Grid-connected inverter phase-locked loop method and system based on model predictive control

Abstract

This invention provides a phase-locked loop (PLL) method and system for grid-connected inverters based on model predictive control (MMC). The method includes: replacing the PI controller with MMC based on the PLL's structural model and its output phase characteristics; selecting a cost function and determining the control set (i.e., the phase change value) within each calculation cycle based on the control step size of the actual control system or the simulation step size of the simulation system; calculating the cost function values ​​under different phase angle changes in the control set; selecting the phase angle increment corresponding to the minimum cost function as the phase angle change value, thus achieving successful phase locking. This invention directly provides a universal basis for selecting control set parameters by combining the control step size or simulation step size, eliminating the need to consider other system parameters besides the step size to obtain the PLL system, thus avoiding cumbersome parameter design; and replacing the PI controller with MMC eliminates the influence of PI parameters on the PLL control performance, achieving the same phase-locking effect as traditional PLLs without the need for PI controllers.

Description

Technical Field

[0001] This invention relates to the technical field of AC power generation, transmission, distribution and consumption, and specifically to a phase-locked loop method and system for grid-connected inverters based on model predictive control. Background Technology

[0002] The high penetration of new energy power generation into the power system has become a major characteristic of the new power system. This portion of electricity needs to be connected to the grid through power electronic conversion equipment such as inverters. Therefore, as the main interface for interaction between new energy and the grid, the stable operation and control performance of the inverter directly determine whether renewable energy can be utilized efficiently.

[0003] Inverters, as crucial devices for power conversion, comprise main circuits and control sections. Common control strategies include vector current loops (VPLs) and phase-locked loops (PLLs). The VPL controls the inverter's output power, while the PLL acquires grid voltage to obtain phase information, synchronizing the inverter's output voltage with the grid voltage. Therefore, the PLL is critical in inverter control. Among various PLLs, the Synchronous Reference Frame Phase-locked Loop (SRF-PLL) is widely used in engineering practice. The SRF-PLL first detects the phase difference between two input signals using a phase detector, converts the phase error into a voltage output, and then uses a proportional-integral (PI) controller to filter noise and high-frequency signals from the phase detector's output voltage. Finally, the output of the loop filter is input to a voltage-controlled oscillator (VCO), ensuring that the frequency of the input signal in the phase detector perfectly matches the frequency of the output signal in the loop filter, thus achieving phase locking.

[0004] In SRF-PLL, a PI controller is used to lock onto the grid frequency and phase, so the PI parameters have a significant impact on the steady-state and dynamic performance of the phase-locked loop. However, due to the continuous changes in system parameters and operating conditions, the PI in SRF-PLL cannot guarantee system stability reliably. Increasing the PLL bandwidth by designing the PI parameters can achieve rapid synchronization between the inverter and the grid, but it increases the risk of subsynchronous / supersynchronous oscillations, threatening the stable operation and grid connection of the inverter. Conversely, reducing the PLL bandwidth by designing the PI parameters will decrease the phase-locking speed and dynamic response speed, and may even affect low-voltage ride-through. Therefore, adjusting the PI parameters requires extensive theoretical design and practical testing, combined with the actual system model and parameters, to achieve rapid phase-locking while ensuring system stability. The parameter selection process is quite complex.

[0005] In summary, most commonly used phase-locked loops (PLLs) are based on PI controllers to achieve phase-locking and synchronization of the power grid. Their control performance is highly dependent on the PI parameters, which makes PI parameter design difficult and makes it impossible to simultaneously achieve both speed and stability.

[0006] Xu, Zhiye. Robust Model Predictive Control for Grid-Connected Inverters Based on Improved Phase-Locked Loop and Inductor Identification [D]. Zhengzhou University of Light Industry, 2022.

[0007] Abstract: Grid-connected inverter technology, as the core of new energy power generation, is increasingly widely used in new energy grid-connected power generation systems. How to quickly and accurately identify circuit parameters and rationally improve inverter control technology has become a key technical challenge that urgently needs to be solved. This paper focuses on robust predictive model control of grid-connected inverters. It improves the robustness of system parameters from multiple aspects, thereby enhancing the control performance of grid-connected inverters. The research content includes: traditional robust model predictive control, improved phase-locked loop (PLL) design without a PI controller, research on improved inductor identification methods, and robust model predictive control based on improved PLL and inductor identification. Addressing the problems of traditional PLLs relying on PI controllers, difficult debugging, and weak adaptability to frequency changes, an improved PI-controller-free PLL is proposed. It requires no PI controller adjustment, has a very fast dynamic convergence speed, high estimation accuracy, and exhibits strong adaptability to sudden changes in grid frequency.

[0008] This paper improves upon the iterative search strategy based on the bisection method by performing multiple bisections within the range of 0 to 2π, selecting the phase angle with the smaller cost function as the optimal phase estimation angle each time, achieving an accuracy of π / 2n+1 in n iterations. However, this paper searches for the phase angle rather than the phase change, resulting in a larger search range. Furthermore, obtaining the phase angle at each moment requires multiple bisections and calculations, which increases computational load, reduces operational efficiency, and places higher demands on the computing speed of the controller hardware to achieve the required accuracy.

[0009] Therefore, a new technical solution is needed to improve the above-mentioned technical problems. Summary of the Invention

[0010] To address the shortcomings of existing technologies, the purpose of this invention is to provide a phase-locked loop method and system for grid-connected inverters based on model predictive control.

[0011] According to the present invention, a phase-locked loop method for grid-connected inverters based on model predictive control is provided, the method comprising the following steps:

[0012] Step S1: Combining the structural model of the phase-locked loop and the phase characteristics of its output, model predictive control is used to replace PI control;

[0013] Step S2: Select the cost function and determine the control set, i.e. the magnitude of the phase change value, within each calculation cycle based on the control step size of the actual control system or the simulation step size of the simulation system.

[0014] Step S3: Calculate the cost function values ​​under different phase angle changes in the control set, select the phase angle increment corresponding to the minimum cost function as the phase angle change value, and perform phase locking.

[0015] Preferably, step S1 includes phase-locked loop modeling and output phase angle characteristic analysis:

[0016] The grid voltage signal input to the phase-locked loop (PLL) can be written in abc coordinate system as follows:

[0017]

[0018] Among them, u a u b and u c Let U be the three-phase grid voltage (abc), U1 be the amplitude of the input voltage signal, and θ be the phase of the grid voltage.

[0019] After Clark transformation, the grid voltage in the two-phase stationary αβ coordinate system can be written as:

[0020]

[0021] Among them, u α u β The voltage components are the αβ coordinate axes;

[0022] The voltage in the synchronously rotating dq coordinate system is written as:

[0023]

[0024] Among them, u d u q Let θ be the voltage component of the dq coordinate axis. PLL This is the phase angle output of the phase-locked loop.

[0025] Preferably, step S2 includes the implementation process of a grid-connected inverter phase-locked loop system based on model predictive control:

[0026] When the phase-locked loop system can successfully lock the phase, the magnitude of the q-axis component of the grid voltage is 0. Based on this characteristic of the phase-locked loop, the cost function g in model predictive control is defined as the magnitude of the q-axis component of the grid voltage.

[0027] The power grid phase angle changes continuously from 0 to 2π within one power frequency cycle, i.e., 0.02s. If the phase-locked loop can successfully lock the phase without any sudden changes in phase angle, the formula for calculating the phase angle increment within one control step is:

[0028]

[0029] Among them, T s To calculate the step size or simulation step size, derta_theta is the phase angle increment within a control step size; the phase angle increment derta_theta is 100πT. s If the grid voltage leads or lags in phase at a certain moment, the phase angle will change abruptly, corresponding to a vertical shift in the phase-locked loop (PLL) output phase curve. During this transient process, if the PLL phase angle change within a control step is still 100πT... s Phase-locked loop fails; add an integer multiple of derta_theta to the control set; add the 0 option to the control set.

[0030] Preferably, in step S3, the control set parameters are selected, that is, the phase angle change value within one control step is: [0, 400πT] s 200πT s 100πT s -100πT s -200πT s -400πT s There are a total of 7 sets of data. In each control step, the q-axis component magnitude value corresponding to the above 7 phase angle increments is calculated, that is, the cost function value. The phase angle change corresponding to the minimum cost function is selected and added to the original phase angle to obtain the phase-locked loop phase angle output in this control step.

[0031] Preferably, the αβ axis components of the input grid voltage are used to initialize T. s The control set parameters are [0, π / 100, π / 200, π / 400, -π / 400, -π / 200, -π / 100], g_opt is 1e10, and x_opt is 1. The q-axis component magnitude corresponding to different phase angle increments in the control set is calculated by formula (3), which is the cost function value. If the cost function value corresponding to the i-th phase angle increment is less than the existing optimal cost function value, it is used to replace the optimal cost function value, and the optimal control set index x_opt is recorded as i. The x_opt-th phase angle increment in the control set parameters is added to the original output phase angle, which can be used as the output phase angle of the phase-locked loop in the current calculation cycle, and the phase is successfully locked.

[0032] The present invention also provides a grid-connected inverter phase-locked loop system based on model predictive control, the system comprising the following modules:

[0033] Module M1: Combining the structural model of the phase-locked loop and the phase characteristics of its output, model predictive control is used to replace PI control;

[0034] Module M2: Selects the cost function and determines the control set, i.e. the magnitude of the phase change value, within each calculation cycle based on the control step size of the actual control system or the simulation step size of the simulation system.

[0035] Module M3: Calculates the cost function values ​​under different phase angle changes in the control set, selects the phase angle increment corresponding to the minimum cost function as the phase angle change value, and performs phase locking.

[0036] Preferably, module M1 includes a phase-locked loop model and output phase angle characteristic analysis:

[0037] The grid voltage signal input to the phase-locked loop (PLL) can be written in abc coordinate system as follows:

[0038]

[0039] Among them, u a u b and u c Let U be the three-phase grid voltage (abc), U1 be the amplitude of the input voltage signal, and θ be the phase of the grid voltage.

[0040] After Clark transformation, the grid voltage in the two-phase stationary αβ coordinate system can be written as:

[0041]

[0042] Among them, u α u β The voltage components are the αβ coordinate axes;

[0043] The voltage in the synchronously rotating dq coordinate system is written as:

[0044]

[0045] Among them, u d u q Let θ be the voltage component of the dq coordinate axis. PLL This is the phase angle output of the phase-locked loop.

[0046] Preferably, module M2 includes the implementation process of a grid-connected inverter phase-locked loop system based on model predictive control:

[0047] When the phase-locked loop system can successfully lock the phase, the magnitude of the q-axis component of the grid voltage is 0. Based on this characteristic of the phase-locked loop, the cost function g in model predictive control is defined as the magnitude of the q-axis component of the grid voltage.

[0048] The power grid phase angle changes continuously from 0 to 2π within one power frequency cycle, i.e., 0.02s. If the phase-locked loop can successfully lock the phase without any sudden changes in phase angle, the formula for calculating the phase angle increment within one control step is:

[0049]

[0050] Among them, T s To calculate the step size or simulation step size, derta_theta is the phase angle increment within a control step size; the phase angle increment derta_theta is 100πT. s If the grid voltage leads or lags in phase at a certain moment, the phase angle will change abruptly, corresponding to a vertical shift in the phase-locked loop (PLL) output phase curve. During this transient process, if the PLL phase angle change within a control step is still 100πT... s Phase-locked loop fails; add an integer multiple of derta_theta to the control set; add the 0 option to the control set.

[0051] Preferably, in module M3, the control set parameters are selected, that is, the phase angle change value within one control step is: [0, 400πT] s 200πT s 100πT s -100πT s -200πT s -400πT s There are a total of 7 sets of data. In each control step, the q-axis component magnitude value corresponding to the above 7 phase angle increments is calculated, that is, the cost function value. The phase angle change corresponding to the minimum cost function is selected and added to the original phase angle to obtain the phase-locked loop phase angle output in this control step.

[0052] Preferably, the αβ axis components of the input grid voltage are used to initialize T. s The control set parameters are [0, π / 100, π / 200, π / 400, -π / 400, -π / 200, -π / 100], g_opt is 1e10, and x_opt is 1. The q-axis component magnitude corresponding to different phase angle increments in the control set is calculated by formula (3), which is the cost function value. If the cost function value corresponding to the i-th phase angle increment is less than the existing optimal cost function value, it is used to replace the optimal cost function value, and the optimal control set index x_opt is recorded as i. The x_opt-th phase angle increment in the control set parameters is added to the original output phase angle, which can be used as the output phase angle of the phase-locked loop in the current calculation cycle, and the phase is successfully locked.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] 1. This invention introduces Model Predictive Control (MPC) into the implementation process of phase-locked loop (PLL), and achieves fast phase-locking of grid voltage by solving for the optimal result within a finite time and a finite control set;

[0055] 2. This invention replaces PI with model predictive control, avoiding the cumbersome design of PI parameters, and rationally designs the control set parameters according to the control step size, simplifying the calculation complexity and improving the operating efficiency.

[0056] 3. This invention combines the control step size or simulation step size to directly provide a universal basis for selecting control set parameters. The phase-locked loop system can be obtained without considering other system parameters besides the step size, thus avoiding cumbersome parameter design.

[0057] 4. This invention uses model predictive control to replace the PI controller, eliminating the influence of PI parameters on the performance of the phase-locked loop control. Even without PI, it can achieve the same phase-locking effect as the traditional phase-locked loop, and can achieve accurate phase-locking even if there is a sudden change in the grid phase. Attached Figure Description

[0058] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0059] Figure 1 This is a voltage vector diagram in different coordinate systems according to the present invention;

[0060] Figure 2 This is a block diagram of the model predictive control phase-locked loop control of the present invention;

[0061] Figure 3 This is a flowchart of the phase-locked loop system based on model predictive control according to the present invention;

[0062] Figure 4 This is a diagram showing the phase-locked loop (PLL) results of the present invention.

[0063] Figure 5 This is a diagram showing the phase-locked loop results of the present invention under the condition of a sudden change in phase angle. Detailed Implementation

[0064] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.

[0065] Example 1:

[0066] According to the present invention, a phase-locked loop method for grid-connected inverters based on model predictive control is provided, the method comprising the following steps:

[0067] Step S1: Combining the structural model of the phase-locked loop and the phase characteristics of its output, model predictive control is used to replace PI control;

[0068] Phase-locked loop (PLL) model and output phase angle characteristic analysis:

[0069] The grid voltage signal input to the phase-locked loop (PLL) can be written in abc coordinate system as follows:

[0070]

[0071] Among them, u a u b and u c Let U be the three-phase grid voltage (abc), U1 be the amplitude of the input voltage signal, and θ be the phase of the grid voltage.

[0072] After Clark transformation, the grid voltage in the two-phase stationary αβ coordinate system can be written as:

[0073]

[0074] Among them, u α u β The voltage components are the αβ coordinate axes;

[0075] The voltage in the synchronously rotating dq coordinate system is written as:

[0076]

[0077] Among them, u d u q Let θ be the voltage component of the dq coordinate axis. PLL This is the phase angle output of the phase-locked loop.

[0078] Step S2: Select the cost function and determine the control set, i.e. the magnitude of the phase change value, within each calculation cycle based on the control step size of the actual control system or the simulation step size of the simulation system.

[0079] Implementation process of grid-connected inverter phase-locked loop system based on model predictive control:

[0080] When the phase-locked loop system can successfully lock the phase, the magnitude of the q-axis component of the grid voltage is 0. Based on this characteristic of the phase-locked loop, the cost function g in model predictive control is defined as the magnitude of the q-axis component of the grid voltage.

[0081] The power grid phase angle changes continuously from 0 to 2π within one power frequency cycle, i.e., 0.02s. If the phase-locked loop can successfully lock the phase without any sudden changes in phase angle, the formula for calculating the phase angle increment within one control step is:

[0082]

[0083] Among them, T s To calculate the step size or simulation step size, derta_theta is the phase angle increment within a control step size; the phase angle increment derta_theta is 100πT. s If the grid voltage leads or lags in phase at a certain moment, the phase angle will change abruptly, corresponding to a vertical shift in the phase-locked loop (PLL) output phase curve. During this transient process, if the PLL phase angle change within a control step is still 100πT... s Phase-locked loop fails; add an integer multiple of derta_theta to the control set; add the 0 option to the control set.

[0084] Step S3: Calculate the cost function values ​​under different phase angle changes in the control set, select the phase angle increment corresponding to the minimum cost function as the phase angle change value, and perform phase locking.

[0085] Select the control set parameters, i.e., the phase angle change value within one control step: [0, 400πT] s 200πT s 100πT s -100πT s -200πT s -400πT s There are a total of 7 sets of data. In each control step, the q-axis component magnitude value corresponding to the above 7 phase angle increments is calculated, that is, the cost function value. The phase angle change corresponding to the minimum cost function is selected and added to the original phase angle to obtain the phase-locked loop phase angle output in this control step.

[0086] Input the αβ axis components of the mains voltage and initialize T. s The control set parameters are [0, π / 100, π / 200, π / 400, -π / 400, -π / 200, -π / 100], g_opt is 1e10, and x_opt is 1. The q-axis component magnitude corresponding to different phase angle increments in the control set is calculated by formula (3), which is the cost function value. If the cost function value corresponding to the i-th phase angle increment is less than the existing optimal cost function value, it is used to replace the optimal cost function value, and the optimal control set index x_opt is recorded as i. The x_opt-th phase angle increment in the control set parameters is added to the original output phase angle, which can be used as the output phase angle of the phase-locked loop in the current calculation cycle, and the phase is successfully locked.

[0087] Example 2:

[0088] Example 2 is a preferred example of Example 1, and is used to illustrate the present invention in more detail.

[0089] This invention also provides a grid-connected inverter phase-locked loop system based on model predictive control, the system comprising the following modules:

[0090] Module M1: Combining the structural model of the phase-locked loop and the phase characteristics of its output, model predictive control is used to replace PI control;

[0091] Phase-locked loop (PLL) model and output phase angle characteristic analysis:

[0092] The grid voltage signal input to the phase-locked loop (PLL) can be written in abc coordinate system as follows:

[0093]

[0094] Among them, u a u b and u c Let U be the three-phase grid voltage (abc), U1 be the amplitude of the input voltage signal, and θ be the phase of the grid voltage.

[0095] After Clark transformation, the grid voltage in the two-phase stationary αβ coordinate system can be written as:

[0096]

[0097] Among them, u α u β The voltage components are the αβ coordinate axes;

[0098] The voltage in the synchronously rotating dq coordinate system is written as:

[0099]

[0100] Among them, u d u q Let θ be the voltage component of the dq coordinate axis. PLL This is the phase angle output of the phase-locked loop.

[0101] Module M2: Selects the cost function and determines the control set, i.e. the magnitude of the phase change value, within each calculation cycle based on the control step size of the actual control system or the simulation step size of the simulation system.

[0102] Implementation process of grid-connected inverter phase-locked loop system based on model predictive control:

[0103] When the phase-locked loop system can successfully lock the phase, the magnitude of the q-axis component of the grid voltage is 0. Based on this characteristic of the phase-locked loop, the cost function g in model predictive control is defined as the magnitude of the q-axis component of the grid voltage.

[0104] The power grid phase angle changes continuously from 0 to 2π within one power frequency cycle, i.e., 0.02s. If the phase-locked loop can successfully lock the phase without any sudden changes in phase angle, the formula for calculating the phase angle increment within one control step is:

[0105]

[0106] Among them, T s To calculate the step size or simulation step size, derta_theta is the phase angle increment within a control step size; the phase angle increment derta_theta is 100πT. s If the grid voltage leads or lags in phase at a certain moment, the phase angle will change abruptly, corresponding to a vertical shift in the phase-locked loop (PLL) output phase curve. During this transient process, if the PLL phase angle change within a control step is still 100πT... s Phase-locked loop fails; add an integer multiple of derta_theta to the control set; add the 0 option to the control set.

[0107] Module M3: Calculates the cost function values ​​under different phase angle changes in the control set, selects the phase angle increment corresponding to the minimum cost function as the phase angle change value, and performs phase locking.

[0108] Select the control set parameters, i.e., the phase angle change value within one control step: [0, 400πT] s 200πT s 100πT s -100πT s -200πT s -400πT s There are a total of 7 sets of data. In each control step, the q-axis component magnitude value corresponding to the above 7 phase angle increments is calculated, that is, the cost function value. The phase angle change corresponding to the minimum cost function is selected and added to the original phase angle to obtain the phase-locked loop phase angle output in this control step.

[0109] Input the αβ axis components of the mains voltage and initialize T. s The control set parameters are [0, π / 100, π / 200, π / 400, -π / 400, -π / 200, -π / 100], g_opt is 1e10, and x_opt is 1. The q-axis component magnitude corresponding to different phase angle increments in the control set is calculated by formula (3), which is the cost function value. If the cost function value corresponding to the i-th phase angle increment is less than the existing optimal cost function value, it is used to replace the optimal cost function value, and the optimal control set index x_opt is recorded as i. The x_opt-th phase angle increment in the control set parameters is added to the original output phase angle, which can be used as the output phase angle of the phase-locked loop in the current calculation cycle, and the phase is successfully locked.

[0110] Example 3:

[0111] Example 3 is a preferred example of Example 1, and is used to illustrate the present invention in more detail.

[0112] This invention introduces Model Predictive Control (MPC) into the implementation process of a phase-locked loop (PLL). By solving for the optimal result within a finite time and a finite control set, it achieves rapid phase-locking of the grid voltage. This invention replaces PI control with MPC, avoiding the cumbersome design of PI parameters. Furthermore, by rationally designing the control set parameters according to the control step size, it simplifies computational complexity and improves operational efficiency.

[0113] To achieve fast and accurate phase-locking of the inverter while maintaining stable system operation and avoiding oscillation, the PI parameters in the traditional phase-locked loop need to be precisely designed and subjected to extensive practical testing. The parameter selection process is quite complex, and changes in system parameters may affect the steady-state and dynamic performance of the phase-locked loop, causing the original PI parameters to fail.

[0114] This invention establishes a phase-locked loop (PLL) system model, defines a cost function, and provides a universal basis for selecting control set parameters based on the influence of control step size on the output results. The PLL system for grid-connected inverters based on model predictive control (MMC) can achieve accurate and rapid phase-locking with the grid voltage. Comparing the MMC-based PLL system with traditional PLL systems, this invention achieves the same phase-locking effect without a PI controller, accurately tracking even sudden changes in grid voltage phase angle and exhibiting good transient characteristics. This invention provides a novel design and application approach for PLL systems in grid-connected inverters, offering the following advantages:

[0115] This method introduces model predictive control into the phase-locked loop (PLL) system of a grid-connected inverter. By combining the control step size with the possible phase angle changes of the PLL, a finite control set is formed, which avoids large-scale blind searching and improves the operating efficiency of the PLL.

[0116] The proposed model predictive control-based grid-connected inverter phase-locked loop system replaces the PI controller with model predictive control, eliminating the influence of PI parameters on the phase-locked loop control performance. Phase-locking can be achieved without considering other system parameters besides the control step size, avoiding the cumbersome design process of PI parameters and improving the phase-locking speed.

[0117] The technical solution claimed in this invention is a phase-locked loop system for grid-connected inverters based on model predictive control.

[0118] Compared to traditional phase-locked loops (PLLs), the grid-connected inverter PLL system based on model predictive control (MMDC) proposed in this invention achieves phase-locking functionality using MMDC, enabling accurate synchronization even without a PI controller. This invention first combines the structural model of the PLL and its output phase characteristics, replacing PI with MMDC. Then, a suitable cost function is selected, and based on the control step size of the actual control system or the simulation step size of the simulation system, the possible magnitudes of the control set, i.e., the phase change values, are determined for each calculation cycle. Next, the cost function values ​​under different phase angle changes are calculated for each control set, and the phase angle increment corresponding to the minimum cost function is selected as the phase angle change value, ultimately achieving successful phase-locking. The specific implementation methods are as follows:

[0119] A. Phase-locked loop model and output phase angle characteristic analysis:

[0120] The grid voltage signal input to the phase-locked loop can be written in abc coordinates as:

[0121]

[0122] Among them, u a u b and u c Let U be the three-phase grid voltage (a, b, c), U1 be the amplitude of the input voltage signal, and θ be the phase of the grid voltage. After Clark transformation, the grid voltage in two-phase stationary αβ coordinates can be written as:

[0123]

[0124] Among them, u α u β Let be the voltage components along the αβ coordinate axes. Then, the voltage in the synchronously rotating dq coordinate system can be written as:

[0125]

[0126] Among them, u d u q Let θ be the voltage component of the dq coordinate axis. PLL This represents the output phase angle of the phase-locked loop. The relationship between the grid voltage and its position on the αβ and dq axes is as follows: Figure 1 As shown, the bold symbol u represents the rotation vector of the grid voltage along the stationary αβ coordinate axis. Figure 1 The expression for the grid voltage in the dq coordinate system can also be derived.

[0127] If the d-axis in the dq coordinate system completely coincides with the rotating vector u of the grid voltage, i.e., θ = θ PLLAt this point, the projection of the grid voltage rotation vector u onto the d-axis is the grid voltage amplitude U1, and the projection onto the q-axis is 0. This indicates that the phase-locked loop (PLL) can lock the grid frequency and accurately track the grid voltage phase angle, meeting the inverter's requirements for the PLL. Therefore, this invention proposes a grid-connected inverter PLL system based on model predictive control, the control structure block diagram of which is shown below. Figure 2 As shown in the figure, theta is the output phase angle of the model predictive control phase-locked loop in the current control cycle, which will be retained as theta_old in the next control cycle and fed back to the MPC.

[0128] B. Implementation process of grid-connected inverter phase-locked loop system based on model predictive control:

[0129] When a phase-locked loop (PLL) system can successfully lock the phase, the magnitude of the q-axis component of the grid voltage is 0. Based on this characteristic of the PLL, the cost function g in model predictive control can be defined as the magnitude of the q-axis component of the grid voltage. The smaller the projection of the grid voltage onto the q-axis, the smaller the angle difference between the dq coordinate system (based on the PLL output phase angle) and the rotating vector u of the grid voltage, which means a better phase-locking effect.

[0130] The phase angle of the power grid changes continuously from 0 to 2π within one power frequency cycle, i.e., 0.02s. However, most practical control systems are discrete controllers with control step sizes. Therefore, under normal circumstances, if the phase-locked loop (PLL) can successfully lock the phase without any sudden phase angle changes, the formula for calculating the phase angle increment within one control step is:

[0131]

[0132] Among them, T s To calculate the step size or simulation step size, derta_theta is the phase angle increment within a control step. The phase angle increment derta_theta is then calculated to be 100πT. s If the grid voltage leads or lags in phase at a certain moment, the phase angle will change abruptly, corresponding to a vertical shift in the phase-locked loop (PLL) output phase curve. During this transient process, if the PLL phase angle change within a control step is still 100πT... s This will lead to phase-locked loop (PLL) failure. In this case, it is desirable for the PLL output phase angle to change rapidly, relock the grid voltage, and quickly enter a new steady state. Therefore, an integer multiple of derta_theta should be added to the control set. Simultaneously, considering the initialization of the PLL system, a 0 option should be added to the control set.

[0133] In summary, the control set parameters, i.e., the possible phase angle changes within one control step, are selected as: [0, 400πT] s 200πT s 100πT s -100πTs -200πT s -400πT s There are a total of 7 sets of data. Within each control step, the magnitude of the q-axis component corresponding to the 7 phase angle increments is calculated, i.e., the cost function value. The phase angle change corresponding to the minimum cost function is selected and added to the original phase angle to obtain the output phase-locked loop phase angle within this control step. If the control step size changes or the phase-locked loop control accuracy requirements change, only the control set parameters need to be modified.

[0134] The implementation flowchart of the grid-connected inverter phase-locked loop system based on model predictive control is as follows: Figure 3 As shown. Figure 3 In the middle, T s To control the step size, g_opt is the optimal value of the cost function, x_opt is the index of the optimal control set, g is the cost function value, theta_old is the output phase angle of the model predictive control phase-locked loop in the previous calculation cycle, and theta is the output phase angle of the model predictive control phase-locked loop in the current control cycle. The input is the αβ axis components of the grid voltage, and T is initialized. s The control set parameters are [0, π / 100, π / 200, π / 400, -π / 400, -π / 200, -π / 100], g_opt is 1e10, and x_opt is 1. Then, the q-axis component magnitude corresponding to different phase angle increments in the control set is calculated using formula (3), i.e., the cost function value. If the cost function value corresponding to the i-th phase angle increment is less than the existing optimal cost function value, it is used to replace the optimal cost function value, and the optimal control set index x_opt is recorded as i. Finally, the x_opt-th phase angle increment in the control set parameters is added to the original output phase angle, which can then be used as the output phase angle of the phase-locked loop in the current calculation cycle, achieving successful phase locking.

[0135] This solution proposes a phase-locked loop (PLL) system for grid-connected inverters based on model predictive control. Comparing the proposed PLL structure with traditional PLL structures, it replaces the PI controller with model predictive control, achieving the same performance as a traditional PLL without considering system parameters. The beneficial effects of this invention are:

[0136] This method combines the control step size or simulation step size to directly provide a universal basis for selecting control set parameters. It can obtain the phase-locked loop system without considering other system parameters besides the step size, thus avoiding cumbersome parameter design.

[0137] By replacing the PI controller with model predictive control, the influence of PI parameters on the performance of the phase-locked loop (PLL) control is eliminated. Even without a PI controller, the same phase-locking effect as a traditional PLL can be achieved, ensuring accurate phase-locking even in the event of sudden changes in grid phase.

[0138] Traditional SRF-PLLs use PI controllers to track and synchronize the grid voltage phase. This invention proposes a model predictive control-based grid-connected inverter phase-locked loop system. First, it analyzes the synchronization mechanism of traditional PLL systems. Then, based on this, it selects a suitable cost function and determines the control set parameters based on the control step size. Finally, within each step size, it selects the control set parameters corresponding to the minimum cost function, ultimately achieving successful phase-locking of the grid voltage. To demonstrate the advantages of this invention, experimental verification is conducted. Under normal conditions, the phase angle between the grid voltage waveform and the output phase angle of the model predictive control-based PLL is as follows: Figure 4 As shown. From Figure 4 It can be seen that the output of the phase-locked loop of this invention changes from 0 to 2π within one power frequency cycle, accurately calibrating the grid voltage phase and achieving successful phase locking. When the grid voltage undergoes a sudden change, i.e., at 0.06s, the grid voltage lags by π / 6, the phase angle between the grid voltage waveform and the output of the phase-locked loop based on model predictive control is as follows: Figure 5 As shown. From Figure 5 It can be seen that at 0.06s, the output phase angle of the phase-locked loop of this invention recovers to a steady state after fluctuation and can still successfully lock phase. The experiment verifies the correctness of the theory and proves that the grid-connected inverter phase-locked loop system based on model predictive control can achieve the same phase-locking effect as the traditional phase-locked loop even without a PI controller. Even if there are sudden changes in the power grid, it can successfully lock phase and has good transient performance.

[0139] Compared with traditional PLL structures, the model predictive control-based PLL structure for grid-connected inverters proposed in this invention achieves accurate phase-locking without PI control, and its output results are identical to those of traditional PLLs. Furthermore, by incorporating the concept of control step size, control set parameters are determined, avoiding large-scale blind searches and reducing computational complexity. This invention provides a novel design and application approach for PLL systems in grid-connected inverters.

[0140] Those skilled in the art can understand this embodiment as a more specific description of Embodiment 1 and Embodiment 2.

[0141] Those skilled in the art will understand that, besides implementing the system and its various devices, modules, and units provided by this invention in the form of purely computer-readable program code, the same functions can be achieved entirely through logical programming of the method steps, making the system and its various devices, modules, and units of this invention function in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, the system and its various devices, modules, and units provided by this invention can be considered as a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered as structures within the hardware component; alternatively, the devices, modules, and units for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0142] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A phase-locked loop method for grid-connected inverters based on model predictive control, characterized in that, The method includes the following steps: Step S1: Combining the structural model of the phase-locked loop and the phase characteristics of its output, model predictive control is used to replace PI control; Step S2: Select the cost function and determine the control set, i.e. the magnitude of the phase change value, within each calculation cycle based on the control step size of the actual control system or the simulation step size of the simulation system. Step S3: Calculate the cost function values ​​under different phase angle changes in the control set, select the phase angle increment corresponding to the minimum cost function as the phase angle change value, and perform phase locking; Step S2 includes the implementation process of a grid-connected inverter phase-locked loop system based on model predictive control: When the phase-locked loop system can successfully lock the phase, the grid voltage... The magnitude of the axis component is 0; based on this characteristic of the phase-locked loop, the cost function in model predictive control is defined. g For the grid voltage Axial component magnitude; The power grid phase angle changes continuously from 0 to 2π within one power frequency cycle, i.e., 0.02s. If the phase-locked loop can successfully lock the phase without any sudden changes in phase angle, the formula for calculating the phase angle increment within one control step is: (4) wherein, is a calculation step or simulation step, is a corresponding phase angle increment for one control step; the phase angle increment is 100 ; If the grid voltage leads or lags in phase at a certain moment, the phase angle will change abruptly, corresponding to a vertical shift in the phase-locked loop (PLL) output phase curve. During this transient process, if the PLL phase angle change within a control step is still 100°... Phase-locked loop failed; add to control set Integer multiples of the specified value; add the 0 option to the control set.

2. The model predictive control based phase locked loop method for grid-tied inverters according to claim 1, wherein, Step S1 includes phase-locked loop model and output phase angle characteristic analysis: The grid voltage signal inputted to the phase-locked loop is in abc and written in coordinates as (1) in, , and for abc Three-phase grid voltage, The amplitude of the input voltage signal. The phase of the grid voltage; After the Clark transformation, the grid voltage in two-phase stationary αβ coordinates is written as: (2) wherein , is αβ a voltage component of the coordinate axis; In synchronous rotation dq The voltage in the coordinate system is written as: (3) wherein , is dq the voltage component of the coordinate axis, is the phase angle of the phase-locked loop output.

3. The grid-connected inverter phase-locked loop method based on model predictive control according to claim 1, characterized in that, In step S3, the control set parameters are selected, that is, the phase angle change value within one control step is: [0, 400]. 200 100 -100 -200 -400 A total of 7 sets of data were collected; within each control step, the corresponding values ​​of the 7 phase angle increments were calculated. The axis component magnitude, i.e. the cost function value, is selected by adding the phase angle change corresponding to the minimum cost function value to the original phase angle to obtain the phase-locked loop phase angle output within this control step.

4. The grid-connected inverter phase-locked loop method based on model predictive control according to claim 2, characterized in that, Input grid voltage αβ Axis components, initialization T s The value is 25 μs, and the control set parameters are [0, π / 100, π / 200, π / 400, -π / 400, -π / 200, -π / 100]. g_opt For 1e10, x_opt =1; The formula (3) is used to calculate the phase angle increments corresponding to different phase angles in the control set. q The magnitude of the axis component, i.e., the cost function value, if the first... i If the cost function value corresponding to a phase angle increment is less than the existing optimal cost function value, then the optimal cost function value is replaced by the original value, and the optimal control set index is adjusted accordingly. x_opt Recorded as i ; the first parameter in the control set x_opt The phase angle increment is added to the original output phase angle, which can then be used as the output phase angle of the phase-locked loop in the current calculation cycle, thus successfully locking the phase.

5. A model predictive control based phase-locked loop system for a grid-connected inverter, characterized in that, The system includes the following modules: Module M1: Combining the structural model of the phase-locked loop and the phase characteristics of its output, model predictive control is used to replace PI control; Module M2: Selects the cost function and determines the control set, i.e. the magnitude of the phase change value, within each calculation cycle based on the control step size of the actual control system or the simulation step size of the simulation system. Module M3: Calculates the cost function values ​​under different phase angle changes in the control set, selects the phase angle increment corresponding to the minimum cost function as the phase angle change value, and performs phase locking; The module M2 includes the implementation process of a grid-connected inverter phase-locked loop system based on model predictive control: When the phase-locked loop system can successfully phase-locked, the modulus of the axis component of the grid voltage is 0; according to the characteristics of the phase-locked loop, the cost function in model predictive control is defined as the modulus of the g axis component of the grid voltage ​ The power grid phase angle changes continuously from 0 to 2π within one power frequency cycle, i.e., 0.02s. If the phase-locked loop can successfully lock the phase without any sudden changes in phase angle, the formula for calculating the phase angle increment within one control step is: (4) in, To calculate or simulate the step size, The phase angle increment is the phase angle increment within a control step. 100 If the grid voltage leads or lags in phase at a certain moment, the phase angle will change abruptly, corresponding to a vertical shift in the phase-locked loop (PLL) output phase curve. During this transient process, if the PLL phase angle change within a control step remains 100°... Phase-locked loop failed; add to control set Integer multiples of the specified value; add the 0 option to the control set.

6. The grid-connected inverter phase-locked loop system based on model predictive control according to claim 5, characterized in that, Module M1 includes a phase-locked loop model and output phase angle characteristic analysis: The grid voltage signal input to the phase-locked loop is in abc Coordinates are written as: (1) in, , and for abc Three-phase grid voltage, The amplitude of the input voltage signal. The phase of the grid voltage; After Clark transformation, the grid voltage remains stationary in both phases. αβ The voltage in the coordinate system is written as: (2) in, , for αβ Voltage components of the coordinate axes; In synchronous rotation dq Voltage in a coordinate system is written as: (3) in, , for dq Voltage components of the coordinate axes, This is the phase angle output of the phase-locked loop.

7. The grid-connected inverter phase-locked loop system based on model predictive control according to claim 5, characterized in that, In module M3, the control set parameters are selected, that is, the phase angle change value within one control step is: [0, 400]. 200 100 -100 -200 -400 A total of 7 sets of data were collected; within each control step, the corresponding values ​​of the 7 phase angle increments were calculated. The axis component magnitude, i.e. the cost function value, is selected by adding the phase angle change corresponding to the minimum cost function value to the original phase angle to obtain the phase-locked loop phase angle output within this control step.

8. The grid-connected inverter phase-locked loop system based on model predictive control according to claim 6, characterized in that, Input grid voltage αβ Axis components, initialization T s The value is 25 μs, and the control set parameters are [0, π / 100, π / 200, π / 400, -π / 400, -π / 200, -π / 100]. g_opt For 1e10, x_opt =1; The formula (3) is used to calculate the phase angle increments corresponding to different phase angles in the control set. q The magnitude of the axis component, i.e., the cost function value, if the first... i If the cost function value corresponding to a phase angle increment is less than the existing optimal cost function value, then the optimal cost function value is replaced by the original value, and the optimal control set index is adjusted accordingly. x_opt Recorded as i ; the first parameter in the control set x_opt The phase angle increment is added to the original output phase angle, which can then be used as the output phase angle of the phase-locked loop in the current calculation cycle, thus successfully locking the phase.

Images