Flow battery PCS dynamic intelligent alternation control method and flow battery PCS dynamic intelligent alternation control device

Abstract

The invention discloses a flow battery PCS dynamic intelligent rotation control method and device, and belongs to the technical field of flow batteries, and the control method comprises the following steps: S1, multi-parameter real-time monitoring and fusion; s2, dynamic allocation and execution; and S3, fault tolerance and self-healing. The device comprises a sensor network, an edge computing unit and a PCS controller, wherein the sensor network, the edge computing unit and the PCS controller are connected in sequence. According to the dynamic intelligent alternating control method, load balancing and efficiency optimal distribution of multiple groups of PCS units are realized, the calling strategy of the PCS is optimized, the problem that the difference between long-time shutdown of equipment and long-time operation of the equipment is too large due to a traditional mode is solved, the service life of the whole equipment is prolonged, and the service life of the whole equipment is prolonged. The system design can be applied to a large-scale energy storage power station and is beneficial to equipment of the whole station.

Description

Technical Field

[0001] This invention belongs to the field of flow battery technology, specifically relating to a dynamic intelligent switching control method and device for flow battery PCS. Background Technology

[0002] Existing patent CN102122826A discloses a large-capacity battery energy storage bidirectional converter, which has the following problems:

[0003] 1. Fixed rotation defects: Existing PCS mostly adopts a fixed sequence or time rotation strategy, which cannot be dynamically adjusted according to the real-time status of the battery stack (such as SOC, temperature, internal resistance), resulting in some battery stacks being overloaded or idle for a long time, accelerating uneven aging.

[0004] 2. Insufficient trade-offs: Traditional methods struggle to balance efficiency and battery life under real-time operating conditions, especially in dynamic load scenarios where over-discharge or under-charging is likely to occur.

[0005] 3. Lack of synergy: Existing technologies lack the ability to synergistically optimize multiple parameters (such as SOC equalization, temperature difference, and response speed), resulting in limited overall system performance.

[0006] 4. Poor fault tolerance: When some battery stacks fail, traditional rotation strategies cannot be adjusted quickly, affecting system reliability. Summary of the Invention

[0007] To address the aforementioned problems, this invention proposes a dynamic intelligent switching control method for a flow battery PCS, comprising the following steps:

[0008] S1. Real-time monitoring and fusion of multiple parameters;

[0009] S2, Dynamic Allocation and Execution;

[0010] S3, Fault Tolerance and Self-Healing.

[0011] Furthermore, in step S1, the SOC, temperature, internal resistance and electrolyte pressure data of each battery stack are collected by various sensors, normalized and then used to establish a digital model. The data is then input into the model for evaluation.

[0012] Furthermore, step S2 includes the following steps:

[0013] S21, design weighting coefficient, calculate the real-time priority score of each battery stack;

[0014] S22, Multi-objective optimization decision-making;

[0015] S23, Dynamic Allocation and Execution.

[0016] Further, in step S21, weighting coefficients are designed, and the real-time priority score of each battery stack is calculated:

[0017] Priority = α·(1-|SOC) avg -SOC i ∣)+β·(1-T max T i )+γ·SOH i r

[0018] The weighting coefficients are dynamically adjusted based on grid demand.

[0019] -Priority: Priority

[0020] -α: SOC equilibrium weight;

[0021] -β: Temperature weighting;

[0022] -γ: Health score weight;

[0023] -avg: average value;

[0024] -max: Maximum value;

[0025] -i: Current count.

[0026] Furthermore, in step S22, multi-objective optimization decision-making is performed, and an objective function is established:

[0027] Minimize i=1∑n(w1·Agingi+w2·Lossi)

[0028] -Minimize: Minimize the objective function;

[0029] -w1, w2: These are the weight coefficients of the two objectives (and usually w1+w2=1);

[0030] -Agingi: The aging metric for the i-th system / component;

[0031] -Lossi: The "loss" metric for the i-th system / component;

[0032] Constraints: SOC balance, safe temperature range, power requirement matching;

[0033] The optimal rotation sequence is solved using a genetic algorithm or particle swarm optimization.

[0034] Furthermore, in step S23, dynamic allocation and execution are performed:

[0035] High-priority battery stacks are assigned to current high-load tasks, while low-priority stacks are assigned to light-load or standby states; the "cross-stack energy transfer" mode is automatically triggered based on SOC differences to achieve SOC balancing.

[0036] Furthermore, S3's fault tolerance and self-healing:

[0037] Real-time monitoring of fault signals: voltage drop, temperature exceeding limits, triggering a three-level response:

[0038] Level 1: Reduced power operation;

[0039] Level 2: Switch to backup battery stack;

[0040] Level 3: Isolate the faulty unit and trigger an alarm.

[0041] An apparatus based on the above-described flow battery PCS dynamic intelligent rotation control method includes a sensor network, an edge computing unit, and a PCS controller, wherein the sensor network, the edge computing unit, and the PCS controller are connected in sequence.

[0042] Furthermore, the sensor network includes a SOC sensor, a temperature sensor, an internal resistance sensor, and an electrolyte pressure sensor, used to collect various parameters of the battery stack in real time.

[0043] Furthermore, the edge computing unit deploys a dynamic optimization algorithm to process and analyze the collected data, generate control commands in real time, and send them to the PCS controller for execution. The PCS controller switches the charging and discharging states of the battery stack according to the received control commands to achieve load balancing and optimal efficiency allocation.

[0044] The beneficial effects of this invention are as follows:

[0045] This invention achieves load balancing and optimal efficiency allocation of multiple PCS units through a dynamic intelligent rotation control method, optimizes the PCS calling strategy, solves the problem of excessively large lifespan differences between long-term equipment downtime and long-term operation caused by traditional methods, and extends the overall service life of the equipment. This system design can be applied to large-scale energy storage power stations and is beneficial to the entire station equipment. Attached Figure Description

[0046] Figure 1 The process of this invention Figure 1 ;

[0047] Figure 2 The process of this invention Figure 2 . Detailed Implementation

[0048] Example 1

[0049] To facilitate understanding of the technical means and objectives of this invention, the following detailed description, in conjunction with specific embodiments, further elucidates the invention: a dynamic intelligent rotation control method for flow battery PCS, aiming to solve the problems of uneven battery life and low efficiency caused by the rigidity of existing rotation strategies. The technical solution involves real-time acquisition of parameters such as SOC, temperature, internal resistance, and grid demand of the battery stack, combined with a dynamic priority algorithm and a multi-objective optimization model, to generate adaptive rotation commands. For example... Figure 1 As shown, a modular design is adopted, including data monitoring, intelligent decision-making, dynamic allocation, and fault-tolerant control modules, achieving the following innovations:

[0050] Dynamic priority assessment: Dynamically adjust the operation priority based on the health status of the battery stack and the operating requirements.

[0051] Multi-objective optimization: Simultaneously optimize battery life, system efficiency, and response speed.

[0052] Fault self-healing mechanism: Real-time detection of anomalies and switching to backup units to ensure continuous system operation.

[0053] This invention is applicable to large-scale flow battery energy storage power stations, and is expected to improve the overall lifespan of the system by more than 20% and reduce operation and maintenance costs by 30%.

[0054] The technical solution is as follows:

[0055] 1. Real-time status awareness:

[0056] Collect operating data from each PCS and calculate its health status (HI) and remaining life (RUL) based on long-term operating temperature and internal resistance.

[0057] Monitor the power fluctuation trend of the power grid and predict the power change trend within the next 5 minutes based on the power grid load curve (ARIMA model).

[0058] 2. Dynamic weight allocation:

[0059] Construct an evaluation matrix that includes efficiency weight (40%), heat load weight (30%), historical failure rate (20%), and economic weight (10%).

[0060] Adjust the weighting ratio based on real-time data (e.g., increase the heat load weighting to 50% in high-temperature environments).

[0061] 3. Optimize decision generation:

[0062] Input constraints (total power demand, maximum output limit per PCS, system efficiency threshold);

[0063] The optimal PCS combination and power allocation scheme are solved using the PSO algorithm, with the objective function being to minimize system losses and lifetime degradation costs.

[0064] 4. Seamless switching control:

[0065] Pre-synchronization control technology is adopted to adjust the phase and amplitude of the PCS to be connected before switching, so as to achieve zero-impact switching;

[0066] Perform a hot backup standby or deep cooling process on the decommissioned PCS unit.

[0067] 1. System Architecture

[0068] Hardware components:

[0069] Sensor network: Collects data on the battery stack's SOC, temperature, voltage, current, and electrolyte flow rate.

[0070] Edge computing unit: Deploys dynamic optimization algorithms to generate control commands in real time.

[0071] PCS controller: Executes rotation commands to switch the charge and discharge states of the battery stack.

[0072] 2. Software Logic:

[0073] Build a digital twin model to simulate the dynamic behavior of the battery stack;

[0074] Predict battery state of health (SOH) degradation trends based on machine learning.

[0075] Example 2

[0076] A dynamic intelligent switching control method for a flow battery PCS includes the following steps:

[0077] S1. Real-time monitoring and fusion of multiple parameters

[0078] S2. Dynamic Allocation and Execution

[0079] S3. Fault Tolerance and Self-Healing

[0080] Furthermore, S1 multi-parameter real-time detection and fusion

[0081] Data on SOC, temperature, internal resistance, and electrolyte pressure of each battery stack are collected using various sensors. After normalization, a digital model is established, and the data is input into the model for evaluation.

[0082] Furthermore, S2 makes the corresponding priority scheduling algorithm more flexible and scientific, which can prevent some devices from being never scheduled and also prevent some devices from monopolizing operation for a long time.

[0083] S21. Design weighting coefficients and calculate the real-time priority score for each battery stack:

[0084] Priority = α·(1-|SOC) avg -SOCi ∣)+β·(1-T max T i )+γ·SOH i r

[0085] The weighting coefficients are dynamically adjusted according to grid demand (such as frequency regulation, peak shaving and valley filling).

[0086] -Priority: Priority

[0087] -α: SOC equilibrium weight;

[0088] -β: Temperature weighting;

[0089] -γ: Health score weight;

[0090] -avg: average value;

[0091] -max: Maximum value;

[0092] -i: Current count.

[0093] S22, Multi-objective optimization decision-making

[0094] Establish the objective function:

[0095] Minimize i=1∑n(w1·Agingi+w2·Lossi)

[0096] -Minimize: Minimize the objective function;

[0097] -w1, w2: These are the weight coefficients of the two objectives, respectively;

[0098] -Agingi: The aging metric for the i-th system / component;

[0099] -Lossi: The "loss" metric for the i-th system / component;

[0100] Constraints: SOC balance, safe temperature range, and power requirement matching.

[0101] like Figure 2 As shown, the optimal rotation sequence is solved using a genetic algorithm or particle swarm optimization.

[0102] S23, Dynamic Allocation and Execution

[0103] Assign high-priority battery stacks to current high-load tasks, and low-priority stacks to light-load or standby states;

[0104] The "cross-reactor energy transfer" mode is automatically triggered based on the SOC difference to achieve SOC balancing.

[0105] S3 Fault Tolerance and Self-Healing

[0106] Real-time monitoring of fault signals (such as voltage drops, temperature exceeding limits) triggers a three-level response:

[0107] Level 1: Reduced power operation;

[0108] Level 2: Switch to backup battery stack;

[0109] Level 3: Isolate the faulty unit and trigger an alarm.

[0110] Example 3

[0111] A dynamic intelligent switching control method for a flow battery PCS includes the following steps:

[0112] S1. Real-time monitoring and fusion of multiple parameters;

[0113] S2, Dynamic Allocation and Execution;

[0114] S3, Fault Tolerance and Self-Healing.

[0115] In step S1, data on the SOC, temperature, internal resistance, and electrolyte pressure of each battery stack are collected using various sensors. After normalization, a digital model is established, and the data is input into the model for evaluation.

[0116] Step S2 includes the following steps:

[0117] S21, design weighting coefficient, calculate the real-time priority score of each battery stack;

[0118] S22, Multi-objective optimization decision-making;

[0119] S23, Dynamic Allocation and Execution.

[0120] In step S21, weighting coefficients are designed, and the real-time priority score of each battery stack is calculated.

[0121] Priority = α·(1-|SOC) avg -SOC i ∣)+β·(1-T max T i )+γ·SOH i r

[0122] The weighting coefficients are dynamically adjusted based on grid demand.

[0123] -Priority: Priority

[0124] -α: SOC equilibrium weight;

[0125] -β: Temperature weighting;

[0126] -γ: Health score weight;

[0127] -avg: average value;

[0128] -max: Maximum value;

[0129] -i: Current count.

[0130] In step S22, multi-objective optimization decision-making is performed, and an objective function is established:

[0131] Minimize i=1∑n(w1·Agingi+w2·Lossi)

[0132] -Minimize: Minimize the objective function;

[0133] -w1, w2: These are the weight coefficients of the two objectives, respectively;

[0134] -Agingi: The aging metric for the i-th system / component;

[0135] -Lossi: The "loss" metric for the i-th system / component;

[0136] Constraints: SOC balance, safe temperature range, power requirement matching;

[0137] The optimal rotation sequence is solved using a genetic algorithm or particle swarm optimization.

[0138] In step S23, dynamic allocation and execution are performed:

[0139] High-priority battery stacks are assigned to current high-load tasks, while low-priority stacks are assigned to light-load or standby states; the "cross-stack energy transfer" mode is automatically triggered based on SOC differences to achieve SOC balancing.

[0140] In step S3, fault tolerance and self-healing are described below:

[0141] Real-time monitoring of fault signals: voltage drop, temperature exceeding limits, triggering a three-level response:

[0142] Level 1: Reduced power operation;

[0143] Level 2: Switch to backup battery stack;

[0144] Level 3: Isolate the faulty unit and trigger an alarm.

[0145] An apparatus for the above-mentioned flow battery PCS dynamic intelligent rotation control method includes a sensor network, an edge computing unit, and a PCS controller, which are connected in sequence.

[0146] The sensor network includes a SOC sensor, a temperature sensor, an internal resistance sensor, and an electrolyte pressure sensor, which are used to collect various parameters of the battery stack in real time.

[0147] The edge computing unit deploys a dynamic optimization algorithm to process and analyze the collected data, generate control commands in real time, and send them to the PCS controller for execution. The PCS controller switches the charging and discharging states of the battery stack according to the received control commands to achieve load balancing and optimal efficiency allocation.

[0148] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A dynamic intelligent switching control method for a flow battery PCS, characterized in that, Includes the following steps: S1. Real-time monitoring and fusion of multiple parameters; S2, Dynamic Allocation and Execution; S3, Fault Tolerance and Self-Healing.

2. The flow battery PCS dynamic intelligent switching control method as described in claim 1, characterized in that, In step S1, data on SOC, temperature, internal resistance, and electrolyte pressure of each battery stack are collected through various sensors, normalized, and then used to establish a digital model. The data is then input into the model for evaluation.

3. The flow battery PCS dynamic intelligent switching control method as described in claim 2, characterized in that, Step S2 includes the following steps: S21, design weighting coefficient, calculate the real-time priority score of each battery stack; S22, Multi-objective optimization decision-making; S23, Dynamic Allocation and Execution.

4. The flow battery PCS dynamic intelligent switching control method as described in claim 3, characterized in that, In step S21, weighting coefficients are designed, and the real-time priority score of each battery stack is calculated: Priority=α·(1-∣SOC avg -SOC i ∣)+β·(1-T max T i )+γ·SOH i The weighting coefficients are dynamically adjusted based on grid demand. -Priority: Priority -α: SOC equilibrium weight; -β: Temperature weighting; -γ: Health score weight; -avg: average value; -max: Maximum value; -i: Current count.

5. The flow battery PCS dynamic intelligent switching control method as described in claim 3, characterized in that, In step S22, multi-objective optimization decision-making is performed, and an objective function is established: Minimize i=1∑n(w1·Agingi+w2·Lossi) -Minimize: Minimize the objective function; -w1, w2: These are the weight coefficients of the two objectives, respectively; -Agingi: The aging metric for the i-th system / component; -Lossi: The "loss" metric for the i-th system / component; Constraints: SOC balance, safe temperature range, power requirement matching; The optimal rotation sequence is solved using a genetic algorithm or particle swarm optimization.

6. The flow battery PCS dynamic intelligent switching control method as described in claim 3, characterized in that, In step S23, dynamic allocation and execution are performed: High-priority battery stacks are assigned to current high-load tasks, while low-priority stacks are assigned to light-load or standby states; the "cross-stack energy transfer" mode is automatically triggered based on SOC differences to achieve SOC balancing.

7. The flow battery PCS dynamic intelligent switching control method as described in claim 1, characterized in that, In step S3, fault tolerance and self-healing are described below: Real-time monitoring of fault signals: voltage drop, temperature exceeding limits, triggering a three-level response: Level 1: Reduced power operation; Level 2: Switch to backup battery stack; Level 3: Isolate the faulty unit and trigger an alarm.

8. An apparatus for a dynamic intelligent switching control method for a flow battery PCS according to any one of claims 1 to 7, characterized in that, It includes a sensor network, an edge computing unit, and a PCS controller, which are connected in sequence.

9. The apparatus for the dynamic intelligent switching control method of the flow battery PCS according to claim 8, characterized in that, The sensor network includes a SOC sensor, a temperature sensor, an internal resistance sensor, and an electrolyte pressure sensor, used to collect various parameters of the battery stack in real time.

10. The apparatus for the dynamic intelligent switching control method of the flow battery PCS according to claim 8, characterized in that, The edge computing unit deploys a dynamic optimization algorithm to process and analyze the collected data, generate control commands in real time, and send them to the PCS controller for execution. The PCS controller switches the charging and discharging states of the battery stack according to the received control commands to achieve load balancing and optimal efficiency allocation.

Images