on different devices such as switches, disconnectors, relays, temperature controllers, control valves, sensors, speed controllers, PLCs ( Programmable Logic Controllers), frequency converters, lighting control systems and any other device that can act on an energy distribution network, an energy generation system (renewable or non-renewable), as well as internal energy networks of any type. of consumer, proceeding to its activation/deactivation based on the optimization strategy developed by the system, which is in connection with the interaction agents (D), Interface (E) and with any control device with the capacity to collect data (TO). 2. - System for comprehensive energy management according to claim 1 characterized in that the structured data (3) are transmitted through various communication technologies (4), such as the Internet of Things (IoT), 5G or communication systems short range such as Bluetooth or Zigbee. 3. - System for comprehensive energy management according to claim 1 characterized in that the unstructured data (1) is transmitted through communication technologies (2) such as FTP, SFTP, API, SDX, Microsoft One Drive, Amazon S3, etc. . 4 - System for comprehensive energy management according to claim 2 or 3 characterized in that a data analysis platform with AI (6) is implemented in at least one server provided with processing means that carries out the following actions: • Data processing: Various data analysis techniques such as machine learning, deep learning, natural language processing and statistical analysis are used to extract useful insights from the data. • Creating AI models: Machine learning and deep learning tools are used to create custom AI models that can be used to predict, classify, or cluster data. • AI model evaluation: Created AI models are evaluated using cross-validation and other techniques to ensure that they are producing accurate and reliable results. • Deployment of AI models: AI models are deployed in a production environment so that they can be used by end users. • Monitoring and maintenance of AI models: AI models are constantly monitored to ensure they are producing accurate and up-to-date results, and adjustments and improvements are made as necessary. 5. - System for comprehensive energy management according to any of the previous claims, characterized in that the Data lake storage subsystem (B) can use a distributed architecture, which comprises a master node (14) that coordinates the work and several slave nodes (15). that process the data. 6. - System for comprehensive energy management according to claim 2 and 3 characterized in that from the structured data sources (3) and transmitted Through various communication technologies (4), as well as from unstructured data sources (1) and transmitted through communication technologies (2), a process of data extraction, loading and transformation (ELT) takes place ( Extraction, Load and Transformation), where the ELT process is a data integration process to transfer raw data from a source server to the database subsystem (B) on a destination server and then prepare the information for subsequent uses . 7. - System for comprehensive energy management according to claim 3 characterized in that an ETL (Extract, Transform, Load) process takes place from the sources of structured data (3) and transmitted through communication technologies (4), referring to a data integration process in which the transformation takes place on an intermediate server before it is uploaded to the destination. 8. - Procedure for comprehensive energy management according to the system according to any of claims 1 to 7, characterized in that it comprises the stages of: - Capture of data from: - structured data such as: smart meters, energy monitoring systems, remote management systems, submetering meters, energy bill analysis, renewable energy monitoring systems, electrical distribution network monitoring systems. - Unstructured data from some of the following or combinations thereof: Free text data, image data, audio data, video data, social network data. - Storage of captured data in distributed and/or non-distributed architecture databases. - Data analysis with an Artificial Intelligence platform where the tasks of: o Data processing: Various data analysis techniques, such as machine learning, deep learning, natural language processing, and statistical analysis, are used to extract useful insights from the data. o Creating AI models: Machine learning and deep learning tools are used to create custom AI models that can be used to predict, classify, or group data. o Evaluation of AI models: AI models created are evaluated using cross-validation and other techniques to ensure that they are producing accurate and reliable results. o Deployment of AI models: AI models are deployed in a production environment so that they can be used by end users. - Presentation of results through interaction agents based on AI - Sending action orders to a control signal generation subsystem to act on different devices such as switches, disconnectors, relays, temperature controllers, control valves, sensors, speed controllers, PLCs (Programmable Logic Controllers), converters frequency, lighting control systems and any other device that can act on an energy distribution network, an energy generation system (renewable or non-renewable), accumulation as well as the internal energy networks of any type of consumer, proceeding to activate/deactivate it based on the optimization strategy developed by the system.