Micro-grid flow battery energy storage peak regulation method based on predictive control

Through the microgrid liquid flow battery energy storage peak regulating method based on prediction control, the dynamic behavior of the microgrid system is managed and the current and voltage interaction is optimized, and the problems of uncertainty in supply and demand and difficulty in peak load regulation are solved, and the power grid is improved Efficiency and stability, adapt to renewable energy volatility and reduce operating costs.

CN119029973APending Publication Date: 2024-11-26WUXI UNIV
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411081229.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-07
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

Microgrid systems face the problems of uncertainty in supply and demand and difficulties in peak load regulation. The existing research mainly focuses on photovoltaics, wind energy and fossil fuels, ignoring the improvement of biomass resources to the sustainability and efficiency of microgrids.

Method used

Adopting the microgrid liquid flow battery energy storage peak regulating method based on prediction control, through system modeling, predictive control and optimization, combined with particle swarm algorithm to optimize controller parameters, realize the interaction management of dynamic current and voltage, and optimize pump power and energy efficiency. , reduce the power grid energy loss.

Benefits of technology

It effectively solves the problems of inefficient power grid, low sustainability and unbalanced supply and demand, improves grid stability and flexibility, adapts to the volatility of renewable energy generation, reduces grid energy loss and reduces operating costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119029973A_ABST
    Figure CN119029973A_ABST
Patent Text Reader

Abstract

The invention discloses a micro-grid flow battery energy storage peak regulation method based on predictive control, relates to the technical field of micro-grid energy storage, and aims to provide a micro-grid flow battery energy storage peak regulation method based on a traditional fixed control method by performing verification analysis through system modeling and simplification, predictive control and optimization, performance evaluation and optimization and simulation model construction. A strategy of a prediction control method is introduced, so that the problems of low power grid efficiency, low sustainability, lack of elasticity and the like caused by excessive power grid transmission energy loss and uncertainty of supply and demand are effectively solved; compared with a traditional fixed control method, the predictive control method provided by the invention has excellent intelligence and flexibility, and can seamlessly adapt to the dynamic power demand of the micro-grid system; by remarkably improving the VRFB efficiency of the vanadium redox flow battery and reducing the operation cost, the method of the embodiment is expected to improve the sustainability and elasticity of micro-grid infrastructures.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of microgrid energy storage technology, and in particular to a microgrid liquid flow battery energy storage peak regulation method based on predictive control. Background Art

[0002] Traditional centralized power grids have long been criticized for being inefficient, characterized by large energy losses during long-distance transmission, and requiring large investments in infrastructure, especially in remote mountainous areas of developing countries. Rural electrification is essential for improving living standards and promoting economic growth, but it is often hampered by financial constraints, which affect living standards and economic growth.

[0003] Currently, challenges faced by microgrid systems include uncertainties in both supply and demand, as well as the various specifications inherent in microgrid configurations. One of the key obstacles is the effective regulation of peak loads, which requires effective strategies to manage fluctuations in energy demand. Existing research in the field of microgrid systems has mainly focused on the use of photovoltaics, wind energy, and fossil fuels, while biomass resources have received relatively limited attention, ignoring the improvement of biomass resources on the sustainability and efficiency of microgrids. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a microgrid liquid flow battery energy storage peak regulation method based on predictive control, comprising the following steps:

[0005] S1. System modeling and simplification: Assume that the electrolyte, electrodes and membrane have uniform physical properties, and assume that the diffusion of ions through the membrane is uniform, and consider the permeability of the membrane; divide the vanadium redox flow battery into positive and negative half-cells, consider the charging and discharging processes respectively, and establish a half-cell model based on this; use the Nernst equation to calculate the open circuit voltage; establish a mathematical model to consider the charge state changes of the two half-cells to capture the charge state changes caused by the diffusion of vanadium ions; consider concentration overpotential, formation potential, battery voltage and charge state; model the pump power and energy efficiency; and finally optimize the operating parameters of the pump;

[0006] S2. Predictive control and optimization: Establish a mathematical model to describe the dynamic behavior of the system, adjust the control input based on the predictions by predicting the future behavior of the system, and use the optimization algorithm to design the controller to optimize the system according to its performance indicators; use different sensors to measure various parameters in the system, and pass the data into the control algorithm to update the state estimation and prediction model of the system; combine predictive control with feedback control to correct the system when deviations occur, and observe the energy conversion process in the system through efficiency indicators; use particle swarm algorithm in predictive control to optimize controller parameters to maximize the performance indicators of the system; describe the interaction between dynamic current and voltage and determine the charging and discharging operation modules; calculate and determine the charging state; finally, design experiments for surface fitting;

[0007] S3. Performance evaluation and optimization of flow battery energy storage systems implementing predictive control;

[0008] S4. Build a simulation model for verification analysis.

[0009] The technical solution further defined in the present invention is:

[0010] Furthermore, in step S1, it is assumed that the electrolyte, the electrode and the membrane have uniform physical properties, including density and conductivity; the vanadium redox flow battery is divided into positive and negative half-cells, and the charging and discharging processes are considered respectively, and a half-cell model is established based on this, as shown in the following formula:

[0011] V 3+ +e - →V 2+

[0012]

[0013] During the charging process, the positive half-cell The ions are oxidized, generating electrons that migrate to the negative half-cell, where they promote V 3+ Reduction of ions, thereby converting chemical energy into electrical energy;

[0014] The open circuit voltage is calculated using the Nernst equation as shown below:

[0015]

[0016] Among them, F represents the Faraday constant, C represents the ion concentration, cs represents the battery stack configuration, and n represents the number of electron transfers in the electrochemical reaction.

[0017] In the aforementioned microgrid liquid flow battery energy storage peak-shaving method based on predictive control, in step S1, a mathematical model is established to consider the charge state changes of the two half-cells to capture the charge state changes caused by the diffusion of vanadium ions, including considering the changes in ion concentration during charging and discharging. The specific formula is as follows:

[0018]

[0019] Among them, SOC- represents the reduction in the battery's state of charge, SOC + represents the increase in the state of charge of the battery; tk, tk2, tk3, tk4 and tk5 represent different storage tanks; W CS Indicates battery capacity, C CS represents the ion concentration, Q CS Indicates the battery charging capacity, C tki represents the ion concentration in the storage tank, C csi Indicates the ion concentration in the electrolyte during charging, Φ i represents the battery molar flux, S i represents the battery internal source term, Φ i and S i Affects the change of concentration over time; t represents time; the value of variable i is 2, 3, 4 or 5, representing different types of vanadium ions;

[0020] Finally, the open circuit voltage η oa It is derived from the product of current density and resistivity, including ohmic and activation overpotential, as follows:

[0021] η oa =η ohm +η act =ir

[0022] Among them, η ohm represents the ohmic loss of the battery, η act It represents the chemical activity loss of the battery, and ir represents the product of the current and the internal resistance of the battery.

[0023] In the aforementioned microgrid liquid flow battery energy storage peak regulation method based on predictive control, in step S1, concentration overpotential, formation potential, battery voltage and charging state are considered: concentration overpotential, formation potential, battery voltage and charging state are used as optimization variables to maximize the performance and efficiency of the system, as shown in the following formula:

[0024]

[0025] Among them, η con represents the concentration overpotential, R represents the ideal gas constant, T represents the temperature, Q cs represents the vanadium concentration in the electrolyte, A flrepresents the electrode surface area, C rea represents the rate constant of the reversible reaction at the electrode; and manages the factors affecting the cell voltage by the following equation:

[0026] E chcell =E OCV +η oa +η con

[0027] E discell =E OCV -η oa -η con

[0028] E stack =N cell ·E cell

[0029] Among them, E chcell Indicates the chemical energy of the battery cell during the charging process, E OCV Represents the battery open circuit voltage energy, η oa Represents the battery open circuit voltage, η con Indicates other losses of the battery, E discell Indicates the energy released by the battery cell during the discharge process, E stack Represents the total energy of the entire battery stack, N cell Indicates the number of cell stacks contained within the module.

[0030] In the aforementioned microgrid liquid flow battery energy storage peak regulation method based on predictive control, in step S1, pump power and energy efficiency modeling is performed: the flow of electrolyte in the system is analyzed, and the influence of the power consumption of the pump on the energy efficiency of the system is evaluated, as shown in the following formula:

[0031]

[0032] Where Δp represents the pressure rise or pressure difference of the pump, Q represents the flow rate, and η pump Indicates the efficiency of the pump;

[0033] Optimize the operating parameters of the pump. The specific process is shown in the following formula:

[0034] Δp=Δp pipe +Δp cell

[0035]

[0036] Where μ represents the dynamic viscosity of the electrolyte, ρ represents the density of the electrolyte, ∈ represents the electrode porosity, f represents the Darcy friction coefficient, and K represents the Kozeny-Karman constant.

[0037] In the aforementioned microgrid liquid flow battery energy storage peak-shaving method based on predictive control, in step S2, various parameters in the system, including voltage, current and temperature, are measured by different sensors, and the data are input into the control algorithm to update the state estimation and prediction model of the system; predictive control is combined with feedback control to correct when the system deviates; the energy conversion process in the system is observed through efficiency indicators, and the efficiency indicators include charging efficiency, discharging efficiency and total efficiency. The specific process is shown in the following formula:

[0038]

[0039] Among them, η represents the efficiency of the system process, E storage Indicates the total energy stored during the charging process, E output Indicates the total energy actually released during the discharge process, E release Indicates the total energy released by the discharge, E input Represents the total energy of charging input, η ch and η dis Indicates the charge and discharge efficiency;

[0040] In predictive control, the particle swarm algorithm is used to optimize the controller parameters to maximize the performance indicators of the system. In predictive control, the parameters of the controller, including gain and weight, are adjusted to minimize the control error or optimize the performance indicators of the system. The convergence of the particle swarm algorithm is analyzed to ensure that the particle swarm algorithm can find a solution that is close enough to the optimal solution within a limited number of iterations, as shown in the following formula:

[0041]

[0042] Where J represents the objective function; represents the input efficiency at time step i+1; η i+1 represents the efficiency at time step i+1; λ represents the weight coefficient, which is used to adjust parameters and penalty terms; Q i+1 and Q i Respectively represent the state variables of the system at time steps i and i+1;

[0043] Rated power P power The power plant size represented is calculated iteratively using the following formula:

[0044] iP power >P load :P ch =P power -P load

[0045] iP power <P load :P dis =Pload -P power

[0046] η ch η dis ∫ t1 P ch dt-∫ t2 P ch dt<ω

[0047] Among them, P power Indicates the output power of the battery, P load represents the load power, ω represents the energy threshold, ∫ t1 P ch dt and ∫ t2 P ch dt represents the power integration during the charging and discharging process.

[0048] In the aforementioned microgrid liquid flow battery energy storage peak regulation method based on predictive control, in step S2, the interaction between dynamic current and voltage and the determination of charging and discharging operation modules are described: How the current in the energy storage system changes with the dynamic change of the stack voltage is described by the following formula:

[0049] N ch ·P ch <P ch η ch <(N ch +1)P ch

[0050] Then, by considering the maximum power demand, battery charge state and energy conversion efficiency, the optimal operation module under different conditions is determined, as shown in the following formula:

[0051]

[0052] Among them, N ch and N dis Represent the number of charge and discharge cycles, P ch and P dis Represent the power of charging and discharging, P stack and E stack Represent the power and energy of the stack respectively.

[0053] In the aforementioned microgrid liquid flow battery energy storage peak regulation method based on predictive control, in step S2, the charging state is calculated and determined by the following formula:

[0054]

[0055] Among them, SOC ch and SOC dis Respectively represent the state of charging and discharging, ENinitial ,EN ch ,EN dis and EN tot They represent initial energy, charging energy, discharging energy and total energy respectively;

[0056] Experiments are designed to collect input and output data of the system in order to use the surface fitting method to estimate the dynamic characteristics of the system, including charge and discharge efficiency and electrolyte flow rate. The specific fitting process is shown in the following formula:

[0057]

[0058] Among them, η fit represents the fitting efficiency; p mn represents the coefficients in the fitting model, which is used to fit the actual data; m and n are exponents, which are used to determine the power of the charge state and current in the polynomial.

[0059] In the aforementioned microgrid liquid flow battery energy storage peak-shaving method based on predictive control, step S3 specifically includes the following sub-steps:

[0060] S3.1. Establish a numerical simulation model: Use physical equations to establish a numerical simulation model to describe the dynamic behavior of the entire battery system; then use numerical methods to discretize and solve the model equations to update the system state and control input at each time step;

[0061] S3.2. Define performance indicators: Define performance indicators to evaluate the performance of the system. The performance indicators include energy efficiency, response time, and stability. When defining, consider system constraints. System constraints include current limit, voltage limit, and temperature limit.

[0062] S3.3. System performance evaluation: Evaluate the performance of the flow battery energy storage system implementing predictive control, including load balancing effect, energy cost savings, and system stability;

[0063] S3.4. Control strategy optimization: Adjust and optimize the predictive control strategy based on the performance evaluation results.

[0064] In the aforementioned microgrid liquid flow battery energy storage peak-shaving method based on predictive control, step S4 specifically includes the following sub-steps:

[0065] S4.1. Using MATLAB / Simulink platform, establish a comprehensive model of flow battery energy storage system, which includes battery model, load model and controller model;

[0066] S4.2, parameter setting and initialization: set the parameters of the simulation experiment, including battery parameters, load demand and energy price, and initialize the battery charging state and controller parameters;

[0067] S4.3, importing load and energy prices: importing historical load and energy price data and future data generated based on the prediction results as inputs for the simulation experiment;

[0068] S4.4. Implement predictive control strategies based on load and energy price forecasts, implement predictive control strategies by writing control algorithms, implement control strategies based on real-time input data, and adjust the charge and discharge state of the flow battery to meet load demand and minimize energy costs;

[0069] S4.5. Use Simulink to simulate the real-time system operation process, observe the system operation status, and perform performance evaluation on the results of simulation experiments, including load balancing effect, energy cost saving, and system stability;

[0070] S4.6. Based on the performance evaluation results, optimize and adjust the predictive control strategy until the expected performance indicators are achieved.

[0071] The beneficial effects of the present invention are:

[0072] (1) In this invention, on the basis of the traditional fixed control method, a predictive control method strategy is introduced, which effectively solves the negative effects caused by excessive transmission energy loss in the power grid and uncertainty in both supply and demand, such as low grid efficiency, low sustainability, and lack of flexibility. The fixed control method is compared with the predictive control method of the present invention. The results show that the predictive control method has superior intelligence and flexibility and can seamlessly adapt to the dynamic power demand of the microgrid system. In addition, it shows strong adaptability to different levels of renewable energy penetration and only experiences minimal performance degradation.

[0073] (2) The present invention performs outstandingly in solving the problems of excessive energy loss in power grid transmission and imbalance between supply and demand. While enhancing the stability of the power system, it seamlessly adapts to the inherent volatility of renewable energy generation and the dynamic power demand of the microgrid system, effectively reducing power grid energy loss, optimizing efficiency, and reducing operating costs. Therefore, the method of the present invention is expected to be widely used in future power systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 It is a schematic diagram of the overall process of the present invention;

[0075] Figure 2 Graph showing the variation of pump power with different electrolyte flow rates in an embodiment of the present invention;

[0076] Figure 3 Graph showing changes in concentration overpotential with different electrolyte flow rates and currents in an embodiment of the present invention;

[0077] Figure 4 A schematic diagram of a predictive control model in an embodiment of the present invention;

[0078] Figure 5 A schematic diagram of load and charge / discharge power in an embodiment of the present invention;

[0079] Figure 6 Graph showing changes in efficiency and electrolyte flow rate at different SOCs in an embodiment of the present invention;

[0080] Figure 7 This is a schematic diagram of electricity prices and electricity consumption during off-peak periods, transition periods, and peak periods in an embodiment of the present invention. DETAILED DESCRIPTION

[0081] This embodiment provides a microgrid liquid flow battery energy storage peak regulation method based on predictive control, such as Figure 1 As shown, the following steps are included

[0082] S1. System modeling and simplification, including the following sub-steps:

[0083] S1.1. Determine the basic assumptions: Assume that the electrolyte, electrodes, and membrane have uniform physical properties, such as density, conductivity, etc.; and assume that the diffusion of ions through the membrane is uniform, and consider the permeability of the membrane.

[0084] S1.2. Establish an electrochemical model: Divide the vanadium redox flow battery into positive and negative half-cells, consider the charging and discharging processes respectively, and establish a half-cell model based on this, as shown in the following formula:

[0085] V 3+ +e - →V 2+

[0086]

[0087] During the charging process, the positive half-cell The ions undergo oxidation, generating electrons that migrate to the negative half-cell, where they contribute to V 3+ Reduction of ions; this conversion, which converts chemical energy into electrical energy, is essential to the functioning of batteries, allowing for efficient energy storage and distribution.

[0088] S1.3. Use the Nernst equation to calculate the open circuit voltage and understand the electrochemical characteristics of the battery. The Nernst equation is shown below:

[0089]

[0090] Among them, F represents the Faraday constant, C represents the ion concentration, cs represents the battery stack configuration, and n represents the number of electron transfers in the electrochemical reaction.

[0091] S1.4. Modeling the state of charge (SOC): Establish a mathematical model that considers the SOC changes of the two half-cells to capture the SOC changes caused by the diffusion of vanadium ions, including considering the changes in ion concentration during charging and discharging. The specific formula is as follows:

[0092]

[0093]

[0094] Among them, SOC- represents the reduction in the battery's state of charge, SOC + represents the increase in the state of charge of the battery; tk, tk2, tk3, tk4 and tk5 represent different storage tanks, which are important components of the vanadium redox flow battery (VRFB) system and contain electrolytes that are critical for energy storage; W CS Indicates battery capacity, C CS represents the ion concentration, Q CS Indicates the battery charging capacity, C tki represents the ion concentration in the storage tank, C csi Indicates the ion concentration in the electrolyte during charging, Φ i represents the battery molar flux, S i represents the battery internal source term, Φ i and S i Affects the change of concentration over time; t represents time; the value of variable i is 2, 3, 4 or 5, representing different types of vanadium ions.

[0095] Finally, the open circuit voltage η oa It is derived from the product of current density and resistivity, including ohmic and activation overpotential, as follows:

[0096] η oa =η ohm +η act =ir

[0097] Among them, η ohm represents the ohmic loss of the battery, η act It represents the chemical activity loss of the battery, and ir represents the product of the current and the internal resistance of the battery.

[0098] S1.5. Consider concentration overpotential, formation potential, battery voltage and SOC: Consider system parameters such as formation potential and SOC in modeling; formation potential affects the open circuit voltage of the battery, while SOC reflects the charge and discharge state of the battery, thereby determining its impact on system performance; concentration overpotential, formation potential, battery voltage and SOC are used as optimization variables to maximize the performance and efficiency of the system; concentration overpotential is caused by the concentration difference between the bulk electrolyte and the electrode surface, which greatly affects the system efficiency and operation; it essentially represents the resistance encountered by ions during migration and shapes the kinetics of electrochemical reactions; managing this overpotential is crucial to improving the performance and life of VRFB, as shown in the following formula:

[0099]

[0100] Among them, η con represents the concentration overpotential, R represents the ideal gas constant, T represents the temperature, Q cs represents the vanadium concentration in the electrolyte, A fl represents the electrode surface area, C rea It represents the rate constant of the reversible reaction at the electrode.

[0101] Battery voltage is a determining factor in the charge and discharge performance of an electrochemical system and plays a vital role in its efficiency. In energy storage systems, including VRFB, managing the factors that affect battery voltage is critical to optimizing system operation and maximizing its benefits. This is shown in the following equation:

[0102] E chcell =E OCV +η oa +η con

[0103] E discell =E OCV -η oa -η con

[0104] E stack =N cell ·E cell

[0105] Among them, E chcell Indicates the chemical energy of the battery cell during the charging process; E OCV Represents the open circuit voltage energy of the battery; η oa Indicates the battery open circuit voltage; η con Indicates other battery losses; E discell Indicates the energy released by the battery cell during the discharge process; E stack Represents the total energy of the entire battery stack; N cellN represents the number of unit stacks contained in the module, which is a key factor affecting the overall capacity and function of the VRFB system. cell The number of battery stacks directly affects the system's energy storage capacity, power output, and scalability, while also influencing considerations such as system complexity, cost, and maintenance requirements.

[0106] S1.6. Modeling pump power and energy efficiency: Analyze the flow of electrolyte in the system and evaluate the impact of pump power consumption on system energy efficiency, such as considering the flow and pressure characteristics of the pump, as shown in the following formula:

[0107]

[0108] Where Δp represents the pressure rise (or pressure difference) of the pump, Q represents the flow rate, and η pump Indicates the efficiency of the pump.

[0109] S1.7、Optimize the operating parameters of the pump: Figure 2 and Figure 3 As shown in the figure, the pump power changes with different electrolyte flow rates and concentration overpotentials with different electrolyte flow rates and currents. By analyzing the changes in power loss with electrolyte flow rate and current and pump power and concentration overpotential, the pump operating parameters are optimized to maximize the efficiency of the system and reduce energy loss. The specific process of optimizing the pump is shown in the following formula:

[0110] Δp=Δp pipe +Δp cell

[0111]

[0112] Among them, μ represents the dynamic viscosity of the electrolyte, ρ represents the density of the electrolyte, ∈ represents the electrode porosity, f represents the Darcy friction coefficient, and K represents the Kozenye-Carman constant; they jointly determine the pressure loss of the VRFB system. These parameters significantly affect the electrolyte flow dynamics and the overall system performance.

[0113] S2, predictive control and optimization, specifically includes the following sub-steps:

[0114] S2.1. Establish a predictive control model: Establish a Figure 4 The predictive control model shown is used to describe the dynamic behavior of the system. The future behavior of the system is predicted and the control input is adjusted based on these predictions. An optimization algorithm is used to design the controller to optimize the system according to its performance indicators.

[0115] S2.2. Implement data acquisition and feedback: Use different sensors to measure various parameters in the system, such as voltage, current, temperature, etc., and pass the data into the control algorithm to update the system's state estimation and prediction model; combine predictive control with feedback control to correct the system when deviations occur; feedback control usually involves measuring the difference between the actual system state and the predicted state, and adjusting the control input based on these differences.

[0116] The efficiency of a VRFB system is critical to its practical application and economic feasibility. Efficiency indicators such as charging efficiency, discharging efficiency and total efficiency provide insights into the energy conversion process within the system; charging efficiency compares the stored energy to the total energy consumed during the charging process, while discharging efficiency compares the transferred energy to the total energy released during the discharging process. The specific process is shown in the following formula:

[0117]

[0118] Among them, η represents the efficiency of the system process, E storage Indicates the total energy stored during the charging process, E output Indicates the total energy actually released during the discharge process, E release Indicates the total energy released by the discharge, E input Represents the total energy of charging input, η ch and η dis Indicates the charge and discharge efficiency.

[0119] S2.3, Particle Swarm Optimization (PSO) integration: Use the PSO algorithm to optimize controller parameters in predictive control to maximize the system performance indicators; in predictive control, adjust the controller parameters, such as gain or weight, to minimize the control error or optimize the system performance indicators; analyze the convergence of the PSO algorithm to ensure that the particle swarm algorithm can find a solution that is close enough to the optimal solution within a limited number of iterations, as shown in the following formula:

[0120]

[0121] Where J represents the objective function; represents the input efficiency at time step i+1; η i+1 represents the efficiency at time step i+1; λ represents the weight coefficient, which is used to adjust parameters and penalty terms; Q i+1 and Q i Represent the state variables of the system at time steps i and i+1 respectively.

[0122] The objective function J and weight coefficient λ of PS0 are the key to determine the microgrid scale peak load control strategy, where the rated power Ppower The power plant size represented by is calculated iteratively using the following formula, satisfying a smaller ω value to ensure accuracy:

[0123] iP power >P load :P ch =P power -P load

[0124] iP power <P load :P dis =P load -P power

[0125] η ch η dis ∫ t1 P ch dt-∫ t2 P ch dt<ω

[0126] Among them, P power Indicates the output power of the battery, P load represents the load power, ω represents the energy threshold, ∫ t1 P ch dt and ∫ t2 P ch dt represents the power integration during the charging and discharging process.

[0127] S2.4, describing the interaction between dynamic current and voltage and determining the charging and discharging operation modules, describes how the current changes with the dynamic change of the stack voltage in the energy storage system through the following equation, and then reflects the complex interaction between electrical characteristics and operating parameters:

[0128] N ch ·P ch <P ch η ch <(N ch +1)P ch

[0129] Then, by considering the maximum power demand, battery charge state and energy conversion efficiency, the optimal operation module under different conditions is determined to improve the efficiency and reliability of the system, as shown in the following formula:

[0130]

[0131] Among them, N ch and N dis Represent the number of charge and discharge cycles, P ch and P dis Represent the power of charging and discharging, Pstack and E stack Represent the power and energy of the stack respectively.

[0132] S2.5, calculate and determine the SOC, so as to gain a deeper understanding of the system's operating status, develop more effective charging and discharging strategies and battery management systems, and optimize their performance, as shown in the following formula:

[0133]

[0134] Among them, SOC ch and SOC dis Respectively represent the state of charging and discharging, EN initial ,EN ch ,EN dis and EN tot They represent initial energy, charging energy, discharging energy and total energy respectively.

[0135] S2.6. Finally, design an experiment for surface fitting: Design an experiment to collect the input and output data of the system so as to use the surface fitting method to estimate the dynamic characteristics of the system, such as charge and discharge efficiency and electrolyte flow rate; by selecting an appropriate fitting function and adjusting its parameters, make it as close as possible to the actual data. The specific fitting process is shown in the following formula:

[0136] η fit =∑p mn SOC m I n (0≤m≤5; 0≤n≤5; m+n≤5)

[0137] Among them, η fit represents the fitting efficiency; p mn represents the coefficients in the fitting model, which is used to fit the actual data; m and n are exponents, which are used to determine the powers of SOC and current in the polynomial.

[0138] S3. Evaluate and optimize the performance of the flow battery energy storage system implementing predictive control, including the following sub-steps:

[0139] S3.1. Establish a numerical simulation model: Use physical equations to establish a numerical simulation model to describe the dynamic behavior of the entire battery system including various parts such as batteries, controllers and the environment; then use numerical methods (such as finite element method or finite difference method) to discretize and solve the model equations to update the system state and control input at each time step.

[0140] S3.2. Define performance indicators: Define performance indicators such as energy efficiency, response time and stability to evaluate the performance of the system. When defining, it is necessary to consider system constraints such as current limit, voltage limit and temperature limit to meet the physical characteristics and safety requirements of the system.

[0141] S3.3 System performance evaluation: The performance of the flow battery energy storage system implementing predictive control shall be evaluated in terms of load balancing effect, energy cost savings and system stability.

[0142] S3.4. Control strategy optimization: Based on the performance evaluation results, the predictive control strategy is adjusted and optimized to improve the peak load regulation effect and economy of the system.

[0143] S4. Construct a simulation model for verification analysis, which specifically includes the following steps:

[0144] S4.1. Use the MATLAB / Simulink platform to establish a comprehensive model of the liquid flow battery energy storage system, which includes a battery model, a load model, and a controller model.

[0145] S4.2, parameter setting and initialization: set the parameters of the simulation experiment, including battery parameters, load demand, and energy price, and initialize the battery charging state and controller parameters.

[0146] S4.3. Importing load and energy prices: Importing historical load and energy price data and future data generated based on the prediction results as inputs to the simulation experiment.

[0147] S4.4. Implement predictive control strategies based on load and energy price forecasts. By writing control algorithms and implementing control strategies based on real-time input data, the charge and discharge states of the flow battery are adjusted to meet load demand and minimize energy costs.

[0148] S4.5. Use Simulink to simulate the real-time system operation process, observe the system operation status, and perform performance evaluation on the results of simulation experiments such as load balancing effect, energy cost savings, and system stability.

[0149] S4.6. Based on the performance evaluation results, optimize and adjust the predictive control strategy to improve the peak load regulation effect and economy of the system until the expected performance indicators are achieved.

[0150] In this embodiment, actual electricity price data of a certain park during peak, flat and valley periods are used, as shown in Table 1 below.

[0151] Table 1

[0152] Low period Transition period Peak Electricity price ($ / kWh) 0.0417 0.1118 0.1896 Electricity consumption (kWh) -11869.3 833.98 9184.93

[0153] Through the MATLAB / Simulink platform, a comprehensive model of the flow battery energy storage system is established, which includes a battery model, a load model, a controller model, etc.

[0154] Set parameters and initialize, import battery parameters, load demand, and electricity price data as shown in Table 1; import historical load and price data and future data generated based on prediction results as input for the experiment.

[0155] Based on load and energy price forecasts, predictive control strategies are implemented. Predictive control models such as Figure 4 shown.

[0156] Write control algorithms to implement control strategies based on real-time input data to adjust the charge and discharge state of the flow battery to meet load demand and minimize energy costs.

[0157] Analyze the system energy loss and explore the change of power loss with electrolyte flow rate and current. Figure 2 and Figure 3 As shown, it can be obtained that: in the later stage of charging (SOC is 90%), the concentration overpotential increases with the current and decreases with the flow rate; optimizing the flow rate alone cannot minimize both the pump power and the concentration overpotential; therefore, the focus should be on mitigating the combined impact of these factors to improve the performance of VRFB.

[0158] Consider the cost economics, such as the impact of electrolyte costs on the investment cost of VRFB, such as Figure 5 As shown; it can be seen that profitability has the greatest impact on the VRFB investment payback period, followed by electrolyte cost; it is worth noting that a 45% reduction in electrolyte cost can shorten the investment payback period to 10.70 years, while a corresponding 45% increase in profitability can shorten the investment payback period to 12.08 years; therefore, implementing a suitable time-of-use electricity price scheme or reducing electrolyte prices can significantly improve the economic feasibility of VRFB.

[0159] The efficiency of the control strategy is optimized by surface fitting method and compared with the traditional control method. Figure 6 As shown, the average efficiency of the whole charging process is 90.78% for the traditional method and 91.12% for the method of this embodiment. In the later stage of charging, the efficiency of the method of this embodiment is obviously higher than that of the traditional method, and the control of the electrolyte flow rate is more precise.

[0160] According to the performance evaluation results, the predictive control strategy is optimized and adjusted to achieve peak load regulation; Figure 7Figure 2 shows the result of the peak load control strategy, where the original microgrid load profile is represented by the purple line; using this method, the initial rated power of the power generation system is set to 6.12MW, represented by the red line, below this line, the green bar represents the charging area, indicating the power input from the grid, and above this line, the purple bar represents the discharging area, reflecting the power supply to the grid; despite the energy loss, the total energy intake during charging (green bar) exceeds the energy released (purple bar).

[0161] In summary, the present embodiment method uses rigorous analysis and cutting-edge methods to provide a deep understanding of the complex design, seamless operation and cost-effectiveness of the biomass integrated microgrid system; compared with the traditional fixed control method, the proposed predictive control method has superior intelligence and flexibility, and can seamlessly adapt to the dynamic power demand of the microgrid system; by significantly improving the efficiency of vanadium redox flow battery (VRFB) and reducing operating costs, the present embodiment method is expected to improve the sustainability and resilience of microgrid infrastructure.

[0162] In addition to the above embodiments, the present invention may also have other implementation modes. Any technical solution formed by equivalent replacement or equivalent transformation falls within the protection scope required by the present invention.

Claims

1. A microgrid liquid flow battery energy storage peak regulation method based on predictive control, characterized in that: The following steps are involved: S1. System modeling and simplification: Assume that the electrolyte, electrodes and membrane have uniform physical properties, and assume that the diffusion of ions through the membrane is uniform, and consider the permeability of the membrane; divide the vanadium redox flow battery into positive and negative half-cells, consider the charging and discharging processes respectively, and establish a half-cell model based on this; use the Nernst equation to calculate the open circuit voltage; establish a mathematical model to consider the charge state changes of the two half-cells to capture the charge state changes caused by the diffusion of vanadium ions; consider concentration overpotential, formation potential, battery voltage and charge state; model the pump power and energy efficiency; and finally optimize the operating parameters of the pump; S2. Predictive control and optimization: Build a mathematical model to describe the dynamic behavior of the system, predict the future behavior of the system and adjust the control input based on these predictions, and use optimization algorithms to design controllers to optimize the system according to its performance indicators; use different sensors to measure various parameters in the system and pass the data into the control algorithm to update the system's state estimation and prediction model; combine predictive control with feedback control to correct the system when deviations occur, and observe the energy conversion process in the system through efficiency indicators; use particle swarm optimization in predictive control to optimize controller parameters to maximize the system's performance indicators; describe the interaction between dynamic current and voltage and determine the charging and discharging operation modules; Calculate and determine the charging state; finally design an experiment to perform surface fitting; S3. Performance evaluation and optimization of flow battery energy storage systems implementing predictive control; S4. Build a simulation model for verification analysis.

2. According to a microgrid liquid flow battery energy storage peak-shaving method based on predictive control according to claim 1, it is characterized in that: In step S1, it is assumed that the electrolyte, the electrode and the membrane have uniform physical properties, including density and conductivity; the vanadium redox flow battery is divided into positive and negative half-cells, and the charging and discharging processes are considered respectively, and a half-cell model is established based on this, as shown in the following formula: V 3+ +e - →V 2+ During the charging process, the positive half-cell The ions are oxidized, generating electrons that migrate to the negative half-cell, where they promote V 3+ Reduction of ions, thereby converting chemical energy into electrical energy; The open circuit voltage is calculated using the Nernst equation as shown below: Among them, F represents the Faraday constant, C represents the ion concentration, cs represents the battery stack configuration, and n represents the number of electron transfers in the electrochemical reaction.

3. A method for peak-shaving energy storage of liquid flow batteries in a microgrid based on predictive control according to claim 1, characterized in that: In step S1, a mathematical model is established to consider the charge state changes of the two half-cells to capture the charge state changes caused by the diffusion of vanadium ions, including the changes in ion concentration during the charging and discharging process. The specific formula is as follows: Among them, SOC- represents the reduction in the battery's state of charge, SOC + represents the increase in the state of charge of the battery; tk, tk2, tk3, tk4 and tk5 represent different storage tanks; W CS Indicates battery capacity, C CS represents the ion concentration, Q CS Indicates the battery charging capacity, C tki represents the ion concentration in the storage tank, C csi Indicates the ion concentration in the electrolyte during charging, Φ i represents the battery molar flux, S i represents the battery internal source term, Φ i and S i Affects the change of concentration over time; t represents time; the value of variable i is 2, 3, 4 or 5, representing different types of vanadium ions; Finally, the open circuit voltage η oa It is derived from the product of current density and resistivity, including ohmic and activation overpotential, as follows: or oa =the ohm +n act =ir Among them, η ohm represents the ohmic loss of the battery, η act It represents the chemical activity loss of the battery, and ir represents the product of the current and the internal resistance of the battery.

4. A method for peak-shaving energy storage of liquid flow batteries in a microgrid based on predictive control according to claim 1, characterized in that: In step S1, concentration overpotential, formation potential, battery voltage and state of charge are considered: concentration overpotential, formation potential, battery voltage and state of charge are used as optimization variables to maximize the performance and efficiency of the system, as shown in the following formula: Among them, η con represents the concentration overpotential, R represents the ideal gas constant, T represents the temperature, Q cs represents the vanadium concentration in the electrolyte, A fl represents the electrode surface area, C rea represents the rate constant of the reversible reaction at the electrode; and manages the factors affecting the cell voltage by the following equation: E chcell =E OCV +n oa +n con E discell =E OCV -or oa -or con AND stack =N cell ·AND cell Among them, E chcell Indicates the chemical energy of the battery cell during the charging process, E OCV Represents the battery open circuit voltage energy, η oa Represents the battery open circuit voltage, η con Indicates other losses of the battery, E discell Indicates the energy released by the battery cell during the discharge process, E stack Represents the total energy of the entire battery stack, N cell Indicates the number of cell stacks contained within the module.

5. A method for peak-shaving energy storage of liquid flow batteries in a microgrid based on predictive control according to claim 1, characterized in that: In step S1, pump power and energy efficiency modeling is performed: the flow of electrolyte in the system is analyzed, and the influence of pump power consumption on system energy efficiency is evaluated, as shown in the following formula: Where Δp represents the pressure rise or pressure difference of the pump, Q represents the flow rate, and η pump Indicates the efficiency of the pump; Optimize the operating parameters of the pump. The specific process is shown in the following formula: Δp=Δp pipe +Δp cell Where μ represents the dynamic viscosity of the electrolyte, ρ represents the density of the electrolyte, ∈ represents the electrode porosity, f represents the Darcy friction coefficient, and K represents the Kozeny-Karman constant.

6. A method for peak-shaving energy storage of liquid flow batteries in a microgrid based on predictive control according to claim 1, characterized in that: In step S2, various parameters in the system, including voltage, current and temperature, are measured by different sensors, and the data are input into the control algorithm to update the state estimation and prediction model of the system; the prediction control is combined with the feedback control to correct the system when deviation occurs; the energy conversion process in the system is observed by the efficiency index, which includes the charging efficiency, the discharging efficiency and the total efficiency. The specific process is shown in the following formula: Among them, η represents the efficiency of the system process, E storage Indicates the total energy stored during the charging process, E output Indicates the total energy actually released during the discharge process, E release Indicates the total energy released by the discharge, E input Represents the total energy of charging input, η ch and η dis Indicates the charge and discharge efficiency; In predictive control, the particle swarm algorithm is used to optimize the controller parameters to maximize the performance indicators of the system. In predictive control, the parameters of the controller, including gain and weight, are adjusted to minimize the control error or optimize the performance indicators of the system. The convergence of the particle swarm algorithm is analyzed to ensure that the particle swarm algorithm can find a solution that is close enough to the optimal solution within a limited number of iterations, as shown in the following formula: J=min((ηr i+1 -or i+1 +λ(Q i+1 -Q i ) 2 ) 2 ) Where J represents the objective function; ηr i+1 represents the input efficiency at time step i+1; η i+1 represents the efficiency at time step i+1; λ represents the weight coefficient, which is used to adjust parameters and penalty terms; Q i+1 and Q i Respectively represent the state variables of the system at time steps i and i+1; Rated power P power The power plant size represented is calculated iteratively using the following formula: i£P power >P load :P ch =P power -P load i£P power <P load :P dis =P load -P power η ch η dis ∫ t1 P ch dt-∫ t2 P ch dt<ω Among them, P power Indicates the output power of the battery, P load represents the load power, ω represents the energy threshold, ∫ t1 P ch dt and ∫ t2 P ch dt represents the power integration during the charging and discharging process.

7. A method for peak-shaving energy storage of liquid flow batteries in a microgrid based on predictive control according to claim 1, characterized in that: In step S2, the interaction between dynamic current and voltage and the determination of charging and discharging operation modules are described: how the current in the energy storage system changes with the dynamic change of the stack voltage is described by the following formula: N ch ·P ch <P ch or ch <(N ch +1)P ch Then, by considering the maximum power demand, battery charge state and energy conversion efficiency, the optimal operation module under different conditions is determined, as shown in the following formula: Among them, N ch and N dis Represent the number of charge and discharge cycles, P ch and P dis Represent the power of charging and discharging, P stack and E stack Represent the power and energy of the stack respectively.

8. A method for peak-shaving energy storage using liquid flow batteries in a microgrid based on predictive control according to claim 1, characterized in that: In step S2, the charging state is calculated and determined by the following formula: Among them, SOC ch and SOC dis Respectively represent the state of charging and discharging, EN initial ,EN ch ,EN dis and EN tot They represent initial energy, charging energy, discharging energy and total energy respectively; Experiments are designed to collect input and output data of the system in order to use the surface fitting method to estimate the dynamic characteristics of the system, including charge and discharge efficiency and electrolyte flow rate. The specific fitting process is shown in the following formula: η fit ∑p mn SOC m I n (0≤m≤5\0≤n≤5\m+n≤5) Among them, η fit represents the fitting efficiency; p mn represents the coefficients in the fitting model, which is used to fit the actual data; m and n are exponents, which are used to determine the power of the charge state and current in the polynomial.

9. A method for peak-shaving energy storage using liquid flow batteries in a microgrid based on predictive control according to claim 1, characterized in that: The step S3 specifically includes the following sub-steps: S3.

1. Establish a numerical simulation model: Use physical equations to establish a numerical simulation model to describe the dynamic behavior of the entire battery system; then use numerical methods to discretize and solve the model equations to update the system state and control input at each time step; S3.

2. Define performance indicators: Define performance indicators to evaluate the performance of the system. The performance indicators include energy efficiency, response time, and stability. When defining, consider system constraints. System constraints include current limit, voltage limit, and temperature limit. S3.

3. System performance evaluation: Evaluate the performance of the flow battery energy storage system implementing predictive control, including load balancing effect, energy cost savings, and system stability; S3.

4. Control strategy optimization: Adjust and optimize the predictive control strategy based on the performance evaluation results.

10. A method for peak-shaving energy storage using liquid flow batteries in a microgrid based on predictive control according to claim 1, characterized in that: The step S4 specifically includes the following sub-steps: S4.

1. Using MATLAB / Simulink platform, establish a comprehensive model of flow battery energy storage system, which includes battery model, load model and controller model; S4.2, parameter setting and initialization: set the parameters of the simulation experiment, including battery parameters, load demand and energy price, and initialize the battery charging state and controller parameters; S4.3, importing load and energy prices: importing historical load and energy price data and future data generated based on the prediction results as inputs for the simulation experiment; S4.

4. Implement predictive control strategies based on load and energy price forecasts, implement predictive control strategies by writing control algorithms, implement control strategies based on real-time input data, and adjust the charge and discharge state of the flow battery to meet load demand and minimize energy costs; S4.

5. Use Simulink to simulate the real-time system operation process, observe the system operation status, and perform performance evaluation on the results of simulation experiments, including load balancing effect, energy cost saving, and system stability; S4.

6. Based on the performance evaluation results, optimize and adjust the predictive control strategy until the expected performance indicators are achieved.

Citation Information

Cited By

  • Flow control system of flow battery

    CN120103880A