An integrated distributed energy storage UPS system

Abstract

This invention discloses an integrated distributed energy storage UPS system, relating to the field of distributed energy storage technology. The invention includes steps S1: real-time data acquisition; S2: parameter normalization processing; S3: dynamic weight calculation; S4: dynamic load command allocation; S5: multi-objective coordinated control execution; S6: status feedback and parameter self-tuning; and S7: seamless fault switching mechanism. This invention constructs a time-varying weight function by real-time evaluation of core parameters such as the state of charge, health, temperature, and internal resistance of each energy storage unit. This function can dynamically adjust the output ratio of each unit, achieving intelligent load allocation that matches the aging state of the equipment and environmental conditions. By introducing a health degradation factor, the system can quantify the nonlinear degradation characteristics of the battery pack, prioritizing units in good condition to bear the base load, while providing intermittent rest periods for high-loss units.

Description

Technical Field

[0001] This invention belongs to the field of distributed energy storage technology, and in particular relates to an integrated distributed energy storage UPS system. Background Technology

[0002] Distributed energy storage is an energy storage method that involves installing energy storage devices in different locations. It allows for flexible energy storage and release based on user needs and grid conditions. This approach improves energy utilization efficiency, enhances grid stability and reliability, and reduces energy loss during transmission.

[0003] A UPS (Uninterruptible Power Supply) system is a device that provides a continuous and stable power supply to a load when the power grid fails or malfunctions. It mainly consists of batteries, inverters, and chargers, and can quickly switch to a backup power source to ensure the normal operation of equipment and prevent data loss and equipment damage.

[0004] Existing technologies mostly employ static allocation strategies based on fixed ratios or single parameters, which are difficult to effectively address the dynamic characteristic differences in multi-unit collaborative operation. Fixed allocation modes ignore the state differences between units, causing high-health units and aging units to bear the same load for a long time, accelerating the irreversible performance degradation of local units. Especially in scenarios of sudden load changes, the current sharing controller cannot dynamically adjust the output ratio, causing high-internal-resistance units to operate continuously under overload, triggering a vicious cycle of increased local temperature rise and sudden capacity drop. To address the above problems, the following solutions are proposed. Summary of the Invention

[0005] The purpose of this invention is to provide an integrated distributed energy storage UPS system. By constructing a weighting function with time-varying characteristics, the output ratio of each unit can be dynamically adjusted, solving the problem that existing technologies cannot effectively cope with the dynamic characteristic differences in the collaborative operation of multiple units.

[0006] To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution:

[0007] This invention relates to an integrated distributed energy storage UPS system, comprising:

[0008] Step S1, Real-time Data Acquisition: By deploying a BMS system and sensors, real-time data on output current, SOC, SOH, temperature, internal resistance, and bus voltage of each energy storage unit are acquired.

[0009] Step S2, Parameter Normalization Processing: The collected SOC, internal resistance, and temperature parameters are standardized, and a health decay factor is constructed.

[0010] Step S3, Dynamic Weight Calculation: The load allocation weight of each energy storage unit is generated by formula and the weight is normalized.

[0011] Step S4, Dynamic Load Command Allocation: Generate target current based on total demand current and the weight of each unit, and add SOC balancing adjustment and slope limiting;

[0012] Step S5: Multi-objective coordinated control execution: Target current command is issued through a dual-loop control strategy, and current sharing deviation is monitored in real time;

[0013] Step S6, Status Feedback and Parameter Self-Tuning: Dynamically adjust the weighting coefficients based on the output error and perform boundary checks;

[0014] Step S7, Seamless Fault Switching Mechanism: Monitor the unit status in real time, set the weight to zero and start pre-charging of the backup unit when a fault occurs.

[0015] Furthermore, step S1, real-time data acquisition, specifically includes the following steps:

[0016] Step S11: Install a Hall current sensor at the output terminal of each energy storage unit to collect the output current value I of each unit in real time. i (t);

[0017] Step S12: Periodically obtain the SOC of each energy storage unit through the BMS system. i (t), SOH i (t), internal temperature T i (t) and equivalent series internal resistance R i (t);

[0018] Step S13: Deploy voltage sensors at the DC busbar to monitor the bus voltage U in real time. bus (t);

[0019] Step S14: Upload the data from each unit to the central controller via the CAN bus, and use a timestamp synchronization mechanism to ensure data timing consistency;

[0020] This design can acquire key parameters such as output current, SOC (State of Charge), SOH (State of Health), and temperature of each energy storage unit in real time, as well as DC bus voltage, providing comprehensive operating status data for the system. High-precision sensors and synchronization mechanisms ensure the timeliness and accuracy of the data, laying the foundation for subsequent analysis.

[0021] Furthermore, step S2, parameter normalization processing, specifically includes the following steps:

[0022] Step S21: Standardize the collected raw parameters to eliminate unit differences between different parameters and provide standardized input for subsequent steps. The formulas used are as follows:

[0023]

[0024] In the formula, SOC′ i (t) represents the normalized charge state of the i-th energy storage unit at time t, SOC i (t) represents the actual charge state of the i-th energy storage unit at time t, SOC min The minimum SOC allowed by the system. max R′ is the maximum SOC allowed by the system. i (t) is the normalized equivalent series resistance, R i (t) is the equivalent series internal resistance, R max ,R min These are the preset upper and lower limits of the internal resistance, T. i ′(t) is the normalized temperature value, T i (t) represents the real-time internal temperature, T rated T represents the upper limit of the rated operating temperature of the energy storage unit. min This is the lowest permissible operating temperature of the system.

[0025] Step S22: Construct a health decay factor that combines the effects of SOH and temperature to reflect the current availability of the cell. The formula is:

[0026] α i (t)=1-λ×(1-SOH i (t))-κ×(T i ′(t)-T opt ) 2 ;

[0027] In the formula, α i (t) is the health decay factor, λ is the SOH decay weighting coefficient, κ is the temperature deviation penalty coefficient, and T opt To achieve the optimal value for normalized temperature, SOH i (t) represents the healthy state, T i ′(t) is the normalized internal temperature;

[0028] This design standardizes the raw data to a uniform range (e.g., 0-1), eliminating dimensional differences and facilitating algorithm processing; at the same time, it calculates the health degradation factor, comprehensively assesses the effects of battery aging and temperature, and provides a quantitative basis for dynamic weight allocation.

[0029] Furthermore, step S3, the dynamic weight calculation, specifically includes the following steps:

[0030] Step S31: Calculate the dynamic weight allocation for each unit. The specific formula is as follows:

[0031]

[0032] In the formula, W i (t) represents the dynamic weight allocation of the i-th energy storage unit at time t, α i (t) is the health decay factor, dI i / dt is the output current change rate, and k1, k2, and k3 are all adjustable coefficients that control the sensitivity of SOC, internal resistance, and current change rate, respectively.

[0033] Step S32: Perform weight normalization, the formula is:

[0034]

[0035] In the formula, W i ′(t) represents the normalized weights, N is the total number of energy storage units in the system, and W j (t) represents the dynamic weight allocation of the j-th energy storage unit at time t;

[0036] This design can calculate the real-time allocation weight of each energy storage unit; by introducing dynamic parameters such as SOC priority coefficient and internal resistance suppression coefficient, it can achieve flexible adjustment of load allocation and optimize system efficiency and battery life.

[0037] Furthermore, step S4, dynamic allocation of load instructions, specifically includes the following steps:

[0038] Step S41: Calculate the total system current demand I total (t), the formula is:

[0039]

[0040] In the formula, P load (t) represents the load power, P loss (t) represents the system power loss, U bus (t) represents the DC bus voltage;

[0041] Step S42: Generate the target current for each unit, using the following formula:

[0042]

[0043] In the formula, Let W be the target output current of the i-th energy storage unit at time t. i ′(t) represents the normalized dynamic weights, β is the SOC equilibrium adjustment coefficient, and SOC avg The average SOC of all energy storage units;

[0044] This design can generate target current commands for each cell based on total demand current and dynamic weights; it incorporates SOC equalization adjustment and slope limiting to avoid the impact of sudden current changes on the battery, while ensuring long-term consistency of SOC for each cell.

[0045] Furthermore, step S5, the execution of multi-objective coordinated control, specifically includes the following steps:

[0046] Step S51: Send the target current command to the local controller of each unit via RS485 bus;

[0047] Step S52: Each unit adopts a dual-loop control strategy, including:

[0048] Outer loop: PI regulator tracking proportionality coefficient K p =0.8, integration time T i =0.1s;

[0049] Inner loop: Hysteresis current control, loop width set to ±2% of rated current;

[0050] Step S53: Compare the deviation between the actual output current of each unit and the target value in real time. When ΔI i When the percentage is >5%, dynamic weight recalculation is triggered.

[0051] This design can send the target current command to each unit controller and use a dual-loop control strategy (outer loop PI regulation + inner loop hysteresis control) to achieve accurate tracking; and correct the deviation in real time through a parallel current sharing mechanism to ensure the accuracy of current distribution.

[0052] Furthermore, step S6, state feedback and parameter self-tuning, specifically includes the following steps:

[0053] Step S61: Acquire the actual output current I i (t) real Deviation from target value ε i (t);

[0054] Step S62: Adjust the weighting coefficients based on the error integral, specifically as follows:

[0055] k1←k1+0.01×∫(SOC′ i (t)×ε i (t))dt;

[0056] k2←k2-0.005×∫(R′ i (t)×ε i (t))dt;

[0057] Step S63: Perform a parameter boundary check every 30 minutes to ensure that k1∈[0.5,1.2], k2∈[0.8,1.5], and k3∈[0.3,0.8];

[0058] This design can monitor the deviation between the actual output and the target, dynamically adjust the key coefficients in the weight calculation formula, periodically check the parameter boundaries to prevent over-adjustment, and achieve adaptive optimization of the algorithm.

[0059] Furthermore, step S7, the seamless fault switching mechanism, specifically includes the following steps:

[0060] Step S71: Monitor the status flags of each unit in real time;

[0061] Step S72: When a faulty cell is detected, reset the weight of the faulty cell to zero within 10ms and reallocate W according to the proportion of the remaining cell capacity. i ′(t), start the backup unit pre-charging, complete the grid connection switching within 200ms, and limit the bus voltage fluctuation range to within ±5% during the switching process;

[0062] This design can detect unit faults (such as overcurrent and communication interruption) in real time, isolate faulty units and redistribute the load within 10ms; and ensure system power supply continuity by pre-charging and fast grid-connected switching of backup units, while avoiding voltage fluctuations.

[0063] Furthermore, the energy storage UPS system includes a data acquisition module, a central control module, a local control module, a dynamic weight calculation module, a fault management module, a backup unit, and an energy dispatch module;

[0064] The output of the data acquisition module is unidirectionally connected to the input of the central control module. The central control module is bidirectionally connected to the local control module, the dynamic weight calculation module, and the energy scheduling module. The output of the central control module is unidirectionally connected to the input of the fault management module. The output of the fault management module is unidirectionally connected to the input of the backup unit. The output of the backup unit is unidirectionally connected to the input of the central control module. The input of the energy scheduling module is unidirectionally connected to an external load.

[0065] Furthermore, the data acquisition module is used to collect the current, voltage, SOC, SOH, and temperature parameters of each energy storage unit in real time to ensure data synchronization and high accuracy;

[0066] The central control module is used to run the dynamic weight allocation algorithm, coordinate the load allocation of each unit, and handle fault switching logic.

[0067] The local control module is used to receive central instructions, execute dual-loop control, and achieve flow equalization.

[0068] The dynamic weight calculation module is used to dynamically allocate the load weight of each unit based on the data parameters collected by the data acquisition module.

[0069] The fault management module is used to detect unit faults and trigger a seamless switch to the backup unit.

[0070] The backup unit is used to quickly take over the load when the main energy storage unit fails, and ensures power supply continuity through pre-charging and grid connection control, while participating in dynamic weight allocation to maintain the overall system balance.

[0071] The energy dispatch module is used to calculate the total system current demand based on the bus voltage and total load demand, and then distribute it to each unit.

[0072] The present invention has the following beneficial effects:

[0073] 1. This invention assesses the core parameters of each energy storage unit in real time, such as state of charge, health, temperature, and internal resistance, and constructs a weighting function with time-varying characteristics. This function can dynamically adjust the output ratio of each unit, achieving intelligent load allocation that matches the aging state of the equipment and environmental conditions. By introducing a health degradation factor, the system can quantify the nonlinear degradation characteristics of the battery pack, prioritizing units in good condition to bear the base load, while providing intermittent rest periods for high-loss units. This design effectively mitigates the local stress accumulation caused by centralized charging and discharging, balances the cyclic aging rate of each unit, and slows down the overall performance degradation of the system. Under complex operating conditions, by synergistically optimizing the suppression of sudden current surges and internal resistance losses, the thermal fatigue damage of power devices is further reduced.

[0074] 2. This invention constructs a hierarchical and progressive fault response system. Through weighted mutation detection and real-time monitoring of status flags, it identifies abnormal units within a microsecond timescale. The weighted reset zeroing strategy and capacity ratio reallocation can maintain bus voltage stability during primary / standby switching, avoiding secondary faults caused by power mutations. The pre-charge collaborative control module achieves smooth grid connection of standby units through dynamic impedance matching technology, eliminating circulating current impacts during switching. This design ensures power supply continuity while maintaining the power supply quality of critical loads through intelligent power rescheduling in degraded mode. The self-repair algorithm in the anomaly recovery phase adopts a progressive parameter rollback strategy, which ensures that the system quickly returns to the optimal state and avoids transient oscillations during the recovery process, improving the system resilience under complex fault scenarios.

[0075] 3. This invention achieves dynamic optimization of the operating strategy through multi-objective optimization, finding the energy distribution path with the lowest loss in real time while meeting load requirements; the time-varying exponential coefficient can autonomously adjust the power distribution tendency according to environmental parameters such as bus voltage fluctuations and temperature gradient changes, establishing a dynamic balance between energy conversion efficiency and equipment safety threshold; the current change smoothing factor suppresses the switching losses caused by frequent power adjustments, while the parameter self-tuning module continuously corrects the weight function through online learning, gradually approaching the optimal operating point of the system; this design enables the system to adaptively match different load characteristics, reduce peak power demand through peak shaving and valley filling strategies, and eliminate local circulating current losses by utilizing the complementary characteristics of energy storage units, thereby improving the overall energy efficiency level.

[0076] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0077] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0078] Figure 1 This is a system framework diagram of an integrated distributed energy storage UPS system according to the present invention. Detailed Implementation

[0079] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0080] Please see Figure 1 As shown, the present invention is an integrated distributed energy storage UPS system, comprising:

[0081] Step S1: Real-time data acquisition:

[0082] Step S11: Install a Hall current sensor at the output terminal of each energy storage unit to collect the output current value I of each unit in real time. i (t);

[0083] Step S12: Periodically obtain the SOC of each energy storage unit through the BMS system. i (t), SOH i (t), internal temperature T i(t) and equivalent series internal resistance R i (t);

[0084] Step S13: Deploy voltage sensors at the DC busbar to monitor the bus voltage U in real time. bus (t);

[0085] Step S14: Upload the data from each unit to the central controller via the CAN bus, and use a timestamp synchronization mechanism to ensure data timing consistency.

[0086] Step S2, parameter normalization processing:

[0087] Step S21: Standardize the collected raw parameters to eliminate unit differences between different parameters and provide standardized input for subsequent steps. The formulas used are as follows:

[0088]

[0089] In the formula, SOC′ i (t) represents the normalized charge state of the i-th energy storage unit at time t, SOC i (t) represents the actual charge state of the i-th energy storage unit at time t, SOC min The minimum SOC allowed by the system. max R′ is the maximum SOC allowed by the system. i (t) is the normalized equivalent series resistance, R i (t) is the equivalent series internal resistance, R max ,R min These are the preset upper and lower limits of the internal resistance, T. i ′(t) is the normalized temperature value, T i (t) represents the real-time internal temperature, T rated T represents the upper limit of the rated operating temperature of the energy storage unit. min This is the lowest permissible operating temperature of the system.

[0090] Step S22: Construct a health decay factor that combines the effects of SOH and temperature to reflect the current availability of the cell. The formula is:

[0091] α i (t)=1-λ×(1-SOH i (t))-κ×(T i ′(t)-T opt ) 2 ;

[0092] In the formula, α i (t) is the health decay factor, λ is the SOH decay weighting coefficient, κ is the temperature deviation penalty coefficient, and T opt To achieve the optimal value for normalized temperature, SOHi (t) represents the healthy state, T i ′(t) is the normalized internal temperature.

[0093] Step S3, Dynamic Weight Calculation:

[0094] Step S31: Calculate the dynamic weight allocation for each unit. The specific formula is as follows:

[0095]

[0096] In the formula, W i (t) represents the dynamic weight allocation of the i-th energy storage unit at time t, α i (t) is the health decay factor, dI i / dt is the output current change rate, and k1, k2, and k3 are all adjustable coefficients that control the sensitivity of SOC, internal resistance, and current change rate, respectively.

[0097] Step S32: Perform weight normalization, the formula is:

[0098]

[0099] In the formula, W i ′(t) represents the normalized weights, N is the total number of energy storage units in the system, and W j (t) represents the dynamic weight allocation of the j-th energy storage unit at time t.

[0100] Step S4: Dynamic allocation of load instructions:

[0101] Step S41: Calculate the total system current demand I total (t), the formula is:

[0102]

[0103] In the formula, P load (t) represents the load power, P loss (t) represents the system power loss, U bus (t) represents the DC bus voltage;

[0104] Step S42: Generate the target current for each unit, using the following formula:

[0105]

[0106] In the formula, Let W be the target output current of the i-th energy storage unit at time t. i ′(t) represents the normalized dynamic weights, β is the SOC equilibrium adjustment coefficient, and SOC avg The average SOC of all energy storage units.

[0107] Step S5: Execution of multi-objective coordinated control:

[0108] Step S51: Send the target current command to the local controller of each unit via RS485 bus;

[0109] Step S52: Each unit adopts a dual-loop control strategy, including:

[0110] Outer loop: PI regulator tracking proportionality coefficient K p =0.8, integration time T i =0.1s;

[0111] Inner loop: Hysteresis current control, loop width set to ±2% of rated current;

[0112] Step S53: Compare the deviation between the actual output current of each unit and the target value in real time. When ΔI i When the percentage is greater than 5%, dynamic weight recalculation is triggered.

[0113] Step S6, Status Feedback and Parameter Self-Tuning:

[0114] Step S61: Acquire the actual output current I i (t) real Deviation from target value ε i (t);

[0115] Step S62: Adjust the weighting coefficients based on the error integral, specifically as follows:

[0116] k1←k1+0.01×∫(SOC′ i (t)×ε i (t))dt;

[0117] k2←k2-0.005×∫(R′ i (t)×ε i (t))dt;

[0118] Step S63: Perform a parameter boundary check every 30 minutes to ensure that k1∈[0.5,1.2], k2∈[0.8,1.5], and k3∈[0.3,0.8].

[0119] Step S7, Seamless Fault Switching Mechanism:

[0120] Step S71: Monitor the status flags of each unit in real time;

[0121] Step S72: When a faulty cell is detected, reset the weight of the faulty cell to zero within 10ms and reallocate W according to the proportion of the remaining cell capacity. i′(t), start the backup unit pre-charging, complete the grid connection switching within 200ms, and limit the bus voltage fluctuation range to within ±5% during the switching process.

[0122] One specific application of this embodiment is:

[0123] I. System Configuration:

[0124] Energy storage unit: Equipped with 6 lithium iron phosphate energy storage units, each with a capacity of 50kWh, a rated voltage of 400V DC, and a maximum output current of 125A;

[0125] Each cell is equipped with a BMS (Battery Management System) to monitor SOC, SOH, temperature and internal resistance in real time;

[0126] Sensors and Communications:

[0127] Current acquisition: A high-precision Hall current sensor (±0.5% accuracy, 10kHz sampling) is used;

[0128] Voltage monitoring: LEM LV25-P voltage sensor (±0.1% accuracy) deployed on DC bus;

[0129] Data transmission: Real-time communication between the central controller and each unit's BMS is achieved via CAN bus, with a synchronization cycle of 100ms;

[0130] Controller:

[0131] Main control unit: An embedded controller based on ARM Cortex-M7, running an adaptive load distribution algorithm;

[0132] Local controller: Each energy storage unit is equipped with an STM32G4 series MCU to perform dual-loop current control;

[0133] II. Operational Procedure:

[0134] Step 1: Data Acquisition and Preprocessing

[0135] Real-time data acquisition: Output current I of each energy storage unit i (t), SOC i (t)(range 20%–95%), temperature T i (t)(threshold 0-50℃), internal resistance R i (t) (range 0.5–2 mΩ), and bus voltage U bus (t);

[0136] Parameter normalization: (Mapped to [0,1]);

[0137]

[0138] Step 2, Dynamic Weight Calculation:

[0139] Weighting formula:

[0140]

[0141] Normalization process:

[0142] Step 3, Load Distribution and Balancing Adjustment:

[0143] Total current demand calculation: based on load power P load (t) = 28kW and bus voltage U bus (t) = 400V:

[0144]

[0145] Target current allocation:

[0146] For unit 1 (W1′(t)=0.18, SOC1′(t)=0.65, the average SOC is... avg =0.7):

[0147]

[0148] Slope limit: Controls the rate of change of current

[0149] Step 4: Coordinate and control execution:

[0150] Dual-loop control strategy:

[0151] Outer loop PI control: K p =0.8,T i =0.1s, tracking the target current;

[0152] Inner loop hysteresis control: loop width ±2.5A (±2% of rated current 125A) to ensure rapid current convergence;

[0153] Current sharing monitoring: If the current deviation between units is >5% (i.e. >6.25A), weight recalculation is triggered;

[0154] Step 5, Fault Switching Mechanism:

[0155] Fault detection: Unit 3 triggered an alarm due to overtemperature (T3=52℃);

[0156] Seamless switching: W3 is set to zero within 10ms, and the remaining cells are reweighted according to capacity;

[0157] The standby unit is pre-charged and connected to the grid within 200ms, with bus voltage fluctuation < ±5% (380V-420V);

[0158] III. Application Scenarios and Effects:

[0159] Scenario: Uninterruptible power supply system for data centers, with load fluctuations ranging from 20kW to 50kW;

[0160] Communication protocol: CAN bus transmission delay <1ms, RS485 is used for local controller command issuance;

[0161] Heat dissipation design: Forced air cooling system to ensure unit temperature <45℃;

[0162] Safety redundancy: The spare unit capacity is 20% of the total demand, supporting N+1 redundancy configuration;

[0163] IV. Conclusion:

[0164] After SOC balancing, the difference is less than 5%, the system efficiency is improved to 96.8%, the switching time is shortened to 200ms, and the voltage sag is reduced by 80%. Through dynamic weight allocation algorithm and SOC balancing strategy, efficient collaborative control of distributed energy storage units is achieved, and excellent stability and reliability are demonstrated in scenarios such as load change and fault switching.

[0165] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0166] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. An integrated distributed energy storage UPS system, characterized in that, The workflow of the energy storage UPS system is as follows: Step S1, Real-time Data Acquisition: By deploying a BMS system and sensors, real-time data on output current, SOC, SOH, temperature, internal resistance, and bus voltage of each energy storage unit are acquired. Step S2, Parameter Normalization Processing: The collected SOC, internal resistance, and temperature parameters are standardized, and a health decay factor is constructed. Step S3, Dynamic Weight Calculation: The load allocation weight of each energy storage unit is generated by formula and the weight is normalized. Step S4, Dynamic Load Command Allocation: Generate target current based on total demand current and the weight of each unit, and add SOC balancing adjustment and slope limiting; Step S5: Multi-objective coordinated control execution: Target current command is issued through a dual-loop control strategy, and current sharing deviation is monitored in real time; Step S6, Status Feedback and Parameter Self-Tuning: Dynamically adjust the weighting coefficients based on the output error and perform boundary checks; Step S7, Seamless Fault Switching Mechanism: Real-time monitoring of unit status; in case of a fault, zero weight is set and backup unit pre-charging is initiated. Step S1, real-time data acquisition, specifically includes the following steps: Step S11: Install a Hall current sensor at the output of each energy storage unit to collect the output current value of each unit in real time. ; Step S12: Periodically obtain the energy storage unit data through the BMS system. , Internal temperature and equivalent series internal resistance ; Step S13: Deploy voltage sensors at the DC busbar to monitor the bus voltage in real time. ; Step S14: Upload the data from each unit to the central controller via the CAN bus, and use a timestamp synchronization mechanism to ensure data timing consistency; Step S2, parameter normalization processing, specifically includes the following steps: Step S21: Standardize the collected raw parameters to eliminate unit differences between different parameters and provide standardized input for subsequent steps. The formulas used are as follows: ； ； ； In the formula, For the first Each energy storage unit at time The normalized charge state, For the first Each energy storage unit at time The actual charge state, For the minimum allowed by the system , The maximum allowed by the system , For the normalized equivalent series internal resistance, For equivalent series internal resistance, These are the preset upper and lower limits of the internal resistance, respectively. This is the normalized temperature value. For real-time internal temperature, This refers to the upper limit of the rated operating temperature of the energy storage unit. This is the lowest permissible operating temperature of the system. Step S22: Construct a health decay factor that combines the effects of SOH and temperature to reflect the current availability of the cell. The formula is: ； In the formula, As a health decay factor, This is the SOH attenuation weighting coefficient. This is the temperature deviation penalty coefficient. To find the optimal normalized temperature value, For a healthy state, Normalized internal temperature; Step S3, the dynamic weight calculation, specifically includes the following steps: Step S31: Calculate the dynamic weight allocation for each unit. The specific formula is as follows: ； In the formula, For the first Each energy storage unit at time Dynamic weight allocation, As a health decay factor, The rate of change of output current. All are adjustable coefficients, which respectively control the sensitivity to SOC, internal resistance and current change rate; Step S32: Perform weight normalization, the formula is: ； In the formula, The normalized weights, This represents the total number of energy storage units in the system. For the first Each energy storage unit at time Dynamically allocated weights.

2. The integrated distributed energy storage UPS system according to claim 1, characterized in that, Step S4, dynamic allocation of load instructions, specifically includes the following steps: Step S41: Calculate the total system current demand The formula is: ； In the formula, For load power, For system power loss, This is the DC bus voltage; Step S42: Generate the target current for each unit, using the following formula: ； In the formula, For the first Each energy storage unit at time The target output current, The normalized dynamic weights This is the SOC equalization adjustment coefficient. The average SOC of all energy storage units.

3. The integrated distributed energy storage UPS system according to claim 2, characterized in that, Step S5, the execution of multi-objective coordinated control, specifically includes the following steps: Step S51: Send the target current command to the local controller of each unit via RS485 bus; Step S52: Each unit adopts a dual-loop control strategy, including: Outer loop: PI regulator tracking proportionality coefficient Integral time ; Inner loop: Hysteresis current control, loop width set to ±2% of rated current; Step S53: Compare the deviation between the actual output current of each unit and the target value in real time. When the value exceeds 5%, dynamic weight recalculation is triggered.

4. The integrated distributed energy storage UPS system according to claim 3, characterized in that, Step S6, state feedback and parameter self-tuning, specifically includes the following steps: Step S61: Acquire the actual output current Deviation from target value ; Step S62: Adjust the weighting coefficients based on the error integral, specifically as follows: ； ； Step S63: Perform a parameter boundary check every 30 minutes to ensure... .

5. The integrated distributed energy storage UPS system according to claim 4, characterized in that, Step S7, the seamless fault switching mechanism specifically includes the following steps: Step S71: Monitor the status flags of each unit in real time; Step S72: When a faulty cell is detected, reset the weight of the faulty cell to zero within 10ms and redistribute it according to the proportion of the remaining cell capacity. The standby unit is pre-charged and the grid connection is completed within 200ms. During the switching process, the bus voltage fluctuation is limited to within ±5%.

6. The integrated distributed energy storage UPS system according to claim 5, characterized in that, The energy storage UPS system includes a data acquisition module, a central control module, a local control module, a dynamic weight calculation module, a fault management module, a backup unit, and an energy dispatch module. The output of the data acquisition module is unidirectionally connected to the input of the central control module. The central control module is bidirectionally connected to the local control module, the dynamic weight calculation module, and the energy scheduling module. The output of the central control module is unidirectionally connected to the input of the fault management module. The output of the fault management module is unidirectionally connected to the input of the backup unit. The output of the backup unit is unidirectionally connected to the input of the central control module. The input of the energy scheduling module is unidirectionally connected to an external load.

7. An integrated distributed energy storage UPS system according to claim 6, characterized in that, The data acquisition module is used to collect the current, voltage, SOC, SOH and temperature parameters of each energy storage unit in real time to ensure data synchronization and high accuracy. The central control module is used to run the dynamic weight allocation algorithm, coordinate the load allocation of each unit, and handle fault switching logic. The local control module is used to receive central instructions, execute dual-loop control, and achieve flow equalization. The dynamic weight calculation module is used to dynamically allocate the load weight of each unit based on the data parameters collected by the data acquisition module. The fault management module is used to detect unit faults and trigger a seamless switch to a backup unit. The backup unit is used to quickly take over the load when the main energy storage unit fails, and ensures power supply continuity through pre-charging and grid connection control, while participating in dynamic weight allocation to maintain the overall system balance. The energy dispatch module is used to calculate the total system current demand based on the bus voltage and total load demand, and then distribute it to each unit.

Images