A DC micro-grid energy storage system performance equalization droop control method

Abstract

The application discloses a DC micro-grid energy storage system performance balancing droop control method, first, based on the ANN battery capacity evaluation model, the remaining capacity of each energy storage unit battery is evaluated online in real time; secondly, the battery performance index is calculated based on the remaining capacity of the battery, and the performance index deviation degree of each energy storage unit battery is calculated; then, the maximum deviation degree of all energy storage unit battery performance indexes is determined to determine the system working mode, and the performance balancing-based droop control is started when the battery performance index deviation is large; next, each energy storage unit corrects the initial droop coefficient according to the battery performance index deviation degree to obtain the optimized droop coefficient, and then the reference value of the output voltage of each energy storage unit battery is obtained from the droop control, and finally, the PWM switching signal is generated through the voltage-current double closed loop control of the energy storage unit interface DC / DC converter to make the output follow the reference value. Through the method, the overall service life of the DC micro-grid energy storage system can be effectively prolonged.

Description

Technical Field

[0001] This invention belongs to the technical field of DC microgrid energy storage systems, and in particular relates to a method for performance droop control of DC microgrid energy storage systems. Background Technology

[0002] In recent years, to address the escalating global energy crisis, the use of clean energy sources such as photovoltaics and wind power has become increasingly widespread. However, renewable energy generation is characterized by randomness, volatility, and intermittency, posing a severe challenge to the power system's source-load balance scheduling and stable operation after being integrated into the power grid. To address these shortcomings of renewable energy generation, microgrids have emerged as a solution. Deploying energy storage systems within microgrids, these systems possess bidirectional power regulation capabilities, allowing them to charge during low load periods and discharge during high load periods to compensate for power imbalances between loads and power sources. Based on the bidirectional power regulation of energy storage systems, combined with various energy management strategies such as multi-energy synergy and load demand response, power fluctuations in the system can be effectively mitigated, reducing the adverse impact of large-scale renewable energy generation on power system stability, improving renewable energy absorption efficiency, and enhancing the power supply-demand balance capability.

[0003] Microgrids are mainly divided into two categories: AC microgrids and DC microgrids. Compared to AC microgrids, DC microgrids do not have frequency stability and synchronization stability issues, nor do they require reactive power regulation. Therefore, DC microgrids are receiving increasing attention from academia and industry, and have broad development prospects in practical applications. DC microgrid energy storage systems are an important component of DC microgrids, comprising energy storage devices and energy storage control systems. Energy storage devices can consist of components such as batteries, capacitors, and supercapacitors. Among these, battery energy storage occupies a dominant position in energy storage systems, especially large-capacity systems, due to its advantages of large storage capacity and good economic efficiency. Research on DC microgrids and DC microgrid energy storage systems has significant practical implications, contributing to the construction of intelligent, efficient, and reliable power systems to meet the ever-increasing energy demands and environmental protection requirements.

[0004] Due to the limited capacity of a single energy storage unit, energy storage systems in DC microgrids typically consist of multiple energy storage units connected in parallel to the DC bus. Therefore, current distribution among these units is a fundamental issue in the energy management of DC microgrid energy storage devices. Among various current distribution methods, droop control has gained widespread application and attention in distributed systems with no or low-speed communication due to its advantages of not relying on high-speed communication and its simplicity of implementation. For DC microgrid energy storage systems, the traditional droop control method is based on battery state of charge (SOC) balancing. This method aims to prioritize charging current allocation to battery cells with lower SOC during a single charge cycle and prioritizing discharging current allocation to battery cells with higher SOC during a single discharge cycle. This achieves SOC balance among the battery cells over a period of time, such as several hours, avoiding overcharging and over-discharging problems of individual battery cells.

[0005] However, during long-term operation, DC microgrid energy storage systems experience varying capacity decay and aging rates due to differences in the electrochemical characteristics, temperature, and actual load of individual battery cells. Over extended periods, key parameters such as remaining capacity, internal impedance, and no-load potential among the cells will show significant differences, leading to a gradual increase in the performance imbalance among battery cells. When the health of individual battery cells becomes too low, the entire energy storage system may be forced out of service for maintenance, reducing its overall service life. Traditional droop control based on SOC equalization aims for SOC equalization within a single charge-discharge cycle of several hours. Assuming that the battery capacity of all energy storage cells remains constant over the long term, even if the number of charge-discharge cycles for each cell remains the same, it cannot prevent differences in battery aging rates, leading to a gradual increase in performance imbalance among the cells. Therefore, it is necessary to consider the long-term goal of balanced battery performance and propose new control methods. Summary of the Invention

[0006] To address the issue that traditional SOC-based droop control does not consider the adverse effects of battery performance differences among energy storage units in a DC microgrid energy storage system on the overall service life of the entire system, this paper proposes to select the remaining battery capacity of an energy storage unit as an evaluation index for its performance status and apply this evaluation index to the optimization calculation of the droop coefficient, thereby achieving the goal of performance balance among energy storage units in a DC microgrid energy storage system.

[0007] To achieve the above objectives, the present invention aims to provide a performance droop control method for a DC microgrid energy storage system, the method comprising:

[0008] Step 1: Build and train a battery capacity assessment model based on artificial neural networks;

[0009] Step 2: Use the battery capacity assessment model to evaluate the remaining capacity of all energy storage units in the DC microgrid energy storage system in real time online, obtain the remaining capacity of each energy storage unit, and calculate the battery performance evaluation index and battery performance index deviation of each energy storage unit. The battery performance evaluation index of each energy storage unit is calculated based on the remaining capacity and rated capacity of the battery of that energy storage unit, and the battery performance index deviation of each energy storage unit is calculated based on its own performance evaluation index and the comprehensive performance evaluation index of all energy storage units.

[0010] Step 3: If the current droop control operating mode is S1, and the absolute value of the deviation of the battery performance index of a certain energy storage unit is greater than or equal to the preset threshold t1, then switch the droop control operating mode of the DC microgrid energy storage system to S2; if the current operating mode is S2, and the absolute value of the deviation of the battery performance index of all energy storage units is less than or equal to the preset threshold t2, then switch the droop control operating mode of the DC microgrid energy storage system to S1. Here, S1 represents droop control based on SOC equalization, and the droop coefficient of each energy storage unit is related to the SOC value of the battery of that energy storage unit; S2 represents droop control based on performance equalization, and the droop coefficient of each energy storage unit is related to the battery performance index of that energy storage unit. The initial operating mode of the DC microgrid energy storage system is S1; the threshold t1 is not less than the threshold t2.

[0011] Step 4: Based on the droop control operating mode determined in Step 3, update the correction coefficient of the droop control, and calculate the final droop control coefficient corresponding to the operating mode based on this correction coefficient. The underlying voltage-current dual closed-loop control of the energy storage unit DC / DC converter is usually adopted. Therefore, the droop control is used as the front-end loop, and the output of the droop control is used as the output voltage reference value of the energy storage unit DC / DC converter. The PWM switching signal is generated by the voltage-current dual closed-loop control of the underlying voltage of the energy storage unit DC / DC converter to realize the current or power control of the DC / DC converter of each energy storage unit interface.

[0012] Furthermore, in the battery capacity assessment model based on artificial neural networks, the neurons in the input layer correspond one-to-one with battery-related parameters, and the neurons in the output layer correspond to the remaining battery capacity. Among these, the battery-related parameters include the battery state of charge, battery output voltage, and battery output current.

[0013] Furthermore, the battery performance indicators are expressed by the formula:

[0014] ,

[0015] Among them, D i In the DC microgrid energy storage system, the first Battery performance indicators of each energy storage unit , This indicates the number of all energy storage units in this energy storage system; Indicates the rated capacity of the battery; This indicates the first value obtained using the battery capacity assessment model. The remaining battery capacity of each energy storage unit.

[0016] Furthermore, the performance deviation of the i-th energy storage unit battery is expressed by the formula:

[0017] ,

[0018] Among them, D avg This represents the battery performance indicators D1, D2, ..., D of all energy storage units in this energy storage system. i ... D N The comprehensive evaluation index is denoted as the global target value of the battery performance index of this energy storage system; ΔD i Indicates the first Battery performance index D of each energy storage unit i The global target value D of the battery performance index of this energy storage system avg The deviation between the target and the target values ​​is considered. A positive deviation indicates that the battery performance of the energy storage unit is better than the target value, while a negative deviation indicates that the battery performance of the energy storage unit is worse than the target value.

[0019] Furthermore, in S2 operating mode, the droop coefficient R of each energy storage unit dpi Expressed as a formula:

[0020] ,

[0021] Among them, R 0i f represents the initial droop factor of the i-th energy storage unit; dpi The coefficient applied to the initial droop factor represents the correction coefficient for droop control obtained by the i-th energy storage unit based on the battery performance balancing target. When the battery performance of all energy storage units reaches equilibrium, each D... i All are the same, i.e., ΔD i =0 f dpi = 1, which means that there is no need to adjust the initial droop coefficient.

[0022] In S2 operating mode, the droop factor of each energy storage unit is expressed by the formula:

[0023] ,

[0024] in, Indicates the first The droop factor ultimately used by each energy storage unit These are real coefficients greater than 0.

[0025] Furthermore, in S1 operating mode, the droop coefficient R of each energy storage unit dpi With the SOC value S of the energy storage unit OCi The relationship between them can be expressed by the following formula:

[0026] ,

[0027] Among them, R 0i still represents the initial droop coefficient of the i-th energy storage unit; f dp (S oci ) is a function with the SOC of the i-th energy storage unit as the independent variable, acting on the initial droop coefficient, and when the SOC of all units is completely equal, that is, for any two energy storage units i and j, S ∈ SOC. oci = S ocj At this time, there is f dp (S oci ) = f dp (S ocj = 1 means that when the SOC of each energy storage battery reaches complete equilibrium, there is no need to adjust the initial droop coefficient.

[0028] The beneficial effects of this invention are:

[0029] This invention evaluates the remaining battery capacity online in real time based on an ANN battery capacity assessment model. Secondly, it calculates the battery performance evaluation index using the remaining capacity of the energy storage unit's battery, and determines the system operating mode based on the maximum deviation of all energy storage unit performance evaluation indices. When the performance index deviation is large, droop control based on performance balancing is initiated. Finally, each energy storage unit establishes performance evaluation indices based on its remaining capacity to achieve performance balancing droop control. This allows energy storage units with poor performance to bear a smaller current during charging and discharging to reduce their capacity degradation rate, while energy storage units with better performance bear a larger current during charging and discharging to increase their capacity degradation rate. This ensures that the performance of each energy storage unit in the DC microgrid energy storage system eventually tends towards equilibrium, reducing the adverse impact of premature failure of poor-performing energy storage units on the overall lifespan of the entire energy storage system, and effectively extending the overall service life of the DC microgrid energy storage system.

[0030] This invention selects the remaining battery capacity as a benchmark to establish performance evaluation indicators for each energy storage unit, and implements droop control based on performance balance based on these performance evaluation indicators. With the goal of converging the deviation of battery performance indicators to 0, droop control is implemented on each energy storage unit in the energy storage system, which can achieve the purpose of balanced and uniform degradation of the performance of each energy storage unit, thereby improving the overall service life of the energy storage system.

[0031] In the droop control process based on performance balancing described in this invention, apart from calculating the global target value of the energy storage unit's battery performance, all other calculations can be performed in the local controller of each energy storage unit. The communication between the local controller and the central controller is limited to uploading the locally evaluated battery performance index results, receiving the global target value of battery performance, and the operating mode switching signal. Since the performance index evaluation results of each energy storage battery change slowly, low-speed communication is sufficient to meet the requirements, and there is no other additional information exchange or communication between the energy storage units.

[0032] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0033] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will now be described in further detail with reference to the accompanying drawings, wherein:

[0034] Figure 1 This is a schematic diagram of the performance droop control method for a DC microgrid energy storage system according to the present invention.

[0035] Figure 2 This is a diagram of the performance droop control structure of the DC microgrid energy storage system in an embodiment of the present invention. Detailed Implementation

[0036] The preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be understood that the preferred embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0037] Combination Figure 1 and Figure 2 A method for performance droop control in a DC microgrid energy storage system includes the following steps:

[0038] Step 1: Build and train a battery capacity assessment model based on an Artificial Neural Network (ANN);

[0039] Step 2: Use the battery capacity assessment model to evaluate the remaining battery capacity of all energy storage units in the DC microgrid energy storage system in real time online, obtain the remaining battery capacity of each energy storage unit, and calculate the battery performance evaluation index and battery performance index deviation of each energy storage unit. The battery performance evaluation index of each energy storage unit is calculated based on the remaining battery capacity and the rated battery capacity of the energy storage unit, and the battery performance deviation index of each energy storage unit is calculated based on its own performance evaluation index and the global target value of battery performance of all energy storage units.

[0040] Step 3: If the current droop control mode is S1, and the battery performance deviation of any energy storage unit is greater than the preset threshold t1, then switch the droop control mode of the DC microgrid energy storage system to S2; if the current droop control mode is S2, and the battery performance deviation of all energy storage units is less than the preset threshold t2, then switch the droop control mode of the DC microgrid energy storage system back to S1. Here, S1 represents droop control based on battery SOC balancing, where the droop coefficient of each energy storage unit is related to the SOC value of its battery; S2 represents droop control based on battery performance balancing, where the droop coefficient of each energy storage unit is related to the battery performance deviation of its battery. The initial operating mode of the DC microgrid energy storage system is S1; the threshold t1 is not less than the threshold t2. If t1 is appropriately greater than t2, a hysteresis switching can be formed, which can avoid repeated and frequent switching of control modes near the equilibrium point.

[0041] Step 4: Based on the working mode determined in Step 3, update the correction coefficient of droop control, and calculate the final droop control coefficient (i.e., droop coefficient) corresponding to this working mode based on this correction coefficient.

[0042] The underlying voltage-current dual closed-loop control of the energy storage unit DC / DC converter typically employs droop control as the front-end loop. Therefore, the output of the droop control serves as the output voltage reference value of the energy storage unit DC / DC converter. The underlying voltage-current dual closed-loop control of the energy storage unit DC / DC converter generates a PWM (Pulse Width Modulation) switching signal to achieve current or power control of each energy storage unit interface DC / DC converter.

[0043] In step 1, the battery capacity assessment model based on artificial neural networks (ANNs) may include an input layer, a hidden layer, and an output layer. The neurons in the input layer correspond one-to-one with battery-related parameters, and the neurons in the output layer correspond to the remaining battery capacity. Among them, battery-related parameters include the battery state of charge (S). OCi Output voltage V bati and output current I batiThat is, input X=[S OCi V bati , I bati ], output Y=[Q bati ],in, , This indicates the total number of energy storage units.

[0044] The artificial neural network model can be a recurrent neural network (RNN) model, a convolutional neural network (CNN) model, a Transformer model, etc. By acquiring the battery state of charge, battery output voltage, battery output current, and remaining battery capacity of each energy storage unit from training samples (i.e., some energy storage units), the battery capacity assessment model can be trained. This training can be performed offline on a powerful computer using historical data collected from the energy storage system.

[0045] After the model training is completed, the battery capacity assessment model can be directly used to perform online real-time battery capacity assessment of energy storage units in DC microgrid systems, i.e., step 2.

[0046] The remaining capacity of the battery is evaluated online in real time using this battery capacity assessment model. The obtained remaining capacity value is used for local calculation of battery performance evaluation indicators. The evaluated battery performance indicators are not only used by the local controller, but are also uploaded to the central controller for global battery performance indicator target values ​​(i.e., the battery performance of all energy storage units in this energy storage system, D1, D2, ..., D...). i ... D N The calculation of comprehensive evaluation indicators and the determination of operating mode switching are performed. The central controller of the DC microgrid energy storage system acquires the performance index values ​​of all energy storage unit batteries, calculates the global battery performance index target value, and finds the maximum absolute deviation between the performance index of each energy storage battery and the global battery performance index target value. This maximum deviation value is compared with the set maximum allowable difference (i.e., the preset threshold t1) to determine whether to switch the system's operating mode. This invention can achieve real-time switching of the system's operating mode by estimating the remaining battery capacity online in real time.

[0047] For the i-th energy storage unit in a DC microgrid energy storage system, the state of charge S of the battery is... OCi Output voltage V bati Output current I bati The remaining capacity Q of the battery can be evaluated online in real time based on the ANN capacity assessment model. bati Therefore, based on the remaining battery capacity Q... bati and battery rated capacity Qrated Calculate battery performance index D i After obtaining the performance indicators of all N energy storage units in the DC microgrid energy storage system, the target value D of the global battery performance indicator for all energy storage units in the system is calculated. avg Then, the performance index D of each energy storage unit battery is calculated. i With the global target value D avg The difference, and calculate these differences ΔD. i The maximum absolute value (i.e., the maximum unsigned difference) ΔD max The condition for switching system operating states is the maximum unsigned difference ΔD. max When the maximum unsigned difference ΔD is greater than or equal to the set upper limit of the allowable difference in performance indicators (i.e., threshold t1), the energy storage system switches to the droop control based on performance balancing proposed in this invention (operating state S2); when the maximum unsigned difference ΔD max When the performance index difference is less than or equal to the lower limit of the set allowable difference (threshold t2), the energy storage system switches back to traditional droop control based on SOC equalization (operating state S1), which can be expressed as:

[0048] ,

[0049] Among these, threshold t1 is not less than threshold t2. If t1 is chosen to be appropriately greater than t2, hysteresis can be formed, avoiding repeated and frequent switching of control modes near the equilibrium point. It should be noted that the maximum unsigned difference ΔD... max A deviation greater than or equal to the threshold t1 indicates that the battery performance index deviation of a certain energy storage unit is greater than or equal to the preset threshold t1, with the maximum unsigned difference ΔD. max If the absolute value of the deviation of the battery performance index of all energy storage units is less than or equal to the preset threshold t2, it means that the absolute value of the deviation is less than or equal to the preset threshold t2. In this invention, the droop control method based on SOC equalization can be any existing droop control method based on SOC equalization.

[0050] For example, when the system operates in S1 mode, the droop factor R of each energy storage unit dpi The relationship between the energy storage unit's SOC value and the SOC value is expressed by the following formula:

[0051] ,

[0052] Among them, R 0i still represents the initial droop coefficient of the i-th energy storage unit; f dp (S ociS is a function with the SOC of the i-th energy storage unit as the independent variable. It acts on the initial droop coefficient and corrects the droop coefficient by multiplying it by the initial droop coefficient. Furthermore, when the SOC of all energy storage units reaches complete equilibrium, that is, for any two energy storage units i and j, S... oci = S ocj At this point, f should be present. dp (S oci ) = f dp (S ocj ) = 1, which means that there is no need to adjust the initial droop coefficient at this time.

[0053] When the system is operating in S1 mode, it continues to obtain the remaining capacity value online in real time through the ANN-based battery capacity evaluation model, and converts the remaining capacity value into battery performance evaluation indicators. Each energy storage unit calculates the deviation of battery performance indicators based on its own battery performance evaluation indicators, judges whether the droop control switching conditions are met, and has the ability to switch to droop control based on performance balancing at any time.

[0054] When the system operates in S2 mode, the battery performance evaluation index is expressed by the formula:

[0055] ,

[0056] Among them, D i In the DC microgrid energy storage system, the first Battery performance indicators of each energy storage unit , This indicates the number of all energy storage units in this energy storage system; Indicates the rated capacity of the battery; This indicates the first value obtained using the battery capacity assessment model. The remaining battery capacity of each energy storage unit.

[0057] In some embodiments, to reduce random errors such as voltage and current sampling glitches, D can be... i Perform outlier removal and low-pass filtering.

[0058] The performance deviation of the i-th energy storage unit battery is expressed by the formula:

[0059] ,

[0060] Among them, D avg The values ​​D1, D2, ..., D represent the battery performance of all energy storage units in this energy storage system. i ... D N The comprehensive evaluation index is denoted as the target value of the global battery performance index (referred to as the "global target value"); ΔD i Indicates the first Battery performance evaluation index D for each energy storage unit i The deviation between the target value and the overall battery performance metric;

[0061] In some embodiments, D avg The options are D1, D2, …, D i ,…, D N The arithmetic mean, i.e.

[0062] .

[0063] In some embodiments, D avg Alternatively, you can choose D1, D2, …, D i ,…, D N Other means, such as the squared mean, can be used to increase the variability of the indicators, i.e.:

[0064] .

[0065] D avg Alternatively, you can directly select a constant value D as needed. c 0 < D c < 1, so that the performance indicators of all energy storage unit batteries are as close as possible to D c convergence.

[0066] Based on the performance evaluation indicators, the droop coefficient is adjusted from the initial droop coefficient to achieve droop control based on performance balance. Since the most direct factor affecting battery capacity degradation during normal charging and discharging is the magnitude of the charging and discharging current, the larger the charging and discharging current, the faster the capacity degradation. Therefore, when there are performance differences between batteries in an energy storage unit, to ensure their performance converges to a consistent level more quickly, batteries with poorer performance (negative deviation) should bear a smaller charging and discharging current, while batteries with better performance (positive deviation) should bear a larger charging and discharging current. Considering that the deviation of the energy storage unit's output voltage from the bus voltage is usually small during normal operation, the charging and discharging power of the energy storage unit is approximately proportional to the magnitude of its charging and discharging current. Therefore, regardless of whether the target quantity for droop control of the energy storage unit's batteries is the usage current or power, the monotonicity of the function of its charging and discharging current and its performance deviation should be set to an increasing function, i.e.

[0067] ,

[0068] Among them, I oi Let ΔD represent the charging and discharging current of the i-th energy storage unit. i This represents the battery performance deviation (or "battery performance index deviation") of the i-th energy storage unit. This indicates differentiation.

[0069] Considering that the deviation of the energy storage unit's output voltage from the bus voltage is usually small during normal operation of a DC microgrid, and based on the droop control characteristics, the magnitude of the energy storage unit's charging and discharging current is approximately inversely proportional to the droop coefficient, i.e.:

[0070] ,

[0071] Among them, R dpi and R dpj This represents the droop coefficient ultimately used by the i-th and j-th energy storage units.

[0072] When the system operates in S2 mode, the droop factor R of each energy storage unit is... dpi This can be expressed by the formula:

[0073] ,

[0074] Then, when the performance of the i-th energy storage unit is better than the global performance target value, i.e., the performance deviation ΔD of that energy storage unit is... i When the value is greater than 0, the droop factor of the cell needs to be reduced to allow the cell to handle a larger charge and discharge current. In this case, the correction factor f is... dpi It should be less than 1. Conversely, when the performance of the i-th energy storage unit is worse than the global performance target value, i.e., the deviation ΔD of the energy storage unit's battery performance index, it is considered a deviation. i When it is less than 0, the correction factor f dpi It should be greater than 1. Therefore, the correction factor f for droop control... dpi Deviation ΔD from the performance index of the energy storage unit battery i The monotonicity of the function formed should be decreasing, that is...

[0075] .

[0076] Because in S2 working mode, f dpi Is with ΔD i Related quantities (also related to D) i (indirectly related), therefore, f dpi It can be written as f dpi (ΔD i ) or f dpi (D i ), R dpi It can also be written as R dpi (ΔD i ) or R dpi (D i ).

[0077] In some embodiments, f dpi It can be chosen as a linear function, that is:

[0078] ,

[0079] in, The coefficients are real coefficients greater than 0. The larger f is dpi The smaller the absolute value, the greater the adjustment strength based on the initial droop coefficient, but... The maximum value needs to be guaranteed within ΔD i Within the permitted operating range, there is always > 0.

[0080] In some embodiments, under the S2 operating mode, the relationship between the droop coefficient of the energy storage unit and the deviation of the battery performance index of the energy storage unit is expressed by the following formula:

[0081] ,

[0082] By f dpi Setting it to an exponential function accelerates the convergence of battery performance deviation values. Where R... 0i Let α represent the initial droop coefficient of the i-th energy storage unit, e represent the natural constant, and α and β be real coefficients greater than 0. α is mainly used to adjust the droop control correction coefficient f when ΔDi is large. dpi The value of ΔD is thus changed when the battery performance index deviation ΔDi is large. i The convergence speed, β is mainly used to adjust the droop control correction coefficient f when ΔDi is small. dpi The value of ΔD thus changes the deviation of the battery performance index ΔDi. When ΔD is small, the deviation is reduced. i The convergence rate. Specifically, when ΔD i When it is large, i.e., ΔD i >>β, f dpi The value of the expression is mainly determined by α, and is close to Therefore, α is mainly used to adjust ΔD i The correction strength of the droop coefficient when it is large; the larger α is, the stronger the correction for the initial droop coefficient R. 0i The stronger the shrinkage effect, the greater the shrinkage effect in ΔD. i When the value of α is larger, the increase in the charging and discharging current distribution value of the energy storage unit is greater. The upper limit of the value of α must ensure that the power of the energy storage unit will not exceed the power limit after the initial droop coefficient correction; ΔD i When it is small, i.e., ΔD i When << β, f dpi The value of the expression is close to 1, which means that the i-th energy storage unit is already close to the global target value, and there is no need to adjust the initial droop coefficient R. 0i Adjustments are made, therefore β affects ΔD i During the convergence to zero, when should the correction force for the droop coefficient be reduced until it no longer needs adjustment? The larger β is, the greater ΔD becomes.i As the coefficient approaches 0, the earlier the correction for droop is reduced, the less force is needed. Appropriately setting β can prevent ΔD from being affected. i Vibration around 0 can damage battery health. According to the above formula, if the remaining capacity and performance of the i-th energy storage unit are smaller, the droop coefficient R of that energy storage unit will be higher. dpi The greater the relative increase of the initial droop coefficient, the smaller the charging and discharging current allocated to the energy storage unit through droop control, thereby slowing down the performance degradation rate of the unit's battery. Therefore, under the proposed performance-balanced droop control method, the performance difference between the batteries of each energy storage unit will converge to 0 after long-term operation, so that the performance of each battery tends to be balanced.

[0083] It should be noted that the remaining battery capacity of a single energy storage unit is a physical quantity that changes very slowly. Except for the global battery performance target value and the switching signals for the two droop control operating modes, which are calculated by the central controller and sent to the local controllers of each energy storage unit in the DC microgrid energy storage system, all other control steps are implemented in the local controllers of each energy storage unit. The communication between the local controller and the central controller only requires uploading the locally evaluated battery performance information and receiving the global battery performance target value and the operating mode switching signal at a low frequency. There is no other additional information exchange and communication between the energy storage units, and the communication needs can be met by relying on the low-speed data bus.

[0084] In summary, the proposed performance droop control method for DC microgrid energy storage systems selects the ratio of the remaining capacity to the rated capacity of the energy storage unit as a performance evaluation index. Based on this, performance droop control is implemented to reduce the adverse effects of the gradual increase in performance imbalance among the batteries of each energy storage unit during long-term operation. This allows the energy storage unit with poor performance to bear a smaller current during charging and discharging to reduce its capacity degradation rate, while the energy storage unit with better performance bears a larger current during charging and discharging to accelerate its capacity degradation rate. This ensures that the performance of each energy storage unit in the DC microgrid energy storage system eventually tends to be balanced, mitigating the adverse effects of excessively rapid degradation rates of locally poor-performing energy storage units on the operation of the energy storage system, and effectively extending the overall service life of the DC microgrid energy storage system.

[0085] Figure 2 The diagram shown is a performance droop control structure diagram of a DC microgrid energy storage system in an embodiment of the present invention. Figure 2 Only the first one is shown and the Two energy storage units are used as examples. Taking the i-th energy storage unit as an example, by collecting relevant operating data of the battery in real time and using an ANN-based battery capacity assessment model, the remaining battery capacity of the energy storage unit can be determined. Among them, the battery capacity assessment model , and The neurons in the input layer are N1, N2, ..., N. k ..., N m Neurons representing hidden layers, This represents the output layer neuron. According to the... The remaining battery capacity of each energy storage unit and battery rated capacity Q rated It can calculate battery performance indicators (such as...) Figure 2 D i According to the first Battery performance index D i The global target value D of all battery performance indicators within the energy storage system avg The difference can be calculated for the first... Individual battery performance deviation ΔD i All battery performance deviations constitute the set {ΔD}. i |1 ≤ i ≤ N}, based on the maximum absolute value ΔD of the elements in this set. max Has the value reached (i.e., is greater than or equal to) the set threshold t1 or the maximum absolute value ΔD? max Whether the system reaches (i.e., is less than or equal to) the set threshold t2 determines whether the operating mode needs to switch from traditional SOC-based droop control S1 to battery performance-based droop control S2, or vice versa. When it is determined that the system needs to switch to or remain in mode S1, traditional SOC-based droop control is used, in which case the droop coefficient is related to the SOC value (e.g., ...). Figure 2 R in dpi (S oci When it is determined that the system needs to switch to or remain in S2 mode, droop control based on battery performance balancing is adopted. In this case, the droop coefficient is related to the battery performance evaluation index (e.g., ...). Figure 2 R in dpi (D i After determining the droop coefficient R... dpi Subsequently, the droop control obtains the output voltage reference value V of the DC / DC converter at the energy storage unit interface based on the IU droop characteristic curve or PU droop characteristic curve determined by the droop coefficient. refi Finally, the energy storage unit interface DC / DC converter implements voltage-current dual closed-loop control to generate the corresponding PWM switching signal d. i This allows for the regulation of the actual output current or power of each energy storage unit.

[0086] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for performance droop control in a DC microgrid energy storage system, characterized in that, include: Step 1: Build and train a battery capacity assessment model based on artificial neural networks; Step 2: Use the battery capacity assessment model to perform online real-time assessment of the remaining capacity of all energy storage units in the DC microgrid energy storage system, obtain the remaining battery capacity of each energy storage unit, and calculate the battery performance index and battery performance index deviation of each energy storage unit based on the remaining battery capacity of each energy storage unit. The battery performance index of each energy storage unit is calculated based on the remaining battery capacity and the rated battery capacity of the energy storage unit, and the battery performance index deviation of each energy storage unit is calculated based on the battery performance index of the energy storage unit and the comprehensive battery performance assessment index of all energy storage units. Step 3: If the current droop control operating mode is S1, and the absolute value of the battery performance index deviation of a certain energy storage unit is greater than or equal to the preset threshold t1, then switch the droop control operating mode of the DC microgrid energy storage system to S2; if the current droop control operating mode is S2, and the absolute value of the battery performance index deviation of all energy storage units is less than or equal to the preset threshold t2, then switch the droop control operating mode of the DC microgrid energy storage system to S1. Here, S1 represents droop control based on SOC equalization, and the droop coefficient of each energy storage unit is related to the SOC value of that energy storage unit; S2 represents droop control based on performance equalization, and the droop coefficient of each energy storage unit is related to the battery performance index of that energy storage unit. The initial operating mode of the DC microgrid energy storage system is S1, and the threshold t1 is not less than the threshold t2. Step 4: Based on the droop control operating mode determined in Step 3, update the correction coefficient of the droop control, and calculate the final droop coefficient corresponding to this operating mode based on the correction coefficient. The energy storage unit DC / DC converter adopts voltage-current dual closed-loop control at the bottom layer, and uses droop control as the front-end loop. Therefore, the droop control outputs the reference value of the output voltage of the energy storage unit DC / DC converter, which generates a PWM switching signal through the voltage-current dual closed-loop control at the bottom layer of the energy storage unit DC / DC converter to realize the current or power control of each energy storage unit DC / DC converter.

2. The method for performance droop control of a DC microgrid energy storage system according to claim 1, characterized in that, The input layer neurons of the battery capacity assessment model based on artificial neural networks correspond one-to-one with battery-related parameters, and the output layer neurons correspond to the remaining battery capacity. Among them, battery-related parameters include battery state of charge, battery output voltage, and battery output current.

3. The method for performance droop control of a DC microgrid energy storage system according to claim 1, characterized in that, Battery performance evaluation indicators are expressed by the following formula: ， Among them, D i In the DC microgrid energy storage system, the first Battery performance indicators for each energy storage unit , This indicates the total number of energy storage units in the energy storage system; Indicates the battery's rated capacity; This indicates the first value obtained using the battery capacity assessment model. The remaining battery capacity of each energy storage unit.

4. The method for performance droop control of a DC microgrid energy storage system according to claim 3, characterized in that, The deviation of the battery performance index for each energy storage unit is expressed by the formula: ， in, This indicates the battery performance indicators of all energy storage units in this energy storage system. , … … The comprehensive evaluation index is denoted as the target value of the battery performance index of this energy storage system; Indicates the first Battery performance indicators of individual energy storage units With respect to the target values ​​of battery performance indicators of this energy storage system The deviation between the target and the target values ​​is considered. A positive deviation indicates that the battery performance of the energy storage unit is better than the target value, while a negative deviation indicates that the battery performance of the energy storage unit is worse than the target value.

5. The method for performance droop control of a DC microgrid energy storage system according to claim 4, characterized in that, In S2 operating mode, the droop factor of each energy storage unit is expressed by the formula: ， Among them, R dpi R represents the droop factor ultimately used by the i-th energy storage unit. 0i f represents the initial droop factor of the i-th energy storage unit. dpi The initial droop coefficient is applied to the droop control, which is the correction coefficient for the droop control of the i-th energy storage unit based on the battery performance balance target. It is determined by the deviation of the performance index of the i-th energy storage unit battery itself. e represents the natural constant, and α and β are real coefficients greater than 0.

6. The method for performance droop control of a DC microgrid energy storage system according to claim 4, characterized in that, In S2 operating mode, the droop factor of each energy storage unit is expressed by the formula: ， in, Indicates the first The droop factor ultimately used by each energy storage unit Indicates the first The initial droop factor of each energy storage unit Acting on the initial droop coefficient, representing the first... The correction coefficient for droop control obtained by each energy storage unit based on the battery performance balance target. These are real coefficients greater than 0.

7. The method for performance droop control of a DC microgrid energy storage system according to claim 1, characterized in that, In S1 operating mode, the relationship between the droop factor of each energy storage unit and the SOC value of that energy storage unit is expressed by the following formula: ， Among them, R dpi R represents the droop factor ultimately used by the i-th energy storage unit. 0i f represents the initial droop factor of the i-th energy storage unit. dp (S oci The coefficient applied to the initial droop coefficient represents the correction coefficient for droop control obtained by the i-th energy storage unit based on the battery SOC equalization target. It is determined by the SOC value of the i-th energy storage unit's battery itself.

Images