Optical storage micro-grid super-spiral sliding mode master-slave voltage control system under ZigBee communication network

Abstract

The invention discloses an optical storage micro-grid super-spiral sliding mode master-slave voltage control system under a ZigBee communication network, and belongs to the technical field of optical storage micro-grid optimization control. The invention aims to provide a master-slave control system suitable for an optical storage micro-grid under a ZigBee communication network, so that the defects of a PI controller in transient response are overcome, and the common buffeting problem in traditional sliding mode control is solved by the super-spiral sliding mode master-slave voltage control system for the optical storage micro-grid under the ZigBee communication network. According to the invention, a super-spiral sliding mode controller of a master-slave controller is firstly designed, and then a master unit voltage controller and a slave unit power tracking controller are respectively designed. The invention provides bus voltage network master-slave control under the Zigbee network for optical storage micro-grid network control, overcomes the defects that the existing control main unit is large in capacity and the control effect is poor under conventional PI control communication delay, improves the voltage control capability of the optical storage micro-grid, and remarkably improves the operation voltage safety of the optical storage micro-grid.

Description

Technical Field

[0001] The invention belongs to the technical field of photovoltaic storage microgrid optimization control. Background Art

[0002] As global attention to climate change continues to increase, my country has proposed the dual carbon policy goals of "carbon peak and carbon neutrality". To achieve this ambitious goal, China is accelerating the transformation of its energy structure, reducing its dependence on fossil fuels, and vigorously developing renewable energy. Against this background, the new energy industry has ushered in unprecedented development opportunities. In particular, the development and utilization of clean energy such as wind power and solar energy have received unprecedented attention. However, due to the intermittent and unstable characteristics of these energy sources, how to ensure their stable output has become a key issue. As a distributed energy solution that integrates photovoltaic power generation, energy storage system and intelligent control technology, the photovoltaic storage microgrid can not only effectively solve the problem of large fluctuations in new energy power generation, but also significantly improve the reliability and efficiency of regional power supply and reduce users' electricity costs. Therefore, in future development, the photovoltaic storage microgrid will surely become an important force in promoting energy transformation and helping to achieve the dual carbon goals. When the photovoltaic storage microgrid operates independently, its bus voltage needs to rely on multiple internal distributed generation to maintain. Bus voltage control is the key to the stable operation of the photovoltaic microgrid in off-grid mode. The photovoltaic storage microgrid voltage control method can be divided into peer control and master-slave control. In peer control, multiple distributed generation units are in the same position, and the power to maintain voltage control is distributed among the distributed generation units through droop control. In master-slave control, there is a master unit responsible for controlling the bus DC voltage, and other distributed generation units serve as slave units and do not participate in system voltage control. The technical difficulty of peer control is the impact of line impedance on power distribution. Since the line impedance is unknown or the system structure may change, the bus voltage control cannot be accurately controlled, resulting in voltage fluctuations. The disadvantage of master-slave control is that it requires a master unit with a larger power capacity, which reduces the economic efficiency.

[0003] With the popularization of communication technology, it has become a trend to improve the voltage control of microgrids with the help of communication. In peer control, communication can be relied on to achieve secondary control of bus voltage and reduce bus voltage deviation. In master-slave control, relying on communication technology, the master unit can cooperate with the slave unit to participate in system voltage control, not just relying on the control capability of the master unit. This reduces the dependence on the capacity of the master unit and alleviates the defects of conventional master-slave control.

[0004] In communication technology, ZigBee has emerged as an advanced wireless communication technology suitable for short-distance and low-speed transmission applications. ZigBee has a large number of transmission nodes, which can achieve rapid monitoring and transmission requirements between different networks in a microgrid environment, ensuring the efficiency and real-time nature of data exchange. In addition, ZigBee also has the characteristics of low power consumption, low cost, and high security, ensuring the safe and stable operation of microgrid communications. However, ZigBee also has certain limitations, such as limited transmission rate, which is not suitable for application scenarios with large amounts of data or high-speed transmission.

[0005] The photovoltaic storage microgrid based on ZigBee communication can transmit information between various distributed power sources. This allows the master unit to lead other slave units to jointly realize bus voltage control, reducing the requirements for the master unit capacity. However, it should be noted that due to the low bandwidth characteristics of ZigBee, its communication delay is usually large, and there may be communication interruptions. Conventional proportional integral control is easily affected by communication delays and cannot adapt well to this working environment. Therefore, it is necessary to design an advanced control system suitable for photovoltaic storage microgrids based on ZigBee communication.

[0006] Patent CN119009926A discloses a method and system for stabilizing the bus voltage of a photovoltaic DC microgrid. The invention relates to the fields of power electronics and control technology. By establishing a generalized reduced-order average model of DC-DC and utilizing feedback linearization technology and a fixed-time adaptive backstepping sliding mode control strategy, the bus voltage can be precisely controlled and rapidly stabilized. However, the invention is mainly aimed at single-unit control, and an improved control method is used to provide a voltage control effect. The method does not involve network control, and there is no coordination problem between multiple units.

[0007] Patent CN118630824A discloses a secondary voltage control method for an isolated microgrid based on a disturbance observer. The method estimates the disturbance by designing a nonlinear disturbance observer, establishes a large signal voltage observation model under disturbance of the microgrid, and controls the voltage of the i-th distributed generation to follow the voltage reference value based on the exponential sliding film surface. The method is an observation-based peer control method under distributed communication mode. Through disturbance observation, the convergence speed and robustness of voltage control in the microgrid control system are improved. However, the method relies on a large disturbance observation model, and the observation information needs to be transmitted between distributed units, which increases the complexity of the algorithm. Moreover, the method is a peer control, there is no main unit, and the voltage control depends on the coordination of each distributed unit. Although the peer control method can reduce the capacity of the single main unit, the peer control is subject to defects such as line impedance, communication loss, and slow dynamic convergence speed, and has no advantages in small-scale microgrids with a small number of distributed units.

[0008] Patent CN113224789B discloses a dynamic master-slave control system and its application method in the secondary control of an isolated microgrid. By dynamically selecting a leading unit and relying on a power reference sent from a communication network, the droop curves of other follower units are adjusted to achieve microgrid power proportion balancing and voltage recovery. Although the invention provides a dynamic master-slave control strategy, its essence is a dynamic droop control. The master unit is a leading unit, and the slave unit is a communication follower unit. It is not a master-slave control mode in the conventional sense. In a droop-controlled microgrid, the power distribution of each unit is affected by the line impedance, and the power output characteristics of the slave unit are not necessarily consistent with the master unit. The power reference value sent by the master unit may not be suitable for the slave unit, which affects the effect of bus voltage control. And the invention does not involve a specific communication network. Further research is needed on how to achieve more efficient voltage energy management in combination with the master-slave control mode of the ZigBee communication network. Summary of the invention

[0009] The purpose of the present invention is to propose a master-slave control system for a photovoltaic microgrid with energy storage under a ZigBee communication network, thereby overcoming the shortcomings of the PI controller in transient response and solving the common jitter problem in traditional sliding mode control.

[0010] Design of super-helical sliding mode controller of master-slave controller in master-slave voltage control system of the present invention: S1. Establish the system state equation: The dynamic equation of the DG unit in the stationary abc coordinate system: Where: V abc is the load voltage; V t,abc is the inverter output voltage; i t,abc is the inverter output current; i L,abc is the load current; L t is the sum of the inductance of the transmission line and the filter; R t is the sum of the inductance and resistance of the transmission line and the filter; R is the load resistance; L is the load inductance; C is the load capacitance, R l is the load inductance branch resistance; The state space equation of DG access system: in: Where: x = [V d , V q , I td , I tq , I Ld , ILq ] T ——state vector; u=[V td , V tq ] T ——Input control vector; y = [V d , V q ] T is the output vector of the main DG unit; y = [I tds , I tqs ] T is the output vector from the DG unit; w(t) is the fluctuation caused by the photovoltaic and energy storage systems; S2, control stability caused by Zigbee communication delay: The state space expression of the system when there is a delay is a small signal model: Where: A, A d is a fixed matrix of appropriate dimension; Φ(t) is the initial condition of the system in the time period t∈[-ρ,0]; τ(t) is the time-varying delay; The time-varying delay τ(t) satisfies the condition: Assume that scalar ρ>0, μ>0, when there exists a symmetric positive definite matrix P=P T >0, Q=Q T >0 and Z=Z T >0, nonsymmetric positive semidefinite matrix: and arbitrary matrices Y and T, the time-delay system (4) is asymptotically stable when the following LMI holds: Where: S3. Steps to solve LMI: S31. Obtain the system matrix from the microgrid parameters to calculate A and A d . S32, to determine whether there is a feasible solution for LMI under a given time delay. S33, use the binary iterative algorithm to find the system's MADB: define an interval [τ u ,τ f ], where τ u represents the stabilization delay, τ f represents unstable delay, then τ n =(τ u +τ f ) / 2, then according to τ n Reduce the delay bound if τ nis a stable delay, then τ u =τ n ; If unstable, then τ f =τ n , iterate until the accuracy requirement is met: τ ac =|τ f -τ u |<ε, where ε is the minimum error.

[0011] The present invention designs the main unit voltage controller based on the master-slave controller super spiral sliding mode controller: The direct-axis component of the load voltage V d and the quadrature axis component V q The corresponding reference value V d,ref 、V q,ref The error between d 、e q As the input of the main controller, it is adjusted accordingly; d, q axis voltage tracking error: Where: V d,ref is the voltage d-axis component reference value; V q,ref is the reference value of the voltage q-axis component; V d (t) is the d-axis component of the output voltage of the main controller; V q (t) is the q-axis component of the output voltage of the main controller; Define auxiliary variable e d,1 =e d , e q,1 =e q A second-order auxiliary system with external disturbances: Where: u i is the controlled signal of the main DG unit in the dq coordinate system; d(x) is the external bounded disturbance; Where: f i and g i as follows The control variable u is designed using the second-order sliding mode control of the superhelical algorithm i , let the sliding surface be s, satisfying the following equation: Where: s is the sliding mode variable; v is the estimated value of the acceleration of the control law, λ and α are the control parameters to be designed; sign(s) is the sign function; Adjust the parameters of the sliding mode controller: Replace the discontinuous sign function sign(s) in the superhelical controller with a continuous hyperbolic tangent function: At this time, formula (12) becomes: in Determines the speed at which the limit is reached; The designed sliding surface is: At this time, the control quantity u i as follows:

[0012] The present invention designs a slave unit power tracking controller based on a master-slave controller super spiral sliding mode controller: The direct axis component I of the current output by each DG unit ds and the quadrature axis component I qs With the corresponding reference value I ds,ref ,I qs,ref The error between ds 、e qs Regulated as input from the controller; d, q axis current tracking error: Where: I ds,ref is the reference value of the current d-axis component; I qs,ref is the reference value of the current q-axis component; I ds (t) is the d-axis component of the current output from the controller; I qs (t) is the q-axis component of the current output from the controller; First-order system with external disturbance: Where: u i is the dq component of the inverter output voltage from the DG; f i and g i as follows: In the first-order sliding mode control system, let the sliding surface be s, satisfying the following equation: At this time, the control quantity u is The expression in the slave controller is: Where u is is the control signal from the DG unit; K iis the sliding gain from the DG unit.

[0013] The present invention is aimed at the network control of photovoltaic and energy-storage microgrids, and proposes a bus voltage network master-slave control under a Zigbee network, which solves the defects of large capacity of the existing control main unit and poor control effect under conventional PI control communication delay, improves the voltage control capability of the photovoltaic and energy-storage microgrid, and significantly improves the operating voltage safety of the photovoltaic and energy-storage microgrid. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is the control structure diagram of the photovoltaic-storage-charging microgrid; Figure 2 ZigBee networking method; Figure 3 It is the main controller structure diagram; Figure 4 This is the block diagram of the slave controller structure; Figure 5 Network schedule in ZigBee; Figure 6 This is a comparison chart of different control methods for the dq components of the main DG output voltage in the ZigBee network; Figure 7 A comparison chart of different control methods for the dq components of the output current from the DG in a ZigBee network; Figure 8 Phase diagram of tracking error between master and slave controllers in ZigBee network; Fig. 9 Three-phase voltage and current components of load in ZigBee network. DETAILED DESCRIPTION

[0015] The present invention proposes a master-slave control system suitable for a photovoltaic storage microgrid under a ZigBee communication network. The control system uses ZigBee to realize the transmission of current reference value information between different distributed power generation. The master controller sends a current reference signal to the slave controller through the ZigBee communication network, realizing the network control of the system. The master unit control adopts a method based on super twisting sliding mode control (STSMC), and the slave unit adopts improved sliding mode control, which is used to accurately adjust the voltage and current of the photovoltaic storage microgrid in a low-bandwidth network environment. This method overcomes the shortcomings of the PI controller in transient response and solves the common jitter problem in traditional sliding mode control (SMC).

[0016] The present invention is aimed at a photovoltaic storage microgrid based on a ZigBee communication network, and adopts a master-slave bus voltage control method based on a communication network. The master unit is responsible for bus voltage control and sends a cooperative current control reference value to the slave unit. The slave unit responds to the master unit current reference value and cooperates with the master unit to maintain a constant bus voltage. The master voltage adopts super-helical sliding mode control, and the slave unit adopts an improved sliding mode control algorithm. The master-slave control system achieves bus voltage control stability under low-bandwidth communication.

[0017] The steps and main features of the present invention are as follows. Step 1: Design the master-slave control architecture of the photovoltaic storage microgrid under the ZigBee communication network; The characteristics are as follows: the wireless communication ZigBee network is used for communication between the power generation units. The optical microgrid adopts AC bus networking. The master unit is a large-capacity photovoltaic storage system, and the slave unit is a distributed photovoltaic control. The master unit realizes constant control of the AC bus voltage, and the slave unit responds to the power command of the master unit through communication. The slave unit receives the power command of the master unit through the ZigBee network, realizes the power control following the master unit, and assists the master unit in DC voltage control.

[0018] Step 2: Design the bus voltage controller of the master unit based on improved spiral sliding mode control. The design features: the master unit adopts bus voltage tracking control, adopts super spiral sliding mode to conventional PI controller, and replaces the discontinuous sign function sign(s) with a continuous hyperbolic tangent function to reduce sliding mode chattering. The master unit sends the slave unit current reference value and transmits it to the slave unit controller through the ZigBee network.

[0019] Step 3: Design a current tracking controller for the slave unit based on sliding mode control The invention is characterized in that: a current reference value is sent from a control response master unit; first-order sliding mode control is used to replace conventional PI control to improve the control stability of the system under control delay.

[0020] In the existing photovoltaic microgrid, bus voltage control is the core control to achieve constant bus voltage and ensure the safety of load power supply. The main problems faced by photovoltaic microgrid bus voltage control are: (1) dynamic consistency of droop control; (2) master-slave coordination under master-slave control; (3) the influence of communication network delay on control effect. The present invention adopts a wireless ZigBee communication network, network master-slave control, improved super-helical sliding mode master unit control and improved sliding mode slave unit control to improve the bus voltage control effect.

[0021] The steps of the present invention are described in detail below with reference to the accompanying drawings and typical examples: Step 1: Master-slave control architecture of photovoltaic-storage-charging microgrid based on ZigBee communication network The present invention designs a light-storage-charging microgrid example based on ZigBee communication network Figure 1 As shown. It includes a photovoltaic array, which is responsible for converting solar energy into DC power. The photovoltaic array consists of multiple photovoltaic modules, each of which can generate electricity through the photovoltaic effect; the energy storage system, including batteries and supercapacitors, is used to store excess electricity or provide additional energy support when needed. Among them, the battery is suitable for energy regulation in a steady state, while the supercapacitor handles transient power requirements; the DC / DC conversion circuit is used to connect the photovoltaic array with the DC bus, and the energy storage system with the DC bus to achieve DC / DC conversion, and can control the direction of power flow; the three-phase full-bridge inverter performs DC / AC conversion, that is, converts the DC power from the DC bus into AC power suitable for grid connection; the main controller uses the super spiral sliding mode control (STSMC) algorithm to accurately adjust the load voltage to ensure that it is maintained near a constant reference value. The controller is also responsible for communicating with other subsystems, such as receiving instructions from the upper dispatch center; the slave controller uses the sliding membrane control algorithm (SMC) to work according to the reference information set by the master controller; the ZigBee communication network uses ZigBee wireless communication equipment to transmit data and commands, such as current reference signals, to form a networked control system. This helps to improve the reliability and flexibility of the system; the transmission line includes the necessary resistance and inductance components to build the actual physical connection, and the impact of these components on the dynamic behavior of the system is also considered in the model.

[0022] The main unit consists of an energy storage system, a supercapacitor, a DC / DC conversion circuit and a three-phase full-bridge inverter circuit. The DC / DC conversion circuit uses a bidirectional Buck / Boost circuit. When the energy storage system is connected to the grid through a bidirectional Buck / Boost circuit, the first stage of the circuit is used for voltage regulation. The bidirectional Buck / Boost circuit consists of an inductor, two switching tubes, two diodes and a filter capacitor. Its core function is to adjust the DC voltage from the battery to a level suitable for grid connection requirements, which can be increased from low voltage to high voltage (boost mode) and can also be reduced from high voltage to low voltage (buck mode). The three-phase full-bridge inverter circuit consists of 12 power switching devices, which are alternately turned on and off in a specific order to convert the DC power regulated by the bidirectional Buck / Boost circuit into a three-phase sinusoidal AC power. In order to achieve high-quality AC output.

[0023] The slave unit consists of a photovoltaic power generation system, an inverter, and a power control circuit. The distributed photovoltaic power generation system will be connected to the grid through a two-stage inverter circuit. The boost converter is the first stage of the two-stage inverter circuit, which consists of an inductor, a switch tube, a diode, and a filter capacitor. The main function of the boost converter is to increase the low DC voltage from the photovoltaic array to a higher DC voltage level in order to provide sufficient voltage support for subsequent DC / AC inversion. The three-phase three-level full-bridge inverter is the second stage of the two-stage inverter circuit and consists of 12 power switching devices. These switches are alternately turned on and off in a specific order to convert DC power into three-phase or multi-phase sinusoidal AC power. Sinusoidal pulse width modulation or space vector pulse width modulation is used to generate a near-ideal sinusoidal waveform, reduce harmonic distortion, and improve power quality.

[0024] ZigBee network is a low-power, short-range wireless communication technology built on the IEEE 802.15.4 standard. It is designed for low-cost and low-data-rate applications. There are three frequency bands available, namely the 2.4GHz ISM band, the 868MHz band in Europe, and the 915MHz band in the United States. The channels available in different frequency bands are 16, 1, and 10, respectively. The 2.4G band is used in China. Its characteristics are that the device can work in sleep mode to extend battery life, and the hardware implementation is simple to reduce costs. A ZigBee network can support up to 65,000 nodes, has a high degree of security and self-organization capabilities, and can automatically discover new nodes and dynamically adjust routes.

[0025] The architecture of the ZigBee network is divided into four layers, each responsible for different functions. The ZigBee protocol stack is composed of the physical layer, media access control layer, network layer, and application layer from bottom to top. The physical layer defines basic parameters such as the selection of radio frequency bands, modulation methods, and data transmission rates. The detailed parameters are shown in Table 1. The MAC layer is responsible for access control and manages shared channel access between multiple nodes through the CSMA / CA mechanism to prevent data packet collisions and optimize communication efficiency. The network layer supports three topological structures: star, tree, and mesh, and is responsible for routing selection, address management, and calculation of the maximum allowable delay boundary. The application layer provides end-to-end data transmission services and processes high-level protocol elements such as cluster link tables (Cluster ID).

[0026] When networking, the host first starts manual networking, reads the local configuration (command 0*D1), then modifies the corresponding fields in DEV_INFO, including the working type, channel number and network number, and saves the modified configuration through command 0*D6. Next, the host sends a reset command (0*D0) and waits for 1 second, then queries the host status (command 0*E8). If the status is 0*02 or 0*00, it means the device is ready; if the status is 0*01, it means the networking is complete. The entire process ensures that the device can successfully join the network and complete the networking after correct configuration. Figure 1 See the ZigBee networking method corresponding to the example. Figure 2 The upper coordination controller (master unit) transmits current reference signals to each slave controller through the ZigBee network. These signals include the reference values ​​of the d-axis and q-axis current components, ensuring that the slave controller can adjust the output current according to the reference value to achieve reasonable power distribution. At the same time, the master controller also sends control signals to the slave controller through ZigBee to instruct it to perform specific operations and receive status feedback from the slave controller.

[0027] Step 3: Master-slave controller super-helical sliding mode controller design First, establish the system state equation. The following formula represents the dynamic equation of the DG unit in the stationary abc coordinate system: Where: V abc ——Load voltage; V t,abc ——Inverter output voltage; i t,abc ——Inverter output current; i L,abc ——Load current; L t ——The sum of the inductance of the transmission line and the filter; R t ——The sum of the inductance and resistance of the transmission line and the filter; R——load resistance; L——load inductance; C——load capacitance. l ——Load inductance branch resistance.

[0028] The state space equation of DG access system: in: Where: x = [V d , V q , I td , I tq , I Ld , I Lq ] T ——state vector; u=[V td , V tq ]T ——Input control vector; y = [V d , V q ] T ——The output vector of the main DG unit; y = [I tds , I tqs ] T ——output vector from DG unit; w(t)——fluctuation caused by PV and energy storage system.

[0029] Then, the control stability caused by Zigbee communication delay is analyzed below. When time delay occurs in the microgrid control system, the system will become a time-varying system. The maximum allowable delay (MADB) is defined as the upper limit of the maximum allowable delay to ensure the stable operation of the system, and is the most common criterion for the stability conclusion of the time-delay system. For the stability analysis of the time-delay system, the most common method is to use the direct construction of the universal function (Lyapunov-Krasovskii, LK) in the time domain and combine it with the linear matrix inequality (Linear Matrix Inequality, LMI) to achieve it. For its positive definite problem, it is generally implemented in the form of state feedback.

[0030] In general, the state-space expression of a system in the presence of delay can be shown as the following small signal model: Where: A, A d ——fixed matrix of appropriate dimension;Φ(t)——initial condition of the system in the time period t∈[-ρ,0];τ(t)——time-varying delay.

[0031] The following conditions are met: Assume that scalar ρ>0, μ>0, when there exists a symmetric positive definite matrix P=P T >0, Q=Q T >0 and Z=Z T >0, nonsymmetric positive semidefinite matrix:

[0032] and arbitrary matrices Y and T, the time-delay system (4) is asymptotically stable when the following LMI holds: Where:

[0033] In order to study the stability of microgrids with communication delay, LMI must be obtained for linear time-varying controllers. In the solution process, time delay can be defined as the only variable in LMI to ensure stability. The binary iterative algorithm is used to calculate MADB in the LMI solution step. The main purpose is to ensure the stable operation of the system when the delay is greater than MADB.

[0034] The solution steps are: (1) Obtain the system matrix from the microgrid parameters to calculate A and A d .

[0035] (2) to determine whether there is a feasible solution for LMI under a given time delay.

[0036] (3) Use the binary iterative algorithm to find the system's MADB: define an interval [τ u ,τ f ], where τ u represents the stabilization delay, τ f represents unstable delay. Then τ n =(τ u +τ f ) / 2, then according to τ n Reduce the delay bound if τ n is a stable delay, then τ u =τ n ; If unstable, then τ f =τ n . Iterate until the accuracy requirement is met: τ ac =|τ f -τ u |<ε, where ε is the minimum error.

[0037] Step 2: Design the main unit voltage controller based on improved spiral sliding mode control. The control goal of the main DG unit controller is to control the load voltage V abc By tracking its reference value, the voltage can be quickly adjusted to a stable operating state even if there are fluctuations caused by external interference d(x) and load changes. d and the quadrature axis component V q The corresponding reference value V d,ref 、V q,ref The error between d 、e q As the input of the main controller, the corresponding adjustment is carried out. The structural block diagram of the main controller design is as follows Figure 3 shown.

[0038] The d, q axis voltage tracking error is as follows (9): Where: Vd,ref ——reference value of voltage d-axis component; V q,ref ——reference value of voltage q-axis component; V d (t)——d-axis component of the output voltage of the main controller; V q (t)——q-axis component of the main controller output voltage.

[0039] It can be seen from (9) that the system is a second-order system. By defining the auxiliary variable e d,1 =e d , e q,1 =e q The second-order auxiliary system with external disturbance is expressed as follows: Where: u i ——the controlled signal of the main DG unit in the dq coordinate system; d(x)——external bounded disturbance (|d(x)|≤l g )(where l g is a constant).

[0040] Where: f i and g i as follows

[0041] The present invention uses a second-order sliding mode control (super twisting sliding mode control, STSMC) based on a super twisting algorithm to design the control variable u i , let the sliding surface be s, satisfying the following equation: In the formula: s is the sliding mode variable; v is the estimated value of the acceleration of the control law; λ, α are the control parameters to be designed; sign(s) is the sign function.

[0042] The parameters of the sliding mode controller are adjusted by the following formula:

[0043] To further reduce chattering, the discontinuous sign function sign(s) in the superhelical controller is replaced by a continuous hyperbolic tangent function. In other words, the hyperbolic tangent function is used as an approximation of the sign function. The expression is as follows:

[0044] At this time, formula (12) becomes: in Determines the speed at which the limit is reached.

[0045] The designed sliding surface is:

[0046] At this time, the control quantity u is derived i As shown below:

[0047] Step 3: Design a slave unit power tracking controller based on improved sliding mode control to adjust the output current i of the slave DG unit. t,abcs Track the reference current to reasonably distribute the active and reactive power required in the system. The reference current is transmitted from the upper coordination controller to each slave controller via the communication network.

[0048] The present invention converts the direct axis component I of the current output by each DG unit ds and the quadrature axis component I qs Corresponding to reference value I ds,ref ,I qs,ref The error between ds 、e qs As the input of the slave controller, the block diagram of the slave controller design is as follows Figure 4 shown.

[0049] The d and q axis current tracking errors are given by equation (18): Where: I ds,ref ——reference value of current d-axis component; I qs,ref ——reference value of current q-axis component; I ds (t)——d-axis component of the current output from the controller; I qs (t)——q-axis component of the current output from the controller.

[0050] According to the formula, the DG unit belongs to a first-order system. For a first-order system, the conventional sliding mode controller (SMC) can also show sufficiently good control performance. A first-order system with external disturbances can be expressed as: Where: u i ——dq component of the inverter output voltage from the DG; d(x) is defined in (8).

[0051] f i and g i The calculation is as follows:

[0052] In the first-order sliding mode control system, let the sliding surface be s, satisfying the following equation:

[0053] At this time, the control quantity u is In the slave controller it can be represented as: Where u is ——Control signal from DG unit; K i ——Sliding gain from DG unit.

[0054] Finally, simulation experiments verify the effectiveness and superiority of the invented method. The present invention uses the power simulation toolbox and the communication simulation toolbox to build a master-slave photovoltaic-storage-charging microgrid model. Figure 1 The example builds a microgrid model with a communication network.

[0055] The main controller uses STSMC controller to compare with SMC and PI, and the slave controller uses SMC to compare with PI. The ZigBee wireless communication network is used to analyze its impact on the stability of the system. The specific communication network and microgrid example parameters are shown in Tables 1 and 2.

[0056] Table 1 ZigBee network parameters

[0057] Table 2 Parameters of the photovoltaic-storage-charging microgrid

[0058] In this example, ZigBee has a conflict avoidance feature. Carrier Sense Multiple Access (CSMA / CA). Its characteristic is that when there is a queuing error in the sent data, the control center will wait for a period of time before resending the information, and the slave controller will receive the information with delay. In order to study the performance of the controller, multiple nodes are added to the communication channel. As shown in Table 1, the sum of the transmission delay and sampling time in the wireless network is 2.64ms, in addition to the device acceptance delay, search delay, etc., the total delay is about 30ms. Figure 5 Given in.

[0059] It can be seen that there are obvious queuing errors in the first, third and fifth channels, and this situation increases with the distance. Due to queuing delays, collisions and frequent packet losses, there is increased delay in the network and the value of the reference signal is disturbed.

[0060] The proposed STSMC is compared with SMC and PI respectively. The dq axis components of the main unit voltage of the conventional PI control and the invented control method are shown in Figure 6 As shown, Figure 6 Figure a is a comparison of different control methods for the d-axis component of the main DG output voltage in a ZigBee network, and Figure b is a comparison of different control methods for the q-axis component of the main DG output voltage in a ZigBee network. Figure 6 It can be seen that when there is a large delay in the network, the voltage d-axis component of the PI control before the load is switched fluctuates at ±5V; the q-axis component fluctuates at ±2V; and the response time of the controller becomes longer to 1s. The voltage d-axis component of the SMC fluctuates at ±0.8V; the q-axis component fluctuates at ±0.3V. The voltage d-axis component of the STSMC fluctuates at ±0.05V; the q-axis component fluctuates at ±0.06V, and the fluctuation range remains basically unchanged. At the moment of load switching, the voltage d-axis component of the PI control drops to 295V, and the SMC control drops to 287V. The STSMC control drops to 300V. After analysis, it can be seen that the proposed STSMC control strategy has better resistance performance when facing increased network delays.

[0061] Figure 7 The dq components of the output current from the DG under conventional PI control and the invented control method. The ZigBee network has a large delay, so the current waveform in the slave controller jitters due to interference. By comparing SMC and PI, it can be seen that the response speed of the SMC controller is much faster than that of the PI controller. In this state, the output current waveform under the PI controller loses stability.

[0062] Figure 8 The phase diagram of the tracking error of the master and slave controllers in the ZigBee network is shown. As can be seen from the figure. For the master controller (STSMC), the tracking error and time derivative are not greatly affected and converge to the equilibrium point within a finite time. For the slave controller (SMC), the rate of change of the d-axis component error converges to -10V·s -1 ; The error change rate of the q-axis component converges to -6V·s -1 , the errors of dq axis components are all around 0.1 V. It can be seen that ZigBee time delay has a certain impact on the slave controller.

[0063] Fig. 9 Figure 3. The bus three-phase voltage and load current diagram of the microgrid. Through the coordinated control of the master DG unit controller and the slave controller, the bus voltage is stable, the load current is stable, and a good control effect is achieved. The delay of the Zigee communication network will affect the current distribution of each slave DG. However, the master DG unit controller can make up for the defects caused by the delay of the slave controller, so that the total bus voltage and load current in the microgrid can still reach a stable state.

Claims

1. A super-spiral sliding mode master-slave voltage control system for a photovoltaic energy storage microgrid under a ZigBee communication network, characterized in that: Design of super-helical sliding mode controller for master-slave controller in master-slave voltage control system: S1. Establish the system state equation: The dynamic equation of the DG unit in the stationary abc coordinate system: Where: V abc is the load voltage; V t,abc is the inverter output voltage; i t,abc is the inverter output current; i L,abc is the load current; L t is the sum of the inductance of the transmission line and the filter; R t is the sum of the inductance and resistance of the transmission line and the filter; R is the load resistance; L is the load inductance; C is the load capacitance, R l is the load inductance branch resistance; The state space equation of DG access system: in: Where: x = [V d , V q , I td , I tq , I Ld , I Lq ] T is the state vector; u=[V td , V tq ] T is the input control vector; y=[V d , V q ] T is the output vector of the main DG unit; y = [I tds , I tqs ] T is the output vector from the DG unit; w(t) is the fluctuation caused by the photovoltaic and energy storage systems; S2, control stability caused by Zigbee communication delay: The state space expression of the system when there is a delay is a small signal model: Where: A, A d is a fixed matrix of appropriate dimension; Φ(t) is the initial condition of the system in the time period t∈[-ρ,0]; τ(t) is the time-varying delay; The time-varying delay τ(t) satisfies the condition: Assume that scalar ρ>0, μ>0, when there exists a symmetric positive definite matrix P=P T >0, Q=Q T >0 and Z=Z T >0, nonsymmetric positive semidefinite matrix: and arbitrary matrices Y and T, the time-delay system (4) is asymptotically stable when the following LMI holds: Where: S3. Steps to solve LMI: S31. Obtain the system matrix from the microgrid parameters to calculate A and A d . S32, to determine whether there is a feasible solution for LMI under a given time delay. S33, use the binary iterative algorithm to find the system's MADB: define an interval [τ u ,τ f ], where τ u represents the stabilization delay, τ f represents unstable delay, then τ n =(τ u +τ f ) / 2, then according to τ n Reduce the delay bound if τ n is a stable delay, then τ u =τ n ; If unstable, then τ f =τ n , iterate until the accuracy requirement is met: τ ac =|τ f -τ u |<ε, where ε is the minimum error.

2. The super-spiral sliding mode master-slave voltage control system of a photovoltaic energy storage microgrid under a ZigBee communication network according to claim 1 is characterized in that: Design of the main unit voltage controller based on the master-slave controller super-helical sliding mode controller: The direct-axis component of the load voltage V d and the quadrature axis component V q The corresponding reference value V d,ref 、V q,ref The error between d 、e q As the input of the main controller, it is adjusted accordingly; d, q axis voltage tracking error: Where: V d,ref is the voltage d-axis component reference value; V q,ref is the reference value of the voltage q-axis component; V d (t) is the d-axis component of the output voltage of the main controller; V q (t) is the q-axis component of the output voltage of the main controller; Define auxiliary variable e d,1 =e d , e q,1 =e q A second-order auxiliary system with external disturbances: Where: u i is the controlled signal of the main DG unit in the dq coordinate system; d(x) is the external bounded disturbance; Where: f i and g i as follows The control variable u is designed using the second-order sliding mode control of the superhelical algorithm i , let the sliding surface be s, satisfying the following equation: Where: s is the sliding mode variable; v is the estimated value of the acceleration of the control law, λ and α are the control parameters to be designed; sign(s) is the sign function; Adjust the parameters of the sliding mode controller: Replace the discontinuous sign function sign(s) in the superhelical controller with a continuous hyperbolic tangent function: At this time, formula (12) becomes: in Determines the speed at which the limit is reached; The designed sliding surface is: At this time, the control quantity u i as follows:

3. The super spiral sliding mode master-slave voltage control system of the photovoltaic energy storage microgrid under the ZigBee communication network according to claim 1 is characterized in that: Design of slave unit power tracking controller based on master-slave controller super spiral sliding mode controller: The direct axis component I of the current output by each DG unit ds and the quadrature axis component I qs Corresponding to reference value I ds,ref ,I qs,ref The error between ds 、e qs Regulated as input from the controller; d, q axis current tracking error: Where: I ds,ref is the reference value of the current d-axis component; I qs,ref is the reference value of the current q-axis component; I ds (t) is the d-axis component of the current output from the controller; I qs (t) is the q-axis component of the current output from the controller; First-order system with external disturbance: Where: u i is the dq component of the inverter output voltage from the DG; f i and g i as follows: In the first-order sliding mode control system, let the sliding surface be s, satisfying the following equation: At this time, the control quantity u is The expression in the slave controller is: Where u is is the control signal from the DG unit; K i is the sliding gain from the DG unit.

Images