User-side-oriented source-load-storage integrated comprehensive energy management system

Abstract

The invention belongs to the technical field of energy management, and discloses a user-side-oriented source-load-storage integrated comprehensive energy management system, and the hardware architecture comprises a sensing layer, a communication layer and a platform layer. On-site facilities of a source side, a load side and a storage side in the hardware architecture are accessed to the integrated energy management system based on the data collection and calculation unit, and data analysis and display are performed through the server; the software function architecture comprises a data layer, a service layer, an application layer and a display layer; the software function architecture supports the realization of the functions of stable power utilization, flexible scheduling and intelligent management of the source-load-storage integrated comprehensive energy management system; the system further comprises a terminal sensing module, a data collection and calculation unit and an optimization control strategy of the comprehensive energy management system. By adopting the source-load-storage integrated comprehensive energy management system facing the user side, the power consumption cost of the power consumer is reduced, the monitoring, analysis and management capabilities of the source-load-storage comprehensive energy are improved, and the distributed new energy consumption is promoted.

Description

Technical Field

[0001] The present invention relates to the field of energy management technology, and in particular to a user-side source-load-storage integrated energy management system. Background Art

[0002] Distributed photovoltaics, energy storage systems, flexible loads and other diversified resources are rapidly penetrating industrial and commercial parks and large building clusters. The traditional single energy management model can no longer meet users' comprehensive needs for economy, reliability and greenness. Currently, user-side energy systems generally face three core challenges:

[0003] Inadequate resource coordination and high economic costs: Power generation (photovoltaic), load generation (load), and storage (battery) systems often operate independently, with fragmented data and a lack of unified coordination strategies. This makes it difficult for users to effectively utilize photovoltaic power to offset high grid electricity prices. Energy storage charging and discharging strategies are disconnected from electricity pricing mechanisms, preventing them from fully participating in demand response or ancillary services markets, resulting in high overall electricity costs. Furthermore, the volatility of renewable energy combined with the randomness of loads exacerbates system operational risks.

[0004] Weak monitoring and analysis capabilities: Existing systems are mostly limited to basic data collection and monitoring, lacking the ability to deeply mine massive amounts of heterogeneous energy data. The lack of accurate forecasting of PV output and load characteristics, as well as multi-dimensional integrated analysis of energy efficiency, carbon emissions, and economic efficiency, results in a lack of data support for operational decisions, making it difficult to achieve refined management and explore potential value.

[0005] Bottlenecks in new energy consumption and difficulties in policy adaptation: The intermittent nature of distributed photovoltaics impacts local power grids, and users lack effective flexible regulation methods (such as energy storage and adjustable loads) to smooth fluctuations and increase self-consumption. Furthermore, time-of-use electricity prices, peak electricity prices, and electricity market regulations vary widely across regions, making it difficult for the existing system to quickly adapt and generate optimal economic dispatch strategies, limiting users' ability to participate in market transactions and earn additional benefits.

[0006] While some energy management systems exist on the market, their functionality is often fragmented, focusing on single-device monitoring (such as photovoltaic SCADA systems), basic energy consumption statistics, or a lack of deep integration with economic incentive mechanisms at the control level. In particular, in terms of domestic adaptability, many systems rely on specific commercial operating systems or hardware platforms, posing security and controllable risks. In terms of communication compatibility, they struggle to uniformly connect diverse, multi-protocol distributed devices (such as photovoltaic inverters of different brands, energy storage PCS, smart meters, and controllable loads). In terms of core value creation, they generally lack the ability to intelligently control and coordinate the charging and discharging of multi-purpose energy storage (such as peak-valley arbitrage, demand management, and new energy support) based on multi-source forecasting (photovoltaic output, load demand), combining real-time electricity prices with market rules, and aiming for optimal economic benefits or overall energy efficiency.

[0007] Therefore, it is urgent to build a comprehensive energy management system that integrates source, load and storage on the user side to overcome existing technical problems and promote the consumption of distributed new energy. Summary of the Invention

[0008] The purpose of this invention is to provide an integrated energy management system for user-side source-load-storage integration, which aims to meet the operational needs of user-side source-load-storage integrated energy management, reduce electricity costs for power users, improve the source-load-storage integrated energy monitoring, analysis and management capabilities, and promote the consumption of distributed new energy.

[0009] To achieve the above objectives, the present invention provides a user-side integrated energy management system integrating source, load and storage, including hardware architecture and software functional architecture;

[0010] The hardware architecture includes the perception layer, the communication layer, and the platform layer. The on-site facilities on the source, load, and storage sides of the hardware architecture are connected to the integrated energy management system through a data aggregation computing unit, and data analysis and display are performed through the server.

[0011] The software functional architecture includes the data layer, service layer, application layer, and presentation layer. The software functional architecture supports the stable power consumption, flexible scheduling, and intelligent management functions of the integrated energy management system integrating source, load, and storage.

[0012] The system also includes a terminal perception module, a data aggregation and calculation unit, and an optimization control strategy for the integrated energy management system.

[0013] Preferably, in the hardware architecture, the on-site facilities of the perception layer provide various basic data support for source-load-storage analysis and control through sensors and monitoring equipment;

[0014] The communication layer uses edge control cabinets and switches to achieve secure transmission of data and control instructions between the perception layer and the platform layer;

[0015] The platform layer includes web servers, database servers and monitoring workstation hardware infrastructure to realize data collection, storage, processing, analysis and function display.

[0016] Preferably, in the software functional architecture, the data layer collects data from on-site facilities through sensing terminals, uploads it to the platform, and records the data regularly. At the same time, it receives control instructions from the platform and executes relevant control strategies.

[0017] The service layer is used to implement the analysis and preprocessing of underlying data, including data communication services, data processing services, data storage services, data modeling, and permission services;

[0018] The application layer integrates energy business applications, providing insights into comprehensive energy data while enabling comprehensive analysis and management of sources, loads, and storage, from energy monitoring and energy usage analysis to operational management. The system generates operational strategies, issues control instructions, and receives control feedback, completing the closed loop of control logic.

[0019] The central operation cockpit of the display layer serves as a large-screen display for dynamic three-dimensional visualization; the business access web enables the use of integrated energy business functions.

[0020] Preferably, a terminal sensing module is designed to realize full power sensing and protection of the power generation, storage and consumption equipment circuits, specifically including:

[0021] In the distributed photovoltaic, energy storage and load sectors, we are developing DC and AC measurement and control meters with measurement functions to achieve the multifunctional integration goal of energy calculation, demand calculation, maximum value recording, set value exceeding limit and data freezing.

[0022] Develop intelligent circuit breakers with protection and measurement functions at all levels of AC and DC busbars to achieve safe and stable power consumption in microgrids. Automatic disconnection protection functions are implemented in short-circuit and overload scenarios, taking into account subsequent use, using topology-based intelligent setting coordination technology and intelligent hierarchical reclosing functions.

[0023] In the mains power supply line part, hardware design technology and numerical calculation methods are used to develop a power quality detection device with core functions including harmonic analysis, waveform sampling, voltage swell / sag / interruption recording, flicker monitoring, voltage imbalance measurement, transient waveform capture, event recording, and measurement control. It also displays real-time measurement values, power quality parameters, real-time and captured waveforms, harmonic bar graphs, vector graphs, and key event record contents for online detection and analysis of power quality.

[0024] Preferably, a data aggregation computing unit with local computing capabilities is developed to respond to the changes in the microgrid, maintain the dynamic balance between source and load, and realize local decision-making and processing;

[0025] To meet the needs of local control of the source-load-storage system, the data aggregation and computing unit has the following core capabilities:

[0026] Core capability 1: Equipment information collection and communication functions;

[0027] The data collection and calculation unit collects power information of key nodes of power generation, power consumption, and energy storage in real time. To enable the access of a large number of source, load, and storage devices, the downstream data collection should support Modbus TCP, Modbus RTU, IEC101 / 103 / 104, DNP3.0, CDT, DL / T645, and IEC61850 protocols, and the upstream data transmission should support Modbus Slave, IEC101 / 104, and CDT protocols, and have Ethernet, RS485, and 4G communication capabilities.

[0028] Core capability 2: Facility edge control capability;

[0029] First, it is necessary to sample the equipment in the distribution area, including photovoltaic models, energy storage models, charging models, and load models, and refine the collection and control parameters;

[0030] Then, we study the control strategies under different operation modes, perform operation monitoring and control output, including logic judgment, strategy analysis, power switching and load regulation;

[0031] Core capability 3: secondary development capability of edge control algorithms;

[0032] In view of the diversity of user-side source-load-storage systems, programmable technology is used to perform customized configurations for different scenarios and realize secondary development of key algorithms; various types of microgrid equipment are virtualized as power sources, ordinary switches, loads, and energy storage, and relevant models are established. In actual engineering applications, they are instantiated and freely combined to achieve rapid configuration management of the model.

[0033] Optimally, to achieve the operational goals of stable power consumption, economic dispatch, and intelligent management of the source-load-storage integrated system, based on the customer's distributed photovoltaic power generation characteristics, power consumption behavior, energy storage operation status, and combined with local power consumption policies, an optimized control strategy for the integrated energy management system is proposed, including:

[0034] (1) Renewable energy output and load demand forecasting method based on deep learning;

[0035] (2) Optimizing the configuration of user-side energy storage based on various revenue models for different application scenarios and user needs;

[0036] (3) Coordinated optimization scheduling methods and operation strategies for economic operation and stable energy supply;

[0037] (4) Energy-based distributed storage-load coordinated control method, including distributed photovoltaic power control and microgrid control.

[0038] Preferably, the renewable energy output and load demand forecasting method based on deep learning includes the following:

[0039] First, we used photovoltaic power generation output and user-side power load demand as research objects to conduct a correlation analysis of the model's input and output variables. Using the Pearson correlation coefficient as an indicator, we studied the impact of different input parameters, including solar radiation intensity, temperature and humidity, cloud cover, air quality, user production conditions, weekday / holiday identification, and historical operating data, on the prediction model's output, thereby determining the input variables of the prediction model.

[0040] Secondly, through the method of mechanism + experience + principal component analysis, screening and dimensionality reduction are performed based on the original meteorological data or user historical electricity consumption data. The specific operation method is as follows:

[0041] (1) Preliminary screening of the features contained in the original data through the mechanism + experience method to select the most obvious characteristic parameters;

[0042] (2) Linearly combine the selected features and sort out a set of new features arranged in descending order of importance to ensure that the new features are linearly independent;

[0043] Then, with a clear model structure, a convolutional neural network with the ability to extract spatial local features is combined with a bidirectional long-short-term memory (LSTM) that can simultaneously consider long-term information in the forward and backward directions to process highly dependent, multi-dimensional load characteristic information and obtain a prediction model for photovoltaic power generation and user energy demand.

[0044] Finally, considering the multi-time-scale prediction requirements of the source-load-storage integrated system in different scenarios of planning, scheduling, and operation control, the relevant parameters of the training set samples and the model training algorithm are adjusted to obtain hourly, minute-level, and second-level prediction models respectively.

[0045] Preferably, a method for optimizing user-side energy storage configuration based on various revenue models and considering different application scenarios and user needs includes the following steps:

[0046] (1) First, analyze the revenue sources of user-side energy storage, including the reduction of electricity purchase costs and the profit from providing ancillary services; use energy storage to reduce electricity costs in high-generation and low-storage and maximum demand electricity fee management, and establish relevant mathematical models;

[0047] (2) Secondly, the profitability of distributed energy storage in providing demand response and emergency power supply auxiliary services is studied, and the costs and benefits of energy storage participating in various auxiliary services are evaluated. Based on the analysis of user-side energy storage benefits, a user-side energy storage optimization configuration model considering auxiliary services is established, with the goal of maximizing the net benefit over the entire life cycle of user-side energy storage, while simultaneously satisfying a series of constraints such as energy storage charge state, energy storage power constraints, maximum demand constraints, demand response constraints, and emergency power supply constraints, as shown below:

[0048]

[0049] Where G(x) is the net income over the entire life cycle of energy storage; x is the energy storage capacity configuration parameter; G1 is the peak-valley arbitrage income; G2 is the demand management income; G3 is the demand-side response income; G4 is the emergency power supply income; C1 is the initial investment cost of energy storage; C2 is the operation and maintenance cost over the entire life cycle of energy storage; h i (x)=0,i=1,2,...,m are the equality constraints satisfying energy storage operation and ancillary services; g j (x)≤0,i=1,2,...,p is the inequality constraint that satisfies energy storage operation and ancillary services;

[0050] (3) Finally, with the commercial operation of the user-side energy storage system as the goal, a number of economic indicators, including the initial investment of the energy storage system, net income over the entire life cycle, return on investment, and investment payback period, are comprehensively considered. A comprehensive evaluation index decision model is established based on the hierarchical analysis method, and the comprehensive evaluation index value is used as the planning basis for energy storage user configuration and participation in auxiliary services.

[0051] The preferred coordinated optimization scheduling method and operation strategy for economic operation and stable energy supply includes the following:

[0052] A multi-timescale optimization scheduling strategy is adopted, including day-ahead and intraday scheduling. Day-ahead scheduling uses a scheduling interval of 1 hour to optimize the charge and discharge status of the energy storage battery for the next 24 hours. Intraday scheduling uses a scheduling interval of 15 minutes and a rolling cycle of 4 hours for rolling optimization scheduling.

[0053] First, in the day-ahead dispatch model, using peak-valley arbitrage as an application scenario, a dispatch function is established to minimize the microgrid's operating cost within the dispatch cycle based on short-term renewable energy forecasts and load forecasts. Furthermore, a series of constraints, including system power balance constraints, node voltage constraints, active output constraints of distributed power sources and distributed energy storage, and distributed energy storage state of charge constraints, are considered to formulate a day-ahead dispatch plan for distributed energy storage output, as shown below:

[0054] min C total-1 =C grid +C DG +C ess-G ess (2);

[0055] Where C total-1 is the total operating cost of the entire scheduling cycle; C grid is the cost of electricity interaction with the upper power grid; C DG is the power generation cost of distributed power generation; C ess is the energy storage operation cost of the entire dispatch cycle; G ess The benefits of distributed energy storage throughout the dispatch cycle;

[0056] Secondly, based on the day-ahead dispatch, a study on rolling optimal dispatch within the day is conducted. By making an ultra-short-term forecast of renewable energy generation with a time scale of 4 hours, a dispatch model with the goal of minimizing the microgrid dispatch adjustment cost is constructed. The energy storage battery activation state determined by the day-ahead dispatch is used as a constraint condition, and the power balance constraint and node voltage constraint similar to those of the day-ahead dispatch are considered. The objective function shown in Equation (3) is established as follows:

[0057]

[0058] Where C total-2 is the adjustment cost of the intraday scheduling cycle; w g (t) and w ess (t) are the adjustment costs of distributed energy storage and grid interaction electricity respectively; D g (t) and D ess (t) are the day-ahead dispatch values ​​of distributed energy storage and grid interaction power respectively; D g，i (t) and D ess，i (t) are the daily dispatch values ​​of distributed energy storage and grid interaction electricity, respectively.

[0059] Preferably, the energy-based distributed storage-load coordinated control method includes distributed photovoltaic power control and microgrid control, as shown below:

[0060] (1) Photovoltaic power control: For photovoltaic arrays, when their output voltage changes, the output power will change accordingly. By adjusting the voltage output to achieve the goal of maximum power tracking, a boost circuit is used to adjust the on-off duty cycle of its electronic switch tube as a control method. By constructing a mathematical model related to duty cycle, voltage, and power, a particle swarm and genetic algorithm search method is used to achieve global search and achieve maximum power target tracking.

[0061] (2) Microgrid control: A dual-loop control structure is used to ensure that the inverter output meets the control requirements, improve operating performance and enhance power quality. The outer loop control generates a reference signal for the inner loop control based on the control target, and the inner loop control performs adjustments.

[0062] Therefore, the present invention adopts the above-mentioned integrated energy management system with integrated source, load and storage for the user side, supports the deployment of domestic Linux operating system, supports multiple communication protocols and multiple communication methods, and has core capabilities such as energy operation monitoring, energy consumption analysis, and operation management; supports photovoltaic power generation, energy storage devices, power load and optimization control and other functional modules to integrate distributed photovoltaic historical / forecasted output, energy storage device operating status, load forecast and other data, and combines local price mechanisms and market models to provide operation strategies based on charging and discharging control of multi-purpose energy storage devices to achieve overall operation optimization or maximize economic benefits.

[0063] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 It is the hardware architecture of the integrated energy management system of the present invention;

[0065] Figure 2 It is the software functional architecture of the integrated energy management system of the present invention;

[0066] Figure 3 It is the terminal perception module of the present invention;

[0067] Figure 4 It is the data collection and calculation unit of the present invention;

[0068] Figure 5 This is the research framework of the renewable energy output and load demand forecasting method based on deep learning in this invention;

[0069] Figure 6 This is the technical route of the user-side energy storage optimization configuration method of the present invention considering multiple revenue models;

[0070] Figure 7 This is the research route of the present invention for the coordinated optimization scheduling method and operation strategy for economic operation and stable energy supply;

[0071] Figure 8 It is the dual-loop control structure of the present invention;

[0072] Figure 9 is a bar graph of photovoltaic power generation real-time power trend in an embodiment of the present invention;

[0073] Figure 10 This is the inverter monitoring interface in the embodiment of the present invention;

[0074] Figure 11 It is the photovoltaic power generation online diagnosis interface in the embodiment of the present invention;

[0075] Figure 12 This is a primary wiring diagram in an embodiment of the present invention;

[0076] Figure 13 is an environmental monitoring interface in an embodiment of the present invention;

[0077] Figure 14 It is the power quality monitoring interface in the embodiment of the present invention;

[0078] Figure 15 This is the energy storage system overview interface in the embodiment of the present invention;

[0079] Figure 16 This is the energy storage battery monitoring interface in the embodiment of the present invention;

[0080] Figure 17 This is the PCS monitoring interface in the embodiment of the present invention;

[0081] Figure 18 It is the energy storage system monitoring interface in the embodiment of the present invention;

[0082] Figure 19 This is the energy storage control strategy interface in the embodiment of the present invention;

[0083] Figure 20 This is the load monitoring interface in the embodiment of the present invention;

[0084] Figure 21 It is a trend curve analysis interface in an embodiment of the present invention;

[0085] Figure 22 It is the alarm management interface in the embodiment of the present invention;

[0086] Figure 23 This is the event recording interface in the embodiment of the present invention;

[0087] Figure 24 It is the report management interface in the embodiment of the present invention;

[0088] Figure 25 It is the work order management interface in the embodiment of the present invention;

[0089] Figure 26 This is the system setting interface in the embodiment of the present invention. DETAILED DESCRIPTION

[0090] The technical solution of the present invention is further described below with reference to the accompanying drawings and embodiments.

[0091] The present invention provides an integrated energy management system for user-side source-load-storage integration, which meets the operational needs of customer-side source-load-storage integrated energy management, reduces electricity costs for power users, improves the source-load-storage integrated energy monitoring, analysis and management capabilities, and promotes the consumption of distributed new energy.

[0092] 1. Platform architecture of integrated energy management system.

[0093] 1. Hardware architecture of integrated energy management system.

[0094] Figure 1 shows the hardware architecture of the integrated energy management system, which includes the perception layer, communication layer, and platform layer. Field facilities in the hardware architecture, including the source, load, and storage side, are connected to the integrated energy management system through a data aggregation and computing unit. Data is analyzed and displayed on the server.

[0095] (1) The on-site facilities of the perception layer are fully integrated with energy elements, including photovoltaics, energy storage, charging piles, distribution rooms, etc. Through various sensors and monitoring equipment, various basic data support is provided for source-load-storage analysis and control.

[0096] (2) The communication layer uses edge control cabinets and switches to achieve secure transmission of data and control instructions between the perception layer and the platform layer. Compared with the networking method of traditional terminal devices, the edge control cabinet has been greatly optimized and improved in terms of application operability, combining data collection and monitoring, edge computing, and the optimization and execution of cloud control strategies.

[0097] (3) The platform layer includes hardware infrastructure such as web servers, database servers, and monitoring workstations to realize data collection, storage, processing, analysis, and function display.

[0098] 2. Software functional architecture of integrated energy management system.

[0099] To support the realization of stable power consumption, flexible dispatch and intelligent management of the integrated source-load-storage system, such as Figure 2 The software functional architecture of the integrated energy management system shown includes multiple modules such as data layer, service layer, application layer and display layer.

[0100] (1) The data layer collects data from on-site facilities through sensing terminals and uploads it to the platform, including real-time data and timed recorded data from photovoltaic, energy storage, charging piles, and distribution systems. It also receives control instructions from the platform and executes relevant control strategies.

[0101] (2) The service layer is used to implement the analysis and preprocessing of the underlying data, including data communication services, data processing services, data storage services, data modeling, permission services, etc.

[0102] (3) Integrated energy business applications at the application layer provide insight into the park's panoramic energy data while enabling comprehensive analysis and management of sources, loads, and storage, from energy monitoring and energy usage analysis to operational management. The system generates operational strategies, issues control instructions, and receives control feedback, completing the closed loop of control logic.

[0103] (4) The central operation cockpit of the display layer serves as a large-screen display for dynamic three-dimensional visualization; the business access web realizes the use of integrated energy business functions.

[0104] 2. Data collection and processing.

[0105] 1. Design a terminal sensing module to achieve full power sensing and protection of the power generation, storage, and consumption equipment circuits.

[0106] Terminal perception is the foundation of integrated energy management of sources, loads and storage, and is the data guarantee for optimizing the scheduling and linkage control of source, load and storage systems. For industrial parks or buildings with distributed photovoltaics and energy storage, and in response to the needs of measurement, perception, analysis and protection of different parts such as mains power lines, photovoltaics (source side), power loads (load side), and energy storage (storage side), terminal perception modules with innovative functional applications are developed, such as Figure 3 As shown, full power sensing and protection of the power generation, storage and consumption equipment circuits are achieved.

[0107] Among them, in the distributed photovoltaic, energy storage and load parts, the main research and development is DC and AC measurement and control meters with high-precision measurement functions, which should have high precision and high communication speed, and achieve the goal of multi-functional integration such as electric energy calculation, demand calculation, maximum value recording, set value over-limit, and data freezing.

[0108] Develop intelligent circuit breakers with protection and measurement capabilities at all levels of the AC and DC busbars to ensure safe and stable power consumption in microgrids. Automatic disconnection protection in scenarios like short circuits and overloads will be implemented, and for future use, topology-based intelligent setting coordination technology and intelligent hierarchical reclosing capabilities will be employed.

[0109] In the mains power incoming line part, taking into account the power quality needs of some users (such as data centers and high-tech manufacturing enterprises), it is planned to adopt advanced hardware design technology and numerical calculation methods, follow the latest national standards for power quality and the general requirements of power quality monitoring equipment, and develop power quality detection devices with core functions such as harmonic analysis, waveform sampling, voltage swell / sag / interruption recording, flicker monitoring, voltage imbalance measurement, transient capture of waveforms, event recording, measurement control, etc., and display real-time measurement values, power quality parameters, real-time and captured waveforms, harmonic bar graphs, vector graphs, event records and other key contents for online detection and analysis of power quality.

[0110] 2. Develop a data aggregation computing unit with local computing capabilities to quickly respond to changes in the microgrid and maintain a dynamic source-load balance, thereby achieving local rapid decision-making and avoiding imbalance in the entire microgrid system.

[0111] Source-load-storage microgrids incorporate a large number of distributed power sources, energy storage, and charging facilities, and the power flow direction and source-load balance are dynamically changing. Conventional energy management systems typically employ a centralized decision-making and remote control model. Complex coordinated control calculations for the source-load-storage system are typically performed on the system platform, and control instructions are transmitted to local controllers for execution. This control approach relies on the communication and computing capabilities of the system platform and is unable to meet the distributed computing and local operational control requirements of microgrids.

[0112] In order to solve this problem and quickly respond to the changes in the microgrid and maintain the dynamic balance between source and load, it is necessary to develop a set of data aggregation computing units with local computing capabilities such as Figure 4 As shown, local rapid decision-making and processing can be achieved to avoid imbalance of the entire microgrid system.

[0113] To meet the requirements of local control of the source-load-storage system, the data aggregation and computing unit should have the following core capabilities:

[0114] Core capability 1: Equipment information collection and communication functions.

[0115] The data collection and calculation unit should collect power information from key nodes for power generation, consumption, and energy storage in real time. To enable access to a large number of devices, including sources, loads, and storage, downstream data collection should support common protocols such as Modbus TCP, Modbus RTU, IEC101 / 103 / 104, DNP3.0, CDT, DL / T645, and IEC61850. Upstream data transmission should support common protocols such as Modbus Slave, IEC101 / 104, and CDT, and possess common communication capabilities such as Ethernet, RS485, and 4G.

[0116] Core capability 2: Facility edge control capability.

[0117] First, it is necessary to sample the equipment in the distribution area, including photovoltaic models, energy storage models, charging models, load models, etc., to refine the collection and control parameters.

[0118] Then, we study the control strategies under different operation modes, perform operation monitoring and control output, including the following modules:

[0119] (1) Logical judgment: frequency and voltage stability control algorithm, judge whether the system is unbalanced and initiate relevant emergency plans.

[0120] (2) Strategy analysis: Based on the power output and load conditions, an optimization control strategy is given. For example, peak-valley arbitrage strategy, photovoltaic output fluctuation smoothing control, etc.

[0121] (3) Power switching: according to the strategy results, a part of the power supply is disconnected;

[0122] (4) Load control: Based on the strategy results, a portion of the controllable load is removed.

[0123] Core capability 3: Secondary development capability of edge control algorithms.

[0124] Due to the diverse nature of user-side source-load-storage systems, fixed control strategies cannot adapt to diverse needs (for example, different control strategies are required for 380V buildings, integrated photovoltaic-storage-charging systems, and photovoltaic-storage-direct-flexible systems). Programmable technology should be employed to customize configurations for different scenarios and enable secondary development of key algorithms. To this end, various microgrid devices are virtualized as power sources, standard switches, loads, energy storage, and other devices. Related models are established, and they can be instantiated and freely combined in actual engineering applications to achieve rapid model configuration and management.

[0125] 3. Optimization control strategy of integrated energy management system.

[0126] To achieve the operational goals of stable power consumption, economic dispatch, and intelligent management of the source-load-storage integrated system, we develop a complete set of technical methods for new energy output and load demand forecasting models, energy storage capacity optimization configuration methods, multi-objective optimization dispatch and operation strategies, and coordinated control of the source-load-storage integrated system based on the customer's distributed photovoltaic power generation characteristics, power consumption behavior, energy storage operation status, and in combination with local electricity consumption policies.

[0127] 1. Renewable energy output and load demand forecasting method based on deep learning.

[0128] The research framework of renewable energy output and load demand forecasting method based on deep learning is as follows: Figure 5 As shown. First, a correlation analysis was conducted on the model's input and output variables, focusing on PV power generation output and user-side power load demand. Using the Pearson correlation coefficient as an indicator, the impact of various input parameters, such as solar radiation intensity, temperature and humidity, cloud cover, air quality, user production conditions, weekday / holiday identification (0 / 1 data), and historical operating data, on the prediction model's output was studied to determine the model's effective input variables.

[0129] Secondly, through the mechanism + experience + principal component analysis (PCA) method, the original meteorological data or user historical electricity consumption data are screened and reduced in dimension. The specific operation method is as follows:

[0130] (1) First, the features contained in the original data are preliminarily screened through the mechanism + experience method to select the most obvious feature parameters;

[0131] (2) Then, the filtered features are linearly combined to form a set of new features arranged in descending order of importance, ensuring that the new features are linearly independent.

[0132] The new samples are the mappings of the original features onto the new features, compressing the size of the original data matrix while fully reflecting the system's original key characteristics. Through screening and dimensionality reduction, key meteorological / user features can be extracted while reducing the dimensionality of the data, significantly reducing the computational burden of subsequent forecast model training.

[0133] Then, with a clear model structure, the convolutional neural network (CNN) with the ability to extract spatial local features and the bidirectional long short-term memory (Bi-LSTM) that can simultaneously consider long-term information in the forward and backward directions are combined to effectively process highly dependent and multi-dimensional load characteristic information and obtain an accurate prediction model for photovoltaic power generation and user energy demand.

[0134] Finally, considering the multi-time-scale prediction requirements of the integrated source-load-storage system in different scenarios such as planning, scheduling, and operation control, the relevant parameters of the training set samples and the model training algorithm are adjusted to obtain hourly, minute-level, and second-level prediction models respectively.

[0135] 2. User-side energy storage optimization configuration method considering multiple revenue models.

[0136] The user-side energy storage system can provide auxiliary services to the power market according to the local power policy, forming a new intelligent power consumption model based on efficient energy storage, such as peak-valley arbitrage, demand management, demand response, and emergency power supply, thereby effectively reducing the electricity costs of industrial and commercial users and increasing user benefits. Figure 6 The technical route shown studies the user-side energy storage capacity configuration method that achieves the highest overall benefit.

[0137] (1) First, the revenue sources of user-side energy storage are analyzed, mainly including the reduction of electricity purchase costs and the profit from providing ancillary services. Under the guidance of the local electricity policy of the project, the method of using energy storage to reduce electricity costs in high-generation and low-storage and maximum demand electricity fee management is studied, and the relevant mathematical model is established.

[0138] (2) Secondly, the profitability of distributed energy storage in providing auxiliary services such as demand response and emergency power supply is studied, and the costs and benefits of energy storage participating in various auxiliary services are evaluated. Based on the analysis of user-side energy storage benefits, a user-side energy storage optimization configuration model considering auxiliary services is established, with the goal of maximizing the net benefit over the entire life cycle of user-side energy storage, while simultaneously satisfying a series of constraints such as energy storage charge state, energy storage power constraints, maximum demand constraints, demand response constraints, and emergency power supply constraints, as shown below:

[0139]

[0140] Where G(x) is the net income over the entire life cycle of energy storage; x is the energy storage capacity configuration parameter; G1 is the peak-valley arbitrage income; G2 is the demand management income; G3 is the demand-side response income; G4 is the emergency power supply income; C1 is the initial investment cost of energy storage; C2 is the operation and maintenance cost over the entire life cycle of energy storage; h i (x)=0,i=1,2,...,m are the equality constraints satisfying energy storage operation and ancillary services; g j (x)≤0,i=1,2,...,p is the inequality constraint that satisfies energy storage operation and ancillary services.

[0141] (3) Finally, with the commercial operation of the user-side energy storage system as the goal, a number of economic indicators such as the initial investment of the energy storage system, net income over the entire life cycle, return on investment, and investment payback period are comprehensively considered. A comprehensive evaluation index decision model is established based on the hierarchical analysis method, and the comprehensive evaluation index value is used as the planning basis for energy storage user configuration and participation in auxiliary services.

[0142] 3. Coordinated optimization scheduling methods and operation strategies for economic operation and stable energy supply.

[0143] Based on the research on the optimization configuration method of user-side energy storage considering various revenue models, distributed energy storage is used as the main target of the optimized scheduling of the source-load-storage integrated system to achieve synergistic interaction with renewable energy and electricity loads, thereby reducing the operating cost of the system and improving the operational stability of the power grid. The research route of the coordinated optimization scheduling method and operation strategy for economic operation and stable energy supply is as follows: Figure 7 shown.

[0144] Considering that the forecast error for renewable energy generation increases with the forecast timescale, a multi-timescale optimization scheduling strategy is adopted, which includes both day-ahead and intraday scheduling. Day-ahead scheduling uses a one-hour scheduling interval to optimize the charge and discharge status of the energy storage battery over the next 24 hours. Intraday scheduling uses a 15-minute scheduling interval and a rolling optimization cycle of 4 hours.

[0145] First, in the day-ahead dispatch model, using peak-valley arbitrage as an application scenario, a dispatch function is established to minimize the microgrid's operating cost within the dispatch cycle based on the short-term (24-hour) forecast of renewable energy and load forecast. A series of constraints, including system power balance constraints, node voltage constraints, active output constraints of distributed power sources and distributed energy storage, and distributed energy storage state of charge constraints, are considered to formulate a day-ahead dispatch plan for distributed energy storage output, as shown below:

[0146] min C total-1 =C grid +C DG +C ess -G ess (2);

[0147] Where C total-1 is the total operating cost of the entire scheduling cycle; C grid is the cost of electricity interaction with the upper power grid; C DG is the power generation cost of distributed power generation; C ess is the energy storage operation cost of the entire dispatch cycle; G ess The benefits brought by distributed energy storage throughout the entire dispatch cycle.

[0148] Secondly, based on the day-ahead scheduling, a study on rolling optimization scheduling within the day is conducted. By making an ultra-short-term forecast of renewable energy generation with a time scale of 4 hours, a scheduling model is constructed with the goal of minimizing the microgrid scheduling adjustment cost. The energy storage battery activation state determined by the day-ahead scheduling is used as a constraint condition, and the same power balance constraints, node voltage constraints, and other related constraints as the day-ahead scheduling are considered. The objective function shown in Equation (3) is established as follows:

[0149]

[0150] Where C total-2 is the adjustment cost of the intraday scheduling cycle; w g (t) and w ess (t) are the adjustment costs of distributed energy storage and grid interaction electricity respectively; D g (t) and D ess (t) are the day-ahead dispatch values ​​of distributed energy storage and grid interaction power respectively; D g，i (t) and D ess，i (t) are the daily dispatch values ​​of distributed energy storage and grid interaction electricity, respectively.

[0151] 4. Distributed storage-load coordinated control method based on energy.

[0152] To achieve flexible and stable operation of the integrated power-source-load-storage system, the control system under study primarily includes distributed photovoltaic power control and microgrid control. For distributed photovoltaics, a maximum power point tracking (MPPT) control method is proposed to ensure that the photovoltaic system operates at its maximum power point despite external disturbances such as sunlight and temperature. For microgrids, inverter control is crucial. Since the research objects are primarily grid-connected, a constant power control (PQ) strategy in master-slave mode is employed.

[0153] (1) Photovoltaic power control: For a photovoltaic array, when its output voltage changes, the output power will also change accordingly, and under a single operating condition, there is a unique maximum power point. Therefore, the voltage output can be adjusted to achieve the goal of maximum power tracking. The boost circuit is used as a control method by adjusting the on-off duty cycle of its electronic switch tube. By constructing a mathematical model related to duty cycle, voltage, and power, and using search methods such as particle swarm and genetic algorithms to achieve global search, the maximum power point tracking problem is solved.

[0154] (2) Microgrid control: adopt Figure 8 The dual-loop control structure shown in Figure 1 shows that the outer loop generates a reference signal for the inner loop based on the specific control target. The inner loop performs fine adjustments to ensure that the inverter output meets the specific control requirements, improving operating performance and enhancing power quality.

[0155] Example

[0156] Based on the user-side integrated source-load-storage integrated energy management system proposed in the present invention, this embodiment demonstrates the software functions of the system, as shown below:

[0157] 1. Photovoltaic monitoring: Provides comprehensive monitoring and display of the operation status of all equipment in the photovoltaic system and the operating environment. It monitors the photovoltaic power generation power, photovoltaic panel operating parameters, equipment status information and equipment fault information in real time, and displays the real-time photovoltaic power generation power trend bar graph, such as Figure 9 shown.

[0158] 2. Inverter monitoring: Unified monitoring of the real-time operating parameters and alarm information of the inverter, comparative analysis of the efficiency of each inverter, and providing data support for the optimal solution. Evaluate the consistency and stability of the inverter output power and the branch current of the combiner box, and evaluate the overall output of the smart microgrid inverter and components, such as Figure 10 shown.

[0159] 3. Photovoltaic power generation online diagnosis: Analyze and calculate the voltage, current and power generation efficiency of photovoltaic modules to obtain important operation and maintenance parameters such as equivalent utilization time, discrete rate, carbon emission analysis of each inverter. Through the calculation and analysis function, photovoltaic panel cleaning plans and operation and maintenance inspection plans can be formulated in advance to minimize the failure rate of photovoltaic panels and increase the income of photovoltaic power generation. Figure 11 shown.

[0160] 4. Primary wiring diagram: The primary wiring diagram of the system includes important data display, circuit breaker opening and closing status, photovoltaic panel power generation, inverter operation status and other important remote signaling information. It can also support remote opening and closing operations of circuit breakers and remote adjustment control of inverters, such as Figure 12 shown.

[0161] 5. Environmental monitoring: Real-time monitoring of ambient temperature, light intensity, wind speed, current wind direction and other information, and trend analysis and recording, such as Figure 13 shown.

[0162] 6. Power quality monitoring: High-precision online monitoring and analysis of the power supply quality of the microgrid system, using the iMeter7A device to monitor the power supply quality of the microgrid system in real time, such as Figure 14 shown.

[0163] Monitor the quality of power delivered by distributed power sources to the distribution network, including voltage fluctuations and flicker, voltage deviation, frequency, harmonics, DC current components, and three-phase imbalance. Centrally collect statistics on low-voltage and overvoltage fluctuations, displaying the overall voltage distribution, cumulative low-voltage duration, number of low-voltage days, and overvoltage details. This allows for intuitive display of the real-time voltage curve as a line graph, with diverse display formats, including simultaneous display of three-phase and single-phase voltages.

[0164] The system analyzes voltage conditions in real time every day, calculating the percentages of overvoltage, normal voltage, and undervoltage. It calculates basic three-phase imbalance data based on real-time three-phase voltage and current data, visualizing the distribution of imbalances over time. It also categorizes the severity of the imbalances by displaying them as 15%-30%, 30%-50%, and above 50%.

[0165] The system collects statistics and analyzes the total voltage distortion rate, the percentage of each harmonic voltage component, and the total harmonic current component. It also monitors and collects statistics on the power quality performance indicators of the current injected into the grid during inverter operation, including the total harmonic distortion rate, power factor, three-phase current imbalance, and DC current component. It also controls active power, including its rate of change, continuous and smooth regulation capability, and overfrequency derating control. It also records the inverter's low voltage ride-through, high voltage ride-through, and active power recovery.

[0166] 7. Energy storage system overview: Energy storage monitoring mainly includes energy storage status monitoring, energy storage information monitoring, energy storage operating power monitoring and energy storage charge and discharge monitoring; in addition, the economic benefits and charge and discharge of energy storage can be statistically displayed so that customers can intuitively see the operating status of energy storage, such as Figure 15 shown.

[0167] 8. Energy storage battery monitoring: Real-time measurement of battery electrical and thermal data, including single cell voltage, battery module temperature, battery module voltage, series loop current, insulation resistance and other parameters, such as Figure 16 shown.

[0168] 9. PCS monitoring: The real-time operating parameters and alarm information of the energy storage PCS are uniformly monitored. The PCS operation plan can be flexibly selected according to the current power consumption and climate factors to maximize the role of PCS, such as Figure 17 shown.

[0169] 10. Energy storage system monitoring: In the energy storage system control module, you can view the energy storage and environmental control status, remotely control the opening and closing of switches inside the energy storage container, and perform configuration operations such as charge and discharge mode parameter settings, environmental control parameter settings, and BMS parameter settings. Figure 18 shown.

[0170] 11. Energy storage control strategy: In order to ensure the maximum absorption of photovoltaic power and obtain benefits through energy storage, the energy storage charging and discharging strategy can be customized according to the customer's electricity price and load conditions, such as Figure 19 shown.

[0171] 12. Load monitoring: All loads in the system are monitored uniformly, and important telemetry data such as voltage, current, and power are displayed in real time. Figure 20 shown.

[0172] 13. Trend curve analysis: The system supports query and analysis of historical data, and displays historical data of any period in the form of lists, curves, and bar graphs, showing the maximum and average values ​​of the period, and multi-parameter data comparison, such as Figure 21 View history, add to favorites, and export query results to Excel or PNG images.

[0173] 14. Alarm management: according to the event level, the event display color, the object of the alarm event, the display event fault characteristic value, the event occurrence time and location, the event confirmation status, etc. When the event occurs, the relevant screen is pushed in time and the voice alarm prompt is given according to the event level. The accident location changes color or animation prompts, such as Figure 22 shown.

[0174] The recorded events are stored in the database and come with an event query program that can customize the query of events in any time period. It supports classified retrieval and quick query of events, and the event list can be output to Excel. All events can be automatically printed.

[0175] 15. Event record: record the alarm and maintenance situations that occur during the operation of the power grid, and support query, online processing and log query, such as Figure 23 shown.

[0176] Record the time stamp of the alarm event and a brief description of the event, and classify the importance of the event according to the alarm level. Alarm events can be screened in all directions and dimensions, including by time period, event level, event type, etc. Alarm information can be pushed through message windows and notification methods such as sound and light flashing.

[0177] 16. Report management: The system supports flexible report tools, which can be used to calculate data by day, week, month, quarter, year, or custom time period. It supports multiple query methods such as cross-time period and specified date, and displays average value, difference value, etc. The report format and content can be customized to draw and query peak, valley and average power data. The query results can be exported to Excel tables, such as Figure 24 shown.

[0178] 17. Work Order Management: The fault handling and dispatching process function systematically displays preventive tests, inspection tasks, etc. Operation and maintenance personnel can query the information of processed and unprocessed work orders through the mobile phone. The system supports quick query and screening according to keywords, levels, categories, time and other conditions. The work order management interface of the system, such as Figure 25 shown.

[0179] Maintenance personnel are primarily responsible for following up on received or assigned maintenance work orders, performing emergency repairs and troubleshooting at the corresponding project site, promptly uploading maintenance records and results via the mobile app, and approving any repair work orders posted by the system. Work orders not processed within the specified free approving period will be assigned by dispatchers.

[0180] 18. System settings: In the system settings, you can add system access accounts, assign different permissions to different accounts, manage different permissions for different display screens and data, and perform multiple sub-energy management such as alarm level settings, equipment management, enterprise management, etc. Figure 26 shown.

[0181] Therefore, the present invention adopts the above-mentioned integrated energy management system with integrated source, load and storage for the user side, supports the deployment of domestic Linux operating system, supports multiple communication protocols and multiple communication methods, and has core capabilities such as energy operation monitoring, energy consumption analysis, and operation management; supports photovoltaic power generation, energy storage devices, power load and optimization control and other functional modules to integrate distributed photovoltaic historical / forecasted output, energy storage device operating status, load forecast and other data, and combines local price mechanisms and market models to provide operation strategies based on charging and discharging control of multi-purpose energy storage devices to achieve overall operation optimization or maximize economic benefits.

[0182] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the same. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solutions of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A user-side integrated energy management system integrating source, load and storage, characterized by: Including hardware architecture and software functional architecture; The hardware architecture includes the perception layer, the communication layer, and the platform layer. The on-site facilities on the source, load, and storage sides of the hardware architecture are connected to the integrated energy management system through a data aggregation computing unit, and data analysis and display are performed through the server. The software functional architecture includes the data layer, service layer, application layer, and presentation layer. The software functional architecture supports the stable power consumption, flexible scheduling, and intelligent management functions of the integrated energy management system integrating source, load, and storage. The system also includes a terminal perception module, a data aggregation and calculation unit, and an optimization control strategy for the integrated energy management system.

2. The integrated energy management system for user-side source-load-storage integration according to claim 1 is characterized in that: In the hardware architecture, the on-site facilities of the perception layer provide various basic data support for source-load-storage analysis and control through sensors and monitoring equipment; The communication layer uses edge control cabinets and switches to achieve secure transmission of data and control instructions between the perception layer and the platform layer; The platform layer includes web servers, database servers and monitoring workstation hardware infrastructure to realize data collection, storage, processing, analysis and function display.

3. The user-side integrated energy management system with integrated source, load and storage according to claim 1, characterized in that: In the software functional architecture, the data layer collects data from on-site facilities through sensing terminals, uploads it to the platform, and records the data regularly. It also receives control instructions from the platform and executes relevant control strategies. The service layer is used to implement the analysis and preprocessing of underlying data, including data communication services, data processing services, data storage services, data modeling, and permission services; The application layer integrates energy business applications, providing insights into comprehensive energy data while enabling comprehensive analysis and management of sources, loads, and storage, from energy monitoring and energy usage analysis to operational management. The system generates operational strategies, issues control instructions, and receives control feedback, completing the closed loop of control logic. The central operation cockpit of the display layer serves as a large-screen display for dynamic three-dimensional visualization; the business access web enables the use of integrated energy business functions.

4. The user-side integrated energy management system with integrated source, load and storage according to claim 1 is characterized in that: Design a terminal sensing module to achieve full power sensing and protection for the power generation, storage, and consumption equipment circuits, including: In the distributed photovoltaic, energy storage and load sectors, we are developing DC and AC measurement and control meters with measurement functions to achieve the multifunctional integration goal of energy calculation, demand calculation, maximum value recording, set value exceeding limit and data freezing. Develop intelligent circuit breakers with protection and measurement functions at all levels of AC and DC busbars to achieve safe and stable power consumption in microgrids. Automatic disconnection protection functions are implemented in short-circuit and overload scenarios, taking into account subsequent use, using topology-based intelligent setting coordination technology and intelligent hierarchical reclosing functions. In the mains power supply line part, hardware design technology and numerical calculation methods are used to develop a power quality detection device with core functions including harmonic analysis, waveform sampling, voltage swell / sag / interruption recording, flicker monitoring, voltage imbalance measurement, transient waveform capture, event recording, and measurement control. It also displays real-time measurement values, power quality parameters, real-time and captured waveforms, harmonic bar graphs, vector graphs, and key event record contents for online detection and analysis of power quality.

5. The user-side integrated source-load-storage integrated energy management system according to claim 1 is characterized in that: Develop data aggregation computing units with local computing capabilities to respond to changes in the microgrid, maintain the dynamic balance between source and load, and realize local decision-making and processing; To meet the needs of local control of the source-load-storage system, the data aggregation and computing unit has the following core capabilities: Core capability 1: Equipment information collection and communication functions; The data collection and calculation unit collects power information of key nodes of power generation, power consumption, and energy storage in real time. To enable the access of a large number of source, load, and storage devices, the downstream data collection should support Modbus TCP, Modbus RTU, IEC101 / 103 / 104, DNP3.0, CDT, DL / T645, and IEC61850 protocols, and the upstream data transmission should support Modbus Slave, IEC101 / 104, and CDT protocols, and have Ethernet, RS485, and 4G communication capabilities. Core capability 2: Facility edge control capability; First, it is necessary to sample the equipment in the distribution area, including photovoltaic models, energy storage models, charging models, and load models, and refine the collection and control parameters; Then, we study the control strategies under different operation modes, perform operation monitoring and control output, including logic judgment, strategy analysis, power switching and load regulation; Core capability 3: secondary development capability of edge control algorithms; In view of the diversity of user-side source-load-storage systems, programmable technology is used to perform customized configurations for different scenarios and realize secondary development of key algorithms; various types of microgrid equipment are virtualized as power sources, ordinary switches, loads, and energy storage, and relevant models are established. In actual engineering applications, they are instantiated and freely combined to achieve rapid configuration management of the model.

6. The user-side integrated source-load-storage integrated energy management system according to claim 1 is characterized in that: To achieve the operational goals of stable power consumption, economic dispatch, and intelligent management for the integrated source-load-storage system, based on the customer's distributed photovoltaic power generation characteristics, power consumption behavior, energy storage operating status, and in combination with local power consumption policies, an optimized control strategy for the integrated energy management system is proposed, including: (1) Renewable energy output and load demand forecasting method based on deep learning; (2) Optimizing the configuration of user-side energy storage based on various revenue models for different application scenarios and user needs; (3) Coordinated optimization scheduling methods and operation strategies for economic operation and stable energy supply; (4) Energy-based distributed storage-load coordinated control method, including distributed photovoltaic power control and microgrid control.

7. The user-side integrated source-load-storage integrated energy management system according to claim 6 is characterized in that: The deep learning-based renewable energy output and load demand forecasting method includes the following: First, we used photovoltaic power generation output and user-side power load demand as research objects to conduct a correlation analysis of the model's input and output variables. Using the Pearson correlation coefficient as an indicator, we studied the impact of different input parameters, including solar radiation intensity, temperature and humidity, cloud cover, air quality, user production conditions, weekday / holiday identification, and historical operating data, on the prediction model's output, thereby determining the input variables of the prediction model. Secondly, through the method of mechanism + experience + principal component analysis, screening and dimensionality reduction are performed based on the original meteorological data or user historical electricity consumption data. The specific operation method is as follows: (1) Preliminary screening of the features contained in the original data through the mechanism + experience method to select the most obvious characteristic parameters; (2) Linearly combine the selected features and sort out a set of new features arranged in descending order of importance to ensure that the new features are linearly independent; Then, with a clear model structure, a convolutional neural network with the ability to extract spatial local features is combined with a bidirectional long-short-term memory (LSTM) that can simultaneously consider long-term information in the forward and backward directions to process highly dependent, multi-dimensional load characteristic information and obtain a prediction model for photovoltaic power generation and user energy demand. Finally, considering the multi-time-scale prediction requirements of the source-load-storage integrated system in different scenarios of planning, scheduling, and operation control, the relevant parameters of the training set samples and the model training algorithm are adjusted to obtain hourly, minute-level, and second-level prediction models respectively.

8. The user-side integrated source-load-storage integrated energy management system according to claim 6 is characterized in that: The optimized configuration method for behind-the-meter energy storage, considering various revenue models for different application scenarios and user needs, includes the following steps: (1) First, analyze the revenue sources of user-side energy storage, including the reduction of electricity purchase costs and the profit from providing ancillary services; use energy storage to reduce electricity costs in high-generation and low-storage and maximum demand electricity fee management, and establish relevant mathematical models; (2) Secondly, the profitability of distributed energy storage in providing demand response and emergency power supply auxiliary services is studied, and the costs and benefits of energy storage participating in various auxiliary services are evaluated. Based on the analysis of user-side energy storage benefits, a user-side energy storage optimization configuration model considering auxiliary services is established, with the goal of maximizing the net benefit over the entire life cycle of user-side energy storage, while simultaneously satisfying a series of constraints such as energy storage charge state, energy storage power constraints, maximum demand constraints, demand response constraints, and emergency power supply constraints, as shown below: Where G(x) is the net income over the entire life cycle of energy storage; x is the energy storage capacity configuration parameter; G1 is the peak-valley arbitrage income; G2 is the demand management income; G3 is the demand-side response income; G4 is the emergency power supply income; C1 is the initial investment cost of energy storage; C2 is the operation and maintenance cost over the entire life cycle of energy storage; h i (x)=0,i=1,2,...,m are the equality constraints satisfying energy storage operation and ancillary services; g j (x)≤0,i=1,2,...,p is the inequality constraint that satisfies energy storage operation and ancillary services; (3) Finally, with the commercial operation of the user-side energy storage system as the goal, a number of economic indicators, including the initial investment of the energy storage system, net income over the entire life cycle, return on investment, and investment payback period, are comprehensively considered. A comprehensive evaluation index decision model is established based on the hierarchical analysis method, and the comprehensive evaluation index value is used as the planning basis for energy storage user configuration and participation in auxiliary services.

9. The user-side integrated source-load-storage integrated energy management system according to claim 6 is characterized in that: Coordinated optimization scheduling methods and operational strategies for economic operation and stable energy supply include the following: A multi-timescale optimization scheduling strategy is adopted, including day-ahead and intraday scheduling. Day-ahead scheduling uses a scheduling interval of 1 hour to optimize the charge and discharge status of the energy storage battery for the next 24 hours. Intraday scheduling uses a scheduling interval of 15 minutes and a rolling cycle of 4 hours for rolling optimization scheduling. First, in the day-ahead dispatch model, using peak-valley arbitrage as an application scenario, a dispatch function is established to minimize the microgrid's operating cost within the dispatch cycle based on short-term renewable energy forecasts and load forecasts. Furthermore, a series of constraints, including system power balance constraints, node voltage constraints, active output constraints of distributed power sources and distributed energy storage, and distributed energy storage state of charge constraints, are considered to formulate a day-ahead dispatch plan for distributed energy storage output, as shown below: min C total-1 =C grid +C DG +C ess -G ess (2); Where C total-1 is the total operating cost of the entire scheduling cycle; C grid is the cost of electricity interaction with the upper power grid; C DG is the power generation cost of distributed power generation; C ess is the energy storage operation cost of the entire dispatch cycle; G ess The benefits of distributed energy storage throughout the dispatch cycle; Secondly, based on the day-ahead dispatch, a study on rolling optimal dispatch within the day is conducted. By making an ultra-short-term forecast of renewable energy generation with a time scale of 4 hours, a dispatch model with the goal of minimizing the microgrid dispatch adjustment cost is constructed. The energy storage battery activation state determined by the day-ahead dispatch is used as a constraint condition, and the power balance constraint and node voltage constraint similar to those of the day-ahead dispatch are considered. The objective function shown in Equation (3) is established as follows: Where C total-2 is the adjustment cost of the intraday scheduling cycle; w g (t) and w ess (t) are the adjustment costs of distributed energy storage and grid interaction electricity respectively; D g (t) and D ess (t) are the day-ahead dispatch values ​​of distributed energy storage and grid interaction power respectively; D g，i (t) and D ess，i (t) are the daily dispatch values ​​of distributed energy storage and grid interaction electricity, respectively.

10. The user-side integrated source-load-storage integrated energy management system according to claim 6, characterized in that: The energy-based distributed load-storage coordinated control method includes distributed photovoltaic power control and microgrid control, as shown below: (1) Photovoltaic power control: For photovoltaic arrays, when their output voltage changes, the output power will change accordingly. By adjusting the voltage output to achieve the goal of maximum power tracking, a boost circuit is used to adjust the on-off duty cycle of its electronic switch tube as a control method. By constructing a mathematical model related to duty cycle, voltage, and power, a particle swarm and genetic algorithm search method is used to achieve global search and achieve maximum power target tracking. (2) Microgrid control: A dual-loop control structure is used to ensure that the inverter output meets the control requirements, improve operating performance and enhance power quality. The outer loop control generates a reference signal for the inner loop control based on the control target, and the inner loop control performs adjustments.

Images