A hybrid energy storage collaborative control method and system for a direct current microgrid

Abstract

The application relates to the technical field of direct-current micro-grid control, in particular to a hybrid energy storage collaborative control method and system for a direct-current micro-grid, which comprises the following steps: collecting global operation state information such as a direct-current bus voltage value, photovoltaic power generation unit output power, wind power generation unit output power, main load power demand, and residual power of a super capacitor and residual power of a battery in real time, generating a standardized operation state vector through normalization and feature extraction; inputting the vector into a hybrid energy storage distribution model based on an improved adaptive fuzzy control algorithm to calculate expected output power of the battery and the super capacitor; combining the residual power of the energy storage to generate a control instruction set containing power limitation and charging and discharging states; and executing the instruction to collaboratively control the power of the battery and the super capacitor and stabilize the direct-current bus voltage. Through data standardization and an optimized control algorithm, the method improves state representation accuracy and power distribution rationality and improves the collaborative control effect of the hybrid energy storage.

Description

Technical Field

[0001] This invention relates to the field of DC microgrid control technology, and in particular to a hybrid energy storage collaborative control method and system for DC microgrids. Background Technology

[0002] Conventional DC microgrid hybrid energy storage control methods directly collect raw parameters such as DC bus voltage, output power of photovoltaic and wind power generation, load power demand, and remaining power of batteries and supercapacitors. The output power of batteries and supercapacitors is determined through conventional fuzzy control or fixed power allocation logic, and power control commands are directly output to regulate the energy storage units.

[0003] The original operating parameters exhibit significant differences in physical dimensions and numerical ranges. Parameters that have not undergone standardization reduce the accuracy of system state characterization and fail to fully reflect the system's energy balance and real-time energy storage status. Conventional fuzzy control algorithms have limited adaptive adjustment capabilities, resulting in poor matching between power allocation results and real-time microgrid operating conditions. Control commands only include power output values, without incorporating power constraints and charging / discharging state limitations based on remaining energy storage capacity. This makes energy storage units prone to unreasonable charging / discharging conditions, leading to significant fluctuations in DC bus voltage and poor consistency in the coordinated operation of hybrid energy storage.

[0004] By standardizing data processing, the accuracy of system state quantification is improved, the adaptive capability of fuzzy control algorithm is optimized, and comprehensive control commands including power limitation and charging / discharging status are formulated to reduce DC bus voltage fluctuations and improve the stability of hybrid energy storage collaborative operation. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and propose a hybrid energy storage collaborative control method and system for DC microgrids.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a hybrid energy storage collaborative control method for DC microgrids, comprising:

[0007] Real-time acquisition of global operating status information of DC microgrid, including DC bus voltage value, output power of photovoltaic power generation unit, output power of wind power generation unit, main load power demand, and remaining power of supercapacitor and battery;

[0008] The global operating status information is normalized and feature extracted to generate a standardized operating status vector of the DC microgrid. The standardized operating status vector is used to characterize the current energy balance and energy storage status of the system.

[0009] The standardized operating state vector is input into a pre-constructed hybrid energy storage allocation model, which is based on an improved adaptive fuzzy control algorithm to calculate the expected output power of the battery and the expected output power of the supercapacitor.

[0010] Based on the expected output power of the battery and the expected output power of the supercapacitor, and combined with the remaining capacity of the supercapacitor and the remaining capacity of the battery, a set of hybrid energy storage control commands including power limiting and charge / discharge status is generated.

[0011] The hybrid energy storage control command set is executed to coordinate the power output control of batteries and supercapacitors in the DC microgrid, so as to maintain the DC bus voltage stable within a preset range.

[0012] As a further aspect of the present invention, the global operating state information is normalized and feature extracted to generate a standardized operating state vector for the DC microgrid, including:

[0013] Based on the rated value of the DC bus voltage, the total installed power of the photovoltaic power generation unit and the wind power generation unit, the peak value of the main load power demand, and the rated capacity of the battery and the supercapacitor, the DC bus voltage value, the output power of the photovoltaic power generation unit, the output power of the wind power generation unit, the main load power demand, the remaining power of the supercapacitor and the remaining power of the battery are converted into per-unit values ​​to generate a set of per-unit data.

[0014] Calculate the power imbalance at the current moment from the set of per-unit data. The power imbalance is the sum of the output power of the photovoltaic power generation unit and the output power of the wind power generation unit minus the main load power demand.

[0015] From the set of per-unit data, calculate the battery's charge deviation and the supercapacitor's charge deviation, whereby the charge deviation is the degree to which the current remaining charge of the energy storage device deviates from half of its rated capacity.

[0016] The DC bus voltage rate of change is obtained by performing a first-order differential calculation on the per-unit data of the DC bus voltage.

[0017] The power imbalance, the battery charge deviation, the supercapacitor charge deviation, and the DC bus voltage change rate are combined to form the standardized operating state vector.

[0018] As a further aspect of the present invention, the hybrid energy storage allocation model is based on an improved adaptive fuzzy control algorithm, comprising:

[0019] The improved adaptive fuzzy control algorithm includes a core fuzzy inference mechanism and an online fuzzy rule adjustment mechanism;

[0020] The core fuzzy inference mechanism takes the standardized operating state vector as input and outputs the preliminary battery power allocation factor and supercapacitor power allocation factor by querying the initial fuzzy rule base.

[0021] The fuzzy rule adjustment mechanism continuously monitors the actual fluctuation trajectory of the DC bus voltage and the actual charging and discharging response of the energy storage device, and calculates the steady-state error and overshoot of the bus voltage.

[0022] Based on the steady-state error and overshoot of the bus voltage, the conclusion part of the corresponding rule entry in the initial fuzzy rule base is dynamically modified to form an adaptively updated fuzzy rule base for use in the next control cycle.

[0023] When the hybrid energy storage allocation model is running, it comprehensively calls the calculation results of the core fuzzy inference mechanism and the fuzzy rule adjustment mechanism to finally generate the expected output power of the battery and the expected output power of the supercapacitor.

[0024] As a further aspect of the present invention, the core fuzzy inference mechanism takes the standardized operating state vector as input, and outputs preliminary battery power allocation factors and supercapacitor power allocation factors by querying the initial fuzzy rule base, including:

[0025] The power imbalance, DC bus voltage change rate, battery charge deviation, and supercapacitor charge deviation in the standardized operating state vector are converted into corresponding fuzzy sets and membership values ​​through their respective predefined fuzzy membership functions.

[0026] The multiple membership values ​​obtained after transformation are used to calculate the matching degree according to the rule preconditions stored in the initial fuzzy rule base, and all fuzzy rules with a matching degree greater than zero are activated.

[0027] For each activated fuzzy rule, its premise matching degree is weighted and calculated with the predefined reference values ​​of battery power allocation factor and supercapacitor power allocation factor in the rule conclusion.

[0028] The weighted calculation results of all activated fuzzy rules are then defuzzified for the battery power allocation factor and the supercapacitor power allocation factor to obtain the preliminary battery power allocation factor and the preliminary supercapacitor power allocation factor.

[0029] As a further aspect of the present invention, the fuzzy rule adjustment mechanism continuously monitors the actual fluctuation trajectory of the DC bus voltage and the actual charging and discharging response of the energy storage device, and calculates the steady-state error and overshoot of the bus voltage, including:

[0030] After each control cycle, record the complete voltage-time series of the DC bus voltage from the moment the command is issued to the moment it recovers to steady state.

[0031] From the voltage-time series, the final steady-state value of the DC bus voltage is extracted, and the absolute value of the difference between the final steady-state value and the rated value of the DC bus voltage is calculated as the steady-state error of the bus voltage in the current period.

[0032] From the voltage-time series, identify the maximum or minimum value of the DC bus voltage during the transition process, and calculate the absolute value of the difference between the maximum or minimum value and the rated value of the DC bus voltage as the overshoot amount of the current period;

[0033] The actual output power curves of the battery and supercapacitor in response to the set of hybrid energy storage control commands are recorded synchronously during the control cycle.

[0034] The calculated steady-state error of the bus voltage, the overshoot, and the actual output power curve are packaged into a performance evaluation data package for the current cycle.

[0035] As a further aspect of the present invention, based on the steady-state error and overshoot of the bus voltage, the conclusion portion of the corresponding rule entry in the initial fuzzy rule base is dynamically modified, including:

[0036] The steady-state error and overshoot of the bus voltage in the performance evaluation data package are compared with preset error thresholds and overshoot thresholds.

[0037] If the steady-state error of the bus voltage or the overshoot exceeds its corresponding threshold, it is determined that the control effect of the previous control cycle did not meet expectations, and the fuzzy rule adjustment process is triggered.

[0038] The fuzzy rule adjustment process traces back the specific fuzzy rule entries that were activated in the previous control cycle based on the actual output power curve.

[0039] The conclusion portion of the activated specific fuzzy rule entry, namely its corresponding battery power allocation factor reference value and supercapacitor power allocation factor reference value, is directionally corrected according to a predetermined step size. The correction direction is to reduce the steady-state error and overshoot of the bus voltage.

[0040] The corrected battery power allocation factor reference values ​​and supercapacitor power allocation factor reference values ​​are updated to the corresponding entries in the initial fuzzy rule base to complete the adaptive adjustment of the fuzzy rules.

[0041] As a further aspect of the present invention, based on the expected output power of the battery and the expected output power of the supercapacitor, and combining the remaining capacity of the supercapacitor and the remaining capacity of the battery, a hybrid energy storage control command set including power limiting and charge / discharge state is generated, including:

[0042] The expected output power of the battery and the expected output power of the supercapacitor are compared with the current maximum allowable charge and discharge power limit of the battery and the current maximum allowable charge and discharge power limit of the supercapacitor, respectively.

[0043] If the absolute value of the desired output power of the battery is greater than the current maximum allowable charge and discharge power limit of the battery, then the actual commanded output power of the battery is limited to its current maximum allowable charge and discharge power limit, and the sign remains unchanged.

[0044] If the absolute value of the expected output power of the supercapacitor is greater than the current maximum allowable charge and discharge power limit of the supercapacitor, then the actual commanded output power of the supercapacitor is limited to its current maximum allowable charge and discharge power limit, and the sign remains unchanged.

[0045] The remaining power of the battery and the remaining power of the supercapacitor are queried respectively to determine whether they have reached the preset protection threshold for full charge or low charge state.

[0046] If the battery's remaining charge reaches the full charge protection threshold and the battery's expected output power is negative, then the battery's actual command output power will be forcibly set to zero.

[0047] If the battery's remaining charge reaches the low-charge protection threshold and the battery's expected output power is positive, then the battery's actual command output power will be forcibly set to zero.

[0048] Perform the same logic checks and power forced zeroing operation on the supercapacitor;

[0049] By combining the actual command output power of the battery after power limit and power protection processing, and the actual command output power of the supercapacitor, a set of hybrid energy storage control commands containing specific power values ​​and charging / discharging direction indicators is generated.

[0050] As a further aspect of the present invention, executing the hybrid energy storage control command set to perform coordinated power output control on the batteries and supercapacitors in the DC microgrid includes:

[0051] The set of hybrid energy storage control commands is sent to the battery management system and the supercapacitor management system via a communication network.

[0052] The battery management system analyzes the target power value and charging / discharging direction of the battery according to the received instructions, and controls the battery's power converter to output or absorb power according to this target power value and direction.

[0053] The supercapacitor management system analyzes the target power value and charging / discharging direction of the supercapacitor according to the received instructions, and controls the supercapacitor's power converter to output or absorb power according to this target power value and direction.

[0054] The actual output power of the battery and supercapacitor is monitored in real time and compared with their respective target power values. The switching state or duty cycle of the power converter is dynamically adjusted through a closed-loop control algorithm so that the actual output power tracks the target power value.

[0055] As a further aspect of the present invention, the method further includes the steps of performing data quality verification and anomaly handling on the global operating status information:

[0056] After collecting the DC bus voltage, photovoltaic power generation unit output power, wind power generation unit output power, main load power demand, and the remaining power of the supercapacitor and the battery, each data point is checked to ensure it is within its physical reasonable range.

[0057] Data that exceeds the reasonable range is marked as abnormal data points and replaced with the corresponding data from the previous effective control cycle or the moving average of the data.

[0058] For data that changes drastically within the reasonable numerical range, calculate the rate of change of its adjacent sampling points. If the rate of change exceeds the safety threshold, perform low-pass filtering smoothing on the data.

[0059] The data that has undergone verification, replacement, and smoothing processes will be used as the valid global running status information.

[0060] As a further aspect of the present invention, the present invention also includes a hybrid energy storage collaborative control system for DC microgrids, the system including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein when the processor executes the computer program, it implements the steps of the hybrid energy storage collaborative control method for DC microgrids described above.

[0061] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0062] The global operational status information is normalized and feature extracted to standardize the numerical ranges and data dimensions of DC bus voltage, photovoltaic power output, wind power output, main load power, remaining supercapacitor charge, and remaining battery charge. This eliminates differences in the dimensions and numerical ranges of various physical parameters, forming a standardized operational status vector. This processing method reduces the impact of raw data fluctuations and redundant information on system status judgment, fully maps the system's energy balance relationship and the state of charge of energy storage units, improves the accuracy and consistency of the quantitative expression of operational status, and ensures the regularity and reliability of the input data for subsequent power allocation calculations.

[0063] An improved adaptive fuzzy control algorithm is used to construct a hybrid energy storage allocation model, enhancing the algorithm's adaptive adjustment capability to changes in microgrid operating conditions and accurately calculating the expected output power of batteries and supercapacitors. Combining the remaining capacity parameters of batteries and supercapacitors, a set of control commands integrating power constraints and charge / discharge states is generated, clarifying the charge / discharge power boundaries and operating modes of the energy storage units. This constrains the output range and operating state of the energy storage units, preventing them from operating under unreasonable charge / discharge conditions, reducing the disturbance impact of power surges on the DC bus, narrowing the DC bus voltage fluctuation range, and keeping the bus voltage within a preset range. The coordination between battery and supercapacitor outputs is strengthened, optimizing the power allocation and switching process of the hybrid energy storage units, improving the stability of the energy storage unit's operating state, and enhancing the overall energy dispatch and voltage control performance of the DC microgrid. Attached Figure Description

[0064] Figure 1 This is a state diagram of a hybrid energy storage collaborative control method for DC microgrids as described in this invention.

[0065] Figure 2 A flowchart illustrating the operation of a hybrid energy storage allocation model based on an improved adaptive fuzzy control algorithm;

[0066] Figure 3 A flowchart for monitoring and calculating steady-state error and overshoot in a fuzzy rule adjustment mechanism. Detailed Implementation

[0067] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0068] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0069] See Figure 1 This invention provides a hybrid energy storage collaborative control method for DC microgrids, the specific method including:

[0070] The system collects real-time global operating status information, including DC bus voltage, photovoltaic power generation unit output power, wind power generation unit output power, main load power demand, and remaining supercapacitor and battery charge. This global operating status information is then normalized and feature extracted to generate a standardized operating status vector representing the system's energy balance and energy storage status. This vector is input into a hybrid energy storage allocation model built based on an improved adaptive fuzzy control algorithm to calculate the expected output power of the battery and the supercapacitor. Based on these two expected power values ​​and the remaining charge of the supercapacitor and battery, a hybrid energy storage control command set containing power limits and charge / discharge status is generated. This command set is then executed to implement coordinated power output control of the battery and supercapacitor, maintaining the DC bus voltage within a preset range.

[0071] In one embodiment of the present invention, based on the rated DC bus voltage, total installed power of photovoltaic and wind power, peak power demand of main load, and rated capacity of batteries and supercapacitors, per-unit conversion is performed on the collected DC bus voltage, photovoltaic output power, wind power output power, main load power demand, remaining supercapacitor capacity, and remaining battery capacity to generate a set of per-unit data. The power imbalance at the current moment is calculated from this set of data; this amount is the sum of photovoltaic output power and wind power output power minus the main load power demand. Simultaneously, the battery capacity deviation and supercapacitor capacity deviation are calculated from the per-unit data; the capacity deviation is defined as the degree to which the current remaining capacity of the energy storage device deviates from half of its rated capacity. A first-order differential operation is performed on the per-unit value of the DC bus voltage to obtain the DC bus voltage change rate. The power imbalance, battery capacity deviation, supercapacitor capacity deviation, and DC bus voltage change rate are combined to form a standardized operating state vector.

[0072] In the specific implementation, the rated DC bus voltage was set at 400V, the total installed capacity of the photovoltaic and wind power generation units was 50kW, the peak power demand of the main load was 60kW, the rated capacity of the battery was 100kWh, and the rated capacity of the supercapacitor was 10kWh. The collected data showed a DC bus voltage of 395V, an output power of 18kW for the photovoltaic power generation unit, an output power of 12kW for the wind power generation unit, a main load power demand of 35kW, a remaining battery capacity of 52kWh, and a remaining supercapacitor capacity of 5.5kWh. Based on a DC bus voltage rating of 400V, the DC bus voltage of 395V is converted to a per-unit value of 0.9875; based on a total installed capacity of 50kW, the output power of the photovoltaic power generation unit of 18kW is converted to a per-unit value of 0.36, and the output power of the wind power generation unit of 12kW is converted to a per-unit value of 0.24; based on a peak main load power demand of 60kW, the main load power demand of 35kW is converted to a per-unit value of 0.5833; based on a battery rated capacity of 100kWh, the remaining battery capacity of 52kWh is converted to a per-unit value of 0.52; based on a supercapacitor rated capacity of 10kWh, the remaining supercapacitor capacity of 5.5kWh is converted to a per-unit value of 0.55, forming a set of per-unit data including voltage, power, and energy per-unit values.

[0073] In practice, the power imbalance at the current moment is calculated from a set of per-unit data. The power imbalance is expressed by the formula:

[0074]

[0075] in: Indicates the power imbalance. The per-unit value representing the output power of a photovoltaic power generation unit. This represents the per-unit value of the output power of a wind power generation unit. This represents the per-unit value of the main load power demand. Substituting the per-unit value data, the calculated power imbalance is 0.0167. The battery charge deviation is calculated from the per-unit value data; the battery charge deviation is defined as the absolute value of the difference between the battery's remaining per-unit charge and 0.5, and the result is 0.02. The supercapacitor charge deviation is defined as the absolute value of the difference between the supercapacitor's remaining per-unit charge and 0.5, and the result is 0.05. Taking the difference between adjacent sampling points of the per-unit value sequence of the DC bus voltage, the DC bus voltage change rate is obtained as -0.002 / s.

[0076] Optionally, the power imbalance can be calculated by directly using the actual physical quantity corresponding to the per-unit data, and then uniformly converted to per-unit values ​​based on the total installed power. Optionally, the calculation of the charge deviation can use a quadratic polynomial mapping method to enhance the expressive power of nonlinear characteristics. It can be understood that the element arrangement order of the standardized operating state vector can be adjusted to any fixed arrangement of power imbalance, DC bus voltage change rate, battery charge deviation, and supercapacitor charge deviation. It can be understood that the calculation of the DC bus voltage change rate can use the three-point central difference method to improve noise immunity. Finally, the power imbalance (0.0167), battery charge deviation (0.02), supercapacitor charge deviation (0.05), and DC bus voltage change rate (-0.002 / s) are sequentially combined to form a four-dimensional standardized operating state vector.

[0077] In one embodiment of the present invention, see [reference] Figure 2 The improved adaptive fuzzy control algorithm incorporates a core fuzzy inference mechanism and an online fuzzy rule adjustment mechanism. The core fuzzy inference mechanism takes a standardized operating state vector as input and converts the power imbalance, DC bus voltage change rate, battery charge deviation, and supercapacitor charge deviation within the vector into fuzzy sets and membership values ​​using their respective predefined fuzzy membership functions. Multiple membership values ​​are matched according to the rule preconditions of the initial fuzzy rule library, activating all fuzzy rules with a matching degree greater than zero. The matching degree of each activated rule is weighted with its conclusion's predefined battery power allocation factor reference value and supercapacitor power allocation factor reference value. Defuzzification is then performed on both weighted results to output preliminary battery power allocation factors and supercapacitor power allocation factors. The fuzzy rule adjustment mechanism continuously monitors the actual fluctuation trajectory of the DC bus voltage and the actual charging and discharging response of the energy storage equipment, calculates the steady-state error and overshoot of the bus voltage, and dynamically modifies the conclusion portion of the corresponding rule entries in the initial fuzzy rule library based on these two factors, forming an adaptively updated fuzzy rule library for use in the next control cycle. The hybrid energy storage allocation model integrates the results of the two mechanisms mentioned above to generate the expected output power of the battery and the expected output power of the supercapacitor.

[0078] In practical implementation, a four-dimensional standardized operating state vector is received. The vector elements are, in order, the power imbalance (0.0167), the DC bus voltage change rate (-0.002 / s), the battery charge deviation (0.02), and the supercapacitor charge deviation (0.05). The core fuzzy inference mechanism inputs the four input quantities into a predefined triangular membership function for fuzzification. The power imbalance (0.0167) is mapped to a "positive small" fuzzy set with a membership value of 0.85; and a "zero" fuzzy set with a membership value of 0.15. The DC bus voltage change rate (-0.002 / s) is mapped to a "negative small" fuzzy set with a membership value of 0.78; and a "zero" fuzzy set with a membership value of 0.22. The battery charge deviation (0.02) is mapped to a "low" fuzzy set with a membership value of 0.92; and a "medium" fuzzy set with a membership value of 0.08. The supercapacitor charge deviation of 0.05 is mapped to a "medium" fuzzy set with a membership value of 0.82; and a "high" fuzzy set with a membership value of 0.18.

[0079] In the specific implementation, the initial fuzzy rule base contains a rule: "If the power imbalance is small positive, the DC bus voltage change rate is small negative, the battery charge deviation is low, and the supercapacitor charge deviation is medium, then the battery power allocation factor reference value is 0.65, and the supercapacitor power allocation factor reference value is 0.75." The matching degree between the current input membership degree and the preconditions of this rule is calculated using the minimum operator, resulting in a matching degree min(0.85, 0.78, 0.92, 0.82) = 0.78. Another rule, "If the power imbalance is zero, the DC bus voltage change rate is zero, the battery charge deviation is medium, and the supercapacitor charge deviation is high, then the battery power allocation factor reference value is 0.45, and the supercapacitor power allocation factor reference value is 0.95," has a matching degree calculated as min(0.15, 0.22, 0.08, 0.18) = 0.08. Both rules have matching degrees greater than zero, therefore both are activated.

[0080] In the specific implementation, the two activated fuzzy rules are weighted. The matching degree of the first rule (0.78) is multiplied by the battery power allocation factor reference value (0.65) to get 0.507, and multiplied by the supercapacitor power allocation factor reference value (0.75) to get 0.585. The matching degree of the second rule (0.08) is multiplied by the battery power allocation factor reference value (0.45) to get 0.036, and multiplied by the supercapacitor power allocation factor reference value (0.95) to get 0.076. Defuzzification uses a weighted average method. The initial battery power allocation factor is equal to (0.507+0.036) / (0.78+0.08)=0.543 / 0.86≈0.631, and the initial supercapacitor power allocation factor is equal to (0.585+0.076) / (0.78+0.08)=0.661 / 0.86≈0.769. The preliminary battery power allocation factor and the preliminary supercapacitor power allocation factor are the outputs of the core fuzzy inference mechanism.

[0081] In some embodiments, before the end of the current control cycle, the fuzzy rule adjustment mechanism extracts the steady-state error of the bus voltage (0.008) and the overshoot (0.005) from historical voltage data. Both are lower than the error threshold (0.01) and the overshoot threshold (0.012), respectively, so no rule adjustment is triggered, and the initial fuzzy rule library is directly retained. In some embodiments, if the steady-state error is 0.018 and exceeds the threshold, the fuzzy rule entries activated in the previous cycle are traced back, the battery power allocation factor reference value is increased from 0.65 to 0.68 by 0.03, and the supercapacitor power allocation factor reference value is decreased from 0.75 to 0.725 by 0.025, and the update is written into the initial fuzzy rule library.

[0082] In one embodiment of the present invention, see [reference] Figure 3 After each control cycle, a complete voltage-time series of the DC bus voltage from command issuance to steady-state recovery is recorded. The final steady-state voltage value is extracted from the series, and the absolute value of its difference from the rated DC bus voltage is calculated as the steady-state error of the bus voltage for that cycle. The maximum or minimum value of the transient process in the series is identified, and the absolute value of its difference from the rated value is calculated as the overshoot for that cycle. The actual output power curves of the battery and supercapacitor in response to the control command are recorded synchronously during that cycle. The steady-state error, overshoot, and power curves are packaged into a performance evaluation data package. The steady-state error and overshoot of this data package are compared with preset error thresholds and overshoot thresholds. If either exceeds the limit, a fuzzy rule adjustment process is triggered, tracing back to the specific fuzzy rule entry activated in the previous cycle. The conclusion part of that entry, namely the battery power allocation factor reference value and the supercapacitor power allocation factor reference value, is corrected in the direction of reducing steady-state error and overshoot by a predetermined step size, and the corrected value is updated to the corresponding entry in the initial fuzzy rule library to complete the adjustment.

[0083] In the specific implementation, assuming the control command for the nth control cycle is issued at t=0s, and the sampling interval is 0.001s, the DC bus voltage-time series from t=0s to t=1.000s is recorded as shown in Table 1. Referring to Table 1, the DC bus voltage values ​​at key time points are displayed.

[0084] Table 1: DC bus voltage values ​​at key time points 0.000 398.00 0.200 402.20 0.500 401.80 1.000 399.90

[0085] The final steady-state value of 399.90V at t=1.000s is extracted from the DC bus voltage-time series. The rated DC bus voltage is 400V. The steady-state error of the bus voltage in the current period is... The maximum value during the transition process was identified at t=0.200s, with a value of 402.20V, representing the overshoot of the current cycle. Synchronously recording during the nth control cycle, the actual output power curve of the battery responding to the hybrid energy storage control command set shows a trend of first rising and then stabilizing, while the actual output power curve of the supercapacitor responding to the hybrid energy storage control command set shows a trend of rapid spike followed by decay. The steady-state error of the bus voltage (0.10V), the overshoot (2.20V), and the actual output power curves of the battery and supercapacitor are packaged into a performance evaluation data package for the nth control cycle.

[0086] In some embodiments, the steady-state error of the bus voltage in the performance evaluation data package, 0.10V, is compared with a preset error threshold of 0.09V; 0.10V is greater than 0.09V. The overshoot, 2.20V, is compared with a preset overshoot threshold of 2.19V; 2.20V is greater than 2.19V. Because both the steady-state error and overshoot of the bus voltage exceed their respective thresholds, it is determined that the control effect of the nth control cycle has not met expectations, triggering the fuzzy rule adjustment process. The fuzzy rule adjustment process, based on the morphological characteristics of the actual output power curves of the battery and the supercapacitor, traces back to the specific fuzzy rule entry activated in the nth control cycle. The prerequisite for this entry includes a fuzzy set where the power imbalance is in the positive range.

[0087] In practice, the conclusion of a specific activated fuzzy rule entry is directionally modified. The original reference value for the battery power allocation factor is 0.70, and the original reference value for the supercapacitor power allocation factor is 0.88. The predetermined step size is 0.015, and the modification direction is to reduce the steady-state error and overshoot of the bus voltage. Specifically, the reference value for the battery power allocation factor is increased by 0.015 to 0.715, and the reference value for the supercapacitor power allocation factor is decreased by 0.010 to 0.870. The modified reference values ​​for the battery power allocation factor (0.715) and the supercapacitor power allocation factor (0.870) are then updated to the corresponding entries in the initial fuzzy rule base, completing the adaptive adjustment of the fuzzy rules.

[0088] In one embodiment of the invention, the expected output power of the battery and the expected output power of the supercapacitor are compared with their respective current maximum allowable charge / discharge power limits. If the absolute value of the battery's expected output power is greater than its power limit, the actual commanded output power of the battery is limited to that limit while maintaining its sign; the power limit of the supercapacitor is handled similarly. The remaining battery charge is queried; if it reaches the full charge protection threshold and the expected power is negative, the actual commanded output power of the battery is set to zero; if it reaches the low charge protection threshold and the expected power is positive, the actual commanded output power of the battery is also set to zero. The same charge protection judgment and power zeroing operation are performed on the supercapacitor. The actual commanded output powers of the battery and supercapacitor after power limit limiting and charge protection processing are combined to generate a hybrid energy storage control command set containing specific power values ​​and charge / discharge direction indicators.

[0089] In practical implementation, the current maximum allowable charge / discharge power limit of the battery is 25kW, the current maximum allowable charge / discharge power limit of the supercapacitor is 80kW, the remaining capacity of the battery is 98kWh, the full-charge protection threshold of the battery is 95kWh, the depletion protection threshold of the battery is 15kWh, the remaining capacity of the supercapacitor is 9.8kWh, the full-charge protection threshold of the supercapacitor is 9.5kWh, and the depletion protection threshold of the supercapacitor is 1.5kWh. The expected output power of the battery calculated by the hybrid energy storage allocation model is -28kW (negative sign indicates charging), and the expected output power of the supercapacitor is 75kW (positive sign indicates discharging).

[0090] In practice, the battery's expected output power (28kW) is compared with its current maximum allowable charge / discharge power limit (25kW). Since 28kW is greater than 25kW, the battery's actual commanded output power is limited to -25kW, with the sign remaining unchanged. The supercapacitor's expected output power (75kW) is compared with its current maximum allowable charge / discharge power limit (80kW). Since 75kW is less than 80kW, the supercapacitor's actual commanded output power remains at 75kW. The battery's remaining capacity is 98kWh, reaching the full-charge protection threshold of 95kWh. Since the battery's expected output power is negative, it meets the forced zeroing condition, and the battery's actual commanded output power is forcibly set from -25kW to 0kW. The supercapacitor's remaining capacity is 9.8kWh, which is below the full-charge protection threshold of 9.5kWh and the depleted state protection threshold of 1.5kWh. Therefore, no zeroing operation is required, and the supercapacitor's actual commanded output power remains at 75kW. The actual command output power of the battery (0kW) and the actual command output power of the supercapacitor (75kW) after comprehensive processing are used to generate a hybrid energy storage control command set. The set contains specific power values ​​and charging / discharging direction indicators. For specific values, please refer to Table 2.

[0091] Table 2: Set of Control Commands for Hybrid Energy Storage Battery 0 stop Supercapacitor 75 Discharge

[0092] In some embodiments, the battery's expected output power is 30kW, its current maximum allowable charge / discharge power limit is 25kW, and its actual commanded output power is limited to 25kW. The battery's remaining capacity is 14kWh, its depletion protection threshold is 15kWh, and its expected output power is positive, meeting the forced zeroing condition; therefore, the battery's actual commanded output power is ultimately set to 0kW. In some embodiments, the supercapacitor's expected output power is -90kW, its current maximum allowable charge / discharge power limit is 80kW, and its actual commanded output power is limited to -80kW. The supercapacitor's remaining capacity is 1.4kWh, its depletion protection threshold is 1.5kWh, and its expected output power is negative (charging), failing to meet the positive discharge zeroing condition; therefore, zeroing operation is not triggered.

[0093] In one embodiment of the present invention, a set of hybrid energy storage control commands is sent to the battery management system and the supercapacitor management system via a communication network. The battery management system parses the commands to obtain the target power value and charging / discharging direction of the battery, and controls the battery power converter to output or absorb power according to the target value and direction. The supercapacitor management system parses the commands to obtain the target power value and charging / discharging direction of the supercapacitor, and controls the supercapacitor power converter to output or absorb power according to the target value and direction. The actual output power of the battery and supercapacitor is monitored in real time and compared with their respective target power values. The switching state or duty cycle of the power converter is dynamically adjusted through a closed-loop control algorithm to make the actual power track the target power. After collecting all global operating status information, each data item is checked to see if it is within the physically reasonable numerical range. Data that exceeds the limit is marked as abnormal and replaced with the corresponding data of the previous valid period or the moving average. For data within the reasonable range but with drastic changes, the rate of change of adjacent sampling points is calculated. If it exceeds the safety threshold, low-pass filtering is performed for smoothing. The data after verification, replacement, and smoothing is used as the valid global operating status information.

[0094] In practical implementation, the hybrid energy storage control command set generated by the upper-level controller includes a target battery power of 20kW (discharging) and a target supercapacitor power of -15kW (charging). This command set is sent via the CAN bus communication network. The battery management system receives the data frame with frame ID 0x601 and parses out the target battery power value of 20kW and the direction indicator "discharging"; the supercapacitor management system receives the data frame with frame ID 0x602 and parses out the target supercapacitor power value of -15kW and the direction indicator "charging".

[0095] In practical implementation, the battery management system controls the battery's bidirectional DC-DC power converter in buck mode, calculating the current loop reference based on the target power value of 20kW, and adjusting the duty cycle of the upper IGBT to ensure the battery's actual output power tracks the target value. The supercapacitor management system controls the supercapacitor's bidirectional DC-DC power converter in boost mode, calculating the inductor current reference based on the target power value of -15kW, and stabilizing the supercapacitor's actual absorbed power at 15kW through phase shift modulation angle adjustment. Real-time monitoring of the battery's actual output power and the supercapacitor's actual absorbed power is performed at 10ms intervals. An incremental PI algorithm is used to dynamically adjust the power converter's switching frequency, with the control quantity update formula as follows:

[0096]

[0097] in: This is the duty cycle adjustment amount for the current cycle. This represents the current power deviation. This represents the power deviation from the previous cycle. The proportional gain is 0.035. The integral gain is 0.028.

[0098] In some embodiments, the communication network uses the Ethernet UDP protocol, the IP address of the battery management system is 192.168.1.101, port 502, and the data packet payload includes a floating-point number of the battery target power and a 1-byte direction code. In some embodiments, the supercapacitor management system uses RS485 serial communication with a baud rate of 115200bps, and the data frame format includes a CRC-16 check field, requesting retransmission if parsing fails. Optionally, the power tracking closed-loop control can use model predictive control instead of the PI algorithm, using the discrete state equation of the power converter to predict the optimal duty cycle for the next cycle. Optionally, the actual power sampling channel can be configured as a dual-redundant ADC for synchronous acquisition, averaging to reduce the impact of noise.

[0099] In the specific implementation, the following data were collected: DC bus voltage 405V, photovoltaic power generation unit output power 28kW, wind power generation unit output power 0kW, main load power demand 42kW, remaining battery capacity 48kWh, and remaining supercapacitor capacity 6.2kWh. The physical validity of each data point was checked: the normal range for DC bus voltage is [380V, 420V], 405V is valid; the upper limit for photovoltaic output power is 50kW, 28kW is valid; the upper limit for wind power output power is 30kW, 0kW is valid; the upper limit for main load power demand is 60kW, 42kW is valid; the remaining battery capacity range is [0, 100]kWh, 48kWh is valid; the remaining supercapacitor capacity range is [0, 10]kWh, 6.2kWh is valid. No data exceeded these ranges, therefore no replacement was performed.

[0100] In the specific implementation, the photovoltaic power generation unit output power was detected to be 26kW in the previous sample and 28kW in the current sample, with a change rate of 2kW / 10ms = 200kW / s, exceeding the safety threshold of 150kW / s. A first-order low-pass filter with a cutoff frequency of 10Hz was applied to the photovoltaic output power sequence, resulting in a smoothed output value of 27.32kW. The remaining supercapacitor charge was sampled three times consecutively as 6.198kWh, 6.199kWh, and 6.201kWh, with a change rate below the threshold, requiring no further processing. The verified DC bus voltage of 405V, the smoothed photovoltaic output power of 27.32kW, the wind power output power of 0kW, the main load power demand of 42kW, the remaining battery charge of 48kWh, and the smoothed remaining supercapacitor charge of 6.21kWh were taken as valid global operating status information.

[0101] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A hybrid energy storage collaborative control method for a direct current microgrid, characterized in that, The method includes: Real-time acquisition of global operating status information of DC microgrid, including DC bus voltage value, output power of photovoltaic power generation unit, output power of wind power generation unit, main load power demand, and remaining power of supercapacitor and battery; The global operating status information is normalized and feature extracted to generate a standardized operating status vector of the DC microgrid. The standardized operating status vector is used to characterize the current energy balance and energy storage status of the system. The standardized operating state vector is input into a pre-constructed hybrid energy storage allocation model, which is based on an improved adaptive fuzzy control algorithm to calculate the expected output power of the battery and the expected output power of the supercapacitor. Based on the expected output power of the battery and the expected output power of the supercapacitor, and combined with the remaining capacity of the supercapacitor and the remaining capacity of the battery, a hybrid energy storage control command set including power limitation and charge / discharge state is generated. The hybrid energy storage control command set is executed to coordinate the power output control of batteries and supercapacitors in the DC microgrid, so as to maintain the DC bus voltage stable within a preset range. The hybrid energy storage allocation model is based on an improved adaptive fuzzy control algorithm, including: The improved adaptive fuzzy control algorithm includes a core fuzzy inference mechanism and an online fuzzy rule adjustment mechanism; The core fuzzy inference mechanism takes the standardized operating state vector as input and outputs the preliminary battery power allocation factor and supercapacitor power allocation factor by querying the initial fuzzy rule base. The fuzzy rule adjustment mechanism continuously monitors the actual fluctuation trajectory of the DC bus voltage and the actual charging and discharging response of the energy storage device, and calculates the steady-state error and overshoot of the bus voltage. Based on the steady-state error and overshoot of the bus voltage, the conclusion part of the corresponding rule entry in the initial fuzzy rule base is dynamically modified to form an adaptively updated fuzzy rule base for use in the next control cycle. When the hybrid energy storage allocation model is running, it comprehensively calls the calculation results of the core fuzzy inference mechanism and the fuzzy rule adjustment mechanism to finally generate the expected output power of the battery and the expected output power of the supercapacitor.

2. The hybrid energy storage collaborative control method for a DC microgrid of claim 1, wherein, The global operating state information is normalized and its features are extracted to generate a standardized operating state vector for the DC microgrid, including: Based on the rated value of the DC bus voltage, the total installed power of the photovoltaic power generation unit and the wind power generation unit, the peak value of the main load power demand, and the rated capacity of the battery and the supercapacitor, the DC bus voltage value, the output power of the photovoltaic power generation unit, the output power of the wind power generation unit, the main load power demand, the remaining power of the supercapacitor and the remaining power of the battery are converted into per-unit values ​​to generate a set of per-unit data. From the set of per-unit data, calculate the power imbalance at the current moment. The power imbalance is the sum of the output power of the photovoltaic power generation unit and the output power of the wind power generation unit minus the main load power demand. From the set of per-unit data, calculate the battery's charge deviation and the supercapacitor's charge deviation, whereby the charge deviation is the degree to which the current remaining charge of the energy storage device deviates from half of its rated capacity. The DC bus voltage rate of change is obtained by performing a first-order differential calculation on the per-unit data of the DC bus voltage. The power imbalance, the battery charge deviation, the supercapacitor charge deviation, and the DC bus voltage change rate are combined to form the standardized operating state vector.

3. The hybrid energy storage collaborative control method for a DC microgrid of claim 2, wherein, The core fuzzy inference mechanism takes the standardized operating state vector as input, queries the initial fuzzy rule base, and outputs preliminary battery power allocation factors and supercapacitor power allocation factors, including: The power imbalance, DC bus voltage change rate, battery charge deviation, and supercapacitor charge deviation in the standardized operating state vector are converted into corresponding fuzzy sets and membership values ​​through their respective predefined fuzzy membership functions. The multiple membership values ​​obtained after transformation are used to calculate the matching degree according to the rule preconditions stored in the initial fuzzy rule base, and all fuzzy rules with a matching degree greater than zero are activated. For each activated fuzzy rule, its premise matching degree is weighted and calculated with the predefined reference values ​​of battery power allocation factor and supercapacitor power allocation factor in the rule conclusion. The weighted calculation results of all activated fuzzy rules are then defuzzified for the battery power allocation factor and the supercapacitor power allocation factor to obtain the preliminary battery power allocation factor and the preliminary supercapacitor power allocation factor.

4. The hybrid energy storage collaborative control method for a DC microgrid of claim 3, wherein, The fuzzy rule adjustment mechanism continuously monitors the actual fluctuation trajectory of the DC bus voltage and the actual charging and discharging response of the energy storage device, and calculates the steady-state error and overshoot of the bus voltage, including: After each control cycle, record the complete voltage-time series of the DC bus voltage from the moment the command is issued to the moment it recovers to steady state. From the voltage-time series, the final steady-state value of the DC bus voltage is extracted, and the absolute value of the difference between the final steady-state value and the rated value of the DC bus voltage is calculated as the steady-state error of the bus voltage in the current period. From the voltage-time series, identify the maximum or minimum value of the DC bus voltage during the transition process, and calculate the absolute value of the difference between the maximum or minimum value and the rated value of the DC bus voltage as the overshoot amount of the current period; The actual output power curves of the battery and supercapacitor in response to the set of hybrid energy storage control commands are recorded synchronously during the control cycle. The calculated steady-state error of the bus voltage, the overshoot, and the actual output power curve are packaged into a performance evaluation data package for the current cycle.

5. The hybrid energy storage collaborative control method for DC microgrid of claim 4, wherein, Based on the steady-state error and overshoot of the bus voltage, the conclusion portion of the corresponding rule entry in the initial fuzzy rule base is dynamically modified, including: The steady-state error and overshoot of the bus voltage in the performance evaluation data package are compared with preset error thresholds and overshoot thresholds. If the steady-state error of the bus voltage or the overshoot exceeds its corresponding threshold, it is determined that the control effect of the previous control cycle did not meet expectations, and the fuzzy rule adjustment process is triggered. The fuzzy rule adjustment process traces back the specific fuzzy rule entries that were activated in the previous control cycle based on the actual output power curve. The conclusion portion of the activated specific fuzzy rule entry, namely its corresponding battery power allocation factor reference value and supercapacitor power allocation factor reference value, is directionally corrected according to a predetermined step size. The correction direction is to reduce the steady-state error and overshoot of the bus voltage. The corrected battery power allocation factor reference values ​​and supercapacitor power allocation factor reference values ​​are updated to the corresponding entries in the initial fuzzy rule base to complete the adaptive adjustment of the fuzzy rules.

6. The hybrid energy storage collaborative control method for a DC microgrid of claim 1, wherein, Based on the expected output power of the battery and the expected output power of the supercapacitor, and combining the remaining capacity of the supercapacitor and the remaining capacity of the battery, a hybrid energy storage control command set including power limiting and charge / discharge state is generated, including: The expected output power of the battery and the expected output power of the supercapacitor are compared with the current maximum allowable charge and discharge power limit of the battery and the current maximum allowable charge and discharge power limit of the supercapacitor, respectively. If the absolute value of the desired output power of the battery is greater than the current maximum allowable charge and discharge power limit of the battery, then the actual commanded output power of the battery is limited to its current maximum allowable charge and discharge power limit, and the sign remains unchanged. If the absolute value of the expected output power of the supercapacitor is greater than the current maximum allowable charge and discharge power limit of the supercapacitor, then the actual commanded output power of the supercapacitor is limited to its current maximum allowable charge and discharge power limit, and the sign remains unchanged. The remaining power of the battery and the remaining power of the supercapacitor are queried respectively to determine whether they have reached the preset protection threshold for full charge or low charge state. If the battery's remaining charge reaches the full charge protection threshold and the battery's expected output power is negative, then the battery's actual command output power will be forcibly set to zero. If the battery's remaining charge reaches the low-charge protection threshold and the battery's expected output power is positive, then the battery's actual command output power will be forcibly set to zero. Perform the same logic checks and power forced zeroing operation on the supercapacitor; By combining the actual command output power of the battery after power limit and power protection processing, and the actual command output power of the supercapacitor, a set of hybrid energy storage control commands containing specific power values ​​and charging / discharging direction indicators is generated.

7. The hybrid energy storage collaborative control method for DC microgrids according to claim 1, characterized in that, Executing the aforementioned set of hybrid energy storage control commands to perform coordinated power output control of batteries and supercapacitors in a DC microgrid includes: The set of hybrid energy storage control commands is sent to the battery management system and the supercapacitor management system via a communication network. The battery management system analyzes the target power value and charging / discharging direction of the battery according to the received instructions, and controls the battery's power converter to output or absorb power according to this target power value and direction. The supercapacitor management system analyzes the target power value and charging / discharging direction of the supercapacitor according to the received instructions, and controls the supercapacitor's power converter to output or absorb power according to this target power value and direction. The actual output power of the battery and supercapacitor is monitored in real time and compared with their respective target power values. The switching state or duty cycle of the power converter is dynamically adjusted through a closed-loop control algorithm so that the actual output power tracks the target power value.

8. The hybrid energy storage collaborative control method for DC microgrid of claim 1, wherein, The method further includes the steps of performing data quality verification and anomaly handling on the global operating status information: After collecting the DC bus voltage, photovoltaic power generation unit output power, wind power generation unit output power, main load power demand, and the remaining power of the supercapacitor and the battery, each data point is checked to ensure it is within its physical reasonable range. Data that exceeds the reasonable range is marked as abnormal data points and replaced with the corresponding data from the previous effective control cycle or the moving average of the data. For data that changes drastically within the reasonable numerical range, calculate the rate of change of its adjacent sampling points. If the rate of change exceeds the safety threshold, perform low-pass filtering smoothing on the data. The data that has undergone verification, replacement, and smoothing processes will be used as the valid global running status information.

9. A hybrid energy storage collaborative control system for DC microgrid, comprising a memory, a processor and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the hybrid energy storage collaborative control method for DC microgrids as described in any one of claims 1 to 8.

Images