A multi-scene adaptive user-side energy storage operation and maintenance benefit linkage optimization device

Abstract

This invention specifically relates to a multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device. This device includes a user-side energy storage cabinet, internally housing an analog data acquisition module, a digital control module, and a power supply module. The analog data acquisition module, digital control module, and power supply module are isolated from each other by metal partitions to reduce electromagnetic interference. This multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device establishes a linkage channel between energy storage device operation data, scenario characteristic parameters, and revenue calculation data, ensuring the safe operation of the lithium iron phosphate energy storage system. It provides technical support for the efficient, economical, and safe operation of user-side energy storage systems and is suitable for the implementation and large-scale promotion of various user-side energy storage projects.

Description

Technical Field

[0001] This invention relates to the field of energy storage and intelligent operation and maintenance technology, and in particular to a multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device. Background Technology

[0002] With the accelerated construction of new power systems, user-side energy storage, as a key hub between distributed energy resources and end-user energy consumption, has expanded its application scenarios from single functions to diversified and complex ones. Currently, in scenarios such as industrial and commercial parks, residential communities, new energy charging stations, and data centers, energy storage systems not only undertake the basic functions of peak shaving and demand management, but are also gradually integrating into the grid ancillary services market, becoming an important vehicle for users to obtain multiple benefits. This transformation is driving user-side energy storage to shift from "policy-driven" to "market-driven," and users are placing higher demands on the operational efficiency, profitability, and life-cycle reliability of energy storage systems.

[0003] The differentiated characteristics of various scenarios (such as load fluctuations, space constraints, and safety requirements) and the diverse types of components (lithium-ion batteries, flow batteries, supercapacitors, etc.) make it difficult for traditional general-purpose operation and maintenance models to meet personalized needs. For example, industrial and commercial scenarios need to ensure equipment stability during high-load periods, community scenarios need to balance safety protection and resident experience, charging station scenarios need to maintain the continuity of fast charging services, and data centers require absolutely reliable power supply under extreme conditions. These differences in needs, coupled with issues such as equipment aging and battery degradation, pose serious challenges to the accuracy, dynamism, and economy of operation and maintenance technologies.

[0004] Currently, existing technologies face the following development bottlenecks: (1) Insufficient scene adaptability Existing operation and maintenance technologies mostly adopt uniform standards, lacking in-depth exploration of scenario requirements. For example, industrial and commercial scenarios have not optimized equipment response strategies for electricity price-sensitive periods, community scenarios have not established quantitative control models for noise and safety, and charging station scenarios have not developed battery protection mechanisms under high load impacts. This "one-size-fits-all" approach leads to a misallocation of operation and maintenance resources, resulting in contradictions of over-maintenance or insufficient protection in some scenarios.

[0005] (2) Lack of linkage between operation and maintenance and revenue Traditional operations and maintenance (O&M) focuses on equipment health as a core indicator, failing to deeply integrate operational data with revenue models. For example, it doesn't adjust charging and discharging strategies based on real-time electricity price fluctuations, nor does it dynamically plan maintenance windows in conjunction with ancillary service market demand, leading to a disconnect between O&M costs and revenue. While some systems do collect data, they lack intelligent analysis tools, making it impossible to transform monitoring information into actionable operational decisions.

[0006] (3) Lagging intelligence and dynamic response Current technologies still rely on fixed-cycle inspections and preset threshold alarms, making it difficult to cope with sudden changes in electricity load and policy adjustments. For example, when electricity pricing policies change temporarily, the system cannot quickly reallocate maintenance resources; it also lacks adaptive balancing strategies to address the problem of increasing battery pack inconsistency. Furthermore, the system has weak dynamic accounting capabilities for maintenance costs and benefits, making it difficult for users to assess long-term investment returns.

[0007] Based on the above, this invention proposes a multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device. Summary of the Invention

[0008] To overcome the shortcomings of existing technologies, this invention provides a simple and efficient multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device.

[0009] This invention is achieved through the following technical solution: A multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device includes a user-side energy storage cabinet, which is equipped with an analog acquisition module, a digital control module and a power supply module. The analog acquisition module, digital control module and power supply module are isolated by metal partitions to reduce electromagnetic interference. The front panel of the user-side energy storage cabinet is equipped with status indicator lights and a debugging serial port, while the rear panel is equipped with sensor interfaces, power interfaces, and communication interfaces. All interfaces are protected against reverse connection. The power module uses a lithium iron phosphate battery; The analog acquisition module receives data from the sensing terminal via the MQTT protocol and forwards it to the cloud; The digital control module calculates the health status based on the characteristics of lithium iron phosphate batteries, uses a multi-objective particle swarm optimization algorithm to output the optimal charging and discharging strategy, and generates a visual report.

[0010] The power module adopts a 9-36VDC wide voltage input, with a built-in DC-DC converter and low dropout regulator, and the output voltage ripple is ≤10mV to ensure the accuracy of lithium iron phosphate battery parameter acquisition and avoid data deviation caused by power supply fluctuations.

[0011] The analog acquisition module also includes sensing layer software deployed on the sensing terminal and edge layer software deployed on the edge gateway; The perception layer software collects the operating parameters of the lithium iron phosphate battery, removes outliers, and sends alarm information to the edge gateway for operating parameters that exceed a custom threshold. The edge layer software receives and stores data from the sensing terminal, analyzes scene labels based on the improved K-means clustering algorithm, and sends real-time data and statistical data to the cloud.

[0012] The processing flow of the perception layer software is as follows: Step S1.1: Collect the voltage, current and temperature parameters of the lithium iron phosphate battery at a fixed frequency of 5Hz, with an analog-to-digital conversion accuracy of 12 bits; Step S1.2: Use the moving average filtering algorithm (window size 5) and the 3σ criterion to remove outliers to ensure data validity; Step S1.3: When the parameters of the lithium iron phosphate battery exceed the custom threshold (such as temperature > 55℃, voltage > 3.65V / cell), immediately trigger a local alarm (relay output controls the alarm light) and send alarm information to the edge gateway at the same time. The edge layer software processing flow is as follows: Step S2.1: Receive data from the sensing terminal via the MQTT protocol, encapsulate it in the format of "Device ID-Timestamp-Parameter Type-Value", and store it in the local SQLite database (data is retained for 7 days). Step S2.2: Based on the improved K-means clustering algorithm, input load fluctuation amplitude, noise value and electricity price sensitivity, and output scene labels for industrial / commercial / community / charging station / data center, with a clustering accuracy of ≥95%; Step S2.3: Send real-time data and statistical data to the cloud; Meanwhile, it has a pre-stored emergency response plan for lithium iron phosphate battery energy storage (such as overheating: cut off the charging and discharging circuit → start the cooling fan → report to the cloud), with a response time of less than 500ms; The edge layer software forwards real-time data to the cloud via UDP at a frequency of once per minute, and forwards statistical data to the cloud via TCP at a frequency of once every 15 minutes, prioritizing the transmission of emergency data.

[0013] The digital control module also includes cloud-based software deployed on the Inspur cloud platform; the cloud-based software receives and stores the operating parameters and business data of the lithium iron phosphate battery, calculates the health of the lithium iron phosphate battery, outputs the optimal charging and discharging strategy using the multi-objective particle swarm optimization algorithm (MOPSO), and generates a visual report.

[0014] The cloud-layer software processes the following steps: Step S3.1: Store the time-series data composed of lithium iron phosphate battery operating parameters into InfluxDB (retain for 1 year), and store the business data composed of scenario tags and revenue data into RDSMySQL (automatically backed up daily and retained for 30 days). Step S3.2: When the average monthly temperature exceeds 35℃, calculate the health status based on the characteristics of lithium iron phosphate batteries. The calculation formula is as follows: SOH = (Current Capacity / Rated Capacity) × 100% × (1 - 0.005 × Average Monthly Temperature - 0.002 × Average Monthly Power) Step S3.3: Use the Multi-Objective Particle Swarm Optimization (MOPSO) algorithm to output the optimal charging and discharging strategy. The objective function is as follows: maxF = (Peak-Valley Arbitrage Revenue + Ancillary Service Revenue) - (Operation and Maintenance Costs + Depreciation Costs) The constraints are: charge / discharge rate ≤ 1C, and maintenance should be avoided during peak hours. Step S3.4: Use Inspur Cloud DataV to build a dedicated dashboard for lithium iron phosphate battery energy storage, including SOH trends, revenue composition and operation and maintenance costs. Generate a PDF report monthly through Inspur Cloud Document Service (Doc Works) and store it in Object Storage Service (OSS) for users to download.

[0015] The cloud-based software interfaces with Inspur Cloud's Unified Identity Authentication Service (IAM), achieving two-way authentication between devices and the cloud based on the SM2 national cryptographic algorithm; it establishes a dedicated channel between the field and the cloud via Inspur Cloud's Direct Connect, with bandwidth ≥100Mbps and latency ≤30ms; it integrates Inspur Cloud's Message Center, supporting alarms from multiple channels including SMS, email, and WeChat Work, automatically generating maintenance work orders when an alarm is triggered; and it calls Inspur Cloud's ModelArts machine learning platform to achieve automatic retraining and deployment of optimized models.

[0016] The beneficial effects of this invention are: the multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device opens up the linkage channel between energy storage equipment operation data, scenario characteristic parameters and revenue calculation data, ensuring the safe operation of lithium iron phosphate energy storage system, providing technical support for the efficient, economical and safe operation of user-side energy storage system, and is suitable for the implementation and large-scale promotion of various user-side energy storage projects. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Appendix Figure 1 This is a schematic diagram of the multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device of the present invention. Detailed Implementation

[0019] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions in the embodiments of this invention will be clearly and completely described below in conjunction with the embodiments of this invention. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.

[0020] This multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device includes The user-side energy storage cabinet contains an analog acquisition module (sensor signal processing), a digital control module (MCU and communication module), and a power supply module (12V / 5V regulated output). The analog acquisition module, digital control module, and power supply module are isolated by metal partitions to reduce electromagnetic interference. The front panel of the user-side energy storage cabinet is equipped with status indicator lights and a debugging serial port, while the rear panel is equipped with sensor interfaces, power interfaces, and communication interfaces. All interfaces are protected against reverse connection. The power module uses a lithium iron phosphate battery; The analog acquisition module receives data from the sensing terminal via the MQTT protocol and forwards it to the cloud; The digital control module calculates the health status based on the characteristics of lithium iron phosphate batteries, uses a multi-objective particle swarm optimization algorithm to output the optimal charging and discharging strategy, and generates a visual report.

[0021] The power module adopts a wide input voltage (9-36VDC) design, with a built-in DC-DC converter and low dropout regulator (LDO, model TPS7A4700), and the output voltage ripple is ≤10mV, so as to ensure the accuracy of lithium iron phosphate battery parameter acquisition and avoid data deviation caused by power supply fluctuations.

[0022] The analog acquisition module also includes sensing layer software deployed on the sensing terminal and edge layer software deployed on the edge gateway; The perception layer software collects the operating parameters of the lithium iron phosphate battery, removes outliers, and sends alarm information to the edge gateway for operating parameters that exceed a custom threshold. The edge layer software receives and stores data from the sensing terminal, analyzes scene labels based on the improved K-means clustering algorithm, and sends real-time data and statistical data to the cloud.

[0023] The processing flow of the perception layer software is as follows: Step S1.1: Collect the voltage, current and temperature parameters of the lithium iron phosphate battery at a fixed frequency of 5Hz, with an analog-to-digital conversion accuracy of 12 bits; Step S1.2: Use the moving average filtering algorithm (window size 5) and the 3σ criterion to remove outliers to ensure data validity; Step S1.3: When the parameters of the lithium iron phosphate battery exceed the custom threshold (such as temperature > 55℃, voltage > 3.65V / cell), immediately trigger a local alarm (relay output controls the alarm light) and send alarm information to the edge gateway at the same time. The edge layer software processing flow is as follows: Step S2.1: Receive data from the sensing terminal via the MQTT protocol, encapsulate it in the format of "Device ID-Timestamp-Parameter Type-Value", and store it in the local SQLite database (data is retained for 7 days). Step S2.2: Based on the improved K-means clustering algorithm, input load fluctuation amplitude, noise value and electricity price sensitivity, and output scene labels for industrial / commercial / community / charging station / data center, with a clustering accuracy of ≥95%; Step S2.3: Send real-time data and statistical data to the cloud; Meanwhile, it has a pre-stored emergency response plan for lithium iron phosphate battery energy storage (such as overheating: cut off the charging and discharging circuit → start the cooling fan → report to the cloud), with a response time of less than 500ms; The edge layer software forwards real-time data to the cloud via UDP at a frequency of once per minute, and forwards statistical data to the cloud via TCP at a frequency of once every 15 minutes, prioritizing the transmission of emergency data.

[0024] The digital control module also includes cloud-based software deployed on the Inspur cloud platform; the cloud-based software receives and stores the operating parameters and business data of the lithium iron phosphate battery, calculates the health of the lithium iron phosphate battery, outputs the optimal charging and discharging strategy using the multi-objective particle swarm optimization algorithm (MOPSO), and generates a visual report.

[0025] The cloud-layer software processes the following steps: Step S3.1: Store the time-series data composed of lithium iron phosphate battery operating parameters into InfluxDB (retain for 1 year), and store the business data composed of scenario tags and revenue data into RDSMySQL (automatically backed up daily and retained for 30 days). Step S3.2: When the average monthly temperature exceeds 35℃, calculate the health status based on the characteristics of lithium iron phosphate batteries. The calculation formula is as follows: SOH = (Current Capacity / Rated Capacity) × 100% × (1 - 0.005 × Average Monthly Temperature - 0.002 × Average Monthly Power) Step S3.3: Use the Multi-Objective Particle Swarm Optimization (MOPSO) algorithm to output the optimal charging and discharging strategy. The objective function is as follows: maxF = (Peak-Valley Arbitrage Revenue + Ancillary Service Revenue) - (Operation and Maintenance Costs + Depreciation Costs) The constraints are: charge / discharge rate ≤ 1C, and maintenance should be avoided during peak hours. Step S3.4: Use Inspur Cloud DataV to build a dedicated dashboard for lithium iron phosphate battery energy storage, including SOH trends, revenue composition and operation and maintenance costs. Generate a PDF report monthly through Inspur Cloud Document Service (Doc Works) and store it in Object Storage Service (OSS) for users to download.

[0026] The cloud-based software interfaces with Inspur Cloud's Unified Identity Authentication Service (IAM), achieving two-way authentication between devices and the cloud based on the SM2 national cryptographic algorithm; it establishes a dedicated channel between the field and the cloud via Inspur Cloud's Direct Connect, with bandwidth ≥100Mbps and latency ≤30ms; it integrates Inspur Cloud's Message Center, supporting alarms from multiple channels including SMS, email, and WeChat Work, automatically generating maintenance work orders when an alarm is triggered; and it calls Inspur Cloud's ModelArts machine learning platform to achieve automatic retraining and deployment of optimized models.

[0027] Example 1 The following explanation will be based on an example of energy storage scenarios in industrial and commercial parks (peak-valley arbitrage + distributed photovoltaic consumption).

[0028] I) Basic Scenarios Energy storage system configuration: 1MWh lithium iron phosphate battery energy storage system, equipped with 500kW PCS, connected to the park's 10kV distribution network; Core Needs: To reduce electricity costs through peak-valley arbitrage (local peak electricity price: RMB 1.2 / kWh, valley electricity price: RMB 0.3 / kWh, flat electricity price: RMB 0.7 / kWh), while simultaneously absorbing 2MW of distributed photovoltaic power generation in the industrial park to avoid curtailment; Interference factors: The industrial park's production load fluctuates greatly (2000-3000kW during the daytime production period, dropping to below 500kW at night), and the photovoltaic output fluctuates due to weather conditions (1800kW at midday on sunny days, dropping to 300kW on cloudy days).

[0029] II. Hardware and Software Deployment Scheme The hardware deployment is as follows: Multi-scenario sensing terminal: Installed in the energy storage cabinet, it is equipped with a battery parameter acquisition unit (BQ79616), an industrial and commercial load sensor (CSNS1500), and a revenue meter (DSZ1352); the load sensor is connected to the park's main distribution cabinet to monitor the production load in real time; the meter simultaneously measures the energy storage charging and discharging amount and the photovoltaic grid-connected power. Edge gateway: Deployed in the park's power distribution room, it connects to the PCS and photovoltaic inverter via RS485, accesses Inspur Cloud via a 4G module, and is configured with a 16GB MicroSD card to cache 7 days of data; Execution control unit: connects the energy storage BMS and PCS, controls the charge and discharge rate (adjustable from 0.2-1C), and simultaneously links the output limiting function of the photovoltaic inverter.

[0030] The software configuration is as follows: Sensing layer software: Set the battery parameter acquisition frequency to 10Hz (higher than the default 5Hz), temperature alarm threshold to 55℃, and overvoltage threshold to 3.65V / cell. Edge layer software: The scene recognition module is configured with an "industrial and commercial" label, and the load fluctuation amplitude threshold is set to 500kW (ΔI>500kW is judged as a sudden load change). The emergency strategy library adds the logic of "energy storage to replenish power when photovoltaic output drops sharply"; Cloud-based software: The monthly average temperature correction coefficient for the health assessment model is α=0.005 (the average daily temperature in the park is 32℃ in summer and 15℃ in winter). The weights of the operation and maintenance-revenue optimization objective function are set to "peak-valley arbitrage revenue accounts for 70%, and photovoltaic consumption accounts for 30%".

[0031] III. Implementation Process of Operation and Maintenance-Revenue Linkage Strategy Data collection and scene recognition (daily 00:00-24:00), the process is as follows: The sensing terminal collects battery voltage, current, and temperature (25-38℃) every minute, the load sensor collects production load every 10 seconds (e.g., 2800kW load at 14:00 and 600kW load at 23:00), and the meter uploads charging and discharging power every minute (e.g., 200MWh charging during off-peak hours and 180MWh discharging during peak hours). The edge gateway runs a scene recognition algorithm every 15 minutes, based on "load fluctuation range of 800kW and electricity price sensitivity of 1.7 (peak-valley difference / flat price)", and stably outputs "industrial and commercial - production period" or "industrial and commercial - non-production period" labels.

[0032] The dynamic optimization strategy is generated as follows: The cloud platform generates a strategy daily at 22:00 based on the next day's photovoltaic output forecast (predicted to be 1600kW at noon the following day via Inspur Cloud's connection to local meteorological data) and production load plan using the MOPSO algorithm. During off-peak hours (00:00-08:00): production load is low (600-800kW), photovoltaic power output is zero, and energy storage is controlled to charge at a rate of 0.5C until the SOC reaches 90% (approximately 8 hours of charging, charging amount of 800kWh). During normal periods (08:00-10:00, 15:00-18:00): Production load is 1500-2000kW, photovoltaic output gradually increases / decreases, and energy storage is controlled to discharge at a 0.5C rate to supplement the load gap, while absorbing photovoltaic power (e.g., at 10:00, photovoltaic output is 800kW, energy storage discharge is 400kW, totaling 1200kW to meet the load). Peak hours (10:00-15:00, 18:00-22:00): Production load is 2000-3000kW, photovoltaic output reaches 1600kW at noon, and energy storage is controlled to discharge at a 0.5C rate (maximum 500kW) to meet the load in coordination with photovoltaic (e.g., at 12:00, photovoltaic output is 1600kW, energy storage discharges 500kW, covering a load of 2100kW), avoiding the need to purchase electricity from the grid; Emergency Adjustment: At 13:00 one day, a sudden rainstorm caused the photovoltaic output to drop sharply from 1600kW to 400kW. Within 100ms, the edge gateway identified the load gap (2800kW-400kW=2400kW, while the grid supply was only 2000kW) and immediately issued an instruction to increase the energy storage discharge rate to 0.5C to make up for the 400kW gap and avoid a power outage in the park.

[0033] The revenue calculation and feedback optimization process is as follows: Daily Operations and Maintenance - Revenue Report: Peak-valley arbitrage revenue = (400kWh × 1.2 yuan / kWh + 400kWh × 0.7 yuan / kWh) - (800kWh × 0.3 yuan / kWh) = 520 yuan; Operations and maintenance costs (regular inspections + battery balancing) 50 yuan; Net revenue for the day: 470 yuan; The actual returns are compared with the predicted returns every week (with a deviation of less than 5%), and the cloud platform automatically fine-tunes and optimizes the model parameters (such as extending the midday discharge duration by 1 hour to further increase the peak discharge volume).

[0034] IV. Implementation Results Economic impact: Monthly net income of RMB 14,100, representing a 56.9% reduction in electricity costs for the industrial park compared to the traditional model; Equipment performance: After 6 months of operation in the first year, the battery SOH decreased from 98% to 96.5% (a degradation rate of 1.5%), which is lower than the industry average degradation rate (2.5%). Energy efficiency: The photovoltaic absorption rate increased from 82% to 98%, reducing curtailed solar power by 6 MWh per month and reducing carbon emissions in the park by 18 tons. Example

[0035] The following explanation will be based on the energy storage scenario of new energy charging stations (load smoothing + fast charging guarantee).

[0036] I) Basic Scenarios Energy storage system configuration: 200kWh lithium iron phosphate battery energy storage system (single cell voltage 3.2V, 300 series connected to form a 960V DC bus), equipped with 200kW PCS, directly connected to 3 120kW fast charging piles in the charging station (total power 360kW). Core requirements: Smooth out the impact of fast charging load (the current reaches 50A when charging a single pile, and the total current is 150A when charging 3 piles at the same time, which can easily cause grid voltage fluctuations) and ensure the continuity of charging services (the peak charging period is from 8:00 am to 10:00 pm every day, serving an average of 100 electric vehicles per day). Interference factors: During peak charging periods, vehicles arrive in large numbers (such as 9:00-11:00 AM and 6:00-8:00 PM, with up to 15 vehicles charging per hour), and the power grid's capacity is limited (it can only support a continuous load of 200kW, exceeding which may trigger overcurrent protection).

[0037] II. Hardware and Software Deployment Scheme The hardware deployment is as follows: Multi-scenario sensing terminal: installed in the energy storage control cabinet, configured with a battery parameter acquisition unit (BQ79616), a charging station impact detector (TA1000-5), and a charging gun status sensor (newly added, monitoring the connection status of the charging guns of 3 charging piles); the impact detector is connected to the main DC bus of the fast charging pile to monitor the rate of change of current; Edge gateway: Deployed in the charging station control room, it connects to 3 fast charging piles via CAN bus to obtain charging demand (such as vehicle battery SOC, target charging amount) in real time, and connects to Inspur Cloud via LoRa module (transmission distance 500m, suitable for open environment of charging station); Execution control unit: connects the energy storage BMS and PCS, supports millisecond-level charge and discharge switching (response time <200ms), and also connects to the power distribution module of fast charging piles to dynamically adjust the output power of a single pile.

[0038] The software configuration is as follows: Sensing layer software: Set the current change rate monitoring threshold to 100A / s (dI / dt>100A / s is judged as a load impact), and set the upper limit of battery charging rate to 0.8C (to avoid battery overheating under high load). Edge layer software: The scene recognition module is equipped with a "charging station" label, and the data collection frequency during peak charging hours (8:00-22:00) is increased to 15Hz. The emergency strategy library adds logic for "energy storage to replenish power when the grid is overcurrent" and "energy storage to maintain power supply to other piles when a single pile fails". Cloud-based software: The monthly average power correction coefficient β=0.002 for the health assessment model (the daily average charging and discharging power of the charging station is 150kW, accounting for 75% of the rated power), and the weights of the operation and maintenance-revenue optimization objective function are set as "charging service continuity accounts for 60%, and load smoothing accounts for 40%".

[0039] III. Implementation Process of Operation and Maintenance-Revenue Linkage Strategy Data collection and scene recognition (daily charging peak hours 8:00-22:00) are performed as follows: The sensing terminal collects the fast charging bus current (e.g., when 3 electric vehicles are connected at 9:30, the current rises from 0 to 150A, with a change rate of 1500A / s), battery temperature (30-42℃ during charging), and charging gun connection status every 100ms (e.g., when the second charging gun is disconnected at 10:00, the status changes to "idle"). The edge gateway runs a scene recognition algorithm every 5 minutes, and outputs the label "charging station - peak period" based on "current change rate 1200A / s, number of charging gun connections 3"; and outputs the label "charging station - off-peak period" during off-peak periods (22:00-8:00 the next day).

[0040] The dynamic optimization strategy is generated as follows: The cloud platform generates a strategy daily at 7:00 AM based on historical charging data (such as the average charging demand of 280kW during last week's morning peak) and the grid power supply limit (200kW): Peak charging hours (8:00-22:00): Load surge response: When 3 charging piles are charging simultaneously (requiring 280kW), the grid can only provide 200kW. The edge gateway immediately instructs the energy storage to discharge at a 0.4C rate (80kW), with a total power supply of 280kW, avoiding the power limitation of the charging pile (in the original mode, the power of a single pile would be reduced to 80kW, and the charging time would be extended by 50%). Dynamic power allocation: At 14:00, the first charging pile vehicle is fully charged and disconnected (demand drops to 180kW). The energy storage automatically reduces the discharge rate to 0.1C (20kW), the grid supply is 160kW, and the total power is maintained at 180kW to avoid excessive discharge of the energy storage. During off-peak charging periods (22:00-8:00 the next day): With ample power supply from the grid, the energy storage is charged at a rate of 0.5C (100kW), and 200kWh is fully charged in 8 hours to reserve electricity for the next day's peak. Emergency Adjustment: At 19:00, the grid voltage suddenly dropped to 0.9pu (rated voltage 0.4kV). The edge gateway triggered the "grid anomaly" emergency strategy, immediately cut off the energy storage discharge, and switched to the "energy storage independent power supply" mode (200kW) to maintain normal charging of 2 charging piles (demand 180kW) until the grid was restored (20 minutes later). During this period, the charging service was not interrupted.

[0041] The revenue calculation and feedback optimization process is as follows: Daily report generated: Monthly net income: 2378 yuan / day × 30 days = 71,300 yuan (after deducting an additional cost of 500 yuan for equipment inspection once a month); Annual net income: RMB 71,300 / month × 12 months - equipment depreciation RMB 150,000 (total price of 1MWh energy storage equipment is RMB 1,500,000, depreciated over 10 years) = RMB 705,600; Annualized rate of return: 705,600 yuan ÷ 1,500,000 yuan = 47.04% (far exceeding the original low-yield model, in line with the level of high-quality projects in the industry).

[0042] Static payback period: Actual initial investment of RMB 1.45 million ÷ annual net income of RMB 705,600 ≈ 2.06 years (ideal situation, not considering electricity price fluctuations and equipment degradation). Dynamic payback period (considering risk factors): The annual equipment depreciation rate is 2% (SOH decreases by 2% annually, and the return in the 5th year drops to 80% of the initial value). Electricity prices increased by 5% annually (with the peak-valley price difference widening year by year); The dynamic payback period is approximately 3.5 years (still better than the industry average of 5-8 years, making it commercially viable).

[0043] Real-world case study: A 1MWh industrial and commercial energy storage project (commissioned in 2023) generated annual peak-valley arbitrage revenue of 480,000 yuan, demand response revenue of 150,000 yuan, and leasing revenue of 120,000 yuan, resulting in an annual net income of 620,000 yuan, which is close to the calculations in this model. Risk boundary test: Even in extreme scenarios (half of photovoltaic curtailment, demand response times reduced to 2 times / month), the annual net income can still reach 450,000 yuan, and the dynamic payback period is extended to 5.2 years, which is still within the reasonable range of the industry.

[0044] This multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device integrates multi-scenario sensing hardware and intelligent algorithm models, opening up a linkage channel between energy storage equipment operation data, scenario characteristic parameters (such as electricity price, electricity load, and environmental conditions) and revenue calculation data. It breaks through the limitations of traditional user-side energy storage operation and maintenance, which are "heavy on equipment monitoring and light on revenue linkage" and "single-scenario adaptation and lack of flexible adjustment in multiple scenarios". It provides technical support for the efficient, economical and safe operation of user-side energy storage systems and is suitable for the implementation and large-scale promotion of various user-side energy storage projects.

[0045] Compared with existing technologies, it has the following characteristics: (1) Parameter coordination is achieved: the hardware loads the monitoring thresholds (temperature, voltage, rate) of lithium iron phosphate batteries by default, and the software synchronously calls the exclusive health assessment model and optimization algorithm, without the need for manual parameter adjustment; (2) Time synchronization was achieved: Time synchronization between sensing terminals, edge gateways and the cloud was achieved through cloud network time service (NTP), with an error of ≤5ms, which can ensure the consistency of timestamps of lithium iron phosphate battery operation data; (3) Secure collaboration is achieved: the hardware integrates a data encryption chip (storage device key), and the software regularly detects security risks through cloud vulnerability scanning service (VSS), ensuring the safe operation of the lithium iron phosphate energy storage system; (4) High scalability: The hardware reserves 8 analog input interfaces and the software adopts a modular architecture. When adding a new scene, only scene recognition parameters and strategy templates need to be added, without modifying the core code.

[0046] The embodiments described above are merely one specific implementation of the present invention. Ordinary changes and substitutions made by those skilled in the art within the scope of the technical solution of the present invention should be included within the protection scope of the present invention.

Claims

1. A multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device, characterized in that: It includes a user-side energy storage cabinet, which contains an analog acquisition module, a digital control module, and a power supply module; the analog acquisition module, digital control module, and power supply module are isolated from each other by metal partitions to reduce electromagnetic interference; The front panel of the user-side energy storage cabinet is equipped with status indicator lights and a debugging serial port, while the rear panel is equipped with sensor interfaces, power interfaces, and communication interfaces. All interfaces are protected against reverse connection. The power module uses a lithium iron phosphate battery; The analog acquisition module receives data from the sensing terminal via the MQTT protocol and forwards it to the cloud; The digital control module calculates the health status based on the characteristics of lithium iron phosphate batteries, uses a multi-objective particle swarm optimization algorithm to output the optimal charging and discharging strategy, and generates a visual report.

2. The multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device according to claim 1, characterized in that: The power module adopts a 9-36VDC wide voltage input, with a built-in DC-DC converter and low dropout regulator, and the output voltage ripple is ≤10mV to ensure the accuracy of lithium iron phosphate battery parameter acquisition and avoid data deviation caused by power supply fluctuations.

3. The multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device according to claim 1, characterized in that: The analog acquisition module also includes sensing layer software deployed on the sensing terminal and edge layer software deployed on the edge gateway; The perception layer software collects the operating parameters of the lithium iron phosphate battery, removes outliers, and sends alarm information to the edge gateway for operating parameters that exceed a custom threshold. The edge layer software receives and stores data from the sensing terminal, analyzes scene labels based on the improved K-means clustering algorithm, and sends real-time data and statistical data to the cloud.

4. The multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device according to claim 3, characterized in that: The processing flow of the perception layer software is as follows: Step S1.1: Collect the voltage, current and temperature parameters of the lithium iron phosphate battery at a fixed frequency of 5Hz, with an analog-to-digital conversion accuracy of 12 bits; Step S1.2: Use the moving average filtering algorithm and the 3σ criterion to remove outliers to ensure data validity. The window size of the moving average filtering algorithm is 5. Step S1.3: When the parameters of the lithium iron phosphate battery exceed the custom threshold, immediately trigger a local alarm and send alarm information to the edge gateway.

5. The multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device according to claim 3, characterized in that: The edge layer software processing flow is as follows: Step S2.1: Receive data from the sensing terminal via the MQTT protocol, encapsulate it in the format of "Device ID-Timestamp-Parameter Type-Value", store it in the local SQLite database, and retain the data for 7 days; Step S2.2: Based on the improved K-means clustering algorithm, input load fluctuation amplitude, noise value and electricity price sensitivity, and output scene labels for industrial / commercial / community / charging station / data center; Step S2.3: Send real-time data and statistical data to the cloud; Meanwhile, a dedicated emergency response plan for lithium iron phosphate battery energy storage is pre-stored, with a response time of less than 500ms.

6. The multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device according to claim 5, characterized in that: The edge layer software forwards real-time data to the cloud via UDP at a frequency of once per minute, and forwards statistical data to the cloud via TCP at a frequency of once every 15 minutes, prioritizing the transmission of emergency data.

7. The multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device according to claim 1, characterized in that: The digital control module also includes cloud-based software deployed on the Inspur cloud platform. The cloud-based software receives and stores the operating parameters and business data of the lithium iron phosphate battery, calculates the health of the lithium iron phosphate battery, outputs the optimal charging and discharging strategy using the multi-objective particle swarm optimization algorithm (MOPSO), and generates a visual report.

8. The multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device according to claim 7, characterized in that: The cloud-layer software processes the following steps: Step S3.1: Store the time-series data composed of lithium iron phosphate battery operating parameters into the InfluxDB database and retain it for 1 year; store the business data composed of scenario tags and revenue data into the RDSMySQL database, automatically back it up daily, and retain it for 30 days. Step S3.2: When the average monthly temperature exceeds 35℃, calculate the health status based on the characteristics of lithium iron phosphate batteries. The calculation formula is as follows: SOH = (Current Capacity / Rated Capacity) × 100% × (1 - 0.005 × Average Monthly Temperature - 0.002 × Average Monthly Power) Step S3.3: The optimal charging and discharging strategy is output using the Multi-Objective Particle Swarm Optimization (MOPSO) algorithm. The objective function is as follows: maxF = (Peak-Valley Arbitrage Revenue + Ancillary Service Revenue) - (Operation and Maintenance Costs + Depreciation Costs) The constraints are: charge / discharge rate ≤ 1C, and maintenance should be avoided during peak hours. Step S3.4: Use Inspur Cloud DataV to build a dedicated dashboard for lithium iron phosphate battery energy storage, including SOH trends, revenue composition and operation and maintenance costs. Generate a PDF report monthly through Inspur Cloud DocWorks document service and store it in object storage OSS for users to download.

9. The multi-scenario adaptive user-side energy storage operation and maintenance revenue linkage optimization device according to claim 7, characterized in that: The cloud-based software interfaces with Inspur Cloud's Unified Identity Authentication Service (IAM) to achieve two-way authentication between devices and the cloud based on the SM2 national cryptographic algorithm. It establishes a dedicated channel between the field and the cloud via Inspur Cloud's Direct Connect dedicated line, with a bandwidth of ≥100Mbps and a latency of ≤30ms. It integrates Inspur Cloud's Message Center, supporting alarms from multiple channels such as SMS, email, and WeChat Work, and automatically generates maintenance work orders when an alarm is triggered. It also calls Inspur Cloud's ModelArts machine learning platform to achieve automatic retraining and deployment of optimized models.

Images