DC power distribution network electric energy quality control method based on storage battery sliding mode control

By introducing battery and sliding mode controlled supercapacitors in the DC distribution network, combining the controllable load and the integrated control of the battery, the anti-interference ability and voltage deviation problems of the DC distribution network are solved, and the power quality and system stability are improved .

CN118970857APending Publication Date: 2024-11-15STATE GRID HUBEI ELECTRIC POWER RES INST +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410885610.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-03
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

The problems of poor anti-interference ability and voltage deviation of DC distribution networks affect the stable operation of loads and may cause impact on large power grids, and the existing technology is difficult to effectively solve.

Method used

Introducing battery energy storage devices and supercapacitors based on slip mode control in the DC distribution network, combining the controllable load and the battery by establishing an equivalent circuit model, a feedforward control system based on disturbance observer and a slip mode control strategy. Control, suppress DC voltage deviation and improve power quality.

Benefits of technology

It effectively improves the power quality of the DC distribution network, enhances anti-interference ability, reduces voltage deviation, extends the service life of the battery, and improves the stability and response speed of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure BDA0004925966820000021
    Figure BDA0004925966820000021
  • Figure BDA0004925966820000022
    Figure BDA0004925966820000022
  • Figure BDA0004925966820000031
    Figure BDA0004925966820000031
Patent Text Reader

Abstract

A method for controlling the electric energy quality of a direct-current power distribution network based on storage battery sliding mode control comprises the steps that an equivalent circuit model of a direct-current power distribution network structure is established, a storage battery energy storage control device and a controllable load model are established, a disturbance observer is introduced to track a system, and real-time adjustment is carried out by rapidly tracking feedback signals of disturbance current; a stabilization circuit; providing a supercapacitor treatment strategy based on sliding mode control, and constructing a supercapacitor series RC model; when DC voltage sag or short-time interruption occurs in the DC power distribution network, the super capacitor needs to be discharged, and when DC voltage sag occurs in the DC power distribution network, the super capacitor needs to be charged. According to the invention, the problems of direct-current voltage sag and sag, direct-current voltage short-time interruption and direct-current transient electric energy quality can be solved, the system response speed is improved, and the transient fluctuation of the direct-current voltage is reduced; the power quality is further improved, the stability of the system is improved, and effective power quality control is ensured.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of energy utilization, and in particular relates to a method for controlling power quality of a direct current distribution network based on battery sliding mode control. Background Art

[0002] With the massive access of photovoltaic, energy storage and electric vehicles, the power grid continues to develop and expand, its energy consumption continues to increase, the power demand of users and the requirements for power quality, economy, power supply reliability and environmental protection continue to increase, and the demand for "plug and play" of distributed power sources and DC loads on the user side is increasing day by day. The DC distribution network has a small damping and poor anti-interference ability, which makes the AC and DC power quality problems have certain differences, causing voltage deviations to threaten the load, especially the stable operation of sensitive loads, and even cause impacts on the large power grid and other unfavorable conditions.

[0003] With the continuous expansion of urban scale and the rapid growth of electricity load, a series of problems such as insufficient power supply capacity and corridor tension have emerged in urban AC distribution networks. At the same time, the load of distribution networks is variable, and the access of new energy sources makes the operation of distribution networks complicated. Whether DC distribution networks have higher energy efficiency levels than AC distribution networks under various working conditions needs further analysis. Therefore, it is very necessary to conduct a comprehensive and quantitative energy efficiency study on the power quality of AC and DC distribution systems. Summary of the invention

[0004] The purpose of the present invention is to solve the deficiencies in the prior art and provide a method for controlling the power quality of a DC distribution network based on battery sliding mode control, thereby solving the problems of anti-interference ability and voltage deviation of the DC distribution network and improving its power quality.

[0005] The present invention adds a battery energy storage device and a supercapacitor device based on sliding mode control to the DC distribution network to improve the power quality and economy of the DC distribution network, while suppressing DC voltage deviation, compensating for the limitation of controllable load power capacity, and increasing the service life of the battery.

[0006] The access end of the DC distribution network of the present invention includes four aspects: distributed power supply, AC / DC load, energy storage device and AC power grid. No matter which access end has non-ideal operating conditions such as switching or failure, it is possible to make the active power of the DC distribution network unbalanced, thereby affecting the DC bus voltage quality. According to the generation mechanism of DC power quality, it can also be divided into steady state and transient state. DC steady-state power quality includes DC voltage fluctuation and flicker, ripple, and DC voltage deviation.

[0007] The technical solution of the present invention is:

[0008] Step 1: Establish an equivalent circuit model of the DC distribution network structure: DC voltage fluctuations are mainly caused by changes in the power of distributed power sources. Due to the intermittent and random nature of distributed power sources, the output power of distributed power sources changes continuously, which in turn affects the stability of the DC voltage. The expression for the fluctuation coefficient of the DC voltage is:

[0009]

[0010] Among them, U1 and U2 are the peak and trough values ​​of the voltage in the cycle, ΔU is the fluctuation amplitude of the voltage, and U n is the rated voltage of the DC bus;

[0011] Step 2, establish a current feedforward control system based on a disturbance observer: establish a battery energy storage control device and a controllable load model; the battery energy storage control device includes the energy storage mathematical model established in step (2-1), the controllable load model established in step (2-2-1) and the battery control system model established in step (2-2-2), and combines them: in view of the voltage deviation problem, coupled with the frequent switching of loads, which leads to the continuous generation of active power imbalance in the DC distribution network, the feedforward control principle based on the disturbance observer is adopted, that is, the disturbance observer is introduced to track the system, and the feedback signal of the disturbance current is quickly tracked to perform real-time adjustment and stabilize the circuit. The battery energy storage part makes up for the insufficient controllable load capacity, and uses the controllable load to reduce the capacity of the energy storage device. As long as the voltage of a node exceeds the limit, it needs to be controlled to meet the power quality requirements of all nodes; when the deviation is small, the controllable load can completely suppress the deviation; when the deviation is large, the controllable load cannot completely suppress the voltage deviation due to capacity limitations. At this time, the battery energy storage device is used to eliminate the voltage deviation;

[0012] In order to evaluate the severity of DC voltage deviation, the deviation percentage D is defined to limit the DC voltage deviation to ±5%, and its expression is as follows:

[0013]

[0014] Step 3, design a supercapacitor management strategy based on sliding mode control and construct a supercapacitor series RC model; when the DC voltage sags or is briefly interrupted in the DC distribution network, the supercapacitor needs to be discharged, and at this time, the DC / DC converter works in Boost mode; when the DC voltage swells in the DC distribution network, the supercapacitor needs to be charged, and at this time, the DC / DC converter works in Buck mode; this control strategy is used to solve DC voltage sags, DC voltage swells, DC voltage short interruptions, and DC transient power quality problems.

[0015] In step 1 of the present invention, an equivalent circuit model of a DC power distribution network structure is established, and subsequent work will be performed based on the equivalent circuit.

[0016] Step 2 of the present invention comprises:

[0017] Step (2-1), establish the energy storage mathematical model;

[0018] Step (2-2), establish a controllable load model and a battery control system model and introduce a disturbance observer;

[0019] The battery energy storage control device includes the energy storage mathematical model established in step (2-1), the controllable load model established in step (2-2-1) and the battery control system model established in step (2-2-2), and combines them.

[0020] The step (2-1) of establishing the energy storage mathematical model in step 2 of the present invention includes:

[0021] (2-1-1) Establishing a mathematical model for energy storage, the decision variables are usually the selection of power generation technology, power generation capacity and transmission capacity within a given time and space range. The established mathematical model for energy storage is:

[0022]

[0023] Among them, x represents the 0-1 variable of the generator set, transmission line and energy storage device to be built; c represents the capacity variable of the generator set, transmission line and energy storage device to be built; u represents the 0-1 variable of the operating status of the thermal power unit; y represents the output variable of the generator set, transmission line and energy storage device; d, z, b, q represent the system coupling constraints and the boundary parameters of the input system and energy storage system constraints respectively; f(x, c, u, y) is the cost function of the system, h1, h2 represent the coupling constraints of the system, are the constraints of the input system and the energy storage system respectively;

[0024] (2-1-2) In the above-mentioned mathematical model of energy storage system, a concept of critical time is introduced. When the required energy supply time is greater than the critical time, the capacity and model configuration of energy storage can be selected based on the energy storage capacity; when the required energy supply time is less than the critical time, the capacity and model configuration of energy storage must be selected based on the peak power rather than the energy storage capacity; the number of charge / discharge times of the energy storage system within a period of event T is:

[0025]

[0026] Where P Char (t),P DisChar (t),P SelfDisChar(t) represent the charge / discharge power and self-discharge power of the energy storage unit respectively; Cap(t), Cap min (t), and Cap max (t) respectively represent the capacity and boundary value of the energy storage unit in the T period; LR Char (t),LR DisChar (t) represents the charge / discharge loss rate within T period;

[0027] In the control process, when the energy storage coefficient is low and the power is insufficient, the energy storage coefficient is small and decreases with the decrease of energy storage, and the charging coefficient is large; when the energy storage coefficient is high and the power is sufficient, the charging coefficient decreases with the increase of energy storage coefficient, and the discharge coefficient K d and charging factor K ch Establish a relationship:

[0028] Discharge coefficient:

[0029]

[0030] Charging factor:

[0031]

[0032] S min ,S low ,S high ,S max They are the minimum value, lower value, upper value and maximum value of the energy storage system respectively; is the maximum value of the virtual droop coefficient.

[0033] The controllable load model and the battery control system model established in step 2 (2-2) of the present invention include:

[0034] (2-2-1) Construct a controllable load model. Add a controllable load to the DC power grid. The power and terminal voltage can be adjusted within a certain range. By controlling the controllable load power, the DC voltage fluctuation is eliminated. The controllable load circuit topology is established.

[0035] During the regulation process, the controllable load can be controlled by changing the duty cycle of the switch tube; assuming that the controllable load is not fully engaged, when the voltage is lower than the lower limit, the terminal voltage is reduced; when the voltage is higher than the upper limit, the controllable load is fully engaged and participates in the regulation to stabilize the DC voltage;

[0036] When the loss of the controllable load circuit is neglected, the equivalent resistance of the controllable load can be obtained as:

[0037]

[0038] Where R is the load terminal resistance. The controllable load adopts feedforward double closed-loop control based on disturbance observer, which balances the power of DC distribution network by controlling the change of load power to suppress DC voltage deviation, but the power capacity of controllable load limits the control ability of DC voltage, so it needs to be controlled in coordination with the battery.

[0039] Battery energy storage is generally divided into two types: centralized energy storage and distributed energy storage. For the DC voltage deviation problem, if only battery energy storage is used, the required battery capacity is large, and the number of charge and discharge times will also increase, which is not conducive to the economy of the energy storage device. Due to its capacity limitation, the controllable load cannot completely suppress the DC voltage deviation. Therefore, the controllable load and the battery are comprehensively controlled to control the DC deviation, and the corresponding system is constructed based on the current feedforward control. The calculation formula for the filter inductance in the converter is obtained as follows:

[0040] (2-2-2) Establish a battery control system model. According to relevant principles and calculations, the calculation formula for the filter inductance in the converter is obtained as follows:

[0041]

[0042] Where f is the switching frequency, which is 10kHz, I 0min is the minimum average current, which is 5A; U0 is the node voltage, which is 200V; and U is the maximum voltage, which is 400V.

[0043] By adding corresponding feedforward control to the battery, according to control theory and calculation, the second-order damping ζ and natural frequency ω of the system can be calculated: n for:

[0044]

[0045] K PWM is the PWM equivalent gain of the battery, K p is the proportionality coefficient, K i is the integral coefficient. Through these two indicators, the main time domain indicators of the control system can be determined. For this design, the control system needs to have a certain bandwidth when designing parameters. Then the expression of bandwidth frequency is:

[0046]

[0047] Adding feedforward control requires adding additional current sensors to measure the disturbance current i0, which will increase the system cost and is not conducive to the expansion and plug-and-play of the power grid. Therefore, a disturbance observer is introduced to quickly track the disturbance i0, and the dynamic equation of the system is:

[0048]

[0049] In the disturbance observer, C0 is the DC bus capacitor, L0 is the filter inductor, and u sc ,u i are the battery and DC grid voltages, i and i0 are the battery current and disturbance current, respectively. After calculation, it can be written as:

[0050]

[0051] The disturbance observer model is designed as:

[0052]

[0053] Where m is the convergence coefficient. Then the final expression of the disturbance observer is:

[0054]

[0055] Thus, a battery energy storage control device is obtained;

[0056] (2-2-3) In order to solve the DC voltage deviation problem, a comprehensive control based on controllable load and battery is proposed.

[0057] The controllable load circuit topology model established in the controllable load model (2-2-1) of the present invention can adjust the terminal voltage and power within a certain range, thereby eliminating voltage fluctuations. In this model: U i , U c are input voltage and output voltage respectively; i2, i L are input current and output current respectively; i1 is DC bus current; U L is the load voltage; L is the filter inductor; R is the load equivalent resistance; when the controllable load does not participate in the regulation, the switch tube S1 is turned on and S2 is turned off. At this time, U c =U i ; When participating in regulation, the controllable load is controlled by changing the duty cycle D of the switch tube.

[0058] The present invention (2-2-3) aims at the DC voltage deviation problem and proposes a comprehensive control based on controllable load and battery:

[0059] Adopting feedforward control based on disturbance observer, battery energy storage makes up for the lack of controllable load capacity, using controllable load to reduce the capacity of energy storage equipment, and reduce its charge and discharge times, thus increasing service life and improving the power quality of each node in the DC distribution network;

[0060] Among them, as long as one of the node voltage exceeds the limit, it needs to be controlled to meet the power quality requirements of all nodes. When the deviation is small, the controllable load can completely suppress the deviation; when the deviation is large, the controllable load cannot completely suppress the voltage deviation due to capacity limitations, and the battery energy storage device will eliminate the voltage deviation. The voltage is higher than the upper limit, that is, max{U0,U1,U2}>1.05Ue, where U0,U1,U2 are the voltages of nodes 0, 1, and 2 respectively, and U e is the rated voltage of the DC distribution network. Given a reference value U a =1.05U e ; The voltage is lower than the lower limit, that is, max{U0,U1,U2}<0.95U e When U a =0.95U e .

[0061] In short, step 2 first adds a controllable load to the DC power grid, which can adjust the power and terminal voltage within a certain range, eliminates DC voltage fluctuations by controlling the controllable load power, and establishes a controllable load circuit topology. And by changing the duty cycle of the switch tube during the adjustment process, the controllable load can be controlled. Assuming that the controllable load is not fully put into use, the terminal voltage is reduced when the voltage is lower than the lower limit; when the voltage is higher than the upper limit, the controllable load is fully put into use and participates in the adjustment to stabilize the DC voltage and achieve the purpose.

[0062] On the basis of step 2, the present invention designs and adds corresponding supercapacitors in the energy storage unit to further improve the system.

[0063] The control and management strategy in step 3 of the present invention includes: in order to manage problems including DC voltage sag, DC voltage swell, and short-term interruption of DC voltage, and to further improve the quality of electric energy, it is required to improve the rapidity and accuracy of the control system. On the basis of step 2, a supercapacitor is added to form a hybrid energy storage system; the high-speed charging and discharging characteristics of the supercapacitor can effectively improve the speed and accuracy of the control; and a management strategy for a hybrid energy storage system of batteries and supercapacitors based on sliding mode control is proposed. In the hybrid energy storage system, the battery part and the supercapacitor part respectively adopt single-loop constant current control, and use a low-pass filter for power distribution, which can effectively suppress system power fluctuations and achieve bus voltage stability. And energy management is performed on the supercapacitor to ensure the stable operation of the system. This strategy mainly divides the working range of the supercapacitor into a discharge lower limit area, a discharge warning area, a normal working area, a charging warning area, and a charging upper limit area, so as to take appropriate charge and discharge control measures under different conditions to extend the system life and ensure the optimization of system performance.

[0064] The model formula of supercapacitor is:

[0065]

[0066] Where C is the capacitance of the supercapacitor, and the rated voltage U d , E is the storage energy of the supercapacitor, U sc is the voltage value of the supercapacitor, which must meet the maximum operating voltage U scmax The rated voltage is not exceeded, which is 400V. In the supercapacitor RC model, the resistor R1 is equivalent to the internal resistance of the supercapacitor, which is 0.04Ω. The supercapacitor RC model can accurately reflect the electrical characteristics of the supercapacitor during the charging and discharging process, which is convenient for analyzing the charging and discharging process and energy storage efficiency of the supercapacitor.

[0067] During the discharge process, the energy loss of the capacitor is:

[0068]

[0069] Where i is the capacitor current value, R1 is the resistance, that is, the internal resistance of the supercapacitor; the transmission efficiency of the supercapacitor is:

[0070]

[0071] Where: E is the total energy of the supercapacitor; Assume that the terminal voltage of the supercapacitor changes in the range of [u s1 ,u s2 ], the supercapacitor capacitance is C, take 5F. Maximum operating voltage U scmax Take it as 400V, the minimum working voltage U scmin is 125V. Then the power of the discharge process satisfies the following relationship:

[0072]

[0073] Where P max is the maximum required power, which is 100kW; the period T max Assumed to be 1s. Then the discharge energy loss can be expressed as:

[0074]

[0075] U sM is the intermediate voltage, which is defined as the average voltage maintained by the supercapacitor for a long time. It can be understood as the voltage when the supercapacitor is in standby mode. Therefore, the discharge efficiency of the supercapacitor is:

[0076]

[0077] Similarly, the charging power is:

[0078]

[0079] The charging energy loss can be expressed as:

[0080]

[0081] Therefore, the charging efficiency of the supercapacitor is:

[0082]

[0083] According to the characteristics of sliding mode control, the system is set as follows:

[0084]

[0085] Where: f(x,t), g(x,t) are the smooth vector fields of the control system, is the state variable vector, u is the control input vector, which can be expressed as:

[0086] a=a eq +Δa(Equation 26)

[0087] a eq For equivalent control, the system state variable vector can be made to perform sliding mode motion on the sliding surface. Δa is the switch control variable, which is used to make the system state variable reach the sliding surface within a finite time at any position in the state space.

[0088] When the DC voltage of the DC distribution network drops or is interrupted for a short time, the supercapacitor needs to be discharged. At this time, the DC / DC converter works in the Boost mode. The mathematical model of the Boost mode can be expressed as:

[0089]

[0090] In the circuit topology diagram of the supercapacitor, U sc , I represent the battery terminal voltage and current in the supercapacitor topology respectively; U dc , I0 respectively represent the voltage and current on the DC grid side in the circuit topology; L is the filter inductor, C1 is the DC bus capacitor, and C is the supercapacitor capacitor. D1 is the input duty cycle of the charging mode control, which can be written as follows in combination with formula (26):

[0091] D1=D 1eq +ΔD1 (Equation 28)

[0092] D 1eq is the equivalent control of D1, and ΔD1 is the switch control variable. When the DC voltage of the DC distribution network rises temporarily, the supercapacitor needs to be charged. At this time, the DC / DC converter works in Buck mode. The mathematical model of Buck mode can be expressed as:

[0093]

[0094] In the formula, D2 is the control input duty cycle, which can be written as follows according to formula (26):

[0095] D2=D 2eq +ΔD2 (Equation 30)

[0096] After the power quality event ends, the initial voltage of the supercapacitor is restored. If the remaining power of the supercapacitor is not enough to solve the power quality problem in the next extreme situation, it is necessary to restore the initial state, as shown in the following formula:

[0097]

[0098] The lower limit of capacitor voltage is taken as U sL =220V, upper limit value U sH =380V.

[0099] If the supercapacitor voltage is lower than the lower limit, it needs to be charged immediately, but the power quality of the DC distribution network must not be affected during the charging process. Therefore, when the DC voltage is lower than the lower limit, the supercapacitor cannot be charged; if the DC voltage is higher than the lower limit, the supercapacitor is charged when the load is small or the electricity price is low.

[0100] If the supercapacitor voltage is higher than the upper limit, it needs to be discharged immediately, but the power quality of the DC distribution network must not be affected during the discharge process. Therefore, when the DC voltage is higher than the upper limit, the supercapacitor cannot be discharged; if the DC voltage is lower than the upper limit, the supercapacitor is discharged when the load is large.

[0101] The present invention aims at the problem of DC power quality, especially the problem of DC voltage deviation. Based on the control strategy of controllable load and battery energy storage, the present invention designs a comprehensive control device of battery and controllable load. The feedforward control principle based on disturbance observer and the battery energy storage device are adopted to suppress DC voltage deviation, make up for the limitation of controllable load power capacity, reduce the number of charge and discharge times of battery energy storage, and increase its service life. The present invention can improve the system response speed and reduce the transient fluctuation of DC voltage. The supercapacitor based on sliding mode control can further improve the power quality and increase the stability of the system, ensuring the effective control. BRIEF DESCRIPTION OF THE DRAWINGS

[0102] Figure 1 An equivalent circuit diagram model of the DC distribution network structure of the present invention;

[0103] Figure 2 This is the Thevenin equivalent model diagram of the battery;

[0104] Figure 3 The circuit topology diagram for controllable load control;

[0105] Figure 4 It is the overall flow chart of integrated control of controllable load and battery sliding mode control;

[0106] Figure 5 It is the RC model of supercapacitor series connection;

[0107] Figure 6 This is the circuit topology diagram of the supercapacitor. DETAILED DESCRIPTION

[0108] The control method of the present invention first conducts a detailed analysis and research on the types of DC power quality. The DC power quality problems are mainly divided into voltage deviation and transient problems. For the DC voltage deviation problem, a control strategy based on controllable load and battery energy storage is provided.

[0109] Step 1, establish an equivalent circuit model of the DC distribution network. Since the equivalent model is not an infinite power source, the rapid and continuous changes in the source and load power in the DC power grid and the changes in line parameters will cause DC voltage fluctuations. DC voltage fluctuations are a random voltage change, which is a continuous change within a certain amplitude fluctuation range under the premise of system stability. DC voltage fluctuations are mainly caused by changes in distributed power sources. Distributed power sources generally use maximum power tracking technology. Due to the intermittent and random nature of distributed energy, the output power of distributed power sources is constantly changing, which in turn affects the stability of the DC voltage. The DC voltage deviation is analyzed and the basic circuit structure is established. The circuit diagram of the designed DC distribution network equivalent circuit model is shown in the figure. Figure 1 As shown, the analysis will be based on this. The rapid and continuous changes in the source and load power and the changes in line parameters in the DC distribution network will cause DC voltage fluctuations. DC voltage fluctuations are a kind of random voltage changes. Under the premise of system stability, the continuous changes within a certain amplitude fluctuation range. The DC voltage fluctuation coefficient expression is:

[0110]

[0111] Among them, U1 and U2 are the peak and trough values ​​of the voltage in the cycle, ΔU is the fluctuation amplitude of the voltage, and U n is the rated voltage of the DC bus;

[0112] Due to the imbalance of active power and line voltage drop, the DC bus voltage will deviate from the reference value, so the DC voltage deviation is an important indicator of DC power quality.

[0113] The distribution network includes photovoltaic and wind turbine load control, which are connected to the DC distribution network through DC / DC converters and AC / DC converters respectively. The converters use constant DC voltage control, and the DC output voltage U G=400V; R, r1, r2, r3 are line resistances, each 0.28Ω. U0, U1, U2 are node voltages of the three parts, R L1 ,R L2 is the current load equivalent resistance, which is 10Ω, i W ,i PV is the output current of photovoltaic and wind turbine. The node voltage U2 is:

[0114]

[0115] in:

[0116]

[0117] It has been found through testing that in normal system operation, changes in the load and line parameters in the DC power grid will cause drastic deviations in the DC voltage. The heavier the load and the greater the line resistance, the greater the deviation in the DC voltage, causing it to exceed the corresponding limit.

[0118] Step 2: In order to solve the problem of voltage deviation and the continuous imbalance of active power in the DC distribution network caused by frequent switching of loads, a comprehensive control device for batteries and controllable loads is established, and a current feedforward control system based on a disturbance observer is required. The feedforward control principle based on the disturbance observer is adopted, that is, the disturbance observer is introduced to track the system, and the feedback signal of the disturbance current is quickly tracked to make real-time adjustments and stabilize the circuit. The battery energy storage device can suppress the DC voltage deviation and make up for the limitation of the power capacity of the controllable load; the controllable load can reduce the capacity requirements of the battery, improve the economy, and reduce the number of charge and discharge times of the battery energy storage, thereby increasing its service life.

[0119] When adding a corresponding feedforward control system to the battery, the deviation limit should be limited to a very small range for loads with high power quality requirements, while the deviation limit can be larger for non-sensitive loads with low power quality requirements. In order to evaluate the severity of the DC voltage deviation, the deviation percentage D is defined. Generally, the DC voltage deviation limit is limited to ±5%, and its expression is as follows:

[0120]

[0121] Follow-up work will be carried out based on this.

[0122] 2-1 Design the structure of the battery energy storage device: The traditional method will increase the number of charge and discharge times, which is not conducive to the economy of the energy storage device. The controllable load cannot completely suppress the DC voltage deviation due to its capacity limitation. The energy storage device plays an important role in the distributed power generation system, among which the battery energy storage device is the most widely used in the field of energy storage. Therefore, a controllable load control strategy based on battery energy storage is adopted.

[0123] 2-1-1 The optimization algorithm used in the system is the most cost-effective technology choice under the constraints of energy demand and other constraints. Its decision variables are usually the selection of power generation technology, power generation capacity and transmission capacity within a given time and space. The established energy storage mathematical model is:

[0124]

[0125] Where x represents the 0-1 variable of the generator set, transmission line and energy storage device to be built. c represents the capacity variable of the generator set, transmission line and energy storage device to be built. u represents the 0-1 variable of the operating status of the thermal power unit. y represents the output variable of the generator set, transmission line and energy storage device. d, z, b, q represent the boundary parameters of the system coupling constraints and the input system and energy storage system constraints respectively. f(x, c, u, y) is the cost function of the system, h1, h2 represent the coupling constraints of the system, are the constraints of the input system and the energy storage system respectively.

[0126] 2-1-2 introduces the concept of critical time. When the required energy supply time is greater than the critical time, the capacity and model configuration of the energy storage can be selected based on the energy storage capacity; when the required energy supply time is less than the critical time, the capacity and model configuration of the energy storage must be selected based on the peak power rather than the energy storage capacity. The number of charge / discharge times of the energy storage system within an event T is:

[0127]

[0128] Where P Char (t),P DisChar (t),P SelfDisChar (t) represent the charge / discharge power and self-discharge power of the energy storage unit respectively; Cap(t), Cap min (t), and Cap max (t) respectively represent the capacity and boundary value of the energy storage unit in the T period; LR Char (t),LR DisChar (t) represents the charge / discharge loss rate within T period.

[0129] In the control process, when the energy storage coefficient is low and the power is insufficient, the energy storage coefficient is small and decreases with the decrease of energy storage, and the charging coefficient is large; when the energy storage coefficient is high and the power is sufficient, the charging coefficient decreases with the increase of energy storage coefficient, and the discharge coefficient K d and charging factor K ch Establish a relationship:

[0130] Discharge coefficient:

[0131]

[0132] Charging factor:

[0133]

[0134] S min ,S low ,S high ,S max They are the minimum value, lower value, upper value and maximum value of the energy storage system respectively; is the maximum value of the virtual droop coefficient.

[0135] 2-2 Design the battery control system model, first obtain the Thevenin equivalent model of the battery Figure 2 As shown. It can be seen that its equivalent model is composed of a voltage source E and an internal resistance R in series, and the equivalent model can accurately simulate the charging and discharging characteristics of the battery energy storage device.

[0136] Battery energy storage is generally divided into two types: centralized energy storage and distributed energy storage. For the DC voltage deviation problem, the controllable load cannot completely suppress the DC voltage deviation due to its capacity limitation. Therefore, the controllable load and the battery are comprehensively controlled to manage the DC deviation, and the corresponding system is constructed based on the current feedforward control.

[0137] 2-2-1 Construct a controllable load model. Assume that there is a controllable load in the DC grid. The power and terminal voltage can be adjusted within a certain range. The DC voltage fluctuation can be eliminated by controlling the controllable load power. Establish a controllable load circuit topology, such as Figure 3 shown.

[0138] In the figure, U i , U c are input voltage and output voltage respectively; i2, i L are input current and output current respectively; i1 is DC bus current; U L is the load voltage; L is the filter inductor; R is the load equivalent resistance. When the controllable load does not participate in the regulation, the switch tube S1 is turned on and S2 is turned off. At this time, U c =U i; When participating in regulation, the controllable load is controlled by changing the switch duty cycle D. Assuming that the controllable load is not fully engaged, when the voltage is lower than the lower limit, the terminal voltage is reduced; when the voltage is higher than the upper limit, the controllable load is fully engaged, participates in regulation, and stabilizes the DC voltage.

[0139] When the loss of this circuit is neglected, the equivalent resistance of the controllable load is:

[0140]

[0141] Where R is the load terminal resistance. The controllable load adopts feedforward double closed-loop control based on disturbance observer, which balances the power of DC distribution network by controlling the change of load power to suppress DC voltage deviation, but the power capacity of controllable load limits the control ability of DC voltage, so it needs to be controlled in coordination with the battery.

[0142] 2-2-2 Construct the basic structural model of the battery. Battery energy storage is generally divided into two types: centralized energy storage and distributed energy storage. For the DC voltage deviation problem, if only battery energy storage is used, the required battery capacity is large, and the number of charge and discharge times will also increase, which is not conducive to the economy of the energy storage device. The controllable load cannot completely suppress the DC voltage deviation due to its capacity limitation. Therefore, the controllable load and the battery are comprehensively controlled to control the DC deviation, and the corresponding system is constructed based on the current feedforward control. According to relevant principles and calculations, the calculation formula for the filter inductance in the converter is obtained as:

[0143]

[0144] Where f is the switching frequency, which is 10kHz in this design, and I 0min is the minimum average current, which is 5A; U0 is the node voltage, which is selected to be 200V; U is the maximum voltage, which is 400V in the design.

[0145] According to control theory and calculation, the second-order damping ζ and natural frequency ω of the system can be calculated: n for:

[0146]

[0147] K PWM is the PWM equivalent gain of the battery, K p is the proportionality coefficient, K i is the integral coefficient. Through these two indicators, the main time domain indicators of the control system can be determined. For this design, the control system needs to have a certain bandwidth when designing parameters. PWM When the value is 1, ζ is 0.707, and L is 2mH, the calculation formula for the bandwidth frequency is:

[0148]

[0149] Adding feedforward control requires adding additional current sensors to measure the disturbance current i0, which will increase the system cost and is not conducive to the expansion and plug-and-play of the power grid. Therefore, the disturbance observer is introduced to quickly track the disturbance i0, and the dynamic equation of the system is:

[0150]

[0151] In the disturbance observer, C0 is the DC bus capacitor, L0 is the filter inductor, and u sc ,u i are the battery and DC grid voltages, i and i0 are the battery current and disturbance current, respectively. After calculation, it can be written as:

[0152]

[0153] The disturbance observer model is designed as:

[0154]

[0155] Where m is the convergence coefficient and u is the control vector. Then the final expression of the disturbance observer is:

[0156]

[0157] This results in a corresponding battery model structure.

[0158] 2-2-3 Aiming at the DC voltage deviation problem, a comprehensive control based on controllable load and battery is proposed. Feedforward control based on disturbance observer is adopted, and battery energy storage makes up for the lack of controllable load capacity. The controllable load is used to reduce the capacity of energy storage equipment and reduce its charge and discharge times, thereby increasing its service life and improving the power quality of each node of the DC distribution network.

[0159] Among them, as long as one of the node voltage exceeds the limit, it needs to be controlled to meet the power quality requirements of all nodes. When the deviation is small, the controllable load can completely suppress the deviation; when the deviation is large, the controllable load cannot completely suppress the voltage deviation due to capacity limitations, and the battery energy storage device will eliminate the voltage deviation. When the voltage exceeds the upper limit, that is, max{U0,U1,U2}>1.05Ue, where U0,U1,U2 are respectively Figure 1 The voltage of the middle nodes 0, 1, and 2, U e is the rated voltage of the DC distribution network. Given a reference value U a =1.05U e ; The voltage is lower than the lower limit, that is, max{U0,U1,U2}<0.95U e When the reference value U a =0.95U e .

[0160] According to the above, the overall control flow chart of the controllable load and battery sliding mode control in the present invention can be obtained as follows: Figure 4 shown.

[0161] Step 3, design a supercapacitor management strategy based on sliding mode control, that is, construct Figure 5 The supercapacitor series RC model shown in the figure; when the DC voltage sags or is interrupted for a short time in the DC distribution network, the supercapacitor needs to be discharged, and the DC / DC converter works in Boost mode; when the DC voltage rises temporarily in the DC distribution network, the supercapacitor needs to be charged, and the DC / DC converter works in Buck mode. This control strategy can be used to solve the problems of DC voltage sag, DC voltage rise, DC voltage short-term interruption, and DC transient power quality.

[0162] Based on step 2, a supercapacitor based on sliding mode control is designed and added to the energy storage unit. In addition to voltage deviation, DC voltage sag, DC voltage swell, and DC voltage short-term interruption, DC transient power quality problems are also major problems in DC distribution networks. This requires the rapidity and accuracy of the control system. Sliding mode control has fast control speed and high tracking accuracy, and has great advantages in managing transient DC power quality problems. Therefore, a supercapacitor DC transient power quality management strategy based on sliding mode control is proposed.

[0163] In the proposed hybrid energy storage system, the battery and supercapacitor use single-loop constant current control to ensure the stable operation of the system, and use a low-pass filter for power distribution, which can effectively suppress system power fluctuations, achieve bus voltage stability, and manage the energy of the supercapacitor. The battery mainly stores a large amount of electrical energy, and the supercapacitor, as an auxiliary energy source, is responsible for providing fast response and high current peak.

[0164] This strategy mainly divides the working range of the supercapacitor into the lower limit of discharge, the discharge warning zone, the normal working zone, the charging warning zone and the upper limit of charge, so as to take appropriate charge and discharge control measures in different states to extend the system life and ensure the optimization of system performance. The battery is used to store electrical energy and continuously provide reliable power to the load; the supercapacitor is used as an auxiliary energy source. When the load requires more energy, the supercapacitor will release energy to meet the immediate needs of the system.

[0165] The model formula of supercapacitor is:

[0166]

[0167] Where C is the capacitance of the supercapacitor, and the rated voltage U dis 400V, E is the storage energy of the supercapacitor, U sc is the voltage value of the supercapacitor, which must meet the maximum operating voltage U scmax Does not exceed the rated voltage, working voltage U sc It is a very important parameter in the process of charging and discharging. It not only represents the heat loss during charging and discharging, but also varies with the size of the charging and discharging current. Especially when discharging at a high current, the power loss will become very large, affecting the efficiency of the supercapacitor. Figure 5 As shown in the figure, the resistor R1 is equivalent to the internal resistance of the supercapacitor, which is 0.04Ω. The RC model can accurately reflect the electrical characteristics of the supercapacitor during the charging and discharging process, which is convenient for analyzing the charging and discharging process and energy storage efficiency of the supercapacitor.

[0168] During the discharge process, the energy loss of the capacitor is:

[0169]

[0170] In the formula, i is the capacitor current value, and R1 is the resistance, where R1 is equivalent to the internal resistance of the supercapacitor. It is a very important parameter in the charging and discharging process. It not only represents the heat loss during charging and discharging, but also varies with the size of the charging and discharging current. The voltage drop caused is also different. Especially when discharging at a large current, the power loss will become very large, affecting the efficiency of the supercapacitor, so further design is needed.

[0171] The transfer efficiency of the supercapacitor is:

[0172]

[0173] Where: E is the total energy of the supercapacitor. Assume that the terminal voltage of the supercapacitor changes in the range of [u s1 ,u s2 ], the supercapacitor capacitance is C, take 5F. Maximum operating voltage U scmax Take it as 400V, the minimum working voltage U scmin is 125V. Then the power of the discharge process satisfies the following relationship:

[0174]

[0175] P max is the maximum required power, which is 100kW. Period T max Assumed to be 1s. Then the discharge energy loss can be expressed as:

[0176]

[0177] U sMis the intermediate voltage, which is defined as the average voltage maintained by the supercapacitor for a long time. It can be understood as the voltage when the supercapacitor is in standby mode. Therefore, the discharge efficiency of the supercapacitor is:

[0178]

[0179] Similarly, the charging power is:

[0180]

[0181] The charging energy loss can be expressed as:

[0182]

[0183] Therefore, the charging efficiency of the supercapacitor is:

[0184]

[0185] Sliding mode control is a nonlinear control method that was first developed in the control of variable structure systems. Sliding mode control includes a time-varying discontinuous state feedback control law, which can switch between a continuous structure and another continuous structure at a high frequency according to the position of the variable in the state space, and can reach the switching surface (also called the sliding surface) within a limited time, and present sliding mode motion on the sliding surface, with the purpose of making the dynamics of the state variable reach the preset expected dynamics. In view of this feature, the system is set as:

[0186]

[0187] Where: f(x,t), g(x,t) are the smooth vector fields of the control system, is the state variable vector, u is the control input vector, which can be expressed as:

[0188] a=a eq +Δa(Equation 26)

[0189] a eq is the equivalent control variable, and the system state variable vector can be made to perform sliding mode motion on the sliding surface. Δa is the switch control variable, and its function is to make the system state variable reach the sliding surface within a finite time at any position in the state space.

[0190] When the DC voltage of the DC distribution network rises temporarily, the supercapacitor needs to be discharged. At this time, the DC / DC converter works in the Boost mode. The mathematical model of the Boost mode can be expressed as:

[0191]

[0192] The circuit topology of the supercapacitor is shown in the figure below: Figure 6 As shown, U sc , I represent the battery terminal voltage and current in the supercapacitor topology respectively; U dc , I0 respectively represent the voltage and current on the DC grid side in the circuit topology; L is the filter inductor, C1 is the DC bus capacitor, and C is the supercapacitor capacitor. D1 is the input duty cycle of the charging mode control, which can be written as follows in combination with formula (26):

[0193] D1=D 1eq +ΔD1 (Equation 28)

[0194] D 1eq is the equivalent control of D1, and ΔD1 is the switch control variable. When the DC voltage drops or is interrupted for a short time in the DC distribution network, the supercapacitor needs to be charged. At this time, the DC / DC converter works in Buck mode. The mathematical model of Buck mode can be expressed as:

[0195]

[0196] Where D2 is the discharge mode control input duty cycle, which can be written as:

[0197] D2=D 2eq +ΔD2 (Equation 30)

[0198] After the power quality event ends, the initial voltage of the supercapacitor is restored. If the remaining power of the supercapacitor is not enough to solve the power quality problem in the next extreme situation, it is necessary to restore the initial state, as shown in the following formula:

[0199]

[0200] The lower limit of capacitor voltage is taken as U sL =220V, upper limit value U sH =380V.

[0201] If the supercapacitor voltage is lower than the lower limit, it needs to be charged immediately, but the power quality of the DC distribution network must not be affected during the charging process. Therefore, when the DC voltage is lower than the lower limit, the supercapacitor cannot be charged; if the DC voltage is higher than the lower limit, the supercapacitor is charged when the load is small or the electricity price is low.

[0202] If the supercapacitor voltage is higher than the upper limit, it needs to be discharged immediately, but the power quality of the DC distribution network must not be affected during the discharge process. Therefore, when the DC voltage is higher than the upper limit, the supercapacitor cannot be discharged; if the DC voltage is lower than the upper limit, the supercapacitor is discharged when the load is large.

[0203] The voltage deviation problem can be effectively suppressed through the comprehensive control strategy of controllable load and battery energy storage; at the same time, the DC transient problem can be solved through the supercapacitor management strategy based on sliding mode control. It can effectively improve the working efficiency and power quality of the DC distribution network.

Claims

1. A method for controlling power quality of a DC distribution network based on battery sliding mode control, comprising the following steps in sequence: Step 1: Establish an equivalent circuit model of the DC distribution network structure: The DC voltage fluctuation coefficient expression is: Among them, U1 and U2 are the peak and trough values ​​of the voltage in the cycle, ΔU is the fluctuation amplitude of the voltage, and U n is the rated voltage of the DC bus; Step 2: Establish a battery energy storage control device and a controllable load model, introduce a disturbance observer to track the system, track the feedback signal of the disturbance current, so as to make real-time adjustments and stabilize the circuit; the battery energy storage control device makes up for the shortage of the controllable load capacity, and uses the controllable load to reduce the capacity of the energy storage device. When the deviation is small, the controllable load model can completely suppress the deviation; when the deviation is large, the controllable load model cannot completely suppress the voltage deviation due to capacity limitations, and the battery energy storage control device is used to eliminate the voltage deviation; To evaluate the severity of the DC voltage deviation, define the deviation percentage D, and limit the DC voltage deviation to ±5%, which is expressed as follows: Step 3, design a supercapacitor management strategy based on sliding mode control and construct a supercapacitor series RC model; when the DC voltage of the DC distribution network temporarily sags or is briefly interrupted, the supercapacitor needs to be discharged and the DC / DC converter works in Boost mode; when the DC voltage of the DC distribution network temporarily rises, the supercapacitor needs to be charged and the DC / DC converter works in Buck mode.

2. According to the method for controlling power quality of a DC distribution network based on battery sliding mode control according to claim 1, step 2 comprises: Step (2-1), establish the energy storage mathematical model; Step (2-2), establish a controllable load model and a battery control system model and introduce a disturbance observer; The battery energy storage control device includes the energy storage mathematical model established in step (2-1), the controllable load model established in step (2-2-1) and the battery control system model, and combines them.

3. According to the method for controlling the power quality of a DC distribution network based on battery sliding mode control in claim 1, the step (2-1) of establishing a mathematical model of energy storage in step 2 comprises: (2-1-1) Establishing a mathematical model for energy storage: Among them, x represents the 0-1 variable of the generator set, transmission line and energy storage device to be built; c represents the capacity variable of the generator set, transmission line and energy storage device to be built; u represents the 0-1 variable of the operating status of the thermal power unit; y represents the output variable of the generator set, transmission line and energy storage device; d, z, b, q represent the system coupling constraints and the boundary parameters of the input system and energy storage system constraints respectively; f(x, c, u, y) is the cost function of the system, h1, h2 represent the coupling constraints of the system, are the constraints of the input system and the energy storage system respectively; (2-1-2) In the above energy storage mathematical model, when the required energy supply time is greater than the critical time, the capacity and model configuration of the energy storage can be selected based on the energy storage capacity; when the required energy supply time is less than the critical time, the capacity and model configuration of the energy storage must be selected based on the peak power rather than the energy storage capacity; the number of charge / discharge times of the energy storage system within an event T is: Where P Char (t),P DisChar (t),P SelfDisChar (t) represent the charge / discharge power and self-discharge power of the energy storage unit respectively; Cap(t), Cap min (t), and Cap max (t) respectively represent the capacity and boundary value of the energy storage unit in the T period; LR Char (t),LR DisChar (t) represents the charge / discharge loss rate within T period; In the control process, when the energy storage coefficient is low and the power is insufficient, the energy storage coefficient is small and decreases with the decrease of energy storage, and the charging coefficient is large; when the energy storage coefficient is high and the power is sufficient, the charging coefficient decreases with the increase of energy storage coefficient, and the discharge coefficient K d and charging factor K ch Establish a relationship: Discharge coefficient: Charging factor: S min ,S low ,S high ,S max are the minimum, lower, upper and maximum values ​​of the energy storage system respectively; K Emax is the maximum value of the virtual droop coefficient.

4. According to the method for controlling power quality of a DC distribution network based on battery sliding mode control in claim 2, the controllable load model and the battery control system model established in step (2-2) of step 2 include: (2-2-1) Construct a controllable load model. Add a controllable load to the DC power grid. The power and terminal voltage can be adjusted within a certain range. By controlling the controllable load power, the DC voltage fluctuation is eliminated. The controllable load circuit topology is established. During the regulation process, the controllable load can be controlled by changing the duty cycle of the switch tube; assuming that the controllable load is not fully engaged, when the voltage is lower than the lower limit, the terminal voltage is reduced; when the voltage is higher than the upper limit, the controllable load is fully engaged and participates in the regulation to stabilize the DC voltage; When the loss of the controllable load circuit is neglected, the equivalent resistance of the controllable load can be obtained as: Where R is the load terminal resistance; the controllable load adopts feedforward double closed-loop control based on disturbance observer, and balances the power of DC distribution network by controlling the change of load power to suppress DC voltage deviation. The power capacity of controllable load limits the control ability of DC voltage, and needs to be controlled in coordination with the battery; (2-2-2) The battery control system model is established and the calculation formula of the filter inductance in the converter is obtained as follows: Where f is the switching frequency, I0min is the minimum average current, U0 is the node voltage, and U is the maximum voltage; By adding corresponding feedforward control to the battery, the second-order damping ζ and natural frequency ω of the system n for: K PWM is the PWM equivalent gain of the battery, K p is the proportionality coefficient, K i is the integral coefficient. The main time domain indicators of the control system can be determined by these two indicators. The expression of bandwidth frequency is: The disturbance observer is introduced to quickly track the disturbance current i0, and the dynamic equation of the system is: In the disturbance observer, C0 is the DC bus capacitor, L0 is the filter inductor, and u sc ,u i are the battery and DC grid voltages respectively, i and i0 are the battery current and disturbance current respectively; the expression is: The disturbance observer model of the designed system is: Where m is the convergence coefficient; the final expression of the disturbance observer is: Thus, a battery energy storage control device is obtained; (2-2-3) Based on comprehensive control of controllable load and battery.

5. The method for controlling power quality of a DC distribution network based on battery sliding mode control according to claim 3 is characterized in that: The controllable load circuit topology model established in the controllable load model (2-2-1) can adjust the terminal voltage and power within a certain range, and thereby eliminate voltage fluctuations; wherein: U i , U c are input voltage and output voltage respectively; i2, i L are input current and output current respectively; i1 is DC bus current; U L is the load voltage; L is the filter inductor; R is the load equivalent resistance; when the controllable load does not participate in the regulation, the switch tube S1 is turned on and S2 is turned off. At this time, U c =U i ; When participating in regulation, the controllable load is controlled by changing the duty cycle D of the switch tube.

6. The method for controlling power quality of a DC distribution network based on battery sliding mode control according to claim 4 is characterized in that: The comprehensive control of the controllable load and the battery described in (2-2-3) is as follows: By using feedforward control based on disturbance observer, the battery storage makes up for the lack of controllable load capacity, uses controllable load to reduce the capacity of energy storage equipment, and reduces its charge and discharge times; As long as the voltage of a node exceeds the limit, it needs to be controlled to meet the power quality requirements of all nodes; when the deviation is small, the controllable load can completely suppress the deviation; when the deviation is large, the controllable load cannot completely suppress the voltage deviation due to capacity limitations, and the battery energy storage device will eliminate the voltage deviation; the voltage is higher than the upper limit, that is, max{U0,U1,U2}>1.05U e When, U0, U1, and U2 are the voltages of nodes 0, 1, and 2 respectively, and U e is the rated voltage of the DC distribution network; given reference value U a =1.05U e ; The voltage is lower than the lower limit, that is, max{U0,U1,U2}<0.95U e When U a =0.95U e .

7. The method for controlling the power quality of a DC distribution network based on battery sliding mode control according to claim 1 is characterized in that Based on step 2, a supercapacitor is added to the energy storage unit.

8. The method for controlling the power quality of a DC distribution network based on battery sliding mode control according to claim 1 is characterized in that The control and management strategy in step 3 includes: adding supercapacitors to the system to form a hybrid energy storage system based on steps 1 and 2; In this hybrid energy storage system, the battery system and supercapacitor adopt single-loop constant current control respectively, and use low-pass filter for power distribution, which effectively suppresses system power fluctuations, realizes bus voltage stability, and manages the energy of supercapacitors to ensure stable operation of the system; this strategy divides the working range of supercapacitors into discharge lower limit area, discharge warning area, normal working area, charging warning area and charging upper limit area, so as to take appropriate charge and discharge control measures in different states to extend system life and ensure the optimization of system performance; The model formula of supercapacitor is: Where C is the capacitance of the supercapacitor and the rated voltage U is d , E is the storage energy of the supercapacitor, U sc is the voltage value of the supercapacitor, which must meet the maximum operating voltage U scmax The rated voltage is not exceeded, which is 400V. In the supercapacitor RC model, the resistor R1 is equivalent to the internal resistance of the supercapacitor, which is 0.04Ω. The supercapacitor RC model can accurately reflect the electrical characteristics of the supercapacitor during the charging and discharging process, which is convenient for analyzing the charging and discharging process and energy storage efficiency of the supercapacitor. During the discharge process, the energy loss of the capacitor is: Where i is the capacitor current value, R1 is the resistance, that is, the internal resistance of the supercapacitor; the transmission efficiency of the supercapacitor is: Where: E is the total energy of the supercapacitor; Assume that the terminal voltage of the supercapacitor changes in the range of [u s1 ,u s2 ], the supercapacitor capacitance is C, take 5F; the maximum working voltage is U scmax Take it as 400V, the minimum working voltage U scmin is 125V; then the power of the discharge process satisfies the following relationship: Where P max is the maximum required power, which is 100kW; the period T max Assuming 1s; then the discharge energy loss can be expressed as: U sM is the intermediate voltage, which is the average voltage maintained by the supercapacitor for a long time; therefore, the discharge efficiency of the supercapacitor is: Similarly, the charging power is: The charging energy loss is expressed as: Therefore, the charging efficiency of the supercapacitor is: According to the characteristics of sliding mode control, the system is set as follows: Where: f(x,t), g(x,t) are the smooth vector fields of the control system, is the state variable vector, u is the control input vector, which can be expressed as: a = a eq + Δa (Equation 26) a eq is the equivalent control variable, which makes the system state variable vector perform sliding mode motion on the sliding surface; Δa is the switch control variable, which is used to make the system state variable reach the sliding surface in a finite time at any position in the state space; When the DC voltage drops or is interrupted for a short time in the DC distribution network, the supercapacitor needs to be discharged and the DC / DC converter works in the Boost mode. The mathematical model of the Boost mode is expressed as: In the circuit topology diagram of the supercapacitor, U sc , I represent the battery terminal voltage and current in the supercapacitor topology respectively; U dc , I0 respectively represent the voltage and current on the DC grid side in the circuit topology; L is the filter inductor, C1 is the DC bus capacitor, and C is the supercapacitor capacitor; D1 is the input duty cycle of the charging mode control, which can be written as follows in combination with formula (26): D1=D 1eq +ΔD1 (Equation 28) D 1eq is the equivalent control of D1, ΔD1 is the switch control variable; When the DC voltage of the DC distribution network rises temporarily, the supercapacitor needs to be charged. At this time, the DC / DC converter works in Buck mode. The mathematical model of Buck mode is expressed as: Where D2 is the input duty cycle of the charging mode control, which can be written as: D2=D 2eq +ΔD2 (Equation 30) After the power quality event ends, the initial voltage of the supercapacitor is restored. If the remaining power of the supercapacitor is not enough to solve the power quality problem in the next extreme situation, it is necessary to restore the initial state, as shown in the following formula: The lower limit of capacitor voltage is taken as U sL =220V, upper limit value U sH =380V; If the supercapacitor voltage is lower than the lower limit, it needs to be charged immediately. When the DC voltage is lower than the lower limit, the supercapacitor cannot be charged. If the DC voltage is higher than the lower limit, the supercapacitor is charged when the load is small or the electricity price is low. If the supercapacitor voltage is higher than the upper limit, it needs to be discharged immediately. When the DC voltage is higher than the upper limit, the supercapacitor cannot be discharged. If the DC voltage is lower than the upper limit, the supercapacitor is discharged when the load is large.