Hybrid energy storage dc microgrid and power distribution method thereof

Abstract

The present application relates to a kind of hybrid energy storage DC microgrid and its power distribution method, to construct hybrid energy storage DC microgrid with photovoltaic power generation unit and hybrid energy storage module, photovoltaic unit is tracked maximum power point by disturbance observation method and generates power deviation, hybrid energy storage module is based on this power deviation, utilize the power dynamic allocation of multiple branch dynamic filtering and SOE adaptive model predictive control;Microgrid is connected to distribution network through inverter grid connection, inverter uses constant power control under synchronous rotating coordinate system, based on multiple branch dynamic filtering and energy state adaptive model predictive control mechanism carries out power dynamic allocation, realizes stable energy output.The present application significantly improves the precision and adaptability of power distribution, improves the energy balance ability and energy storage unit life under multiple working conditions, has high reliability and stability under dynamic disturbance and abnormal working condition, especially suitable for energy management and system stability control under new energy access scene.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power supply or distribution circuit devices or systems; energy storage systems, and particularly relates to a hybrid energy storage DC microgrid and a power distribution method thereof. BACKGROUND

[0002] With large-scale integration of intermittent renewable energy such as photovoltaic and wind power into the power grid, DC microgrid as a key component in the field of smart grid and green energy has attracted widespread attention due to its low loss, high energy efficiency and excellent renewable energy integration capability. However, the output power of renewable energy is prone to fluctuation, which can easily cause voltage disturbance of the system bus, and further affect the safe and stable operation of the microgrid. Therefore, energy storage technology has become a key link to ensure the stable operation of the microgrid.

[0003] Energy storage technology can be generally divided into energy storage and power storage according to functional characteristics. Energy storage such as lithium battery and lead-acid battery has a large energy density, but its response speed is slow, and it is suitable for long-period power fluctuation. Power storage such as super capacitor (SC), superconducting energy storage system (SMES) and flywheel energy storage has high power density and fast response capability, and can effectively respond to transient power fluctuation, but its storage capacity is relatively small, and it is not suitable for long-term charging and discharging. Among them, SC and flywheel energy storage also have defects such as high self-discharge and mechanical wear, which affect their service life and reliability. In contrast, SMES has millisecond-level response and extremely low energy loss, and is suitable for ideal energy storage unit for short-term high-frequency power regulation, so it has more technical advantages in dynamic power application scenarios. The hybrid energy storage system (HESS) composed of lithium battery and SMES is gradually attracting widespread attention as an emerging solution. HESS can realize coordinated management of power and energy at different time scales by combining the two types of energy storage technology, thereby improving the stability of the system.

[0004] In the power distribution of hybrid energy storage system, the existing technology mostly uses state of charge (SOC) as a parameter for control method, and calculates and distributes based on capacity. This type of distribution method is difficult to accurately reflect the remaining available energy under the nonlinear change of voltage, and is prone to problems of inaccurate energy management and insufficient utilization of energy storage. In contrast, state of energy (SOE) can directly represent the remaining energy of the energy storage unit, avoiding the estimation error caused by voltage change, and helping to improve the power distribution accuracy and capacity utilization efficiency. However, the existing technology mostly uses static strategies such as fixed ratio or single time constant low-pass filtering, generally ignores the dynamic change and predictive optimization of SOE, and is prone to power imbalance, energy bias accumulation and shortening of unit life, which makes it difficult to balance system stability and life management.

[0005] Therefore, there is an urgent need for a power distribution control method capable of multi-time scale decomposition of power fluctuations, dynamic perception of energy state of energy storage, adaptive adjustment of distribution weight, and global optimization between power balance and energy balance, to fully exert the synergistic advantages of hybrid energy storage system. SUMMARY

[0006] The present application solves the problems in the prior art and provides a hybrid energy storage DC microgrid and a power distribution method thereof, which realizes high power distribution accuracy, strong energy balance capability and high system operation reliability through multi-time scale power decomposition, SOE adaptive weight adjustment and hierarchical MPC optimization.

[0007] The technical scheme adopted by the present application is a hybrid energy storage DC microgrid power distribution method, which constructs a hybrid energy storage DC microgrid with a photovoltaic power generation unit and a hybrid energy storage module. The photovoltaic power generation unit tracks the maximum power point with the perturbation and observation method to generate a power deviation. The hybrid energy storage module performs power dynamic distribution based on the power deviation and the integration of multi-branch dynamic filtering and energy state adaptive model predictive control mechanism.

[0008] Preferably, the hybrid energy storage module includes a lithium battery energy storage module and a superconducting magnetic energy storage module.

[0009] Preferably, based on the perturbation and observation method, the actual output power of the photovoltaic power generation unit is sampled in real time, and the power deviation from the grid-connected instruction power is calculated.

[0010] The multi-branch dynamic filter is used to extract the power components of corresponding time scales respectively.

[0011] Based on the energy state of the superconducting magnetic energy storage module, the weights of each filtering channel are adaptively adjusted to obtain low-frequency power components and high-frequency power components, which are respectively used as the responses of the lithium battery energy storage module and the superconducting magnetic energy storage module, to realize the first power distribution based on frequency characteristics.

[0012] Preferably, based on the first power distribution, a bidirectional energy state compensation mechanism is introduced.

[0013] If the energy state of the lithium battery energy storage module or the superconducting magnetic energy storage module is lower than the set threshold, the power distribution proportion of the module is reduced, and part of the power is transferred to the other energy storage module; if the energy state of the lithium battery energy storage module or the superconducting magnetic energy storage module is in the normal range, the normal distribution proportion is maintained.

[0014] The compensated power reference value is output.

[0015] Preferably, the compensated power reference value is input into a model predictive control optimizer to establish a quadratic programming problem, construct an objective function containing double weight coefficients, and solve the optimal power command of the lithium battery energy storage module and the superconducting magnetic energy storage module, and control them respectively.

[0016] Preferably, the objective function is associated with the deviation between the command power and the actual power and the deviation between the expected value and the actual value of the state of energy.

[0017] Preferably, the lithium battery energy storage module calculates the current target according to the reference power value and the output voltage of itself, and realizes current closed-loop control by adopting current inner loop PI regulation.

[0018] Preferably, a voltage-based deadbeat tracking strategy is designed in cooperation with the superconducting magnetic energy storage module, and the duty cycle of the chopper is adjusted in real time by a PI controller.

[0019] A hybrid energy storage DC microgrid, the microgrid comprises a photovoltaic power generation unit and a hybrid energy storage module; the microgrid is connected to a power distribution network through an inverter, the inverter adopts constant power control in a synchronous rotating coordinate system to realize stable energy output; the microgrid adopts the hybrid energy storage DC microgrid power distribution method to realize power dynamic distribution based on multi-branch dynamic filtering and SOE adaptive model predictive control mechanism.

[0020] The application relates to a hybrid energy storage DC microgrid and a power distribution method thereof, wherein the hybrid energy storage DC microgrid is constructed by a photovoltaic power generation unit and a hybrid energy storage module, the photovoltaic unit tracks a maximum power point by a perturb and observe method and generates a power deviation, the hybrid energy storage module realizes power dynamic distribution by using multi-branch dynamic filtering and SOE adaptive model predictive control based on the power deviation; the microgrid adopting the hybrid energy storage DC microgrid power distribution method comprises a photovoltaic power generation unit and a hybrid energy storage module; the microgrid is connected to a power distribution network through an inverter, the inverter adopts constant power control in a synchronous rotating coordinate system, realizes power dynamic distribution based on multi-branch dynamic filtering and SOE adaptive model predictive control mechanism, and realizes stable energy output.

[0021] The application has the following beneficial effects:

[0022] (1) A hybrid energy storage power distribution method is proposed, which fuses multi-branch dynamic filtering and SOE adaptive weight adjustment, compared with a traditional single filtering strategy, can accurately decompose power fluctuations on three time scales of fast, medium and slow, and adjust the distribution proportion in real time according to the energy state of the superconducting magnetic energy storage unit, thereby significantly improving the accuracy and adaptability of power distribution.

[0023] (2) On the basis of one-time power distribution, a bidirectional SOE compensation mechanism is introduced, which can automatically realize power bidirectional transfer when the energy of the energy storage unit is insufficient, avoid long-term high-load operation of a single energy storage unit, effectively suppress energy bias accumulation, and improve the energy balance ability and energy storage unit life of the system under multiple working conditions;

[0024] (3) In the power optimization link, a hierarchical weighted MPC method is used, power tracking error and SOE prediction deviation are respectively given independent weights, short-term dynamic performance and medium and long-term energy management are considered, and global optimization control is realized combined with power boundary and SOE safety constraint to ensure that the system has high reliability and stability under dynamic disturbance and abnormal working conditions;

[0025] (4) The physical characteristics of SMES, such as fast response and low energy loss, are used to propose a power control strategy based on voltage zero-error tracking, which can complete the zero-error response within milliseconds when the bus voltage fluctuates, which is significantly better than the dynamic performance of the existing super capacitor scheme, shortens the voltage recovery time by more than 80%, and effectively supports the impact of new energy output fluctuation on the microgrid system;

[0026] (5) It is especially suitable for energy management and system stability control in the scene of new energy access. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 is the flow chart of the method of the present application;

[0028] Figure 2 is the system architecture of the hybrid energy storage DC microgrid in the present application;

[0029] Figure 3 is the structure block diagram of the power distribution strategy of the multi-branch dynamic filtering and SOE adaptive model predictive control in the present application;

[0030] Figure 4 is the power and bus voltage of each unit in the embodiment working condition 1 of the present application;

[0031] Figure 5 is the SOE of HESS in the embodiment working condition 1 of the present application;

[0032] Figure 6 is the power and bus voltage of each unit in the embodiment working condition 2 of the present application.

[0033] Figure 7 is the SOE of HESS in the embodiment working condition 2 of the present application. DETAILED DESCRIPTION

[0034] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0035] The present application relates to a hybrid energy storage DC microgrid power distribution method, in particular, based on multi-branch dynamic filtering and energy state SOE adaptive model predictive control.

[0036] In the process of specific implementation, as shown in Figure 1 The method comprises the following steps:

[0037] (1) Construct a hybrid energy storage DC microgrid with a photovoltaic power generation unit and a hybrid energy storage module;

[0038] (2) The photovoltaic power generation unit tracks the maximum power point with the perturbation and observation method to generate a power deviation;

[0039] (3) The hybrid energy storage module performs power dynamic distribution based on the power deviation, fusing multi-branch dynamic filtering and energy state (SOE) adaptive model predictive control (MPC) mechanism;

[0040] The method finally realizes efficient, dynamic and state-aware power distribution.

[0041] The specific steps are described below.

[0042] (1) Construct a hybrid energy storage DC microgrid with a photovoltaic power generation unit and a hybrid energy storage module;

[0043] As shown in Figure 2 The hybrid energy storage DC microgrid includes a photovoltaic power generation unit, a lithium battery energy storage module, a superconducting magnetic energy storage module (SMES), a DC / DC converter, a chopper and a grid-connected inverter, etc.

[0044] The lithium battery energy storage module and the superconducting magnetic energy storage module (SMES) are respectively coupled with the DC bus through independently configured bidirectional energy conversion channels to form a hybrid energy storage module (HESS) with multi-time scale response characteristics. The photovoltaic power generation unit realizes maximum power point tracking with the perturbation and observation algorithm. The HESS performs power dynamic distribution through multi-branch dynamic filtering and adaptive model predictive control (MPC) mechanism considering energy state (SOE) to support bus voltage stability. The microgrid is connected to the distribution network through the grid-connected inverter. The inverter realizes stable energy output through constant power control in the synchronous rotating coordinate system.

[0045] (2) Photovoltaic power generation unit tracks the maximum power point with perturbation and observation method, generates power deviation;

[0046] Based on the perturbation and observation method, the actual output power of the photovoltaic power generation unit is sampled in real time, and the power deviation of the grid-connected instruction power is calculated as the response reference of the HESS;

[0047] The power components corresponding to the time scale are extracted by multi-branch dynamic filtering, that is, the power deviation signal is sequentially input into the dynamic filter of the fast, medium and slow channels for frequency domain decoupling processing;

[0048] Based on the energy state of the superconducting magnetic energy storage module, the weights of each filter channel are adaptively adjusted to obtain the low-frequency power component and the high-frequency power component, that is, the weights of the filter channels are calculated according to the SOE value of the SMES to obtain the low-frequency power component and the high-frequency power component, as shown in Fig. 2; as the response of the lithium battery energy storage module and the superconducting magnetic energy storage module respectively, the frequency characteristic based power distribution is realized to meet the dynamic adaptation requirements of the energy type and power type energy storage units. Figure 3

[0049] Specifically, the low-frequency power component at this time is distributed to the lithium battery energy storage module, and the high-frequency component is distributed to the SMES, and the specific steps are as follows:

[0050] The difference between the photovoltaic output and the grid-connected power is the unbalanced power:

[0051] (1)

[0052] In the formula, P grid is the grid-connected power, P pv is the output power of the photovoltaic power generation unit, P HESS is the power response of the hybrid energy storage module, which is negative when charging and positive when discharging, P B is the output power of the lithium battery energy storage module, P M is the output power of the SMES;

[0053] To realize the frequency decomposition of the power fluctuation signal, a second-order Butterworth filter is used as the core unit in each channel of the multi-branch dynamic filter, which belongs to an infinite impulse response digital filter and has the characteristics of maximum flatness and no ripple in the passband amplitude frequency response, which can effectively suppress the components above the specified cutoff frequency while keeping the amplitude characteristic smooth, meeting the accuracy and stability requirements of hybrid energy storage power distribution. First, a second-order Butterworth analog low-pass prototype is constructed, and its normalized transfer function is:

[0054] (2)

[0055] Where s is the Laplace operator; damping coefficient​ To ensure the amplitude-frequency response flat in the passband, and filter pole is uniformly distributed in the left half of the unit circle, to ensure system stability;

[0056] The above analog filter is mapped to the digital domain by bilinear transformation, to obtain the transfer function form of the second-order digital filter:

[0057] (3)

[0058] Wherein, The coefficient calculated according to the cutoff frequency, z is the complex variable of Z transform, indicating the analysis position of signal spectrum on the Z plane, z −1 Corresponding to a sampling period delay of the digital filter.

[0059] The time domain difference equation of the digital filter is:

[0060] y [ n ] = b 0 x [ n ] + b 1 x [ n − 1 ] + b 2 x [ n − 2 ] − a 1 y [ n − 1 ] − a 2 y [ n − 2 ] (4)

[0061] Wherein, x [ n ] The input signal is, y [ n ] The filter output signal is.

[0062] In the multi-branch dynamic filtering structure of the application, different cutoff frequencies f c1 , f c2 , f c3 Are set respectively to construct dynamic filters corresponding to fast, medium and slow channels to obtain corresponding filtering results y1, y2, y3, and calculate the weight of the filtering channel according to the SOE value of SMES:

[0063] When the SMES energy is sufficient, the high-speed channel weight w1 is increased;

[0064] When the SMES energy is insufficient, the low-speed channel weight w3 is increased;

[0065] The remaining part is allocated to the medium-speed channel weight w2.

[0066] In the specific implementation of the present application, the time constant of the fast channel is in the range of [0.05s, 0.2s], the corresponding frequency is greater than 5Hz, mainly responding to sudden power disturbance, the time constant of the medium-speed channel is in the range of [0.5s, 2s], the corresponding frequency is 0.5~5Hz, responsible for medium-rate power fluctuation, and the time constant of the slow channel is ≥ 5s, the corresponding frequency is less than 0.5Hz, mainly used for long-term energy balance. The multi-branch dynamic filter adopted decomposes the power deviation signal into three channels of fast, medium and slow time scales, among which the fast channel extracts high-frequency, short-time dynamic disturbance, mainly used to suppress sudden power fluctuations, the medium-speed channel extracts medium-frequency components, as a transitional adjustment part, used to alleviate the dynamic difference between fast and slow power distribution, and the slow channel extracts low-frequency components, reflecting the long-term energy balance demand. After completion, the high-frequency power component (mainly provided by the fast channel) is preferentially allocated to the superconducting magnetic energy storage module (SMES), and the low-frequency power component (mainly provided by the slow channel) is preferentially allocated to the lithium battery energy storage module (LiB); the medium-speed channel power is dynamically allocated between the two types of energy storage units according to the SOE weight, realizing smooth transition. The above time constant range can be set according to the power fluctuation spectrum characteristics of the system and the response speed of the energy storage unit, to ensure the balance between power decoupling effect and energy storage energy utilization rate.

[0067] After weight normalization, low-frequency power component and high-frequency power component are obtained, the low-frequency power is allocated to the lithium battery, and the high-frequency power is allocated to the SMES, completing a power distribution, meeting,

[0068] (5)

[0069] (6)

[0070] In the formula, P Bref is the three-channel filtered power reference value of the lithium battery energy storage module, P Mref is the filtered active power reference value of the SMES.

[0071] Further, based on the first power distribution, a bidirectional energy state compensation mechanism is introduced, as shown in Figure 3 ;

[0072] If the energy state SOE of the lithium battery energy storage module or the superconducting magnetic energy storage module is lower than the set threshold, the power distribution ratio of the module is reduced, and part of the power is transferred to the other energy storage module; if the energy states of the lithium battery energy storage module or the superconducting magnetic energy storage module are in the normal range, the normal distribution ratio is maintained;

[0073] The compensation factor is calculated in the form of a segmented continuous function, which ensures smooth transition of the power distribution ratio and avoids system impact caused by sudden power change;

[0074] Considering the charge-discharge characteristics of each energy storage unit and the power demand, the SOE of the lithium battery energy storage module and the SMES needs to be dynamically maintained in a high range, ensuring that both types of energy storage units have a capacity margin for fast response to power commands. According to the SOE levels (SOE b , SOE m ) of the lithium battery and the SMES, the compensation factors c B , c SMES for the lithium battery and the SMES can be obtained respectively:

[0075] (7)

[0076] (8)

[0077] The bidirectional SOE compensation mechanism corrects the power reference value of the hybrid energy storage module (lithium battery energy storage module and superconducting magnetic energy storage module) to obtain a time-varying power reference value for subsequent MPC optimization. This signal serves as a target trajectory in the prediction time domain, ensuring the smoothness and energy balance of power distribution. In practical applications, the bidirectional SOE compensation mechanism enables mutual energy regulation between the lithium battery energy storage module and the superconducting magnetic energy storage module. When the SOE of the SMES is below the set threshold, the controller actively increases the discharge proportion of the lithium battery to transfer more low-frequency power components to the SMES for charging, gradually increasing the SOE of the SMES and approaching the target value. Conversely, when the SOE of the lithium battery is low, the discharge proportion of the lithium battery is reduced, and the power component borne by the SMES is increased, thereby preventing excessive discharge of the lithium battery. This achieves bidirectional energy compensation between the two types of energy storage units, avoids long-term high-load operation of a single energy storage unit, and improves the energy balance capability and energy storage life of the system.

[0078] When the SOE of both the lithium battery and the SMES is below the minimum safety threshold, the controller will automatically prohibit both from discharging further, stop the bidirectional compensation, and revert to the safe power output command, thereby preventing excessive discharge of the energy storage devices and ensuring the safety of the system. In this case, the power balance of the microgrid will be borne by the external grid or load regulation, ensuring the safe and stable operation of the overall system.

[0079] After the bidirectional SOE compensation mechanism is corrected, the power reference value P Bf of the lithium battery energy storage module and the power reference value P Mf of the SMES satisfy,

[0080]

[0081]

[0082] wherein, is the maximum charge-discharge power of the SMES.

[0083] (3) Hybrid energy storage module fuses multi-branch dynamic filtering and energy state (SOE) adaptive model predictive control (MPC) mechanism to perform power dynamic allocation based on the power deviation;

[0084] To balance the short-term dynamic performance and long-term energy balance of the system, the compensated power reference value is input into the model predictive control optimizer to establish a quadratic programming problem, construct a target function containing double weight coefficients, and solve the optimal power instructions of the lithium battery storage module and the superconducting magnetic storage module, and control them respectively;

[0085] A quadratic programming target function containing two parts of weights is established, which is related to the deviation between the instruction power and the actual power, and the deviation between the expected value and the actual value of the energy state, including:

[0086] 1) Power item weighted part: taking λ B and λ M as weights, the deviation between the actual power and the reference power of the lithium battery and SMES is punished;

[0087] 2) SOE deviation item weighted part: taking λ SOE_b and λ SOE_m as weights, the deviation between the predicted SOE and the target SOE is punished;

[0088] The target function is:

[0089] m i n [ Lambda B ( P B [ k ] − P B f ) 2 + Lambda M ( P M [ k ] − P M f ) 2 + ∑ N p k = 1 Phi Lambda S O E _ b ( S O E b _ p r e d [ k ] − S O E b _ t a r g e t ) 2 + ∑ N p k = 1 Phi Lambda S O E _ m ( S O E m _ p r e d [ k ] − S O E m _ t a r g e t ) 2 ] (9)

[0090] Where P B and P Bf are the output power and power reference value of the lithium battery in the prediction time domain (Np steps); P M and P Mf are the output power and power reference value of the SMES in the prediction time domain (Np steps);

[0091] Regarding the output power in the prediction time domain, SOE reflects the remaining energy in the rated energy of the energy storage unit. The SOE expression based on the power integration method is:

[0092] (10)

[0093] represents the initial energy state of the energy storage; v and i are the working voltage and output current of the energy storage; E N and P are the rated energy and output active power of the energy storage, respectively;

[0094] According to the power reference value, the SOE of the lithium battery and the SMES in the future Np steps is predicted respectively:

[0095] S O E b _ p r e d [ k ] = S O E b _ 0 − ∑ j = 1 k Phi Phi P B f [ j ] ⋅ T s E B _ c a p (11)

[0096] S O E m _ p r e d [ k ] = S O E m _ 0 − ∑ j = 1 k Phi Phi P M f [ j ] ⋅ T s E M _ c a p (12)

[0097] where SOE b_pred [k], SOE m_pred [k] represent the SOE of the lithium battery and SMES at the kth step of prediction, respectively; SOE b_0 , SOE m_0 represent the initial state of energy of the lithium battery and SMES, respectively; E B_cap , P Bf are the rated energy and the corrected active reference of the lithium battery, respectively; E M_cap , P Mf are the rated energy and the corrected active reference of the SMES, respectively; is the sampling period, and j is an integer between 1 and k;

[0098] SOE b_target , SOE m_target are the desired SOE of the lithium battery and SMES, respectively.

[0099] The constraints are:

[0100] (13)

[0101] (14)

[0102] (15)

[0103] where, is the maximum charge-discharge power of the lithium battery.

[0104] According to the power upper and lower limits and the low SOE protection constraints applied above, a quadratic programming (QP) solver is called to obtain the optimal power for the next Np steps:

[0105] (16)

[0106] where x is the optimization variable vector, containing the lithium battery and SMES output power values in the prediction horizon (Np steps); H is the quadratic term weight matrix, composed of the power tracking error weight and the SOE deviation weight, corresponding to the quadratic part of the objective function; f is the linear term coefficient vector, reflecting the influence of the power reference trajectory and the SOE deviation on the objective value.

[0107] The upper and lower limits of power constraints and SOE lower limit protection constraints are applied in the optimization process to prevent overcharging or over-discharging of the energy storage unit. When QP optimization is not feasible, the controller will automatically fall back to the reference power value after SOE compensation, ensuring safe operation of the system under abnormal conditions.

[0108] Finally, the optimized power instructions of the lithium battery and SMES are sent to the respective bidirectional converters to complete the execution of the current control period, wherein:

[0109] The lithium battery energy storage module calculates the current target according to the reference power value and its output voltage, and realizes current closed-loop control through current inner loop PI regulation; control switch tubes S1 and S2 are executed;

[0110] The voltage-based tracking strategy is designed in cooperation with the superconducting magnetic energy storage module, and the PI controller adjusts the duty cycle of the chopper in real time to control the switch tubes S3 and S4 to execute; so that the SMES compensates for high-frequency power disturbances while ensuring that the DC bus voltage has no overshoot and ultra-fast dynamic stability.

[0111] Specifically, the specific steps of applying control to the lithium battery energy storage module and SMES are as follows:

[0112] The PI controller of the lithium battery energy storage module obtains the reference current instruction i Bf by calculating the ratio of its power reference value P B to the output voltage u Bref , which is used to adjust the duty cycle of the converter,

[0113] (17)

[0114] According to the power conservation, the power P M of the SMES can be obtained

[0115] (18)

[0116] In the formula, P Mf is the filtered optimal power reference value of the SMES, P c is the average power of the output capacitor of the SMES branch (including the converter), u dc is the voltage of the DC bus, and i M is the output current of the SMES;

[0117] To obtain the average charge and discharge power of the capacitor and maintain the DC voltage at the rated value, a PI controller is introduced, and the average power of the SMES branch output capacitor is represented as:

[0118] (19)

[0119] In the formula, Wc W cref These represent the energy stored in the capacitor and its reference value, respectively. M For capacitance, u dcref This is the reference voltage for the DC bus.

[0120] Transmission power measurement value P c Compared with reference value P Mf The difference P M This indicates the power fluctuation component when the system experiences small disturbances, which needs to be compensated by SMES to reduce fluctuations and improve system stability. The PI controller of SMES outputs an average power control signal to adjust the chopper duty cycle by comparing the error between the actual energy and the target energy of its branch capacitor.

[0121] This invention also relates to a hybrid energy storage DC microgrid, the microgrid comprising a photovoltaic power generation unit and a hybrid energy storage module; the microgrid is connected to the distribution network via an inverter, the inverter employing constant power control in a synchronous rotating coordinate system to achieve stable energy output; the microgrid uses the aforementioned hybrid energy storage DC microgrid power allocation method, based on a multi-branch dynamic filtering and energy state adaptive model predictive control mechanism for dynamic power allocation.

[0122] To enable those skilled in the art to better understand the present invention, the numerical example analysis includes the following components:

[0123] according to Figure 2 A simulation model of a hybrid energy storage DC microgrid was built in Matlab / Simulink.

[0124] To verify the superiority of the proposed control strategy in power fluctuation frequency division and SOE collaborative management, the operating state of the microgrid under power surge conditions was simulated and analyzed, and compared with the traditional low-pass filter power allocation method. The designed operating condition 1 is as follows: simulation duration is [0s, 5s], and the light intensity increases from 800W / m² at 1s. 2 Rise to 1000W / m 2 It decreased to 600W / m in 2 seconds. 2 Within 3 seconds, the grid-connected power decreased from 20kW to 10kW. The power of each unit and the bus voltage are as follows: Figure 4 As shown, HESS's SOE is as follows Figure 5 As shown.

[0125] like Figure 4 As shown, the output power (P) of the photovoltaic unit in 0-2s pv ) equals grid-connected power (P) grid ), SMES absorbs electrical energy by charging together with the lithium battery, and P in 2-3 seconds pv Less than Pgrid The SMES and the lithium battery jointly discharge to maintain system power balance, and the HESS is again in a charging state at 3-5s. Both methods can realize power fluctuation frequency division and distribution. Taking the sudden change of light intensity at 1s as an example, under the power distribution control of the application, when the light intensity suddenly changes at 1s, the output power of the photovoltaic unit rises from 20kW to 25.6kW, the grid-connected power is maintained at 20kW, the lithium battery slowly responds to the low-frequency power component and reaches the maximum value at the steady state, the SMES quickly responds to the high-frequency power component and the response power gradually decreases to zero at the steady state, and the DC bus voltage is basically maintained at 700V. Compared with the low-pass filtering power distribution method, the power distribution method of the application can adaptively adjust the power distribution weight according to the SOE of each energy storage while suppressing power fluctuations, preferentially increase the output ratio of the lithium battery, and reduce the discharge power of the SMES. In addition, since the multi-branch dynamic filtering is adopted, it can be observed that the high-frequency response speed of the SMES in the method of the application is faster, so the HESS power response curves under the two methods present obvious differences.

[0126] The control strategy of the application utilizes multi-branch dynamic filtering to perform multi-frequency band decomposition on the power deviation: the low-frequency component is slowly responded by the lithium battery, and the lithium battery bears the main power output at the steady state; the high-frequency component is quickly suppressed by the SMES within milliseconds, and gradually falls back to zero after the disturbance subsides; the bus voltage fluctuation amplitude is less than ±0.2% of the rated value, and is stabilized at 700V. Compared with the traditional low-pass filtering method, the multi-branch dynamic filtering and SOE adaptive weight adjustment strategy of the application can accurately decompose power fluctuations on three types of time scales: fast, medium and slow, and combine the SOE dynamic adjustment of the SMES to realize accurate matching of different frequency band disturbances and energy storage characteristics, realize the reduction of SMES discharge depth and the extension of its service life without reducing the fluctuation suppression efficiency.

[0127] For the SOE cooperative control problem of the HESS, different initial conditions are set: the initial SOE of the lithium battery is higher than the SOE of the SMES (i.e. SOE b =60%, SOE m =36%). The SOE of the HESS is as follows: Figure 5As shown, when the SMES SOE is lower than the target value, the SOE of the SMES under low-pass filter control gradually decreases, while the control strategy of the application actively increases the discharge proportion of the lithium battery, transfers more low-frequency power components to the SMES for charging, and makes the SOE quickly rise and approach the set target value in the disturbance recovery process. This bidirectional SOE compensation mechanism not only effectively suppresses the accumulation of energy bias and prolongs the service life of the SMES, but also enables the system to maintain good energy balance ability under multiple working conditions. In the power optimization link, the application adopts a hierarchical weighted MPC method, assigns independent weights to the power tracking error and SOE prediction deviation respectively, and realizes global optimization by combining power boundaries and SOE safety constraints, so that short-term dynamic performance and medium and long-term energy management are taken into account. Compared with the low-pass filter method, the strategy has obvious advantages in power distribution stability and SOE collaborative regulation.

[0128] To verify the effectiveness of the SMES in improving the overall performance of the system, the SC is set to perform comparison under the same working condition (working condition 2) and power distribution strategy, and the simulation conditions are consistent with those of working condition 1. The power and bus voltage of each unit in the microgrid are as shown in Figure 6 It can be observed from Figure 6 that the maximum overshoot of the bus voltage under SC control is 25.4% in the initial stage, and the steady-state time is 0.95s; while the SMES under the strategy of the application realizes no overshoot and millisecond-level stability, and can complete the bus voltage tracking without error in only 0.15s. When the grid-connected power suddenly changes at 3s, the response speed of the high-frequency power component of the SMES is obviously better than that of the SC, and the suppression process is more stable.

[0129] The SOE of the HESS is as shown in Figure 7 It can be seen that the SOE change trends of the two kinds of energy storage are close, but the voltage fluctuation under the SC scheme is larger and the frequency division effect is worse. This result shows that the application not only utilizes the physical characteristics of the SMES such as fast response speed and low energy loss, but also proposes a power control strategy based on voltage tracking without error, which shortens the bus voltage recovery time by more than 80%, effectively improving the stability and dynamic response capability of the microgrid under the condition of renewable energy output fluctuation.

[0130] Those skilled in the art will appreciate that embodiments of the application can be provided as methods, systems, or computer program products. Therefore, the application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage, etc.) containing computer-usable program code.

[0131] The present application is described in reference to the flowchart illustrations and / or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams.

[0132] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams.

[0133] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams. Figure 1 one or more functions specified in the flowchart illustrations and / or block diagrams.

[0134] While preferred embodiments of the application have been described, modifications and variations can be apparent to those skilled in the art once aware of the general underlying concepts. Accordingly, the appended claims intend to embrace all such modifications and variations as fall within the scope of the application.

[0135] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims and their equivalents, the application can be practiced otherwise than as specifically described.

Claims

1. A hybrid energy storage DC microgrid power distribution method, characterized in that: A hybrid energy storage DC microgrid is constructed by a photovoltaic power generation unit and a hybrid energy storage module; Based on the perturb and observe method, the actual output power of the photovoltaic power generation unit is sampled in real time, and the power deviation of the actual output power from the grid-connected instruction power is calculated; Based on the power deviation, the hybrid energy storage module fuses a multi-branch dynamic filter and an energy state adaptive model predictive control mechanism to perform power dynamic allocation: The multi-branch dynamic filter is used to extract power components of corresponding time scales respectively; Based on the energy state of the superconducting magnetic energy storage module, the weights of each filter channel are adaptively adjusted to obtain low-frequency power components and high-frequency power components, which are used as the responses of the lithium battery energy storage module and the superconducting magnetic energy storage module respectively, to realize primary power allocation based on frequency characteristics; Based on the primary power allocation, a bidirectional energy state compensation mechanism is introduced; If the energy state of the lithium battery energy storage module or the superconducting magnetic energy storage module is lower than a set threshold, the power allocation proportion of the module is reduced, and part of the power is transferred to the other energy storage module; If the energy states of the lithium battery energy storage module and the superconducting magnetic energy storage module are both in a normal range, the normal allocation proportion is maintained; The compensated power reference value is outputted; The compensated power reference value is inputted into a model predictive control optimizer to establish a quadratic programming problem, construct an objective function containing double weight coefficients, solve the optimal power instructions of the lithium battery energy storage module and the superconducting magnetic energy storage module, and control the two modules respectively; The objective function is associated with the instruction power and the actual power deviation, and the expected value and the actual value deviation of the energy state, Including: Power term weighting part: with λ B , λ M as the weight, to punish the deviation of the actual power of lithium battery and SMES from the reference power; SOE bias term weighting portion: with λ SOE_b , λ SOE_m as weights, penalizing the deviation between the predicted SOE and the target SOE.

2. The hybrid energy storage DC microgrid power distribution method of claim 1, wherein: The hybrid energy storage module includes a lithium battery energy storage module and a superconducting magnetic energy storage module. 3.The hybrid energy storage DC microgrid power distribution method of claim 1, wherein: The lithium battery energy storage module calculates a current target according to a reference power value and its own output voltage, and realizes current closed-loop control by using current inner loop PI regulation. 4.The hybrid energy storage DC microgrid power distribution method of claim 1, wherein: A voltage-based deadbeat tracking strategy is designed in cooperation with the superconducting magnetic energy storage module, and the duty cycle of a chopper is adjusted in real time by a PI controller.

5. A hybrid energy storage DC microgrid, characterized in that: The microgrid includes a photovoltaic power generation unit and a hybrid energy storage module; the microgrid is connected to a power distribution network through an inverter connected in parallel to the grid, and the inverter uses constant power control in a synchronous rotating coordinate system to realize stable energy output; the microgrid uses the hybrid energy storage DC microgrid power distribution method of any one of claims 1-4 to perform power dynamic allocation based on a multi-branch dynamic filter and an energy state adaptive model predictive control mechanism.

Images