Photovoltaic energy storage intelligent management system based on micro-service architecture

Abstract

The invention relates to the technical field of photovoltaic energy storage, and discloses a photovoltaic energy storage intelligent management system based on a micro-service architecture, which comprises the following modules: an environmental parameter acquisition module in which G is irradiance, [delta] T is temperature difference, RH is humidity, [alpha], [beta] and [gamma] are attenuation coefficients, dynamic influence of environment on power generation efficiency is quantified, and the environmental parameter acquisition module is used for acquiring environmental parameters; outputting the power correction factor to the cooperative control module; a power generation efficiency correction module; an edge calculation optimization micro-service module; an optical-storage-network cooperative control micro-service module; a self-adaptive prediction micro-service module; the dynamic influence of environmental factors on the power generation efficiency is accurately quantified through a nonlinear coupling model; a power correction factor can be calculated in real time, the influence of complex working conditions such as environment humidity sudden rise and irradiance instantaneous change in cloudy weather on photovoltaic efficiency is effectively captured, a scientific basis is provided for system decision making, and it is ensured that anomaly detection and energy efficiency optimization are close to actual working conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic energy storage, and specifically to a photovoltaic energy storage intelligent management system based on a microservices architecture. Background Art

[0002] With the accelerating pace of the global energy structure transformation towards clean energy, photovoltaic power generation, as a renewable and pollution-free energy form, has been widely used. However, photovoltaic power generation has the characteristics of intermittency and volatility, and its output power is greatly affected by factors such as weather conditions and day-night alternation, which poses challenges to the stability and reliability of power supply. The introduction of an energy storage system provides an effective way to solve this problem. By storing electrical energy when photovoltaic power generation is excessive and releasing electrical energy when power generation is insufficient, it can smooth the power output and improve the stability and flexibility of the power system.

[0003] After retrieval, the patent with Chinese patent number CN119379040A discloses an energy business Internet intelligent management system based on a microservices architecture, including an Internet intelligent management system. The Internet intelligent management system includes a planning module, a monitoring module, an analysis module, an optimization module, a prediction module, and an interaction module. The planning module includes a basic data acquisition module and a marking module. The basic data acquisition module is electrically connected to the photovoltaic system and is also electrically connected to the marking module. In the process of using the present invention, the Internet intelligent management system can finely process various parameter data generated in the energy business to ensure that the obtained abnormal parameter data is more reasonable, accurate, and has a short processing time. It adopts multi-stage processing and sets a time threshold, thereby ensuring the energy storage of new energy and avoiding waste of resources.

[0004] Although the above system collects objective parameters such as light and temperature, it only performs data fusion through a simple "influence threshold" and lacks the ability to dynamically model the non-linear environmental impact. When the environmental humidity suddenly rises, resulting in the attenuation of the photovoltaic panel efficiency, and the instantaneous change of irradiance under cloudy weather, no quantitative correlation model is established, resulting in abnormal detection and energy efficiency optimization deviating from the actual working conditions. In the microservices architecture, high-concurrency real-time data streams are generated by terminals such as photovoltaic arrays, energy storage devices, and inverters. The existing technology uses "time thresholds" for segmented processing, but does not solve the distributed computing bottleneck of multi-source heterogeneous data. When the scale of monitoring nodes exceeds ten thousand, the delay of data cleaning, visualization, and cross-module transmission increases significantly, making it difficult to support minute-level fault response, resulting in energy storage scheduling lagging behind power generation fluctuations. Based on this, the present invention designs a photovoltaic energy storage intelligent management system based on a microservices architecture to solve the above problems. Summary of the Invention

[0005] The object of the present invention is to provide a photovoltaic energy storage intelligent management system based on a microservice architecture, which solves the problem that anomaly detection and energy efficiency optimization in the background technology deviate from the actual working conditions.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: A photovoltaic energy storage intelligent management system based on a microservice architecture, comprising the following modules: An environmental parameter acquisition module, through a non-linear coupling model, the formula is: η = α·G·(1 - β·ΔT)·e^(-γ·RH), where G is the irradiance, ΔT is the temperature difference, RH is the humidity, and α, β, γ are attenuation coefficients, quantifying the dynamic impact of the environment on the power generation efficiency, and outputting a power correction factor to the collaborative control module; A power generation efficiency correction module, deployed at the edge gateway of the station, uses time window slicing to perform distributed cleaning on the concurrent data streams of nodes of photovoltaic modules, energy storage batteries, and inverters, compresses the occupied transmission bandwidth, and realizes anomaly location; An edge computing optimization microservice module, deployed at the edge gateway of the station, receives the concurrent data streams of nodes of photovoltaic modules, energy storage batteries, and inverters. This module uses a time window slicing mechanism to perform distributed data cleaning and compression processing on the data streams; A light-storage-grid collaborative control microservice module, connecting the power generation efficiency correction module and the edge computing optimization microservice module, receives the power correction factor, battery SOC status, grid electricity price signal, and grid load demand; this module dynamically switches the system working mode based on the electricity price signal, battery SOC status, and grid load demand, where: When the sudden drop in photovoltaic power > 30% and SOC ≥ 60%, trigger the "energy storage power supply" mode; When it is the grid valley electricity price period and SOC ≤ 20%, trigger the "low-price power purchase and charging" mode; An adaptive prediction microservice module, connecting the edge computing optimization microservice module, receives the processed historical power generation data and meteorological warning information; this module fuses the historical power generation data and meteorological warning information to generate a power prediction curve for the next 4 hours.

[0007] Preferably, the construction process of the non-linear coupling model is as follows: Step S1.1: Collect the power generation data of photovoltaic modules under different combinations of environmental parameters, including irradiance G, temperature difference ΔT, humidity RH, and the corresponding power generation P; Step S1.2: Take the ratio of the power generation P to the irradiance G as the actual power generation efficiency η of the photovoltaic module, i.e., η = P / G; Step S1.3: Based on the data collected in Step S1.1 and η calculated in Step S1.2, determine the model parameters α, β, γ through optimization algorithm fitting. The optimization algorithm aims to minimize the deviation between the actual power generation efficiency η and the model-predicted power generation efficiency η_pred.

[0008] Preferably, when the edge computing optimization microservice module performs distributed cleaning on the data stream using time window slicing, it specifically includes the following steps: Step S2.1: Dynamically divide the time window into multiple sub-time windows according to the data stream characteristics and real-time requirements; Step S2.2: Perform preliminary cleaning on the data within each sub-time window to remove incorrect or invalid data; Step S2.3: Perform feature extraction on the cleaned data and compress it using a data compression algorithm; Step S2.4: Shard the compressed data, and each shard contains a preset number of data points.

[0009] Preferably, when the optical-storage-network collaborative control microservice module dynamically switches the working mode, it further executes the following control strategies: When the energy storage power supply mode is triggered, in combination with the grid load demand and voltage stability status, the output power of the energy storage system is adjusted in real time; when the low-price power purchase and charging mode is triggered, the rising rate of the charging power is restricted, and the charging strategy is dynamically adjusted according to the real-time grid load status.

[0010] Preferably, when the adaptive prediction microservice module fuses the historical power generation data and meteorological warning information as the input of the model, it uses a multi-source data fusion algorithm, and the specific steps are as follows: Step S4.1: Preprocess the historical power generation data, including data cleaning, missing value filling, and outlier detection; Step S4.2: Obtain the meteorological warning information, which includes the change trends of irradiance, temperature, humidity, and cloud cover within the next 4 hours; Step S4.3: Input the preprocessed historical power generation data and meteorological warning information into a deep learning algorithm to train the prediction model.

[0011] Preferably, the steps further include: Step S4.4: Use the trained prediction model to output the power curve for the next 4 hours; Analyze the fluctuation amplitude and frequency of the power curve; Based on the analysis results, divide the power fluctuation degree into low, medium, and high levels; According to the fluctuation level, determine the safety factor used in the calculation of the reserved capacity in the energy storage buffer capacity management module.

[0012] Preferably, the step further includes: Step S4.5, while considering the charge and discharge efficiency of the energy storage system and the battery life factor, optimize the allocation of the energy storage buffer capacity.

[0013] Preferably, after determining the safety factor and the reserved capacity, the energy storage buffer capacity management module further performs the following optimization operations: Step S4.4.1, allocate a part of the reserved capacity as the standby capacity; Step S4.4.2, allocate the charge and discharge operations of the energy storage capacity according to the power fluctuation period, the fluctuation degree and the fluctuation level; Step S4.4.3, combine the grid load demand and the electricity price signal to dynamically adjust the allocation strategy of the energy storage capacity.

[0014] Preferably, the data compression algorithm adopted by the edge computing optimization microservice module includes at least one of the wavelet transform compression algorithm, the principal component analysis compression algorithm or the sparse representation compression algorithm.

[0015] Preferably, it further includes a visualization and alarm module, configured to: construct a multi-dimensional visualization interface to display the key operation parameters and indicators of the interaction between the photovoltaic array, the energy storage device and the power grid; analyze the operation parameters in real time based on preset rules; when it is detected that the key parameters continuously deviate from the normal range by more than the set threshold, generate and send an alarm message.

[0016] Compared with the prior art, the beneficial effects achieved by the present invention are: 1. In the present invention, the dynamic influence of environmental factors on the power generation efficiency is accurately quantified through the non-linear coupling model. It can calculate the power correction factor in real time, effectively capture the influence of complex working conditions such as the sudden rise of environmental humidity and the instantaneous change of irradiance under cloudy weather on the photovoltaic efficiency, provide a scientific basis for system decision-making, and ensure that the anomaly detection and energy efficiency optimization are close to the actual working conditions.

[0017] 2. In the present invention, through the distributed architecture of the power generation efficiency correction module and the edge computing optimization microservice module, the time window slicing mechanism is used to perform distributed cleaning and compression processing on the data stream. The time window is dynamically divided according to the characteristics of the data stream, which can not only significantly reduce the occupation of the transmission bandwidth, but also solve the distributed computing bottleneck of multi-source heterogeneous data, ensure the efficient operation of the system when the scale of monitoring nodes is large, and support minute-level fault response.

[0018] 3. In the present invention, through the adaptive prediction microservice module, the historical power generation data and the meteorological warning information are fused, the deep learning algorithm is used to generate the power prediction curve for the next 4 hours, and the power fluctuation characteristics are analyzed. Based on the fluctuation level, the safety factor used in the calculation of the reserved capacity in the energy storage buffer capacity management module is determined, realizing the scientific allocation and optimization of the energy storage buffer capacity, and improving the stability and reliability of the system. Description of the Drawings

[0019] Figure 1 It is the architecture diagram of the photovoltaic energy storage intelligent management system based on the microservice architecture of the present invention; Figure 2 It is the flow chart for constructing the non-linear coupling model of the present invention; Figure 3 It is the working flow chart of the edge computing optimized microservice module of the present invention; Figure 4 It is the working flow chart of the adaptive prediction microservice module of the present invention. Specific embodiments

[0020] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0021] Embodiment 1; Please refer to Figures 1-4 , in the embodiment of the present invention, a photovoltaic energy storage intelligent management system based on the microservice architecture, characterized in that it includes the following modules: Environmental parameter acquisition module, through a non-linear coupling model, the formula is: η = α·G·(1 - β·ΔT)·e^(-γ·RH), where G is irradiance, ΔT is temperature difference, RH is humidity, and α, β, γ are attenuation coefficients, quantifying the dynamic impact of the environment on the power generation efficiency, and outputting a power correction factor to the coordinated control module; Power generation efficiency correction module, deployed at the field edge gateway, using time window slicing to perform distributed cleaning on the concurrent data streams of nodes of photovoltaic modules, energy storage batteries, and inverters, compressing the occupied transmission bandwidth, and realizing anomaly location; Edge computing optimized microservice module, deployed at the field edge gateway, receiving the concurrent data streams of nodes of photovoltaic modules, energy storage batteries, and inverters. This module adopts a time window slicing mechanism to perform distributed data cleaning and compression processing on the data stream; Photovoltaic-storage-grid coordinated control microservice module, connecting the power generation efficiency correction module and the edge computing optimized microservice module, receiving the power correction factor, battery SOC status, grid electricity price signal, and grid load demand; this module dynamically switches the system working mode based on the electricity price signal, battery SOC status, and grid load demand, where: When the photovoltaic power drops steeply > 30% and SOC ≥ 60%, trigger the "energy storage power supply" mode; When it is the grid valley electricity price period and SOC ≤ 20%, trigger the "low-price power purchase and charging" mode; The adaptive prediction microservice module is connected to the edge computing optimization microservice module and receives the processed historical power generation data and meteorological warning information. This module fuses the historical power generation data and meteorological warning information to generate a power prediction curve for the next 4 hours.

[0022] The construction process of the non - linear coupling model is as follows: Step S1.1: Collect the photovoltaic module power generation data under different combinations of environmental parameters, including irradiance G, temperature difference ΔT, humidity RH, and the corresponding power generation P; Step S1.2: Take the ratio of the power generation P to the irradiance G as the actual power generation efficiency η of the photovoltaic module, that is, η = P / G; Step S1.3: Based on the data collected in Step S1.1 and η calculated in Step S1.2, determine the model parameters α, β, γ through an optimization algorithm. The optimization algorithm aims to minimize the deviation between the actual power generation efficiency η and the model - predicted power generation efficiency η_pred.

[0023] When the edge computing optimization microservice module performs distributed cleaning on the data stream using time - window slicing, it specifically includes the following steps: Step S2.1: Dynamically divide the time window into multiple sub - time windows according to the data stream characteristics and real - time requirements; Step S2.2: Perform preliminary cleaning on the data within each sub - time window to remove incorrect or invalid data; Step S2.3: Perform feature extraction on the cleaned data and compress it using a data compression algorithm; Step S2.4: Shard the compressed data, and each shard contains a preset number of data points.

[0024] When the optical - storage - grid collaborative control microservice module dynamically switches the working mode, it further executes the following control strategies: When the energy - storage power supply mode is triggered, combine the grid load demand and voltage stability status to adjust the output power of the energy - storage system in real - time; When the low - price power purchase and charging mode is triggered, limit the charging power rising rate and dynamically adjust the charging strategy according to the real - time grid load status.

[0025] The working principle of the embodiments of the present invention is as follows: Each module collaborates closely to achieve efficient and intelligent management of the photovoltaic energy storage system. The environmental parameter acquisition module, based on the non-linear coupling model, accurately quantifies the dynamic impact of environmental factors on power generation efficiency, calculates the power correction factor in real time through a formula, and transmits it to the optical-storage-network collaborative control microservice module. The power generation efficiency correction module and the edge computing optimization microservice module are both deployed on the field edge gateway. The former uses time window slicing to perform distributed cleaning on the concurrent data streams of nodes of photovoltaic modules, energy storage batteries, and inverters, and accurately locates abnormal data. The optical-storage-network collaborative control microservice module receives multiple pieces of information such as the power correction factor, battery SOC status, grid electricity price signal, and grid load demand, and dynamically switches the system working mode according to preset rules. When the grid valley electricity price period and SOC ≤ 20%, it triggers the "low-price power purchase and charging" mode to optimize the electricity cost. The adaptive prediction microservice module fuses the historical power generation data processed by the edge computing optimization microservice module and meteorological warning information, and generates a future 4-hour power prediction curve by means of advanced algorithms, providing a strong basis for the system to make decisions in advance.

[0026] Embodiment 2; Please refer to Figures 1-4 In the embodiments of the present invention, when the adaptive prediction microservice module fuses the input historical power generation data and meteorological warning information in the fusion model, it adopts a multi-source data fusion algorithm. The specific steps are as follows: Step S4.1: Preprocess the historical power generation data, including data cleaning, missing value filling, and outlier detection; Step S4.2: Obtain meteorological warning information, which includes the change trends of irradiance, temperature, humidity, and cloud cover within the next 4 hours; Step S4.3: Input the preprocessed historical power generation data and meteorological warning information into a deep learning algorithm to train the prediction model. Step S4.4: Use the trained prediction model to output the future 4-hour power curve; analyze the fluctuation amplitude and fluctuation frequency of the power curve; divide the power fluctuation degree into low, medium, and high levels based on the analysis results; and determine the safety factor used in the calculation of the reserved capacity in the energy storage buffer capacity management module according to the fluctuation level.

[0027] Step S4.5: Optimize the allocation of the energy storage buffer capacity while considering the charge and discharge efficiency of the energy storage system and the battery life factors. After determining the safety factor and the reserved capacity, the energy storage buffer capacity management module further performs the following optimization operations: Step S4.4.1: Allocate part of the reserved capacity as standby capacity; Step S4.4.2: Allocate the charge and discharge operations of the energy storage capacity according to the power fluctuation period, fluctuation degree, and fluctuation level; Step S4.4.3: Dynamically adjust the allocation strategy of the energy storage capacity in combination with the grid load demand and electricity price signal.

[0028] The data compression algorithms adopted by the edge computing optimization microservice module include at least one of the wavelet transform compression algorithm, the principal component analysis compression algorithm, or the sparse representation compression algorithm. The visualization and alarm module is configured to: construct a multi-dimensional visualization interface to display the key operation parameters and indicators of the interaction between the photovoltaic array, the energy storage device, and the power grid; analyze the operation parameters in real time based on preset rules; and generate and send alarm information when it is detected that the key parameters continuously deviate from the normal range by more than the set threshold.

[0029] The working principle of the embodiment of the present invention is as follows: The adaptive prediction microservice module adopts a multi-source data fusion algorithm to deeply process historical power generation data and meteorological warning information. First, the historical power generation data is finely preprocessed, and then combined with comprehensive meteorological warning information, a prediction model is trained using a deep learning algorithm to output the power prediction curve for the next 4 hours, and its fluctuation characteristics are deeply analyzed. Based on this, the safety factor of the energy storage buffer capacity management module is determined to achieve the scientific allocation and optimization of the energy storage buffer capacity. The edge computing optimization microservice module uses an efficient data compression algorithm to reduce bandwidth occupancy and ensure the rapid transmission and processing of data. The visualization and alarm module constructs a multi-dimensional visualization interface to display the key operation parameters and indicators of the system in real time, accurately analyzes the operation parameters based on preset rules, discovers anomalies in a timely manner, and sends alarm information to ensure the stable operation of the system.

[0030] Embodiment 3; Please refer to Figures 1-4 , a specific embodiment is provided. A certain science and technology park in Shanghai is located at 31.2°N and 121.6°E, with an annual average irradiance of 1,250 kWh / m². The system includes: Photovoltaic system: Monocrystalline silicon modules are used, with a total installed capacity of 200 kW, divided into 4 sub-arrays, equipped with high-efficiency inverters, and the conversion efficiency is ≥98.5%.

[0031] Energy storage system: Lithium-ion battery pack, with a capacity of 300 kWh, a charge-discharge efficiency of 92%, and a SOC working range of 10%-90%.

[0032] Grid interface: 10 kV grid connection point, implementing a time-of-use electricity price policy. The electricity price during peak hours is 1.25 yuan / kWh (8:00-12:00, 18:00-22:00), and the electricity price during valley hours is 0.32 yuan / kWh (22:00-6:00).

[0033] Calibration of the non-linear coupling model parameters: Based on historical data fitting, the attenuation coefficients are α = 0.83, β = 0.0045 / °C, γ = 0.017 / %RH. On a certain day, the measured irradiance is 850 W / m², the temperature difference ΔT = 12 °C, and the humidity RH = 65%. The calculated power generation efficiency correction factor η = 0.83×850×(1 - 0.0045×12)×e^(-0.017×65) = 0.71, which is used to dynamically correct the photovoltaic theory.

[0034] Edge computing optimizes microservice configuration: The time window slice is set to 5 minutes. Within each window, 200 original data points are compressed to 60 through the wavelet transform algorithm, reducing the bandwidth occupancy by 70%. The anomaly location function has detected a 45% sudden drop in the inverter current, automatically marked as "string fault" and triggering an alarm work order.

[0035] Energy storage power supply mode: The trigger condition is that the photovoltaic power drops steeply by > 30% and the SOC ≥ 60%. The response action is that the energy storage output power = load gap × 1.1, with a limit of ±100 kW.

[0036] Low-price power purchase charging mode: The trigger condition is the valley electricity price period and the SOC ≤ 20%. The charging power linearly increases at a rate of ≤ 30 kW / min and is dynamically adjusted according to the real-time grid load.

[0037] Scenario 1: Noon photovoltaic power sudden drop event: Time: 11:45 on July 15, 2025 (peak electricity price period).

[0038] Event: Sudden cloud occlusion causes the photovoltaic power to drop suddenly from 142 kW to 85 kW (a 40% drop).

[0039] Environmental parameter acquisition module: Real-time calculation shows η = 0.69 (due to the RH rising to 80%), and the theoretical photovoltaic output is corrected to 92 kW.

[0040] Coordinated control module: Determines that the power drops steeply by > 30% and the SOC = 68%, triggering the energy storage power supply mode.

[0041] Energy storage action: Outputs 55 kW of compensation power to maintain the balance of the grid load.

[0042] Visualization: The interface pushes an alarm of "abnormal photovoltaic power", and the operation and maintenance personnel respond within 5 seconds.

[0043] Scenario 2: Optimization of energy storage charging during valley period: Time: 23:30 on July 15, 2025 (valley electricity price period).

[0044] System status: SOC = 18%, real-time grid load 180 kW.

[0045] Coordinated control module: Triggers the low-price power purchase charging mode. The charging power starts from 0 kW and rises at a rate of 25 kW / min to 75 kW (the upper limit value).

[0046] After detecting that the grid load > 150 kW, it automatically reduces the speed to 15 kW / min to avoid transformer overload.

[0047] Adaptive prediction and energy storage buffer management, cleaning historical power generation data, filling missing values using linear interpolation, and removing outliers using the 3σ criterion; inputting future 4-hour irradiance prediction (decreasing trend: 820→350W / m²), temperature (29°C→26°C); outputting the power curve and marking the fluctuation level as "high" (fluctuation range ±38kW); taking the full coefficient as 1.5, reserved capacity = 300kWh×1.5×0.3 = 135kWh (accounting for 30% of the total capacity); allocating 50kWh as standby capacity (to cope with sudden power outages), and the remaining 85kWh for suppressing power fluctuations (charge and discharge power limit ±70kW).

[0048] Charging 80kWh during the valley period (cost 25.6 yuan), discharging during the peak period (income 100 yuan), with a daily net income of 74.4 yuan; the photovoltaic fluctuation suppression rate > 88%, the fluctuation range reduced from ±40kW to ±5kW; the fault response time < 5 seconds, with an 80% speed increase compared to the traditional system; the data bandwidth compressed by 70%, and the edge computing delay controlled within 80ms.

[0049] Working principle: The photovoltaic energy storage intelligent management system based on the microservice architecture realizes efficient intelligent management through the close cooperation of each module. The environmental parameter acquisition module uses a non-linear coupling model to calculate the power correction factor, providing a basis for system decision-making. The power generation efficiency correction module and the edge computing optimization microservice module are deployed at the station edge gateway to perform distributed cleaning and compression of data, ensuring data quality and transmission efficiency. The light-storage-network collaborative control microservice module receives various information, dynamically switches the system working mode, and optimizes the electricity cost. The adaptive prediction microservice module integrates historical power generation data and meteorological warning information to generate a power prediction curve and analyze the fluctuation characteristics, which are used for the scientific allocation and optimization of the energy storage buffer capacity. At the same time, the edge computing optimization microservice module uses an efficient data compression algorithm to reduce bandwidth occupancy. The visualization and alarm module displays the key parameters of the system in real time, discovers anomalies in a timely manner, and sends alarm information. Generally speaking, each module cooperates with each other to ensure the stable, efficient, and intelligent operation of the photovoltaic energy storage system and improve economic benefits.

[0050] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirits of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A photovoltaic energy storage intelligent management system based on a microservices architecture, characterized in that It includes the following modules: An environmental parameter acquisition module, through a non-linear coupling model, the formula is: η = α·G·(1 - β·ΔT)·e^(-γ·RH), where G is the irradiance, ΔT is the temperature difference, RH is the humidity, and α, β, and γ are attenuation coefficients, which quantify the dynamic impact of the environment on the power generation efficiency and output a power correction factor to the cooperative control module; A power generation efficiency correction module, deployed at the edge gateway of the station, uses time window slicing to perform distributed cleaning on the concurrent data streams of the nodes of photovoltaic modules, energy storage batteries, and inverters, compresses the occupied transmission bandwidth, and realizes anomaly location; An edge computing optimization microservice module, deployed at the edge gateway of the station, receives the concurrent data streams of the nodes of photovoltaic modules, energy storage batteries, and inverters. This module uses the time window slicing mechanism to perform distributed data cleaning and compression processing on the data streams; A light-storage-network cooperative control microservice module, connects to the power generation efficiency correction module and the edge computing optimization microservice module, and receives the power correction factor, battery SOC status, grid electricity price signal, and grid load demand; based on the electricity price signal, battery SOC status, and grid load demand, this module dynamically switches the system working mode, where: When the sudden drop in photovoltaic power > 30% and SOC ≥ 60%, trigger the "energy storage power supply" mode; When it is the off-peak electricity price period of the grid and SOC ≤ 20%, trigger the "low-price power purchase and charging" mode; An adaptive prediction microservice module, connects to the edge computing optimization microservice module, and receives the processed historical power generation data and meteorological warning information; this module fuses the historical power generation data and meteorological warning information to generate a power prediction curve for the next 4 hours.

2. The photovoltaic energy storage intelligent management system based on the microservice architecture according to claim 1, characterized in that The construction process of the non-linear coupling model is as follows: Step S1.1, collect the power generation data of photovoltaic modules under different combinations of environmental parameters, including irradiance G, temperature difference ΔT, humidity RH, and the corresponding power generation P; Step S1.2, use the ratio of power generation P to irradiance G as the actual power generation efficiency η of the photovoltaic module, that is, η = P / G; Step S1.3, based on the data collected in Step S1.1 and η calculated in Step S1.2, determine the model parameters α, β, and γ through an optimization algorithm. The optimization algorithm aims to minimize the deviation between the actual power generation efficiency η and the model predicted power generation efficiency η_pred.

3. A photovoltaic energy storage intelligent management system based on a microservices architecture according to claim 1, characterized in that, When the edge computing optimization microservice module performs distributed cleaning on the data stream using time window slicing, it specifically includes the following steps: Step S2.1, dynamically divide the time window into multiple sub-time windows according to the data stream characteristics and real-time requirements; Step S2.2, perform preliminary cleaning on the data within each sub-time window to remove incorrect or invalid data; Step S2.3, perform feature extraction on the cleaned data and compress it using a data compression algorithm; Step S2.4, slice the compressed data, and each slice contains a preset number of data points.

4. A photovoltaic energy storage intelligent management system based on a microservices architecture according to claim 1, characterized in that, When the light-storage-network cooperative control microservice module dynamically switches the working mode, it further executes the following control strategies: When the energy storage power supply mode is triggered, in combination with the grid load demand and voltage stability status, the output power of the energy storage system is adjusted in real time; When the low-price power purchase charging mode is triggered, the charging power increase rate is restricted, and the charging strategy is dynamically adjusted according to the real-time load status of the power grid.

5. A photovoltaic energy storage intelligent management system based on a microservice architecture according to claim 1, characterized in that, When the adaptive prediction microservice module fuses the input historical power generation data and meteorological warning information, it adopts a multi-source data fusion algorithm. The specific steps are as follows: Step S4.1: Preprocess the historical power generation data, including data cleaning, missing value filling, and outlier detection; Step S4.2: Obtain meteorological warning information, which includes the change trends of irradiance, temperature, humidity, and cloud cover within the next 4 hours; Step S4.3: Input the preprocessed historical power generation data and meteorological warning information into a deep learning algorithm to train a prediction model.

6. The photovoltaic energy storage intelligent management system based on the microservice architecture according to claim 5, characterized in that The steps further include: Step S4.4: Use the trained prediction model to output the power curve for the next 4 hours; analyze the fluctuation amplitude and fluctuation frequency of the power curve; divide the power fluctuation degree into low, medium, and high levels based on the analysis results; and determine the safety factor used in the reserved capacity calculation in the energy storage buffer capacity management module according to the fluctuation level.

7. A photovoltaic energy storage intelligent management system based on a microservice architecture according to claim 5, characterized in that, The steps further include: Step S4.5: Optimize the allocation of the energy storage buffer capacity while considering the charge and discharge efficiency of the energy storage system and the battery life factor.

8. The photovoltaic energy storage intelligent management system based on the microservice architecture according to claim 6, characterized in that, After determining the safety factor and reserved capacity, the energy storage buffer capacity management module further performs the following optimization operations: Step S4.4.1: Allocate part of the reserved capacity as standby capacity; Step S4.4.2: Allocate the charge and discharge operations of the energy storage capacity according to the power fluctuation period, fluctuation degree, and fluctuation level; Step S4.4.3: Dynamically adjust the allocation strategy of the energy storage capacity in combination with the power grid load demand and electricity price signal.

9. A photovoltaic energy storage intelligent management system based on a microservice architecture according to claim 1, characterized in that, The data compression algorithm adopted by the edge computing optimization microservice module includes at least one of the wavelet transform compression algorithm, principal component analysis compression algorithm, or sparse representation compression algorithm.

10. A photovoltaic energy storage intelligent management system based on a microservice architecture according to claim 1, characterized in that It also includes a visualization and alarm module, configured to: construct a multi-dimensional visualization interface to display the key operation parameters and indicators of the interaction between the photovoltaic array, energy storage device, and power grid; Analyze the operation parameters in real time based on preset rules; when it is detected that the key parameters continuously deviate from the normal range and exceed the set threshold, generate and send alarm information.

Images