Island energy router direct current microgrid cooperative control method and system

Abstract

The application discloses a kind of island energy router direct current micro network's collaborative control method and system, belong to island operation direct current micro grid and energy router control technical field.SOC and output current of energy storage unit are collected;SOC deviation is obtained according to the deviation of SOC and average value;The deviation and output current are input into double-input double-output fuzzy controller, and droop coefficient and inertia adaptive gain are synchronously output;Droop coefficient is adjusted according to droop coefficient to control output power;Inertia adaptive gain is used as the gain coefficient of high-pass filter link, the high-frequency component of output current is extracted, and the virtual capacitance value is determined according to high-frequency component and gain coefficient to adjust output power;The power regulation amount of two branches is superimposed to generate current reference value, which is executed in closed loop after PWM modulation.The application realizes the collaborative control of energy storage SOC equalization and bus voltage inertia enhancement, suppresses the voltage deviation caused by source and load fluctuation, and prolongs the service life of energy storage.

Description

Technical Field

[0001] This invention relates to the field of isolated DC microgrids and energy router control technology, specifically to a collaborative control method and system for isolated energy router DC microgrids. Background Technology

[0002] Islanded energy routers, as core devices in independent microgrids, are used to integrate distributed energy sources such as photovoltaics and energy storage, and to provide stable power to local loads. In islanded operation mode, the system lacks the support of a large power grid, and its voltage stability relies entirely on local distributed power sources and energy storage units. To improve power supply reliability, islanded energy routers often employ a distributed multi-energy storage architecture.

[0003] However, existing technologies have the following main drawbacks: Insufficient DC bus inertia leads to large dynamic voltage deviations and poor stability. Traditional methods such as droop control, voltage closed-loop control, and fixed virtual capacitor control result in a fixed or insufficient equivalent inertia of the system, making it difficult to cope with random fluctuations in photovoltaic power, charging load impacts, and energy storage switching disturbances. This can cause instantaneous power imbalances within the system, leading to severe fluctuations in DC bus voltage and affecting power quality.

[0004] Large differences in the State of Charge (SOC) of multiple energy storage systems exacerbate dispersion and power distribution imbalance. Islanded energy routers often adopt a parallel architecture of multiple energy storage systems to improve reliability. However, the initial SOC of each energy storage system is inconsistent, the charging and discharging rates differ, and the response characteristics are different. Traditional proportional or fixed droop control is prone to long-term over-discharge of those with high SOC, long-term idleness of those with low SOC, overcharging and over-discharging of some energy storage systems, and long-term undercharging of some energy storage systems, resulting in energy storage life decay, decreased system efficiency, and poor operating economy.

[0005] Inertia control and SOC equalization control are independent, poorly coupled, and have complex control structures. Existing inertia enhancement control and SOC equalization control are mostly two independent control loops, making parameter coordination difficult, prone to mutual interference, complex in control structures, and challenging to implement in engineering. In particular, traditional virtual inertia control (such as simulating synchronous motor characteristics) is complex to design, difficult to tune parameters, and does not consider the limitations of the energy storage unit's SOC state on its inertia support capability. For example, strong inertia support at low SOC may exacerbate the risk of battery over-discharge; while charging equalization at excessively high SOC may lead to overcharging.

[0006] Therefore, there is an urgent need for a control method that can comprehensively consider the energy storage SOC state and the system dynamic power demand, and coordinate and optimize SOC balance and voltage inertia support. Summary of the Invention

[0007] The technical problem to be solved by this invention is how to solve the problems of large bus voltage fluctuation, power distribution imbalance and increased SOC dispersion caused by photovoltaic fluctuations, load impacts and differences in SOC of multiple energy storage systems in islanded multi-source DC systems.

[0008] This invention solves the above-mentioned technical problems through the following technical means: a collaborative control method for an islanded energy router DC microgrid, comprising: Collect the state of charge of each distributed energy storage unit and the output current of each energy storage unit; The average state of charge (SOC) of each energy storage unit is calculated based on its SOC, and the SOC deviation of each energy storage unit is obtained based on the difference between its SOC and the average SOC. For each energy storage unit, the state-of-charge deviation and output current of the energy storage unit are used as inputs to a dual-input dual-output fuzzy controller. The dual-input dual-output fuzzy controller outputs the droop coefficient and inertia adaptive gain of the energy storage unit according to a preset fuzzy rule base. The fuzzy rule base is constructed based on a fuzzy subset of the inputs, where the inputs are the state-of-charge deviation and output current. The fuzzy rule base stores the correspondence between the fuzzy subset of the inputs and the droop coefficient and inertia adaptive gain. Based on the droop coefficient and according to the preset droop control relationship, the basic component of the voltage reference for each energy storage unit is obtained. The inertia adaptive gain is used as the gain coefficient of the first-order high-pass filter. The output current of each energy storage unit is input into the first-order high-pass filter. The high-frequency component of the corresponding output current is output by the first-order high-pass filter. The virtual capacitance value of each energy storage unit is determined according to the high-frequency component and the gain coefficient. The inertia compensation voltage of each energy storage unit is obtained according to the virtual capacitance value. The voltage reference base component is superimposed with the inertia compensation voltage to obtain the voltage reference command for each energy storage unit. The voltage reference command generates a current reference value for each energy storage unit, and the current reference value is sent to the current loop of the DC / DC converter of each energy storage unit. Closed-loop control is achieved through PWM modulation.

[0009] Furthermore, in the input of the dual-input dual-output fuzzy controller, the normalized universe of discourse of the charge state deviation is [-1,1], and its fuzzy linguistic variables are set to five levels: VeryLow, Low, Medium, High, and VeryHigh. The membership functions corresponding to each level are triangular membership functions, and the peaks of the five levels of triangles are equally distributed in the universe of discourse [-1,1]. The normalized domain of discourse for the output current of the energy storage unit is [-1,1]. The fuzzy linguistic variables for the output current of the energy storage unit are set to two levels, N and P, where N corresponds to the charging state and P corresponds to the discharging state. The membership functions corresponding to each level are trapezoidal membership functions.

[0010] Furthermore, the fuzzy rules in the fuzzy rule base are executed when the output current is in the discharge state P: If the fuzzy linguistic variable of the charge state deviation is High or VeryHigh, then the output droop coefficient is located in the negative half of the normalized universe of discourse [-1,1], and the output inertia adaptive gain is located in the positive half of the normalized universe of discourse [-1,1]. If the fuzzy linguistic variable of the charge state deviation is Medium, then the output droop coefficient and inertia adaptive gain are both located at the peak of the Medium fuzzy subset in the normalized universe of discourse [-1,1]. If the fuzzy linguistic variable of the charge state deviation is Low or VeryLow, then the output droop coefficient is located in the positive half of the normalized universe of discourse [-1,1], and the output adaptive inertia gain is located in the negative half of the normalized universe of discourse [-1,1].

[0011] Furthermore, the fuzzy rules in the fuzzy rule base are executed when the output current is in charging state N: If the fuzzy linguistic variable of the charge state deviation is High or VeryHigh, then the output droop coefficient is located in the positive half region of the normalized universe of discourse [-1,1], and the output inertia adaptive gain is located in the positive half region of the normalized universe of discourse [-1,1]. If the fuzzy linguistic variable of the charge state deviation is Medium, then the output droop coefficient and inertia adaptive gain are both located at the peak of the Medium fuzzy subset in the normalized universe of discourse [-1,1]. If the fuzzy linguistic variable of the charge state deviation is Low or VeryLow, then the output droop coefficient is located in the negative half of the normalized universe of discourse [-1,1], and the output adaptive inertia gain is located in the negative half of the normalized universe of discourse [-1,1].

[0012] Furthermore, the fuzzy linguistic variables for the droop coefficient and the adaptive inertia gain are both set to five levels: VerySmall, Small, Medium, Large, and VeryLarge. The normalized universe of discourse for both the droop coefficient and the adaptive inertia gain is [-1, 1].

[0013] Furthermore, based on the aforementioned droop coefficient, the droop control equation is used to obtain the first... Modulation voltage reference value of each energy storage unit The droop control equation is expressed as:

[0014] in, This is the bus reference voltage. For the first The output current of each energy storage unit This is the droop coefficient output by the fuzzy controller.

[0015] Furthermore, a first-order high-pass filter is introduced into the droop control equation, and the improved droop control equation is as follows:

[0016] in, To determine the final commanded voltage value of the energy storage unit when introducing droop and high-pass filtering stages, This refers to the droop coefficient output by the fuzzy controller of the energy storage unit. This is the output current of the energy storage unit. Let the transfer function be a first-order high-pass filter. For inertia adaptive gain, The cutoff frequency, This is the Laplace complex frequency domain operator.

[0017] Furthermore, the formula for calculating the virtual capacitance is as follows:

[0018] in, This is the actual supporting capacitor for the DC bus. This is a virtual resistor for the battery port, etc.

[0019] This invention also provides a collaborative control system for an isolated energy router DC microgrid, comprising: The data acquisition module is used to collect the state of charge of each distributed energy storage unit and the output current of each energy storage unit; The state of charge deviation calculation module is used to calculate the average state of charge based on the state of charge of each energy storage unit, and to obtain the state of charge deviation of each energy storage unit based on the difference between the state of charge of each energy storage unit and the average state of charge. The fuzzy control module includes a dual-input dual-output fuzzy controller configured for each energy storage unit. For any energy storage unit, the state-of-charge deviation and output current of the energy storage unit are used as two input quantities of the corresponding dual-input dual-output fuzzy controller. The corresponding dual-input dual-output fuzzy controller outputs the droop coefficient and inertia adaptive gain of the energy storage unit according to a preset fuzzy rule library. The fuzzy rule library is constructed based on fuzzy subsets of the input quantities, where the input quantities are the state-of-charge deviation and output current. The fuzzy rule library stores the correspondence between the fuzzy subsets of the input quantities and the droop coefficient and the inertia adaptive gain. The droop control module is used to obtain the basic component of the voltage reference of each energy storage unit according to the droop coefficient and a preset droop control relationship. The inertia compensation module is used to use the inertia adaptive gain as the gain coefficient of the first-order high-pass filter, input the output current of each energy storage unit into the first-order high-pass filter, output the high-frequency component of the corresponding output current from the first-order high-pass filter, determine the virtual capacitance value of each energy storage unit based on the high-frequency component and the gain coefficient, and obtain the inertia compensation voltage of each energy storage unit based on the virtual capacitance value. The voltage reference command generation module is used to superimpose the basic component of the voltage reference with the inertia compensation voltage to obtain the voltage reference command for each energy storage unit. The closed-loop control module is used to generate current reference values ​​for each energy storage unit according to the voltage reference command, send each current reference value into the current loop of the DC / DC converter of each energy storage unit, and realize closed-loop control through PWM modulation.

[0020] The present invention also provides a processing device, including at least one processor and at least one memory communicatively connected to the processor, wherein: the memory stores program instructions executable by the processor, and the processor can execute the above-described method steps by calling the program instructions.

[0021] The advantages of this invention are: 1) This invention uses inertia adaptive control to dynamically generate virtual capacitors, which significantly improves the equivalent inertia, reduces dynamic deviation, and enhances anti-disturbance capability, effectively solving the problems of low inertia, large voltage fluctuations, and poor stability in isolated DC microgrids.

[0022] 2) This invention dynamically adjusts the droop coefficient based on fuzzy logic to accurately control the charging and discharging rates of each energy storage system, thereby achieving SOC balance among multiple energy storage systems and improving the economic efficiency and reliability of system operation.

[0023] 3) This invention uses the same dual-input dual-output fuzzy controller to synchronously output the droop coefficient and the inertia adaptive gain, so that the SOC equalization control and the inertia enhancement control can complete the decision within the same control cycle. There is no need to set up two independent control loops, which avoids the problems of difficult parameter tuning and mutual interference of control quantities in traditional step control.

[0024] 4) This invention does not require the establishment of an accurate mathematical model of an isolated energy router DC microgrid. It only requires conventional measurement signals. It is highly adaptable to complex operating conditions such as random fluctuations in photovoltaic power, load impacts, and energy storage aging. It has good robustness, simple parameter tuning, and stable and reliable actual operation.

[0025] 5) This invention achieves SOC balance of each energy storage unit while suppressing DC bus voltage fluctuations, thus taking into account system voltage stability, power distribution balance and energy storage operation safety. Attached Figure Description

[0026] Figure 1 This is an overall block diagram of the multi-source DC inertia enhancement and SOC equalization collaborative control of the isolated energy router in Embodiment 1 of the present invention; Figure 2 This is a membership function diagram of the energy storage SOC deviation ΔSOC in Embodiment 1 of the present invention; Figure 3 The system battery current I in Embodiment 1 of the present invention bati Membership function graph; Figure 4 The droop coefficient R in Embodiment 1 of the present invention V Membership function graph; Figure 5 This is a membership function diagram of the adaptive inertia gain K in Embodiment 1 of the present invention; Figure 6(a) is a comparison curve of the virtual resistance R change under the traditional control strategy of Embodiment 1 of the present invention; Figure 6(b) is a comparison curve of the virtual resistance R change under the control strategy of this method in Embodiment 1 of the present invention; Figure 7 This is a simulation verification curve comparing the SOC changes of multiple energy storage units in Embodiment 1 of the present invention. Figure 8 The waveform diagram is a comparison of DC bus voltage for simulation verification of Embodiment 1 of the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] Example 1 like Figure 1 As shown, a collaborative control method for an islanded energy router DC microgrid includes: S1. Collect the state of charge of each distributed energy storage unit and the output current of each energy storage unit.

[0029] Specifically, the state of charge (SOC) of each distributed energy storage unit and the output current of each energy storage unit are collected in real time. Simultaneously, the DC bus voltage is collected. It serves as a feedback reference for the voltage closed loop.

[0030] S2. Calculate the average state of charge (SOC) of each energy storage unit based on its SOC, and obtain the SOC deviation of each energy storage unit based on the difference between its SOC and the average SOC.

[0031] Specifically, based on the SOC value of each energy storage unit, the average SOC of all energy storage units is calculated. :

[0032] in, The total number of energy storage units. For the first The state of charge of each energy storage unit.

[0033] Calculate the first State of charge deviation of individual energy storage units :

[0034] S3. For each energy storage unit, the state-of-charge deviation and output current of the energy storage unit are used as inputs to a dual-input dual-output fuzzy controller. The dual-input dual-output fuzzy controller outputs the droop coefficient and inertia adaptive gain of the energy storage unit according to a preset fuzzy rule library. The fuzzy rule library is constructed based on a fuzzy subset of the inputs, where the inputs are the state-of-charge deviation and output current. The fuzzy rule library stores the correspondence between the fuzzy subset of the inputs and the droop coefficient and inertia adaptive gain.

[0035] Specifically, (1) the input variables of the dual-input dual-output fuzzy controller Energy storage SOC deviation The normalized universe of discourse is [-1, 1]. Fuzzy linguistic variables are set as {VeryLow, Low, Medium, High, VeryHigh}, where VeryLow indicates that the State of Computation (SOC) is much lower than the average; Low indicates that the SOC is slightly lower than the average; Medium indicates that the SOC is close to the average; High indicates that the SOC is slightly higher than the average; and VeryHigh indicates that the SOC is much higher than the average. To balance control accuracy and response speed, in this embodiment, each linguistic variable uses a triangular membership function to describe its mapping relationship with the values ​​in the universe of discourse.

[0036] The standard mathematical expression of the membership function of a triangle within the universe of discourse. The definition is as follows:

[0037] in, For input variables; , , These represent the left endpoint, peak, and right endpoint of the membership function of the triangle, respectively, and satisfy the constraints. .when When the membership degree reaches its maximum value of 1; when or At that time, the membership degree is 0. In this embodiment, to simplify the calculation process and improve the real-time performance of the controller, the three vertices of the membership function of each triangle are in the normalized universe of discourse [ The membership function is as follows: [1,1] are distributed at equal intervals. Figure 2 As shown in Table 1 below: Table 1 Fuzzy subset membership function parameter table Fuzzy subsets Vertex a Vertex b Vertex c VeryLow -1 -1 -0.5 Low -1 -0.5 0 Medium -0.5 0 0.5 High 0 0.5 1 Very High 0.5 1 1 System energy storage unit output current The domain is [-1, 1]. The fuzzy linguistic variables are set to two levels: {N, P}, where N corresponds to the energy storage unit being in a charging state. P corresponds to the energy storage unit being in a discharging state. In this embodiment, both the charging state N and the discharging state P are linear semi-trapezoidal membership functions. The charging state N is a descending semi-trapezoidal function, and its domain is […]. 1,1] within: when When the membership degree is 1; when When the membership degree linearly decreases from 1 to 0; when At that time, the membership degree is 0. The discharge state P is a raised half-trapezoidal shape, within the universe of discourse [-1, 1]: when When the membership degree is 0; when When the membership degree linearly increases from 0 to 1; when When the membership degree is 1, the membership function is as follows: Figure 3 As shown.

[0038] (2) Output variables of a dual-input dual-output fuzzy controller Energy storage droop coefficient The normalized domain is [-1,1]. Fuzzy linguistic variables are set as {VerySmall, Small, Medium, Large, VeryLarge}, where VerySmall represents a very small droop coefficient and maximum discharge output; VeryLarge represents a very large droop coefficient and minimum discharge output. Each level uses a triangular membership function, as shown in the figure. Figure 4 and The parameters of the fuzzy subset membership function are shown in Table 2.

[0039] Table 2 Fuzzy subset membership function parameter table Fuzzy subsets Vertex a Vertex b Vertex c VerySmall -1 -1 -0.5 Small -1 -0.5 0 Medium -0.5 0 0.5 Large 0 0.5 1 VeryLarge 0.5 1 1 Inertia Adaptive Gain The normalized domain is [-1,1]. Fuzzy linguistic variables are set as {VerySmall, Small, Medium, Large, VeryLarge}, where VerySmall represents... Extremely small, small virtual capacitance, weak inertia; VeryLarge indicates Extremely large, large virtual capacitance, strong inertia, all levels use triangular membership functions, such as... Figure 5 As shown, The parameters of the fuzzy subset membership function are shown in Table 3.

[0040] Table 3 Fuzzy subset membership function parameter table Fuzzy subsets Vertex a Vertex b Vertex c VerySmall -1 -1 -0.5 Small -1 -0.5 0 Medium -0.5 0 0.5 Large 0 0.5 1 VeryLarge 0.5 1 1 (3) Fuzzy rule base Fuzzy rules are based on With the output current of the energy storage unit Simultaneously output and Considering There are 5 fuzzy subsets. Given two fuzzy subsets (N, P), a fuzzy rule base of 10 rules is constructed, as shown in Table 4. The core control logic is as follows: Discharge mode ( ) SOC equilibrium: At higher levels, The smaller the value, the weaker the drooping effect and the greater the discharge output. The lower the value, The larger the value, the stronger the drooping effect and the smaller the discharge output, thus avoiding over-discharge of low-charge cells.

[0041] Increased inertia: At higher levels, The larger the value, the greater the virtual capacitance, the stronger the inertia, and the better it suppresses voltage drops caused by sudden load increases. The lower the value, The smaller the value, the weaker the inertia, thus avoiding the impact of large current on low-capacity cells.

[0042] Charging mode ( ) SOC equilibrium: At higher levels, The larger the value, the greater the equivalent internal resistance and the smaller the charging output, thus preventing overcharging; The lower the value, The smaller the value, the lower the equivalent internal resistance, the greater the charging output, and the faster the energy replenishment.

[0043] Increased inertia: At higher levels, The larger the value, the greater the virtual capacitance, which absorbs the surplus power of the bus and suppresses overvoltage; The lower the value, The smaller the value, the weaker the inertia, thus avoiding damage to the energy storage unit from overcharging.

[0044] Table 4 Typical Rules Table

[0045] Typical rules are as follows: Discharge mode ( ) (Very Low): Output (Extreme drooping limits discharge). (Very low inertia, protecting the battery).

[0046] (Low): Output (Large drooping area) (Low inertia).

[0047] (Chinese): Output (Hanging down) (Inertia).

[0048] (High): Output (Small drooping) (Large inertia).

[0049] (Extremely High): Output (Minimal droop, full power discharge). (Extremely large inertia, forcibly supporting the voltage).

[0050] Charging mode ( ) (Very Low): Output (Minimal drooping, rapid energy replenishment). (Very small inertia, reducing losses).

[0051] (Low): Output (Small drooping) (Low inertia).

[0052] (Chinese): Output (Hanging down) (Inertia).

[0053] (High): Output (Large drooping area) (Large inertia).

[0054] (Extremely High): Output (Extreme drooping limits charging) (Extremely large inertia, absorbs overvoltage).

[0055] 4) Defuzzing After matching the above rules using the Mamdani inference method and obtaining the fuzzy inference results, the centroid method is used to apply the results to the output variables. and Defuzzification is performed. Since the output membership function adopts a triangular design, the precise output value after defuzzification is determined by the value corresponding to the apex of the triangle. Ultimately, the fuzzy linguistic variable is transformed into a definite and precise control quantity for subsequent droop control and virtual capacitance calculation.

[0056] S4. Based on the droop coefficient and according to the preset droop control relationship, obtain the basic component of the voltage reference for each energy storage unit.

[0057] Specifically, based on the droop coefficient, the first droop is obtained through the droop control equation. Modulation voltage reference value of each energy storage unit The droop control equation is expressed as:

[0058] in, This is the bus reference voltage. For the first The output current of each energy storage unit This is the droop coefficient output by the fuzzy controller.

[0059] When the control logic indicates that the SOC is high, Smaller When the SOC is relatively high, the discharge output increases, and the SOC decreases rapidly; when the SOC is low, Larger When the SOC is relatively low, the discharge output decreases, and the SOC recovers slowly. During charging, when the SOC is high, When the SOC is relatively high, the energy storage charging output decreases to prevent overcharging; when the SOC is low, The smaller size allows for increased energy storage and charging output, enabling rapid energy replenishment.

[0060] Through the above control, the SOC deviation of multiple energy storage systems can be converged and consistent, effectively avoiding overcharging and over-discharging, and achieving balanced power distribution.

[0061] S5. The inertia adaptive gain is used as the gain coefficient of the first-order high-pass filter. The output current of each energy storage unit is input into the first-order high-pass filter. The high-frequency component of the corresponding output current is output by the first-order high-pass filter. The virtual capacitance value of each energy storage unit is determined according to the high-frequency component and the gain coefficient. The inertia compensation voltage of each energy storage unit is obtained according to the virtual capacitance value.

[0062] Specifically, to achieve adaptive inertia compensation, drawing on the characteristics of high-pass filters that pass high frequencies and block low frequencies, a first-order high-pass filter is introduced into the battery-side droop control equation. The improved battery-side droop control equation is as follows:

[0063] in, To determine the final commanded voltage value of the energy storage unit when introducing droop and high-pass filtering stages, This refers to the droop coefficient output by the fuzzy controller of the energy storage unit. This is the output current of the energy storage unit. Let the transfer function be a first-order high-pass filter. For inertia adaptive gain, The cutoff frequency (rad / s) This is the Laplace complex frequency domain operator.

[0064] When the system is in steady state, the battery output current The changes are slow, with extremely small high-frequency components. The high-pass filter output is approximately zero, and the control equation degenerates into traditional droop control, which does not affect steady-state power distribution. However, when the system experiences drastic fluctuations in source-load power or supercapacitor failure, [the situation changes]. Rapid changes occur, high-frequency components increase significantly, and the output of the high-pass filter increases, thereby dynamically adjusting the voltage reference value, effectively generating a virtual capacitor to replenish the system inertia.

[0065] Virtual Capacitor The calculation formula is:

[0066] in, This is the actual supporting capacitor for the DC bus. This is a virtual resistor for the battery port, etc.

[0067] When the control logic is in operation due to source-load fluctuations, the high-frequency current passes through a high-pass filter, increasing the virtual capacitance and equivalent inertia, rapidly absorbing or releasing power and suppressing voltage fluctuations. In steady state, the high-frequency component is zero, the virtual capacitance is zero, restoring the original capacitor characteristics and not affecting steady-state accuracy. (Inertia adaptive gain) The larger the virtual capacitance, the stronger the system inertia and the smaller the voltage fluctuation; the smaller the inertia adaptive gain, the smaller the virtual capacitance, the weaker the system inertia, and the better the energy storage protection.

[0068] In this invention, inertia adaptive gain The fuzzy controller adaptively outputs based on the real-time state of the energy storage SOC. When the SOC is high, the inertia adaptive gain is adjusted. When the SOC is large, the system inertia is strong, the energy storage discharges more frequently, and voltage fluctuations are suppressed; when the SOC is low, the inertia adaptive gain K is small, the system inertia is weak, the power surge is reduced, and the energy storage is protected.

[0069] S6. Superimpose the base component of the voltage reference with the inertia compensation voltage to obtain the voltage reference command for each energy storage unit.

[0070] S7. Generate current reference values ​​for each energy storage unit according to the voltage reference command, send each current reference value into the current loop of the DC / DC converter of each energy storage unit, and realize closed-loop control through PWM modulation.

[0071] Simulation Experiment An islanded DC microgrid simulation model was built in MATLAB / Simulink, with a DC bus rated voltage of 750V and three sets of lithium battery energy storage. A 20-30kW DC load step disturbance was set at 10s, and the entire simulation lasted 65s. The balancing method of this invention was compared with that of traditional functional regulation.

[0072] As shown in Figures 6(a) and 6(b), although the SOC deviation gradually decreases under the traditional functional regulation equalization control, the convergence speed is slow and equalization is not achieved within the simulation time. However, the SOC of the present invention converges rapidly within 65s, with a steady-state deviation of less than 2%, achieving rapid equalization of multiple energy storage SOCs.

[0073] like Figure 7 , 8 As shown, the inertial control strategy proposed in this method can be evenly distributed according to the SOC state of each energy storage unit. When the SOC is balanced, its inertia is evenly distributed among the energy storage units. Furthermore, traditional inertial control without inertial control results in larger voltage fluctuation peaks and longer recovery times. The voltage fluctuation of this invention is smaller than that of traditional control, with shorter recovery times, significantly enhanced DC inertia, and greatly improved voltage stability.

[0074] Simulation results show that the present invention can simultaneously achieve DC bus voltage stability, rapid balancing of multiple energy storage SOCs, and adaptive inertia support, with all performance indicators significantly superior to traditional control methods.

[0075] Example 2 Based on Embodiment 1, Embodiment 2 also provides a collaborative control system for an islanded energy router DC microgrid, comprising: The data acquisition module is used to collect the state of charge of each distributed energy storage unit and the output current of each energy storage unit.

[0076] The state of charge deviation calculation module is used to calculate the average state of charge based on the state of charge of each energy storage unit, and to obtain the state of charge deviation of each energy storage unit based on the difference between the state of charge of each energy storage unit and the average state of charge.

[0077] The fuzzy control module includes a dual-input dual-output fuzzy controller configured for each energy storage unit. For any energy storage unit, the state-of-charge deviation and output current of the energy storage unit are used as two input quantities of the corresponding dual-input dual-output fuzzy controller. The corresponding dual-input dual-output fuzzy controller outputs the droop coefficient and inertia adaptive gain of the energy storage unit according to a preset fuzzy rule library. The fuzzy rule library is constructed based on fuzzy subsets of the input quantities, where the input quantities are the state-of-charge deviation and output current. The fuzzy rule library stores the correspondence between the fuzzy subsets of the input quantities and the droop coefficient and the inertia adaptive gain. Specifically, in the input quantities of the dual-input dual-output fuzzy controller, the normalized universe of discourse for the charge state deviation is [-1,1]. Its fuzzy linguistic variables are set to five levels: VeryLow, Low, Medium, High, and VeryHigh. The membership functions corresponding to each level are triangular membership functions, and the peaks of the five triangles are equally distributed in the universe of discourse [-1,1].

[0078] The fuzzy rules in the fuzzy rule base are executed when the output current is in the discharge state P: If the fuzzy linguistic variable for the charge state deviation is High or VeryHigh, then the output droop coefficient is located in the negative half of the normalized universe of discourse [-1,1], and the output adaptive inertia gain is located in the positive half of the normalized universe of discourse [-1,1].

[0079] If the fuzzy linguistic variable for the charge state deviation is Medium, then the output droop coefficient and inertia adaptive gain are both located at the peak of the Medium fuzzy subset in the normalized domain [-1,1].

[0080] If the fuzzy linguistic variable of the charge state deviation is Low or VeryLow, then the output droop coefficient is located in the positive half of the normalized universe of discourse [-1,1], and the output adaptive inertia gain is located in the negative half of the normalized universe of discourse [-1,1].

[0081] The fuzzy rules in the fuzzy rule base are executed when the output current is in charging state N: If the fuzzy linguistic variable for the charge state deviation is High or VeryHigh, then the output droop coefficient is located in the positive half-region of the normalized universe of discourse [-1,1], and the output adaptive inertia gain is located in the positive half-region of the normalized universe of discourse [-1,1].

[0082] If the fuzzy linguistic variable for the charge state deviation is Medium, then the output droop coefficient and inertia adaptive gain are both located at the peak of the Medium fuzzy subset in the normalized domain [-1,1].

[0083] If the fuzzy linguistic variable for the charge state deviation is Low or VeryLow, then the output droop coefficient is located in the negative half of the normalized universe of discourse [-1,1], and the output adaptive inertia gain is located in the negative half of the normalized universe of discourse [-1,1].

[0084] The fuzzy linguistic variables for the droop coefficient and the adaptive inertia gain are both set to five levels: VerySmall, Small, Medium, Large, and VeryLarge. The normalized universe of discourse for both the droop coefficient and the adaptive inertia gain is [-1,1].

[0085] The droop control module is used to obtain the base component of the voltage reference for each energy storage unit according to the droop coefficient and a preset droop control relationship.

[0086] Specifically, based on the droop coefficient, the first droop is obtained through the droop control equation. Modulation voltage reference value of each energy storage unit The droop control equation is expressed as:

[0087] in, This is the bus reference voltage. For the first The output current of each energy storage unit This is the droop coefficient output by the fuzzy controller.

[0088] The inertia compensation module is used to use the inertia adaptive gain as the gain coefficient of the first-order high-pass filter, input the output current of each energy storage unit into the first-order high-pass filter, output the high-frequency component of the corresponding output current from the first-order high-pass filter, determine the virtual capacitance value of each energy storage unit based on the high-frequency component and the gain coefficient, and obtain the inertia compensation voltage of each energy storage unit based on the virtual capacitance value.

[0089] Specifically, a first-order high-pass filter is introduced into the droop control equation, and the improved droop control equation is as follows:

[0090] in, To determine the final commanded voltage value of the energy storage unit when introducing droop and high-pass filtering stages, This refers to the droop coefficient output by the fuzzy controller of the energy storage unit. This is the output current of the energy storage unit. Let the transfer function be a first-order high-pass filter. For inertia adaptive gain, The cutoff frequency (rad / s) This is the Laplace complex frequency domain operator.

[0091] The formula for calculating virtual capacitance is:

[0092] in, This is the actual supporting capacitor for the DC bus. This is a virtual resistor for the battery port, etc.

[0093] The voltage reference command generation module is used to superimpose the basic component of the voltage reference with the inertia compensation voltage to obtain the voltage reference command for each energy storage unit.

[0094] The closed-loop control module is used to generate current reference values ​​for each energy storage unit according to the voltage reference command, send each current reference value into the current loop of the DC / DC converter of each energy storage unit, and realize closed-loop control through PWM modulation.

[0095] Example 3 Based on Embodiment 1, Embodiment 3 of the present invention also provides a processing device, including at least one processor and at least one memory communicatively connected to the processor, wherein: the memory stores program instructions executable by the processor, and the processor can execute the method steps of Embodiment 1 by calling the program instructions.

[0096] 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 the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A collaborative control method for an islanded energy router DC microgrid, characterized in that, include: Collect the state of charge of each distributed energy storage unit and the output current of each energy storage unit; The average state of charge (SOC) of each energy storage unit is calculated based on its SOC, and the SOC deviation of each energy storage unit is obtained based on the difference between its SOC and the average SOC. For each energy storage unit, the state-of-charge deviation and output current of the energy storage unit are used as inputs to a dual-input dual-output fuzzy controller. The dual-input dual-output fuzzy controller outputs the droop coefficient and inertia adaptive gain of the energy storage unit according to a preset fuzzy rule base. The fuzzy rule base is constructed based on a fuzzy subset of the inputs, where the inputs are the state-of-charge deviation and output current. The fuzzy rule base stores the correspondence between the fuzzy subset of the inputs and the droop coefficient and inertia adaptive gain. Based on the droop coefficient and according to the preset droop control relationship, the basic component of the voltage reference for each energy storage unit is obtained. The inertia adaptive gain is used as the gain coefficient of the first-order high-pass filter. The output current of each energy storage unit is input into the first-order high-pass filter. The high-frequency component of the corresponding output current is output by the first-order high-pass filter. The virtual capacitance value of each energy storage unit is determined according to the high-frequency component and the gain coefficient. The inertia compensation voltage of each energy storage unit is obtained according to the virtual capacitance value. The voltage reference base component is superimposed with the inertia compensation voltage to obtain the voltage reference command for each energy storage unit. The voltage reference command generates a current reference value for each energy storage unit, and the current reference value is sent to the current loop of the DC / DC converter of each energy storage unit. Closed-loop control is achieved through PWM modulation.

2. The collaborative control method for an islanded energy router DC microgrid according to claim 1, characterized in that, In the input of the dual-input dual-output fuzzy controller, the normalized universe of discourse of the charge state deviation is [-1,1], and its fuzzy linguistic variables are set to five levels: VeryLow, Low, Medium, High, and VeryHigh. The membership functions corresponding to each level are triangular membership functions, and the peaks of the five triangles are equally distributed in the universe of discourse [-1,1]. The normalized domain of discourse for the output current of the energy storage unit is [-1,1]. The fuzzy linguistic variables for the output current of the energy storage unit are set to two levels, N and P, where N corresponds to the charging state and P corresponds to the discharging state. The membership functions corresponding to each level are trapezoidal membership functions.

3. The collaborative control method for an islanded energy router DC microgrid according to claim 2, characterized in that, The fuzzy rules in the fuzzy rule base are executed when the output current is in the discharge state P: If the fuzzy linguistic variable of the charge state deviation is High or VeryHigh, then the output droop coefficient is located in the negative half of the normalized universe of discourse [-1,1], and the output inertia adaptive gain is located in the positive half of the normalized universe of discourse [-1,1]. If the fuzzy linguistic variable of the charge state deviation is Medium, then the output droop coefficient and inertia adaptive gain are both located at the peak of the Medium fuzzy subset in the normalized universe of discourse [-1,1]. If the fuzzy linguistic variable of the charge state deviation is Low or VeryLow, then the output droop coefficient is located in the positive half of the normalized universe of discourse [-1,1], and the output adaptive inertia gain is located in the negative half of the normalized universe of discourse [-1,1].

4. The collaborative control method for an islanded energy router DC microgrid according to claim 2, characterized in that, The fuzzy rules in the fuzzy rule base are executed when the output current is in charging state N: If the fuzzy linguistic variable of the charge state deviation is High or VeryHigh, then the output droop coefficient is located in the positive half region of the normalized universe of discourse [-1,1], and the output inertia adaptive gain is located in the positive half region of the normalized universe of discourse [-1,1]. If the fuzzy linguistic variable of the charge state deviation is Medium, then the output droop coefficient and inertia adaptive gain are both located at the peak of the Medium fuzzy subset in the normalized universe of discourse [-1,1]. If the fuzzy linguistic variable of the charge state deviation is Low or VeryLow, then the output droop coefficient is located in the negative half of the normalized universe of discourse [-1,1], and the output adaptive inertia gain is located in the negative half of the normalized universe of discourse [-1,1].

5. The collaborative control method for an islanded energy router DC microgrid according to claim 1, characterized in that, The fuzzy linguistic variables for the droop coefficient and the adaptive inertia gain are both set to five levels: VerySmall, Small, Medium, Large, and VeryLarge. The normalized universe of discourse for both the droop coefficient and the adaptive inertia gain is [-1, 1].

6. The collaborative control method for an islanded energy router DC microgrid according to claim 1, characterized in that, Based on the aforementioned droop coefficient, the droop control equation is used to obtain the first... Modulation voltage reference value of each energy storage unit The droop control equation is expressed as: in, This is the bus reference voltage. For the first The output current of each energy storage unit This is the droop coefficient output by the fuzzy controller.

7. The collaborative control method for an islanded energy router DC microgrid according to claim 6, characterized in that, By introducing a first-order high-pass filter into the droop control equation, the improved droop control equation is as follows: in, To determine the final commanded voltage value of the energy storage unit when introducing droop and high-pass filtering stages, This refers to the droop coefficient output by the fuzzy controller of the energy storage unit. This is the output current of the energy storage unit. Let the transfer function be a first-order high-pass filter. For inertia adaptive gain, The cutoff frequency, This is the Laplace complex frequency domain operator.

8. The collaborative control method for an islanded energy router DC microgrid according to claim 1, characterized in that, The formula for calculating the virtual capacitance is: in, This is the actual supporting capacitor for the DC bus. This is a virtual resistor for the battery port, etc.

9. A collaborative control system for an isolated energy router DC microgrid, characterized in that, include: The data acquisition module is used to collect the state of charge of each distributed energy storage unit and the output current of each energy storage unit; The state of charge deviation calculation module is used to calculate the average state of charge based on the state of charge of each energy storage unit, and to obtain the state of charge deviation of each energy storage unit based on the difference between the state of charge of each energy storage unit and the average state of charge. The fuzzy control module includes a dual-input dual-output fuzzy controller configured for each energy storage unit. For any energy storage unit, the state-of-charge deviation and output current of the energy storage unit are used as two input quantities of the corresponding dual-input dual-output fuzzy controller. The corresponding dual-input dual-output fuzzy controller outputs the droop coefficient and inertia adaptive gain of the energy storage unit according to a preset fuzzy rule library. The fuzzy rule library is constructed based on fuzzy subsets of the input quantities, where the input quantities are the state-of-charge deviation and output current. The fuzzy rule library stores the correspondence between the fuzzy subsets of the input quantities and the droop coefficient and the inertia adaptive gain. The droop control module is used to obtain the basic component of the voltage reference of each energy storage unit according to the droop coefficient and a preset droop control relationship. The inertia compensation module is used to use the inertia adaptive gain as the gain coefficient of the first-order high-pass filter, input the output current of each energy storage unit into the first-order high-pass filter, output the high-frequency component of the corresponding output current from the first-order high-pass filter, determine the virtual capacitance value of each energy storage unit based on the high-frequency component and the gain coefficient, and obtain the inertia compensation voltage of each energy storage unit based on the virtual capacitance value. The voltage reference command generation module is used to superimpose the basic component of the voltage reference with the inertia compensation voltage to obtain the voltage reference command for each energy storage unit. The closed-loop control module is used to generate current reference values ​​for each energy storage unit according to the voltage reference command, send each current reference value into the current loop of the DC / DC converter of each energy storage unit, and realize closed-loop control through PWM modulation.

10. A processing apparatus, characterized in that, The method includes at least one processor and at least one memory communicatively connected to the processor, wherein the memory stores program instructions executable by the processor, and the processor can execute the method as described in any one of claims 1 to 8 by invoking the program instructions.

Images