A zero-carbon park intelligent system based on AI big model

Abstract

The present invention provides a zero-carbon park intelligent system based on an AI big model, which belongs to the field of energy Internet of Things. The system includes: an energy generation system, including a PVT power generation and heating system, a wind power generation system, an electrochemical and electrothermal phase change energy storage system. The PVT power generation and heating system adopts a transformer big model, and the wind power generation system uses the big model's Encode module and Decode module to predict the power generation; the electrochemical and electrothermal phase change energy storage systems adopt a big model to optimize the charging and discharging strategy; the energy load system and the park energy consumption monitoring system, the AI ​​big model energy operation and maintenance system adjusts the equipment to release and transmit energy according to the big model's predicted power load curve, power consumption, priority and time. The transformer big model realizes the coordinated optimization of each unit and demonstrates its unique value in many aspects such as distribution network prediction and optimization, fault diagnosis and predictive maintenance, and distributed energy management.

Description

Technical Field

[0001] The present invention relates to the technical field of energy Internet of Things, and in particular to a zero-carbon park intelligent system based on an AI large model. Background Art

[0002] With the rapid development of the Internet of Things (IoT) and distributed renewable energy in recent years, especially the unprecedented technological advancements in distributed renewable energy power generation, manufacturing plants in the industrial and commercial sectors, particularly those in the fields of industrial interconnection and energy development, are placing increasing demands on renewable energy flow distribution to meet people's daily needs and energy requirements. With the continuous advancement of technology, the application of large AI models in energy management will become increasingly widespread, bringing revolutionary changes to the energy industry.

[0003] At present, the traditional park energy generation on the market is relatively simple, the energy dispatching Internet of Things is relatively backward, and it has not yet formed a scale. This cannot solve the advantages of the factory's distributed energy diversified power generation, and cannot play the benefits brought by the comprehensive utilization of distributed energy. The system is relatively simple, resources cannot be flexibly allocated and mobilized, and comprehensive energy utilization cannot be interconnected and synchronized. These links have added many problems to the entire energy industry, leading to uneven distribution of new energy resources in factories and energy waste, etc., so that the entire system does not maximize resource sharing and utilization. Summary of the Invention

[0004] The purpose of an embodiment of the present invention is to provide a zero-carbon park intelligent system based on an AI large model, which is used to fully or at least partially solve the problems existing in the above-mentioned prior art, such as the relatively single factory energy generation and the single management system leading to uneven distribution of new energy resources and energy waste, resulting in the lack of maximized resource sharing and utilization in the entire system.

[0005] To achieve the above objectives, an embodiment of the present invention provides a zero-carbon park intelligent system based on an AI big model. The system is embedded with a transformer big model and includes:

[0006] An energy generation system, comprising a PVT power generation and heating system, a photovoltaic power generation system, a wind power generation system, a hydrogen power generation system, an electrochemical energy storage system, and an electrothermal phase change energy storage system, wherein the PVT power generation and heating system uses a transformer large model to predict power generation based on historical meteorological data and power generation data; the wind power generation system uses the Encode module and the Decode module of the transformer large model to predict wind power generation within a period of time based on input historical meteorological data and power generation data; the electrochemical energy storage system is used to store energy during periods of abundant wind and light resources, and to release energy to supply power during periods of resource scarcity and peak electricity consumption; the electrothermal phase change energy storage system is used to convert other forms of energy into thermal energy, and to store the thermal energy through a heat storage medium, and to extract the stored heat for use through heat exchange when needed;

[0007] Energy load system, the energy load system includes a charging pile system and a park factory equipment load system, wherein the charging pile system includes AC charging piles, DC charging piles, and supercharged liquid cooling charging piles, and the park factory equipment load system includes at least central air conditioning load, air compressor load, air energy load, Ma load, and power consumption of heavy-load machines in the factory;

[0008] The park energy consumption supervision system includes an energy consumption monitoring and metering system, an energy auxiliary system, and an industrial Internet system. The energy consumption monitoring and metering system is used to collect, measure, and calculate and analyze the park's energy consumption in real time to achieve energy consumption statistics, energy consumption management and assessment, energy efficiency evaluation, energy quota, and energy-saving services. The energy auxiliary system is used to monitor transformer winding temperature, switchgear busbar temperature, switchgear partial discharge monitoring, and distribution cabinet electrical parameter monitoring. The industrial Internet system is used to comprehensively monitor platform services and data link security, provide time series data analysis, and conduct queries and aggregated analysis for any historical time period.

[0009] The AI ​​large-scale model energy operation and maintenance system includes a park energy dispatching system and an operation and maintenance inspection system. The park energy dispatching system is used to predict the power load curve for a period of time based on the AI ​​large-scale model, and adjust the equipment according to the power consumption in each time period of each day and the pre-set priority and time, so that the equipment layer responds to the energy adjustment instructions issued by the system, realizes the energy release of new energy photovoltaic, wind energy and energy storage media, energy transmission on the incoming line side of the park distribution network, and uses the released energy for other loads.

[0010] Optionally, the photovoltaic power generation system includes: a photovoltaic component, a controller and an inverter, wherein the controller controls the photovoltaic component to convert light energy into direct current, and controls the inverter to convert the direct current into alternating current.

[0011] Optionally, the wind power generation system includes:

[0012] A wind turbine component, comprising at least a wind turbine generator set, a tower, a foundation and cables, wherein the wind turbine component is used to capture wind energy and convert the wind energy into alternating electrical energy;

[0013] A grid-connected control component, comprising at least a rectifier module, a grid-connected controller, a load shed and cables, and configured to control the normal operation of wind turbine components;

[0014] An inverter component, comprising at least a grid-connected inverter and a cable, wherein the inverter component is used to invert the direct current output by the grid-connected control component into alternating current and feed the energy into the grid;

[0015] The unloading component is used to control the wind turbine to brake and stop, so as to ensure the safety of the wind turbine under abnormal working conditions.

[0016] Optionally, the hydrogen power supply system utilizes photovoltaic electricity generated during the daytime peak hours and valley electricity at night to electrolyze water to produce hydrogen, and generates electricity through fuel cells during peak hours of power consumption in the grid.

[0017] Optionally, the electrochemical energy storage system includes:

[0018] a battery pack for DC charging or discharging;

[0019] Energy storage converter, used to control the charging and discharging process of the battery pack and perform AC-DC conversion;

[0020] Battery management system, used to monitor the status information of the battery pack;

[0021] Energy management system, responsible for data collection, network monitoring and energy scheduling.

[0022] Optionally, the energy auxiliary system includes:

[0023] Environmental monitoring subsystem, used to monitor temperature and humidity, gas, noise, dust, water leakage and water level;

[0024] Power equipment monitoring subsystem, used to monitor the operating status of air conditioners, dehumidifiers, fans, lighting, water pumps and fresh air fans;

[0025] Security monitoring subsystem, used for monitoring smoke, infrared, electronic fence, access control and video.

[0026] Optionally, the industrial Internet of Things system adopts a PaaS platform architecture and is configured with:

[0027] Rules engine, used to implement data repair, cleaning, calculation and analysis, circulation, data alarm and multi-scenario linkage;

[0028] Data subscription and open API interface to support third-party application development and integration;

[0029] Multiple device protocol parsing drivers are used to enable access of multiple devices through gateway access;

[0030] Configure alarm rules on the interface for alarm management.

[0031] Optionally, the park energy dispatching system includes a power supply dispatching system that adopts an economic dispatching algorithm and a big data dispatching algorithm for power supply dispatching. When the economic dispatching algorithm is adopted, the microgrid management is completed in combination with time-of-use electricity prices, new energy output patterns, and preset generation, distribution, and use logic; when the big data dispatching algorithm is adopted, on the basis of economic dispatch, the microgrid dispatching granularity is further refined through data aggregation and big data analysis to realize photovoltaic power generation prediction and load prediction.

[0032] Optionally, train a large transformer model according to the following formula:

[0033]

[0034] Where, Represents new energy power generation data, represents energy consumption data, represents the energy data purchased from the park, λ represents the price of new energy power generation, k represents the consumption price, δ represents the price of purchased energy, and F * Represents the integration cost.

[0035] Optionally, power generation can be predicted using the following formula:

[0036]

[0037] Where α, β, ε, η, φ, and ω represent the power generation duration coefficients of each energy time period, and P V 、P T Indicates the power generation of the PVT power generation and heating system, P W Indicates wind power generation power, P ES1 Indicates the electrochemical energy storage release power, P ES2 represents the released power of the phase change energy storage system, P S Indicates the power system input power.

[0038] Through the above technical solution, the management system can reasonably allocate the entire energy resources through the transformer large model, and through the supervision and coordination of new energy power generation and comprehensive utilization of new energy in the industrial Internet, it helps factories meet various energy needs, ensure high penetration of new energy, safe and economic operation of various energy networks, and reliability and comprehensive utilization efficiency of various energy supplies.

[0039] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following detailed description, they are used to explain the embodiments of the present invention, but do not constitute a limitation of the embodiments of the present invention. In the accompanying drawings:

[0041] Figure 1 This is a structural diagram of a zero-carbon park intelligent system based on an AI large model provided by an embodiment of the present invention;

[0042] Figure 2 This is a schematic diagram of the architecture of a zero-carbon park intelligent system based on an AI big model provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The following describes the specific implementation of the embodiment of the present invention in detail with reference to the accompanying drawings. It should be understood that the specific implementation described herein is only used to illustrate and explain the embodiment of the present invention and is not used to limit the embodiment of the present invention.

[0044] Because photovoltaic power generation is daytime and wind power generation is conditionally constrained, we utilize energy storage and phase change energy storage to store unused energy. This stored energy can then be utilized at night or during peak hours when electricity prices are high. Based on daily electricity consumption, large AI models can also predict the load curve for the immediate future. This allows for scheduling based on daily electricity consumption during each time period. For example, if there is no sunshine or wind in the coming days, our energy storage can purchase electricity from the State Grid, store it, and buy it at a low price. If there is any unused electricity, it can be sold externally at a higher price.

[0045] See Figure 1 The figure shows a schematic diagram of the structure of a zero-carbon park intelligent system based on an AI big model provided by an embodiment of the present invention. The system is embedded with a transformer big model, including:

[0046] An energy generation system, comprising a PVT power generation and heating system, a photovoltaic power generation system, a wind power generation system, a hydrogen power generation system, an electrochemical energy storage system, and an electrothermal phase change energy storage system, wherein the PVT power generation and heating system uses a transformer large model to predict power generation based on historical meteorological data and power generation data; the wind power generation system uses the Encode module and the Decode module of the transformer large model to predict wind power generation within a period of time based on input historical meteorological data and power generation data; the electrochemical energy storage system is used to store energy during periods of abundant wind and light resources, and to release energy to supply power during periods of resource scarcity and peak electricity consumption; the electrothermal phase change energy storage system is used to convert other forms of energy into thermal energy, and to store the thermal energy through a heat storage medium, and to extract the stored heat for use through heat exchange when needed;

[0047] Energy load system, the energy load system includes a charging pile system and a park factory equipment load system, wherein the charging pile system includes AC charging piles, DC charging piles, and supercharged liquid cooling charging piles, and the park factory equipment load system includes at least central air conditioning load, air compressor load, air energy load, Ma load, and power consumption of heavy-load machines in the factory;

[0048] The park energy consumption supervision system includes an energy consumption monitoring and metering system, an energy auxiliary system, and an industrial Internet system. The energy consumption monitoring and metering system is used to collect, measure, and calculate and analyze the park's energy consumption in real time to achieve energy consumption statistics, energy consumption management and assessment, energy efficiency evaluation, energy quota, and energy-saving services. The energy auxiliary system is used to monitor transformer winding temperature, switchgear busbar temperature, switchgear partial discharge monitoring, and distribution cabinet electrical parameter monitoring. The industrial Internet system is used to comprehensively monitor platform services and data link security, provide time series data analysis, and conduct queries and aggregated analysis for any historical time period.

[0049] The AI ​​large-scale model energy operation and maintenance system includes a park energy dispatching system and an operation and maintenance inspection system. The park energy dispatching system is used to predict the power load curve for a period of time based on the AI ​​large-scale model, and adjust the equipment according to the power consumption in each time period of each day and the pre-set priority and time, so that the equipment layer responds to the energy adjustment instructions issued by the system, realizes the energy release of new energy photovoltaic, wind energy and energy storage media, energy transmission on the incoming line side of the park distribution network, and uses the released energy for other loads.

[0050] In some embodiments, the PVT power generation and heating system uses the heat energy generated by solar PVT components while generating electricity to provide hot water or heating. It is a high-efficiency cogeneration system that couples PVT components with multi-energy systems, and can simultaneously generate electricity and supply heat under limited space conditions. It can also meet the needs of the park for electricity, hot water, and heating. However, solar power generation is affected by factors such as light intensity, temperature, and cloud thickness, and has large fluctuations. The transformer model can learn historical weather data and power generation data to predict solar power generation and heat supply in the future. In addition, the transformer model can also use satellite remote sensing technology to monitor the light intensity on the surface photovoltaic panels in real time to further improve the accuracy of predictions.

[0051] In some embodiments, the photovoltaic power generation system includes: a photovoltaic component, a controller, and an inverter, wherein the controller controls the photovoltaic component to convert light energy into direct current, and controls the inverter to convert the direct current into alternating current.

[0052] Specifically, a photovoltaic power generation system primarily consists of photovoltaic modules, controllers, inverters, and other accessories (grid-connected systems do not require batteries). Depending on whether or not they rely on the public power grid, they are categorized as off-grid or grid-connected. Off-grid photovoltaic systems are equipped with electrochemical batteries and hydrogen energy storage to ensure stable system power and can supply power to loads even when the photovoltaic system is not generating power at night or on rainy days. The working principle is that photovoltaic modules convert light energy into direct current (DC), which is then converted into AC by an inverter, ultimately enabling power consumption and access to the internet.

[0053] In some embodiments, the wind power generation system includes: a wind turbine component, including at least a wind turbine generator set, a tower, a foundation, and cables, wherein the wind turbine component is used to capture wind energy and convert the wind energy into alternating electrical energy;

[0054] A grid-connected control component, comprising at least a rectifier module, a grid-connected controller, a load shed and cables, and configured to control the normal operation of wind turbine components;

[0055] An inverter component, comprising at least a grid-connected inverter and a cable, wherein the inverter component is used to invert the direct current output by the grid-connected control component into alternating current and feed the energy into the grid;

[0056] The unloading component is used to control the wind turbine to brake and stop, so as to ensure the safety of the wind turbine under abnormal working conditions.

[0057] In some implementations, wind power generation is subject to significant uncertainty due to factors such as wind speed, wind direction, and temperature. The Encode and Decode modules of the Transformer model can be used to analyze and learn from historical meteorological and power generation data to predict future wind power generation, providing a basis for grid energy scheduling within the park.

[0058] Specifically, a wind power generation system: A wind power generation system doesn't directly convert wind energy into electrical energy; it requires mechanical energy transition. Its function is to convert the constantly changing wind energy into alternating current (AC) with a constant frequency and voltage, or direct current (DC) with a constant voltage. Wind turbines convert wind energy into AC electricity. The wind turbine outputs AC with unstable amplitude and frequency. This is rectified by a controller into DC before being fed to an inverter, which converts it into AC with stable amplitude and frequency. After being measured by an electricity meter, the DC power is directly fed into an inverter to convert it into three-phase AC at 380V, 50Hz. It is mainly composed of the following structures: wind turbine part (captures wind energy and converts it into alternating electrical energy; including wind turbines, towers, foundations, cables, etc.), grid-connected control part (controls the safe and normal operation of the wind turbine system, with a built-in rectifier module outputting DC power and limiting the maximum output voltage to protect the back-end inverter; including grid-connected controller, unloader, cables, etc.), inverter part (converts the DC power output by the controller into AC power and feeds the energy into the grid, with step-up transformer isolation; including grid-connected inverter, cables, etc.), unloading part (realizes intelligent control of the wind turbine to brake and stop, ensuring the safety of the wind turbine under abnormal working conditions).

[0059] In some embodiments, the hydrogen power supply system utilizes photovoltaic electricity generated during the daytime peak hours and valley electricity generated at night to electrolyze water to produce hydrogen, and generates electricity through fuel cells during peak hours of grid electricity consumption.

[0060] Specifically, the hydrogen power generation system: By building an efficient, reliable, and stable energy supply system consisting of "renewable energy hydrogen production - hydrogen storage and transportation - hydrogen fuel cell cogeneration," the system will utilize photovoltaic power generated during peak daytime hours and off-peak nighttime electricity to electrolyze water to produce hydrogen. During peak grid demand, fuel cells will generate electricity to ensure sufficient power. Core equipment such as efficient electrolysis hydrogen production systems, fuel cell cogeneration systems, hybrid hydrogen and battery energy storage, and multi-port DC converters will interconnect the production, storage, and consumption links of electricity, hydrogen, and heat energy networks, enabling multifunctional coordinated transformation and deployment, including green electricity hydrogen production, efficient electric-heat-hydrogen cogeneration, flexible vehicle-grid interaction, and long-term off-grid operation, forming a demonstration of an electricity-centric, electric-hydrogen-heat coupled energy internet.

[0061] In some embodiments, the electrochemical energy storage system comprises:

[0062] a battery pack for DC charging or discharging;

[0063] Energy storage converter, used to control the charging and discharging process of the battery pack and perform AC-DC conversion;

[0064] Battery management system, used to monitor the status information of the battery pack;

[0065] Energy management system, responsible for data collection, network monitoring and energy scheduling.

[0066] In some embodiments, chemical, electrothermal phase change energy storage, and hydrogen power generation systems: Under new power systems, unused energy is stored during periods of abundant electricity resources and released to provide power during periods of scarce electricity resources and peak demand, thereby achieving peak load shifting and comprehensive energy utilization. Electrochemical energy storage systems primarily consist of battery packs, power storage converters (PCS), battery management systems (BMS), energy management systems (EMS), and other electrical equipment. The electrothermal phase change energy storage system converts other forms of energy into thermal energy, stores this heat in a well-insulated environment using a specific heat storage medium, and extracts the stored heat for utilization when needed. Hydrogen power generation systems utilize photovoltaic power generated during peak daytime hours and off-peak nighttime electricity to electrolyze water to produce hydrogen. During peak grid demand, fuel cells generate electricity to ensure sufficient power. The transformer model can optimize the energy storage system's charging and discharging strategies, improve the utilization of energy storage equipment, and reduce operating costs. Furthermore, the model can provide the optimal operating strategy for the energy storage system based on factors such as real-time electricity prices and renewable energy generation, maximizing economic benefits. The transformer model can overcome the intermittent and volatile nature of new energy sources, enabling a buffered and smooth transition from source to grid to load, ensuring system efficiency and safety while reducing the cost of clean energy. Key mechanistic changes in electrochemical material systems, such as attenuation and thermal runaway, cannot be observed through online measurement of multiple internal indicators. Therefore, AI data-driven approaches combined with electrochemical material mechanism models are currently a viable approach for battery management.

[0067] In some embodiments, a geothermal heat pump system extracts heat from the ground in winter and stores the building's heat back into the ground in summer, providing both winter and summer heating and cooling sources for heating and air conditioning. In summer, the system transfers heat from the room to the ground, cooling it and storing it for winter use. In winter, a heat pump transfers heat from the soil to the room, heating factory buildings and rooms while also storing cooling for summer use. The earth provides an excellent source of free energy storage, thus achieving seasonal energy conversion. This includes an outdoor geothermal heat exchange system, a geothermal heat pump unit, and an indoor heating and air conditioning terminal system. Geothermal heat pumps primarily come in two types: water-to-water or water-to-air. Heat is transferred between the three systems using water or air as a heat exchange medium. The heat exchange medium between the geothermal heat pump and the ground energy source is water, while the heat exchange medium with the building's heating and air conditioning terminal can be water or air.

[0068] In some embodiments, an energy consumption monitoring and metering system monitors energy media such as water, electricity, gas, cooling, heat, steam, and oil, establishing an energy management and control center to collect, measure, calculate, analyze, and centrally manage energy consumption in real time. Through in-depth data mining and statistical analysis, it displays system load in real time and analyzes system line losses. Energy consumption, metering, analysis, processing, storage, and publishing are performed to implement energy consumption statistics, energy consumption management and assessment, energy efficiency evaluation, energy consumption auditing, energy efficiency publicity, energy consumption quotas, and energy-saving services. The system is also responsible for generating various daily reports, data curves, pie charts, and bar charts. The system collects monitoring and metering data from instruments or systems at all levels of energy consumption in real time, adds time stamps, connects to and stores data in a computer system, and displays enterprise energy data and energy usage through various formats, such as geographic distribution maps, process diagrams, wiring diagrams, trend charts, and infographics, enabling online monitoring and metering. At the same time, it can access the company's production, safety, environmental protection and other data. Through the division and control of system permissions, the scheduling and control functions can be divided into professions (energy, production, environmental protection, safety, etc.), departments, and fields (electricity, water, steam, compressed air, etc.), truly realizing the centralized scheduling and management of enterprise production and comprehensively guaranteeing the company's production operations. The energy consumption measurement data is accurate to every independent space in the building, every branch, workshop, process and team of the industrial enterprise, and the time can be accurate to the minute. The relationship between energy consumption and production output in each link is clearly presented and combined with the performance appraisal of each unit and even individuals. Truly achieve the refinement and comprehensiveness of energy management and put energy conservation and consumption reduction into practice.

[0069] In some embodiments, the energy assistance system includes:

[0070] Environmental monitoring subsystem, used to monitor temperature and humidity, gas, noise, dust, water leakage and water level;

[0071] Power equipment monitoring subsystem, used to monitor the operating status of air conditioners, dehumidifiers, fans, lighting, water pumps and fresh air fans;

[0072] Security monitoring subsystem, used for monitoring smoke, infrared, electronic fence, access control and video.

[0073] Specifically, the intelligent auxiliary monitoring system solution for the park power distribution station / switch station is composed of multiple subsystems, including environmental monitoring subsystem, power equipment monitoring subsystem, video monitoring subsystem, security monitoring subsystem, fire monitoring subsystem, lighting control subsystem, etc. System functions are realized: Dynamic monitoring: transformer winding temperature monitoring, switch cabinet busbar temperature measurement, switch cabinet partial discharge monitoring, feeder power temperature monitoring, distribution cabinet electrical parameter monitoring, etc. Environmental monitoring: temperature and humidity, gas (SF6, O2, etc.), noise, dust, water leakage, water level, etc. Security monitoring: smoke, infrared, electronic fence, access control, video, etc. Equipment control: air conditioners, dehumidifiers, fans, lighting, water pumps, fresh air fans, etc. WeChat push, web page pop-up, SMS alarm, telephone alarm, on-site sound and light alarm, afterglow broadcast alarm, email alarm, etc.

[0074] In some embodiments, the industrial Internet of Things system adopts a PaaS platform architecture and is configured with:

[0075] Rules engine, used to implement data repair, cleaning, calculation and analysis, circulation, data alarm and multi-scenario linkage;

[0076] Data subscription and open API interface to support third-party application development and integration;

[0077] Multiple device protocol parsing drivers are used to enable access of multiple devices through gateway access;

[0078] Configure alarm rules on the interface for alarm management.

[0079] Specifically, the Industrial Internet of Things system utilizes a PaaS platform architecture, offering broad access to industrial terminal devices, comprehensive analysis of industrial communication protocols, remote device configuration and command issuance, and SOE event reporting. It also leverages rule engine technology to implement data repair, cleaning, computational analysis, data transfer, data alarms, and multi-scenario linkage, providing real-time and accurate energy data statistics and analysis, and comprehensive monitoring of platform services and data link security. By digitally modeling device entities and their relationships, device instantiation and grouping management are achieved. Real-time computation and analysis utilizes a rule engine to enable data computation and transfer. Alarm rules are configured through an interface to support alarm management. Time series data analysis is provided, enabling query and aggregate analysis of any historical time period.

[0080] In some embodiments, the power supply dispatching system adopts an economic dispatching algorithm and a big data dispatching algorithm for power supply dispatching. When the economic dispatching algorithm is adopted, the microgrid management is completed in combination with time-of-use electricity prices, new energy output patterns, and preset generation, distribution, and use logic; when the big data dispatching algorithm is adopted, on the basis of economic dispatch, the microgrid dispatching granularity is further refined through data aggregation and big data analysis to realize photovoltaic power generation prediction and load prediction.

[0081] Specifically, the power supply dispatching system collects and monitors data from the factory or industrial park power grid; conducts regional load forecasting and develops power start-up and shutdown plans; compiles inter-provincial and regional power grid energy figures; performs regional power flow, stability, short-circuit current, and offline and online economic operation analysis and calculations; calculates regional power grid relay protection settings; and handles systemic accidents and schedules maintenance within the region. Adjustments are made based on pre-set priorities and timelines. The device layer responds to energy adjustment commands issued by the system, releasing power for other loads. Based on sensor data, the system pre-sets various scenario modes for adjustment. The device layer responds to energy adjustment commands issued by the system, releasing power for other loads. The core of local renewable energy consumption lies in multi-energy dispatching algorithms, including economic dispatching and big data dispatching. Economic dispatching combines time-of-use electricity prices and renewable energy output patterns with pre-defined generation, distribution, and consumption logic to achieve microgrid management. Big data dispatching, based on economic dispatching, uses data aggregation and big data analysis to achieve photovoltaic power generation and load forecasting, further refining microgrid dispatching granularity, optimizing electricity costs, and improving power efficiency.

[0082] In some implementations, the AI ​​large-scale model energy dispatching and operation and maintenance management system includes functions such as park energy dispatching and system operation and maintenance inspection. The park power supply dispatching system mainly collects and monitors data of the park power grid, conducts park load forecasting, and formulates start-up and shutdown plans for production equipment, energy generation equipment, and power system energy park inputs. Adjustments are made according to pre-set priorities and times, and the equipment layer responds to energy adjustment instructions issued by the system to realize the energy release of new energy photovoltaic, wind energy, and energy storage media, energy transmission on the incoming side of the park distribution network, and the use of released energy for other loads. The core of the on-site consumption of new energy in the park is the multi-energy comprehensive dispatching and multi-energy comprehensive utilization algorithm, including through energy economic dispatching algorithms and power generation and power consumption big data dispatching algorithms. According to economic dispatch combined with time-of-use electricity prices and new energy output patterns, the generation, distribution, and consumption logic are preset to complete the management of the park power grid.

[0083] The specific implementation rules for model scheduling are as follows: Since photovoltaic power generation is significantly affected by sunlight and weather, daytime power generation is achieved. PVT photovoltaic thermal systems utilize sunlight to collect ambient heat for heating the park. Wind power generation is also affected by the environment, with wind speed and wind intensity constraining the energy released by the equipment. This requires an AI model to learn factors such as wind power generation start-up time and wind speed. Leveraging the energy collection and release characteristics of electrochemical and physical energy storage, energy storage can be used to store excess renewable energy in the park. Energy can also be purchased at low prices during periods of abundant power in the power system, typically during valleys, for future use in the factory park. Furthermore, energy can be sold at high prices during peak periods of power shortage. Based on daily park electricity consumption, the AI ​​model can also predict the load curve for the immediate future. Scheduling can be coordinated based on power consumption during each time period. For example, if there is no sunshine or wind in the coming days, our energy storage can purchase electricity from the State Grid and store it at a low price. If there is any excess power, it can be sold at a higher price. It can not only realize the coordination of comprehensive energy generation and power consumption of new energy in the park, but also cooperate with the large-scale system dispatching of power system energy, reduce the volatility of the energy system, and enhance the robustness of the system.

[0084] In some embodiments, the transformer large model is trained according to the following formula:

[0085]

[0086] Where, Represents new energy power generation data, represents energy consumption data, represents the energy data purchased from the park, λ represents the price of new energy power generation, k represents the consumption price, δ represents the price of purchased energy, and F * Represents the integration cost.

[0087] In some embodiments, the power generation can be predicted according to the following formula:

[0088] Where α, β, ε, η, φ, and ω represent the power generation duration coefficients of each energy time period, and P V 、P T Indicates the power generation of the PVT power generation and heating system, P W Indicates wind power generation power, P ES1 Indicates the electrochemical energy storage release power, P ES2 represents the released power of the phase change energy storage system, P S Indicates the power system input power.

[0089] Specifically, we first use the transformer model to predict the power generation data, the power generation power P of the PVT power generation and heating system V 、P T , wind power generation power P W , electrochemical energy storage release power P ES1 , Phase Change Energy Storage System P ES2 Power system input power P S etc., using time period energy generation forecast α, β, ε, η, φ, and ω represent the generation duration coefficients for each energy source. The Transformer model continuously learns and predicts the formulas, and by training these coefficients, it derives optimal phased power generation data. α and β are typically affected by sunlight intensity, ε by wind power generation level, η by the release efficiency of chemical energy storage, φ by the release efficiency of phase change media, and ω by the energy load curve of the power system.

[0090] In the process of applying to the comprehensive management of new energy, the transformer model is used to analyze a large amount of energy data. (New energy power generation data , energy consumption data , Park purchased energy data etc.) for prefabrication treatment, ( 、 、 、 Represent the price of new energy power generation, consumption price, external energy price, and integration cost respectively), through the formula Perform training to obtain the optimal data F * In the data processing process, the high-dimensional data sets and data noise characteristics are addressed by using the Encode and Decode modules of the large transformer model and performing effective cluster analysis, cleaning, and denoising autoencoder preprocessing on the extracted data to improve the quality of energy data.

[0091] In addition, due to the complexity of the energy system, large transformer models require constant parameter modification during training, using methods such as regularization and dropout to prevent overfitting and reduce model generalization capabilities. At the same time, cross-validation and hyperparameter adjustment are used to optimize model performance. Furthermore, in order to protect household electricity information and operational data,

[0092] In some implementations, the transformer model uses the following data security technologies: (1) symmetric encryption of energy data is used to ensure data security during transmission and storage; (2) privacy protection technologies such as differential privacy and homomorphic encryption are used to protect enterprise-side privacy data; and (3) strict data access control and auditing systems are established to prevent internal data leaks.

[0093] Because photovoltaic power generation is daytime and wind power generation is conditionally constrained, we utilize energy storage and phase change energy storage to store unused energy. This stored energy can then be utilized at night or during peak hours when electricity prices are high. Based on daily electricity consumption, large AI models can also predict the load curve for the immediate future. This allows for scheduling based on daily electricity consumption during each time period. For example, if there is no sunshine or wind in the coming days, our energy storage can purchase electricity from the State Grid, store it, and buy it at a low price. If there is any unused electricity, it can be sold externally at a higher price.

[0094] In some implementations, the Transformer model can learn information about wind turbine startup times and wind speeds. Leveraging the energy collection and release characteristics of electrochemical and physical energy storage, the model can store excess renewable energy generation in the park through energy storage. Furthermore, during periods of abundant power in the power system, the model can purchase energy at low prices during valleys for future use in the factory park. Furthermore, during periods of power shortage, the model can sell energy at high prices during peak times. Based on daily park electricity consumption, the AI ​​model can also predict the load curve for the immediate future. This allows for scheduling based on daily power consumption. For example, if there is no sunshine or wind in the coming days, the model can purchase electricity from the State Grid, store it at a low price, and sell it to external customers at higher prices. This not only coordinates the integrated generation and consumption of renewable energy in the park, but also supports the overall scheduling of power system energy, reducing energy system volatility and enhancing system stability.

[0095] In some embodiments, see Figure 2As shown, it is a schematic diagram of the architecture of a zero-carbon park intelligent system based on an AI large model provided by an embodiment of the present invention, wherein the yellow part represents the underlying architecture, and the blue part represents the top-level architecture. The top-level architecture includes a factory energy auxiliary system, a factory energy consumption monitoring system, a comprehensive energy supervision and operation system, an industrial Internet system, and an energy consumption metering system. The underlying architecture includes a distribution network system, a load system, a charging pile system, a chemical energy storage system, a wind power generation system, a photovoltaic power generation system, a PVT power generation and heating system, a hydrogen power generation system, an electric thermal phase change energy storage system, and a ground source heat pump system.

[0096] The technical effects achieved by this application are as follows: through the development of micro-energy networks to adapt to the park's new energy power generation resources, mutual conversion between multiple types of energy input and output, distributed energy with mixed storage and comprehensive demand-side response methods are realized, the volatility and randomness of new energy resources are smoothed, the overall power fluctuation level of the distributed power generation system is reduced, and the safe and economic operation of the new energy high-penetration distribution network is guaranteed.

[0097] By analyzing large amounts of historical data, the Transformer model can accurately forecast power demand, power generation, and energy demand. This helps optimize renewable energy production and distribution within the industrial park, power market transactions, and renewable energy grid integration. Furthermore, the Transformer model can be used to optimize power system operations, improving grid efficiency and economic benefits. It can identify early signs of equipment failure from complex data, providing strong support for energy equipment maintenance. Furthermore, the Transformer model can also perform predictive maintenance on equipment, rationalizing maintenance schedules and strategies and reducing operational costs.

[0098] In distributed energy systems, transformer-based large-scale models can help achieve coordinated optimization of various units and improve energy efficiency. The application of large-scale models provides new perspectives and approaches for energy management, not only improving the efficiency and reliability of energy use but also promoting the digital transformation of the energy industry.

[0099] By applying machine learning to large transformer models, their powerful data processing capabilities and predictive accuracy are revolutionizing energy management. Large transformer models have demonstrated their unique value in a variety of areas, including smart distribution network forecasting and optimization, fault diagnosis and predictive maintenance, and distributed energy management. Furthermore, large transformer models have demonstrated significant results in improving the accuracy of wind and solar power generation forecasts and optimizing energy storage systems.

[0100] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0101] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0102] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0103] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0104] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0105] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0106] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0107] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0108] The above are merely embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should all be included within the scope of the claims of the present application.

Claims

1. A zero-carbon park intelligent system based on an AI big model, which is embedded with a transformer big model and is characterized by: include: An energy generation system, comprising a PVT power generation and heating system, a photovoltaic power generation system, a wind power generation system, a hydrogen power generation system, an electrochemical energy storage system, and an electrothermal phase change energy storage system, wherein the PVT power generation and heating system uses a transformer large model to predict power generation based on historical meteorological data and power generation data; the wind power generation system uses the encoder module and decoder module of the transformer large model to predict wind power generation over a period of time based on input historical meteorological data and power generation data; the electrochemical energy storage system and the electrothermal phase change energy storage system use the transformer large model to optimize charging and discharging strategies based on real-time electricity prices and renewable energy power generation; Energy load system, the energy load system includes a charging pile system and a park factory equipment load system, wherein the charging pile system includes AC charging piles, DC charging piles, and supercharged liquid cooling charging piles, and the park factory equipment load system includes at least central air conditioning load, air compressor load, air energy load, Ma load, and power consumption of heavy-load machines in the factory; The park energy consumption supervision system includes an energy consumption monitoring and metering system, an energy auxiliary system, and an industrial Internet system. The energy consumption monitoring and metering system is used to collect, measure, and calculate and analyze the park's energy consumption in real time to achieve energy consumption statistics, energy consumption management and assessment, energy efficiency evaluation, energy quota, and energy-saving services. The energy auxiliary system is used to monitor transformer winding temperature, switchgear busbar temperature, switchgear partial discharge monitoring, and distribution cabinet electrical parameter monitoring. The industrial Internet system is used to comprehensively monitor platform services and data link security, provide time series data analysis, and conduct queries and aggregated analysis for any historical time period. The AI ​​large-scale model energy operation and maintenance system includes a park energy dispatching system and an operation and maintenance inspection system. The park energy dispatching system is used to predict the power load curve for a period of time based on the AI ​​large-scale model, and adjust the equipment according to the power consumption in each time period of each day and the pre-set priority and time, so that the equipment layer responds to the energy adjustment instructions issued by the system, realizes the energy release of new energy photovoltaic, wind energy and energy storage media, energy transmission on the incoming line side of the park distribution network, and uses the released energy for other loads; Train the transformer model according to the following formula: ； Where, Represents new energy power generation data, represents energy consumption data, Indicates the park's purchased energy data, represents the price of new energy power generation, k represents the consumption price, represents the price of purchased energy, represents the integration cost; The power generation is predicted according to the following formula: e=a +b +e +n +φ + ; In the formula, α, β, ε, η, φ, The power generation duration coefficient of each energy time period, 、 Indicates the power generation of the PVT power generation and heating system, Indicates wind power generation power, Indicates the electrochemical energy storage release power, Indicates the power released by the phase change energy storage system, Indicates the power system input power.

2. The intelligent system according to claim 1, characterized in that: The photovoltaic power generation system includes: a photovoltaic component, a controller and an inverter, wherein the controller controls the photovoltaic component to convert light energy into direct current, and controls the inverter to convert the direct current into alternating current.

3. The intelligent system according to claim 1, characterized in that: The wind power generation system comprises: A wind turbine component, comprising at least a wind turbine generator set, a tower, a foundation and cables, wherein the wind turbine component is used to capture wind energy and convert the wind energy into alternating electrical energy; A grid-connected control component, comprising at least a rectifier module, a grid-connected controller, a load shed and cables, and configured to control the normal operation of wind turbine components; An inverter component, comprising at least a grid-connected inverter and a cable, wherein the inverter component is used to invert the direct current output by the grid-connected control component into alternating current and feed the energy into the grid; The unloading component is used to control the wind turbine to brake and stop, so as to ensure the safety of the wind turbine under abnormal working conditions.

4. The intelligent system according to claim 1, characterized in that: The hydrogen energy power supply system utilizes photovoltaic electricity generated during the daytime peak hours and valley electricity generated at night to electrolyze water to produce hydrogen, and generates electricity through fuel cells during peak hours of power consumption in the grid.

5. The intelligent system according to claim 1, characterized in that: The electrochemical energy storage system comprises: a battery pack for DC charging or discharging; Energy storage converter, used to control the charging and discharging process of the battery pack and perform AC-DC conversion; Battery management system, used to monitor the status information of the battery pack; Energy management system, responsible for data collection, network monitoring and energy scheduling.

6. The intelligent system according to claim 1, characterized in that: The energy auxiliary system includes: Environmental monitoring subsystem, used to monitor temperature and humidity, gas, noise, dust, water leakage and water level; Power equipment monitoring subsystem, used to monitor the operating status of air conditioners, dehumidifiers, fans, lighting, water pumps and fresh air fans; Security monitoring subsystem, used for monitoring smoke, infrared, electronic fence, access control and video.

7. The intelligent system according to claim 1, characterized in that: The industrial Internet system adopts the PaaS platform architecture and is configured with: Rules engine, used to implement data repair, cleaning, calculation and analysis, circulation, data alarm and multi-scenario linkage; Data subscription and open API interface to support third-party application development and integration; Multiple device protocol parsing drivers are used to enable access of multiple devices through gateway access; Configure alarm rules on the interface for alarm management.

8. The intelligent system according to claim 1, characterized in that: The park energy dispatching system includes a power supply dispatching system, which adopts an economic dispatching algorithm and a big data dispatching algorithm for power supply dispatching. When the economic dispatching algorithm is adopted, the microgrid management is completed by combining time-of-use electricity prices, new energy output patterns, and preset generation, distribution, and use logic; when the big data dispatching algorithm is adopted, the microgrid dispatching granularity is further refined through data aggregation and big data analysis on the basis of economic dispatch, thereby realizing photovoltaic power generation prediction and load prediction.

Images