Flow rate control method, system and equipment for flow battery pump and medium
By using a flow rate control method for flow batteries and leveraging historical data and intelligent algorithms to dynamically adjust the flow rate, the problem of power instability and lifespan limitation caused by performance differences in multi-cell systems of flow batteries has been solved, achieving efficient and stable power output and long-life operation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHAOXING DAMING ELECTRICITY CONSTRUCT CO LTD
- Filing Date
- 2026-01-28
- Publication Date
- 2026-05-15
AI Technical Summary
The flow rate of flow batteries is uncontrollable during charging and discharging, making it impossible to consider individual performance differences in a multi-cell system. This makes it difficult to balance high performance and long lifespan. Traditional control methods cannot dynamically adjust the flow rate according to the current state of the battery, resulting in unstable power output and limited system operation.
By collecting historical power curve data, the target pump flow rate is calculated. Combined with PID control algorithm and gated recurrent neural network, the channel capacity difference and the weakest unit are identified. The flow rate is dynamically adjusted to match the target power demand, avoid the weakest link effect, protect weak units, and extend battery life.
It achieves efficient and stable power output of flow batteries, reduces power fluctuations, extends battery life, and balances high performance and long life.
Smart Images
Figure CN122051287A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of flow battery optimization control technology, specifically to a flow battery pump flow rate control method, system, device, and medium. Background Technology
[0002] Flow batteries, as an advanced electrochemical energy storage technology, have shown great application potential in large-scale energy storage fields (such as renewable energy grid connection, grid peak shaving and valley filling, and backup power) due to their advantages such as independent design of power and capacity, long cycle life, high safety, and ease of scalability. In a flow battery system, the electrolyte circulates within the stack driven by a pump, and its flow rate directly affects the mass transfer efficiency, internal resistance, and final charge / discharge power output of the battery. Therefore, precise control of the pump flow rate is a key aspect of achieving efficient and stable operation of flow batteries. However, flow batteries still face problems of power output fluctuations and efficiency limitations during actual charge and discharge processes. Traditional pump flow rate control methods typically employ a fixed flow rate-power mapping relationship or a simple open-loop control strategy, failing to fully utilize the system's historical operating data and unable to dynamically adjust according to the battery's current state (such as state of charge, temperature, and aging level). This results in the inability to adjust the flow rate in a timely manner to match actual power demands under different operating conditions, thereby reducing overall energy efficiency and affecting power output stability. Meanwhile, actual flow battery systems typically consist of multiple stack units or channels connected in parallel. Due to manufacturing deviations, differences in operating environments, or uneven degradation during long-term operation, the actual performance (such as maximum safe flow rate and effective capacity) of each unit / channel inevitably varies. When regulating the overall flow rate, related technologies often treat the entire system as a homogeneous whole, ignoring this internal inconsistency. This forces the upper limit of the entire system's operation to be limited by the weakest unit (i.e., the "weakest link" effect). If the flow rate is increased to meet the needs of higher-performance units in pursuit of higher power output, the weaker units will deteriorate faster due to overload, and may even induce side reactions such as gas evolution and membrane fouling. Conversely, if the overall flow rate is reduced to protect the weaker units, the system's potential power output is sacrificed. Summary of the Invention
[0003] The purpose of this application is to address the problem that it is difficult to achieve both high performance and long lifespan during the charging and discharging process of flow batteries due to the uncontrollable flow rate and the inability to coordinate individual performance differences in a multi-cell system. This application proposes a flow battery pump flow rate control method, system, device, and medium. By acquiring historical power curve data of the flow battery, the pump flow rate required to achieve the target power within a specified time period is calculated based on the historical power curve. During the flow rate adjustment process, the shortest-board limitation of the system channel is also considered, thereby reducing the impact of the "weakest link" effect on the overall power output, significantly improving system power stability and operating efficiency, and extending battery lifespan.
[0004] To achieve the above objectives, the technical solutions adopted in the embodiments of this application are as follows: In a first aspect, embodiments of this application provide a flow rate control method for a flow battery pump, the method comprising: Collect historical power curves and pump flow rate information of the flow battery, and determine the target power of the flow battery within a specified time period; Calculate the target pump flow rate required for the flow battery to achieve the target power based on historical power curves and target power; Identify the capacity differences and shortest board unit in each channel of the flow battery, and optimize the pump flow rate of each channel according to the target pump flow rate; The operation of the liquid flow pump is controlled based on the optimized pump flow rate signal. The power output of the liquid flow battery and the pump flow rate are monitored in real time. The target pump flow rate and the optimized pump flow rate signal are dynamically adjusted based on the monitoring results.
[0005] In this scheme, a dynamic mapping between power demand and flow rate is established by calculating the target pump flow rate. This eliminates the reliance on fixed ratios or empirical settings for flow rate, allowing for precise calculation of the required flow rate to achieve the target power based on actual operating history and current targets, thus improving flow rate control accuracy. By identifying channel capacity differences and the weakest unit, the flow rate of each channel is allocated differently, avoiding the "weakest link" problem caused by inconsistent performance in multi-unit systems. While ensuring the safety of the weakest unit, the power potential of the entire system is released, and the overall output capacity is no longer completely limited by a single weak unit. At the same time, overload degradation of the weakest unit is avoided, local accelerated decay is suppressed, and the overall system lifespan is extended. By monitoring actual power and flow rate, the previous calculation results are dynamically corrected to overcome the effects of model bias, environmental disturbances, or aging drift, maintaining high performance and stability.
[0006] Optionally, the acquisition of historical power curves and pump flow rate information of the flow battery includes: The system acquires the operating parameters of the flow battery system for each operating cycle, synchronously collects the pump flow rate information at the corresponding time, aligns the collected operating parameters with the pump flow rate information with timestamps to form multi-dimensional time-series data and labels the corresponding operating modes. A historical operating database containing historical power-flow velocity curves was established based on multidimensional time-series data.
[0007] Optionally, the step of calculating the target pump flow rate required for the flow battery to achieve the target power based on the historical power curve and the target power includes: Based on the historical operation database, historical operation segments that match the current operating conditions are retrieved, and the historical operation segments contain several historical power-flow rate pairs; The correlation between power, state of charge, unit voltage and pump flow rate is calculated based on the historical operating segments, and then compared with the correlation threshold. If the correlation is greater than or equal to the correlation threshold, the target pump flow rate required for the flow battery to achieve the target power is calculated based on the first flow rate acquisition strategy. If the correlation is less than the correlation threshold, the target pump flow rate required for the flow battery to achieve the target power is calculated based on the second flow rate acquisition strategy.
[0008] Optionally, the first flow rate acquisition strategy includes: Based on the historical power-flow rate pairs in the historical operation segments, a flow rate value matching the current target power is determined. Using the reciprocal of the similarity between the historical power-flow rate pairs as a weighting coefficient, a weighted average is calculated on all matched flow rate values to obtain the target pump flow rate; or, A linear regression function is established with power, state of charge, and individual unit voltage as regression variables. The historical operating parameters in the historical operating segment are used as variable values to fit the linear regression function using the least squares method, and the optimal regression variable coefficients are solved. Based on the regression variable coefficients, the target pump flow rate is calculated by combining the target power, current battery state, and single-cell voltage.
[0009] Optionally, the second flow rate acquisition strategy includes: Extract all power, state of charge, unit voltage and pump flow rate from the historical operation database, and normalize the data to obtain time-series feature samples; Based on the flow velocity feature samples, the gated recurrent neural network algorithm is pre-trained to produce a flow velocity model. The pre-trained flow velocity model is then optimized based on the power, state of charge, unit voltage, and pump flow velocity in the historical operation segments to obtain the target flow velocity model. The target pump flow rate is calculated based on the real-time battery status, operating parameters, and target power using the target flow rate model.
[0010] Optionally, the step of identifying the capacity differences and shortest board unit in each channel of the flow battery, and optimizing the pump flow rate of each channel based on the target pump flow rate, includes: Test the maximum safe power handling capacity of each stack unit of the battery at different flow rates; By comparing the load-bearing capacity of all stack units, the unit with the lowest maximum safe load-bearing power is selected as the shortest board unit. Obtain the power deviation between the maximum safe load power of the shortest board unit and the target power; The power deviation and the constraint of the shortest board unit are used as control variables, and the flow rate adjustment is obtained through a PID control algorithm. The pump flow rate optimization amount is obtained based on the sum of the flow rate adjustment amount and the target pump flow rate, and the pump flow rate of each channel is adjusted based on the pump flow rate optimization amount after constraint verification.
[0011] Optionally, the dynamic adjustment of the target pump flow rate and the pump flow rate optimization signal based on the monitoring results includes: The actual power collected in real time is compared with the target power to determine whether the deviation is greater than a preset threshold. If the deviation is greater than or equal to the preset threshold, the target pump flow rate is recalculated based on the deviation signal to obtain a new pump flow rate optimization amount and generate a corresponding pump flow rate optimization signal.
[0012] Secondly, embodiments of this application provide a flow rate control system for a flow battery pump, comprising: The data acquisition module is used to collect historical power curves and pump flow rate information of the flow battery, and to determine the target power of the flow battery within a specified time period. The pump flow rate calculation module is used to calculate the target pump flow rate required for the flow battery to achieve the target power based on historical power curves and target power. The flow rate optimization module is used to identify the capacity differences and the shortest board unit in each channel of the flow battery, and optimize the pump flow rate of each channel according to the target pump flow rate. The pump control module is used to control the operation of the flow pump according to the optimized pump flow rate signal, so as to achieve stable charging and discharging of the flow battery near the target power. The feedback monitoring module is used to monitor the power output of the flow battery and the pump flow rate in real time, and feeds back the monitoring results to the pump flow rate calculation module and the flow rate optimization module to dynamically adjust the target pump flow rate and the pump flow rate optimization signal.
[0013] Thirdly, embodiments of this application provide an electronic device, including: a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; the memory is used to store computer programs; and the processor is used to implement the steps of the method described in the first aspect above when executing the program stored in the memory.
[0014] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the method described in the first aspect above.
[0015] The beneficial effects of this application are: 1. By fitting the relationship between power and flow rate through historical data or by using a gated recurrent neural network algorithm combined with the current battery status to calculate the target pump flow rate, the target power demand can be accurately matched. This avoids the problems of increased electrolyte circulation power consumption and intensified polarization caused by excessive flow rate, as well as the problem of insufficient power output caused by excessive flow rate. Compared with traditional open-loop control, the power fluctuation amplitude can be controlled within a smaller range. 2. By detecting the maximum safe load-bearing capacity of each channel / stack, the shortest board unit is located and its constraints are defined. Then, based on the target pump flow rate, the flow rate is adjusted by PID control combined with the constraints to ensure that all units are within the safe range. This fundamentally overcomes the problem of "overload of the strong and damage of the weak" caused by performance differences in multi-unit systems, avoids side reactions and irreversible loss of active materials caused by the shortest board unit exceeding the flow rate limit, and directly extends the overall service life of the battery. 3. The fast response characteristics of PID control can maximize the approximation of the target power within the constraints without sacrificing the system output performance, while protecting weak units. It achieves a balance between "high-performance output" and "long-life operation". It can also collect the actual power in real time, compare it with the target value, judge the deviation feedback, and optimize the target pump flow rate. It can respond to dynamic conditions such as changes in charge and polarization during battery operation in real time, further ensuring the long-term balance between high performance and long life of the system. Attached Figure Description
[0016] Other features, objects, and advantages of this application will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings. The drawings are for illustrative purposes only and are not intended to limit the scope of this application. Furthermore, the same reference numerals denote the same parts throughout the drawings.
[0017] Figure 1 A flowchart of a flow rate control method for a flow battery pump provided in this application embodiment.
[0018] Figure 2 This is a schematic diagram of a flow control system module for a flow battery pump provided in an embodiment of this application.
[0019] Figure 3 This is a schematic diagram of the collaborative control process of a flow rate control system for a flow battery pump provided in an embodiment of this application. Detailed Implementation
[0020] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of this application is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely one preferred embodiment of this application and are only used to explain this application. They do not limit the scope of protection of this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0021] Example 1: As Figure 1 As shown, a flow rate control method for a flow battery pump includes steps S1-S4, wherein: S1. Collect historical power curves of the flow battery and pump flow rate information, and determine the target power of the flow battery within a specified time period. In an optional embodiment, step S1 involves collecting historical power curves of the flow battery and pump flow rate information, including: The system acquires the operating parameters of the flow battery system for each operating cycle, synchronously collects the pump flow rate information at the corresponding time, aligns the collected operating parameters with the pump flow rate information with timestamps to form multi-dimensional time-series data and labels the corresponding operating modes. A historical operating database containing historical power-flow velocity curves was established based on multidimensional time-series data.
[0022] In some embodiments, the operating parameters of the flow battery system for each operating cycle include at least charging power, discharging power data, battery SOC value, and individual cell voltage parameters. The collected power data, pump flow rate data, and battery state parameters are correlated along the time dimension to establish a historical operating database.
[0023] In some embodiments, the target power includes a numerical value and a duration period, such as "continuously outputting 400kW of power from 14:00 to 16:00" in a power grid peak shaving and valley filling scenario. The target power for charging and discharging can be obtained by receiving external instructions from the power grid dispatch center or by manually setting the target power for a specified time period.
[0024] In this embodiment, by collecting historical power, flow rate, SOC, and individual cell voltage data and establishing a time-series database, and specifying the target power for a given time period, the traditional control mode that relies on experience to set the flow rate is overcome, providing real and comprehensive operational data support for subsequent flow rate calculations. The clearly defined target power can avoid deviations in the direction of flow rate adjustment, reducing power fluctuations and energy losses caused by target ambiguity from the source.
[0025] S2. Calculate the target pump flow rate required for the flow battery to achieve the target power based on the historical power curve and the target power.
[0026] In an optional embodiment, step S2 includes: Based on the historical operation database, historical operation segments that match the current operating conditions are retrieved, and the historical operation segments contain several historical power-flow rate pairs; The correlation between power, state of charge, unit voltage and pump flow rate is calculated based on the historical operating segments, and then compared with the correlation threshold. If the correlation is greater than or equal to the correlation threshold, the target pump flow rate required for the flow battery to achieve the target power is calculated based on the first flow rate acquisition strategy. If the correlation is less than the correlation threshold, the target pump flow rate required for the flow battery to achieve the target power is calculated based on the second flow rate acquisition strategy.
[0027] In some embodiments, a similarity measurement method (such as Euclidean distance, dynamic time warping algorithm, or cosine similarity) is used to retrieve historical operating segments from the historical operating database that are closest to the current operating conditions (SOC, temperature, aging state), and several power-flow rate pairs are matched. Further, based on the data from the historical operating segments, the correlation between power, state of charge, individual cell voltage, and pump flow rate is determined. If a linear correlation is found, a first flow rate acquisition strategy (such as rule-based method or linear regression) is used to obtain the target pump flow rate; if a complex nonlinear relationship exists, a second flow rate acquisition strategy is used to obtain the target pump flow rate.
[0028] In this embodiment, by employing a first flow rate acquisition strategy and a second flow rate acquisition strategy, the target pump flow rate is adaptively and flexibly acquired based on the correlation between power, SOC, individual unit voltage, and pump flow rate under actual operating conditions. When the parameters are linearly correlated, the first flow rate acquisition strategy, which uses a rule-based method or linear regression, is employed, resulting in fast calculation speed and low resource consumption, meeting real-time control requirements. When the parameters exhibit complex nonlinear correlations, the second flow rate acquisition strategy is used, leveraging its time-series modeling capabilities to capture nonlinear relationships between parameters, ensuring calculation accuracy under complex operating conditions. This adaptive mechanism, which selects a simple model for linear scenarios and a complex model for nonlinear scenarios, balances computational efficiency and accuracy, avoiding resource waste caused by using complex models for simple operating conditions and insufficient accuracy caused by using simple models for complex operating conditions.
[0029] In an optional embodiment, the first flow rate acquisition strategy includes: Based on the historical power-flow rate pairs in the historical operation segments, a flow rate value matching the current target power is determined. Using the reciprocal of the similarity between the historical power-flow rate pairs as a weighting coefficient, a weighted average is calculated on all matched flow rate values to obtain the target pump flow rate; or, A linear regression function is established with power, state of charge, and individual unit voltage as regression variables. The historical operating parameters in the historical operating segment are used as variable values to fit the linear regression function using the least squares method, and the optimal regression variable coefficients are solved. Based on the regression variable coefficients, the target pump flow rate is calculated by combining the target power, current battery state, and single-cell voltage.
[0030] In an optional embodiment, the second flow rate acquisition strategy includes: Extract all power, state of charge, unit voltage and pump flow rate from the historical operation database, and normalize the data to obtain time-series feature samples; Based on the flow velocity feature samples, the gated recurrent neural network algorithm is pre-trained to produce a flow velocity model. The pre-trained flow velocity model is then optimized based on the power, state of charge, unit voltage, and pump flow velocity in the historical operation segments to obtain the target flow velocity model. The target pump flow rate is calculated based on the real-time battery status, operating parameters, and target power using the target flow rate model.
[0031] Specifically, core feature parameters are extracted from the historical database, including historical charge / discharge power, battery SOC parameters, and single-cell voltage parameters. These feature parameters are sorted by timestamp and preprocessed, including outlier removal, missing value imputation, and effective operating condition screening, to form time-series feature samples. A lightweight gated recurrent neural network (GRU) algorithm is used to train the time-series feature samples to obtain a basic flow rate model. The GRU network structure includes: a 3-dimensional feature (power, SOC, single-cell voltage) input layer to receive the 3 feature parameters at a single time step; 2 hidden layers to accurately learn the dependency relationship between pump flow rate and the 3 feature parameters, thereby improving the accuracy of the target pump flow rate calculation; a dropout layer to prevent overfitting of the GRU model during training; and a fully connected layer to integrate the time-series features output by the GRU, thereby predicting the pump flow rate under the corresponding operating condition based on the dependency relationship between pump flow rate and the 3 feature parameters.
[0032] Furthermore, after the basic model training is completed, the basic model is tested using the power, state of charge, and cell voltage in historical running segments. The flow rate output by the model is compared with the flow rate corresponding to the power, state of charge, and cell voltage in historical running segments. The parameters of the basic model are adjusted according to the flow rate deviation. For example, if the model is overfitted, the dropout rate of the Dropout layer is increased; if the model is underfitted, the time window of the time series feature samples is increased. Finally, the target flow rate model is obtained.
[0033] It should be noted that the power-flow rate relationship of a flow battery changes with state parameters such as SOC, voltage, and temperature. Using historical data similar to the current battery state as test samples allows the model to focus on the characteristics of the current operating condition, avoiding prediction deviations caused by differences in operating conditions, and making the output pump flow rate more in line with actual needs. Moreover, the distribution of data from similar states is more concentrated and the feature similarity is higher, which can quickly guide the model to learn the core correlation between power and flow rate under that state, reduce gradient oscillations during model training, improve the model's generalization ability under similar operating conditions, ensure stable and reliable prediction results at different operating stages, shorten the number of training iterations, reduce computational resource consumption, and balance the prediction accuracy and efficiency of the target pump flow rate.
[0034] In this embodiment, the historical power, SOC, cell voltage, and pump flow rate data of the flow battery are time-series data, and the parameters change continuously over time and exhibit temporal correlations. GRU can capture the dependencies of parameters across different time dimensions. Since historical data for flow batteries typically covers operating periods of several days to months, ordinary recurrent neural networks (such as RNNs) are prone to gradient vanishing when processing long-term data, leading to model training failure. GRU, through its gating mechanism, filters effective temporal information, enabling stable training on long-term data and ensuring prediction accuracy. Furthermore, GRU has a simpler structure than LSTM, with only two core modules: a reset gate and an update gate. This results in lower computational complexity and faster real-time response. LSTM's gating structure is more complex (input gate, forget gate, output gate), offering slightly higher fitting accuracy but with relatively longer computation time. To balance accuracy and real-time requirements, this embodiment employs GRU.
[0035] S3. Identify the capacity differences and shortest board unit of each channel of the flow battery, and optimize the pump flow rate of each channel according to the target pump flow rate.
[0036] In an optional embodiment, step S3 includes: Test the maximum safe power handling capacity of each stack unit of the battery at different flow rates; By comparing the load-bearing capacity of all stack units, the unit with the lowest maximum safe load-bearing power is selected as the shortest board unit. Obtain the power deviation between the maximum safe load power of the shortest board unit and the target power; The power deviation and the constraint of the shortest board unit are used as control variables, and the flow rate adjustment is obtained through a PID control algorithm. The pump flow rate optimization amount is obtained based on the sum of the flow rate adjustment amount and the target pump flow rate, and the pump flow rate of each channel is adjusted based on the pump flow rate optimization amount after constraint verification.
[0037] In some embodiments, the shortest plate unit refers to the unit with the lowest safety load capacity among all channels / stack sub-units of the flow battery. Its constraints are determined based on physical and electrochemical characteristics, specifically including: maximum allowable flow rate constraints, maximum load capacity constraints, polarization voltage constraints, and temperature constraints.
[0038] Specifically, based on the maximum allowable flow rate of the shortest board unit, the flow rate is dynamically adjusted according to the carrying capacity of each channel to maximize the overall power.
[0039] Specifically, the flow rate adjustment is obtained using a PID control algorithm, as shown below: ; in, Indicates the amount of flow rate adjustment. This represents the power deviation between the maximum safe power capacity of the shortest board element and the target power. For the proportional stage, For the points-based system, This is the differential element. The proportional element is used to quickly output the flow rate adjustment amount based on the current power deviation; for example, the larger the deviation, the greater the flow rate adjustment. The larger the corresponding output flow rate adjustment range, the faster the correction can be achieved; the integral term is used to eliminate steady-state error in long-term operation. For example, when the power deviation persists, the integral term accumulates the deviation and adjusts the flow rate to ensure that the system eventually approaches the target power; the derivative term is used to predict the trend of deviation change and suppress overshoot. For example, when the power deviation increases rapidly, the derivative term outputs the reverse adjustment amount in advance to avoid power fluctuations caused by excessive flow rate adjustment.
[0040] Specifically, the pump flow rate optimization amount = +Target pump velocity. The constraint check is: if the optimized pump velocity is greater than the maximum allowable velocity of the shortest board element, then the optimized pump velocity is set to the maximum allowable velocity of the shortest board element.
[0041] In some possible embodiments, if the battery is configured with a centralized pump, meaning that a main pump is used to drive the electrolyte circulation in all stack channels, the rotational speed of the main pump is adjusted according to the pump flow rate optimization amount, and the overall flow rate is limited to the maximum allowable flow rate constraint of the shortest board unit, and the flow rate of all channels changes synchronously.
[0042] In this embodiment, the proportional component of the PID controller can quickly respond to the deviation between the target power and the actual power, completing flow rate adjustment in milliseconds and improving system stability. The integral component can eliminate accumulated deviations during long-term operation, ensuring that the actual power is basically consistent with the target power during stable system operation, avoiding power drift caused by model errors. The derivative component can effectively suppress overshoot during flow rate adjustment, preventing problems such as fuel cell voltage fluctuations and increased polarization caused by sudden flow rate changes, and reducing the probability of side reactions. Combined with the maximum allowable flow rate constraint of the shortest board unit, the PID control maximizes the system power output within the safe range, avoiding overload damage to the shortest board unit and making full use of the carrying capacity of other units, thereby improving the overall energy efficiency of the system.
[0043] S4. Control the operation of the liquid flow pump according to the optimized pump flow rate signal, monitor the power output of the liquid flow battery and the pump flow rate in real time, and dynamically adjust the target pump flow rate and the optimized pump flow rate signal based on the monitoring results.
[0044] In an optional embodiment, step S4, dynamically adjusting the target pump flow rate and the pump flow rate optimization signal based on the monitoring results, includes: The actual power collected in real time is compared with the target power to determine whether the deviation is greater than a preset threshold. If the deviation is greater than or equal to the preset threshold, the target pump flow rate is recalculated based on the deviation signal to obtain a new pump flow rate optimization amount and generate a corresponding pump flow rate optimization signal.
[0045] In some embodiments, the pump flow rate signal is obtained based on the verified optimized pump flow rate, converted into a control command recognizable by the flow pump, adapted to the flow pump's drive circuit, and driven to run at the optimized flow rate. This adjusts the electrolyte circulation rate, enabling the flow battery to enter a charging and discharging state and achieve power output.
[0046] In this embodiment, by collecting actual power and flow rate in real time and comparing them with the target value to determine the deviation, the target pump flow rate optimization and adjustment and PID adjustment are triggered, forming an adaptive iterative optimization mechanism. This solves the collaborative defects of the performance differences of multiple units changing over time, and can respond in real time to dynamic operating conditions such as changes in SOC and polarization during battery operation. It avoids accelerated battery degradation caused by long-term power deviation, and further ensures the balance between long-term high performance and long life of the system.
[0047] Based on the same inventive concept, this application also provides a flow rate control system for a flow battery pump corresponding to a flow rate control method for a flow battery pump. This system includes: The data acquisition module is used to collect historical power curves and pump flow rate information of the flow battery, and to determine the target power of the flow battery within a specified time period. The pump flow rate calculation module is used to calculate the target pump flow rate required for the flow battery to achieve the target power based on historical power curves and target power. The flow rate optimization module is used to identify the capacity differences and the shortest board unit in each channel of the flow battery, and optimize the pump flow rate of each channel according to the target pump flow rate. The pump control module is used to control the operation of the flow pump according to the optimized pump flow rate signal, so as to achieve stable charging and discharging of the flow battery near the target power. The feedback monitoring module is used to monitor the power output of the flow battery and the pump flow rate in real time, and feeds back the monitoring results to the pump flow rate calculation module and the flow rate optimization module to dynamically adjust the target pump flow rate and the pump flow rate optimization signal.
[0048] In some embodiments, such as Figure 2 As shown, the battery pump flow rate control system also includes a target power setting module for setting a target power and a controller for generating a pump flow rate signal. These modules work together to achieve the flow rate control method for the battery pump.
[0049] As a specific embodiment, combined with Figure 3 As shown, in a renewable energy grid-connected scenario, the specific control flow of the battery pump flow rate control system is as follows: A1. Target power allocation: During a photovoltaic power consumption operation, the dispatch center requires the system to continuously absorb 300kW of power (charging) between 10:00 and 12:00 to consume excess photovoltaic power. A2. Historical Data Retrieval: The system retrieves operational data from the historical database, including charge / discharge power curves and the correlation between pump flow rate and fuel cell stack efficiency for the past three months at different SOC ranges. Statistical analysis reveals that the fuel cell stack efficiency is highest and the pump speed-power curve is most stable when the SOC is between 60% and 80%. When the SOC is below 40% or above 90%, the power regulation efficiency decreases by 12%–18%. A3. Pump speed calculation: The current detected battery SOC is 68% and the voltage is stable at 1.52V / cell. Based on historical data curves, the system uses an interpolation algorithm to predict that the pump speed required to achieve 300kW power is approximately 1200rpm. A4. Flow rate correction and optimization: The pump flow rate calculation module first provides a preliminary pump speed (1200 rpm). Then the system calls the deviation correction factor of historical power data: In the SOC 60%–70% range, there is an average power deviation of +2%; therefore, the system lowers the pump speed to 1180 rpm to avoid overshoot. A5. Control Execution and Feedback: The corrected pump speed command (1180 rpm) is sent from the controller to the liquid flow pump via the pump control module. The pump drives the electrolyte circulation, propelling the fuel cell stack into a stable charging state. The feedback monitoring module detects the actual charging power in real time as 298–302 kW, with a deviation of less than ±1%. If the detected deviation continues to exceed ±2%, the system will again call the flow rate optimization module and the pump flow rate calculation module for joint correction, achieving closed-loop adaptive control.
[0050] In this embodiment, by introducing a prediction and correction mechanism based on historical power data, control errors caused by mismatch between the model and actual operating conditions are avoided, achieving high-precision tracking of the target power. By combining pump speed calculation with the "weakest link" effect optimization, the system is ensured to operate stably within a safe range, avoiding single-stack overload and side reactions, improving energy efficiency, and extending lifespan.
[0051] This application also provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, used to execute a program stored in memory, implements a flow rate control method for a flow battery pump.
[0052] The communication bus mentioned in the above electronic devices can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.
[0053] The communication interface is used for communication between the aforementioned electronic devices and other devices.
[0054] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.
[0055] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.
[0056] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements a flow rate control method for a flow battery pump.
[0057] The above-described embodiments are preferred embodiments of this application and are not intended to limit the specific scope of this application. The scope of this application includes but is not limited to the specific embodiments described above. All equivalent changes made in accordance with the shape, structure, and method of this application are within the protection scope of this application.
Claims
1. A method for controlling the flow rate of a flow battery pump, characterized in that: include: Collect historical power curves and pump flow rate information of the flow battery, and determine the target power of the flow battery within a specified time period; Calculate the target pump flow rate required for the flow battery to achieve the target power based on historical power curves and target power; Identify the capacity differences and shortest board unit in each channel of the flow battery, and optimize the pump flow rate of each channel according to the target pump flow rate; The operation of the liquid flow pump is controlled based on the optimized pump flow rate signal. The power output of the liquid flow battery and the pump flow rate are monitored in real time. The target pump flow rate and the optimized pump flow rate signal are dynamically adjusted based on the monitoring results.
2. The flow rate control method for a flow battery pump according to claim 1, characterized in that: The acquisition of historical power curves and pump flow rate information of the flow battery includes: The system acquires the operating parameters of the flow battery system for each operating cycle, synchronously collects the pump flow rate information at the corresponding time, aligns the collected operating parameters with the pump flow rate information with timestamps to form multi-dimensional time-series data and labels the corresponding operating modes. A historical operating database containing historical power-flow velocity curves was established based on multidimensional time-series data.
3. The flow rate control method for a flow battery pump according to claim 1, characterized in that: The calculation of the target pump flow rate required for the flow battery to achieve the target power based on historical power curves and target power includes: Based on the historical operation database, historical operation segments that match the current operating conditions are retrieved, and the historical operation segments contain several historical power-flow rate pairs; The correlation between power, state of charge, unit voltage and pump flow rate is calculated based on the historical operating segments, and then compared with the correlation threshold. If the correlation is greater than or equal to the correlation threshold, the target pump flow rate required for the flow battery to achieve the target power is calculated based on the first flow rate acquisition strategy. If the correlation is less than the correlation threshold, the target pump flow rate required for the flow battery to achieve the target power is calculated based on the second flow rate acquisition strategy.
4. The flow rate control method for a flow battery pump according to claim 3, characterized in that: The first flow rate acquisition strategy includes: Based on the historical power-flow rate pairs in the historical operation segments, a flow rate value matching the current target power is determined. Using the reciprocal of the similarity between the historical power-flow rate pairs as a weighting coefficient, a weighted average is calculated on all matched flow rate values to obtain the target pump flow rate; or, A linear regression function is established with power, state of charge, and individual unit voltage as regression variables. The historical operating parameters in the historical operating segment are used as variable values to fit the linear regression function using the least squares method, and the optimal regression variable coefficients are solved. Based on the regression variable coefficients, the target pump flow rate is calculated by combining the target power, current battery state, and single-cell voltage.
5. The flow rate control method for a flow battery pump according to claim 3, characterized in that: The second flow rate acquisition strategy includes: Extract all power, state of charge, unit voltage and pump flow rate from the historical operation database, and normalize the data to obtain time-series feature samples; Based on the flow velocity feature samples, the gated recurrent neural network algorithm is pre-trained to produce a flow velocity model. The pre-trained flow velocity model is then optimized based on the power, state of charge, unit voltage, and pump flow velocity in the historical operation segments to obtain the target flow velocity model. The target pump flow rate is calculated based on the real-time battery status, operating parameters, and target power using the target flow rate model.
6. The flow rate control method for a flow battery pump according to claim 3, characterized in that: The process involves identifying the capacity differences and shortest board unit in each channel of the flow battery, and optimizing the pump flow rate of each channel based on the target pump flow rate, including: Test the maximum safe power handling capacity of each stack unit of the battery at different flow rates; By comparing the load-bearing capacity of all stack units, the unit with the lowest maximum safe load-bearing power is selected as the shortest board unit. Obtain the power deviation between the maximum safe load power of the shortest board unit and the target power; The power deviation and the constraint of the shortest board unit are used as control variables, and the flow rate adjustment is obtained through a PID control algorithm. The pump flow rate optimization amount is obtained based on the sum of the flow rate adjustment amount and the target pump flow rate, and the pump flow rate of each channel is adjusted based on the pump flow rate optimization amount after constraint verification.
7. The flow rate control method for a flow battery pump according to claim 1, characterized in that: The dynamic adjustment of the target pump flow rate and the pump flow rate optimization signal based on monitoring results includes: The actual power collected in real time is compared with the target power to determine whether the deviation is greater than a preset threshold. If the deviation is greater than or equal to the preset threshold, the target pump flow rate is recalculated based on the deviation signal to obtain a new pump flow rate optimization amount and generate a corresponding pump flow rate optimization signal.
8. A flow rate control system for a flow battery pump, applicable to the flow rate control method for a flow battery pump as described in any one of claims 1-7, characterized in that: include: The data acquisition module is used to collect historical power curves and pump flow rate information of the flow battery, and to determine the target power of the flow battery within a specified time period. The pump flow rate calculation module is used to calculate the target pump flow rate required for the flow battery to achieve the target power based on historical power curves and target power. The flow rate optimization module is used to identify the capacity differences and the shortest board unit in each channel of the flow battery, and optimize the pump flow rate of each channel according to the target pump flow rate. The pump control module is used to control the operation of the flow pump according to the optimized pump flow rate signal, so as to achieve stable charging and discharging of the flow battery near the target power. The feedback monitoring module is used to monitor the power output of the flow battery and the pump flow rate in real time, and feeds back the monitoring results to the pump flow rate calculation module and the flow rate optimization module to dynamically adjust the target pump flow rate and the pump flow rate optimization signal.
9. An electronic device, characterized in that: include: The system includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus; the memory is used to store computer programs; and the processor, when executing the program stored in the memory, implements the steps of the flow rate control method for a flow battery pump as described in any one of claims 1-7.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of a flow rate control method for a flow battery pump as described in any one of claims 1-7.