Microgrid oscillation suppression control method and system based on adaptive static-dynamic model

Abstract

This application discloses a microgrid oscillation suppression control method and system based on an adaptive static-dynamic model. The method includes: predicting the DC-side capacitor voltage, AC-side three-phase capacitor voltage, and AC-side three-phase inductor current values ​​at the next time step based on the values ​​collected at the current time; using the deviations of the predicted DC-side capacitor voltage, AC-side three-phase capacitor voltage, and AC-side three-phase inductor current values ​​from their expected values ​​as control targets, and establishing a control objective function by combining the weights of each control target calculated according to an adaptive weight function; and using a zero-voltage vector injection modulation method to modulate the voltage vector output by the control objective function to achieve fixed-frequency control. This application adds a DC-side control objective to the model predictive control and uses an adaptive objective function to evaluate the objective function to obtain the optimal voltage vector, and then uses space vector pulse width modulation to obtain the corresponding switching transistor states.

Description

Technical Field

[0001] This application relates to the field of microgrid control technology, specifically to a microgrid oscillation suppression control method and system based on an adaptive static-dynamic model. Background Technology

[0002] In recent years, with the rise of the concept of "energy transition and green development", microgrids composed of distributed generation equipment and power consumption equipment have emerged. They have advantages such as simple structure, high efficiency and high reliability, and are widely used in modern grid-connected autonomous power distribution systems such as electric vehicles, high-speed rail, green buildings, future shipborne aircraft power systems, and data centers.

[0003] The end users of DC microgrids are mainly electronic loads. When these electronic loads are strictly regulated, they may generate a negative impedance effect, manifesting as constant power loads (CPLs), which can affect the stability of the system. At the same time, the interaction between multiple converters in the microgrid can also cause power oscillations and even lead to system instability.

[0004] To overcome this instability, stabilization strategies have been studied in recent years. Traditional methods include passive damping, which increases system damping by connecting additional RC or RL filters. While simple, this increases system cost and size. Another approach involves incorporating various control techniques into the load or converter, known as active damping. This method is more complex but less expensive. Active damping is further divided into linear and nonlinear methods. Linear methods use linear feedback closed-loop control transfer functions, making it easier to stabilize small signals. Nonlinear methods offer robustness and faster dynamic performance over a wide range, making them suitable for large-signal models. With the increasing prevalence of nonlinear control converters, the use of nonlinear control to stabilize microgrids has gained increasing attention. Among nonlinear methods, Model Predictive Control (MPC) is widely used in three-phase rectifiers, three-phase inverters, permanent magnet synchronous motors, and other applications due to its intuitive concept and ability to achieve fast tracking response.

[0005] Finite Control Set Model Predictive Control (FCS-MPC) is an important branch of MPC and the most widely used MPC method. However, traditional FCS-MPC suffers from difficulties in designing weighting factors and inconsistent switching frequencies. The selection of weighting factors in traditional FCS-MPC mainly relies on empirically chosen fixed weighting factors, requiring numerous trials, and the fixed weighting factors weaken the system's dynamic adaptability. In power electronic systems, varying switching frequencies lead to a wide range of harmonic distributions, severely affecting power quality and even causing system resonance instability. Summary of the Invention

[0006] To address the challenges of weight factor design and variable switching frequency in traditional FCS-MPC, this application proposes a microgrid oscillation suppression control method and system based on an adaptive static-dynamic model. This application adds a DC-side control objective to model predictive control and employs an adaptive objective function. The objective function is evaluated to obtain the optimal voltage vector, and space vector pulse width modulation is then used to obtain the corresponding switching state.

[0007] This application is achieved through the following technical solution:

[0008] A microgrid oscillation suppression control method based on an adaptive static-dynamic model, the microgrid oscillation suppression control method comprising:

[0009] Using a pre-built prediction model, based on the sampled current DC-side capacitor voltage, DC-side inductor current, AC-side three-phase capacitor voltage, AC-side three-phase inductor current, and AC-side three-phase output current, the DC-side capacitor voltage, AC-side three-phase capacitor voltage, and AC-side three-phase inductor current values ​​at the next moment are predicted.

[0010] The deviations of the predicted DC-side capacitor voltage, AC-side three-phase capacitor voltage, and AC-side three-phase inductor current values ​​and their expected values ​​are used as control targets. The control target function is established by combining the weights of each control target calculated according to the adaptive weight function.

[0011] A zero-voltage vector injection modulation method is used to modulate the voltage vector output by the control objective function to achieve fixed-frequency control.

[0012] In some implementations, the prediction model construction process specifically includes:

[0013] The mathematical model of the DC-side LC filter is established as follows:

[0014]

[0015] Among them, L dc C dc These are the DC-side inductance and DC-side capacitance values, respectively; v s v dc i pol i dc These are the DC-side input voltage, DC-side capacitor voltage, inverter input current, and DC-side inductor current, respectively.

[0016] The discrete-time model for the DC side is as follows:

[0017]

[0018] Among them, i pol_i and i pol_f T represents the inverter input current at time k and the inverter input current at time k+1, respectively. s v is the sampling period dc (k+1) and v dc (k) represents the DC-side capacitor voltage value at the next moment and the DC-side capacitor voltage value at the current moment, respectively.

[0019] In some implementations, the prediction model construction process further includes:

[0020] The mathematical model of the AC-side LC filter is established as follows:

[0021]

[0022] Where L and C are the AC side inductance and AC side capacitance values, respectively, and v i,αβ v C,αβ i L,αβ and i o,αβ These represent the inverter's three-phase output voltage, three-phase capacitor voltage, three-phase inductor current, and three-phase output current in the α-β two-phase coordinate system, respectively.

[0023] Using the zero-order preserved discretization method, the discrete-time model of the AC side is obtained as follows:

[0024]

[0025] in,

[0026] , , , ;

[0027] Among them, i L,αβ (k+1), i L,αβ(k) represents the AC three-phase inductor current value at the next moment and the AC three-phase inductor current value at the current moment, respectively; v C,αβ (k+1), v C,αβ (k) represents the AC three-phase capacitor voltage value at the next moment and the AC three-phase capacitor voltage value at the current moment, respectively.

[0028] In some implementations, the established control objective function is expressed as:

[0029]

[0030] in,

[0031]

[0032] Where, λ dc , λ i and λ v These are the weighting factors for each control objective item. Indicates the AC side capacitance value. Indicates the reference angular frequency. , , These represent the reference values ​​for the DC-side capacitor voltage, the α component of the AC-side capacitor voltage, and the β component of the AC-side capacitor voltage, respectively. , These represent the α and β components of the AC-side capacitor voltage, respectively. , These represent the α and β components of the AC-side inductor current, respectively.

[0033] In some implementations, the adaptive weighting function includes:

[0034] The adaptive weighting function for the DC-side capacitor voltage is:

[0035]

[0036] The adaptive weighting function for the AC side capacitor voltage is:

[0037]

[0038] The adaptive weighting function for the AC side inductor current is:

[0039]

[0040] in,

[0041]

[0042] , , These represent the allowable voltage error on the DC side, the allowable voltage error on the AC side, and the allowable current error on the AC side, respectively.

[0043] In some embodiments, the zero-voltage vector injection modulation method specifically includes:

[0044] The original eight voltage vectors are reduced in size:

[0045]

[0046] in, Let m be the (i+1)th voltage vector, where i = 0, 1, 2, ..., 7; m is the voltage vector reduction factor.

[0047] A zero-voltage vector of a fixed duration is inserted within each modulation cycle, thereby increasing the duration of the zero-voltage vector's action.

[0048] Secondly, this application proposes a microgrid oscillation suppression control system based on an adaptive static-dynamic model, wherein the microgrid oscillation suppression control system includes:

[0049] The prediction module uses a pre-built prediction model to predict the DC capacitor voltage, DC inductor current, AC three-phase capacitor voltage, AC three-phase inductor current and AC three-phase output current values ​​at the current moment based on the sampled DC capacitor voltage, DC inductor current, AC three-phase capacitor voltage, AC three-phase inductor current and AC three-phase output current values ​​at the next moment.

[0050] An adaptive weight calculation module is used to calculate the control target weights according to an adaptive weight function.

[0051] In addition, there is an objective function evaluation module, which uses the deviations of the predicted DC-side capacitor voltage value, AC-side three-phase capacitor voltage value, and AC-side three-phase inductor current value and their expected values ​​as control targets, and establishes a control objective function by combining the weights of each control target.

[0052] In some embodiments, the microgrid oscillation suppression control system further includes:

[0053] The vector control module employs a zero-voltage vector injection modulation method to modulate the voltage vector output by the control objective function, thereby achieving fixed-frequency control.

[0054] Thirdly, this application proposes a computer device including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described method.

[0055] Fourthly, this application proposes a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0056] This application proposes a microgrid oscillation suppression control method and system based on an adaptive static-dynamic model. It adds a DC-side control objective to traditional model predictive control, designs a static-dynamic adaptive objective function, eliminates the need for selecting fixed weighting factors, and improves the system's adaptability. This allows the system to stabilize the DC microgrid in a short time under large oscillations while maintaining good voltage quality on the AC side, thus improving the system's stability and dynamic performance. Furthermore, a zero-voltage vector injection modulation method is proposed to achieve fixed-frequency control. Attached Figure Description

[0057] The accompanying drawings, which are included to provide a further understanding of the embodiments of this application and form part of this application, do not constitute a limitation on the embodiments of this application. In the drawings:

[0058] Figure 1 This is a flowchart of the method proposed in an embodiment of this application;

[0059] Figure 2 This refers to the inverter output voltage vector.

[0060] Figure 3 This represents the time relationship of the voltage vector action;

[0061] Figure 4 This refers to the SVPWM adjustment process;

[0062] Figure 5 This is a system principle block diagram proposed in the embodiments of this application;

[0063] Figure 6 The integral of the number of switching cycles; where (a) is the integral curve of the number of switching cycles after fixed-frequency processing; and (b) is the integral curve of the number of switching cycles under traditional MPC control.

[0064] Figure 7 For the comparison of steady-state performance and dynamic performance; where (a) is the AC side output waveform and DC side output waveform when using the fixed-frequency stabilization control method (fixed weight factor); (b) is the AC side output waveform and DC side output waveform under the traditional MPC control.

[0065] Figure 8 For FFT analysis comparison; where (a) is the FFT analysis diagram of the output three-phase AC voltage when using the fixed frequency stabilization control method (fixed weight factor); (b) is the FFT analysis diagram of the output three-phase AC voltage under traditional MPC control;

[0066] Figure 9The following are performance comparisons for different DC weighting factors: (a) DC side voltage waveform and AC side capacitor voltage waveform when DC weighting factor is 0; (b) DC side voltage waveform and AC side capacitor voltage waveform when DC weighting factor is 0.3; (c) DC side voltage waveform and AC side capacitor voltage waveform when DC weighting factor is 1; (d) DC side voltage waveform and AC side capacitor voltage waveform when DC weighting factor is 5.

[0067] Figure 10 Changes in the proportion of communication;

[0068] Figure 11 For the comparison of steady-state performance and dynamic performance; wherein, (a) is the AC side output waveform and DC side output waveform when the fixed-frequency stabilization control method (adaptive weight function) proposed in this embodiment is used; (b) is the AC side output waveform and DC side output waveform when the fixed-frequency stabilization control method (fixed weight factor) is used.

[0069] Figure 12 For FFT analysis comparison; where (a) is the FFT analysis diagram of the AC side capacitor voltage before load jump when using the fixed frequency stabilization control method (adaptive weight function) proposed in this embodiment; (b) is the FFT analysis diagram of the AC side capacitor voltage after load jump when using the fixed frequency stabilization control method (adaptive weight function) proposed in this embodiment; (c) is the FFT analysis diagram of the AC side capacitor voltage before load jump when using the fixed frequency stabilization control method (fixed weight factor) proposed in this embodiment; (d) is the FFT analysis diagram of the AC side capacitor voltage after load jump when using the fixed frequency stabilization control method (fixed weight factor) proposed in this embodiment. Detailed Implementation

[0070] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the embodiments and accompanying drawings. The illustrative embodiments and descriptions of this application are only for explaining this application and are not intended to limit this application.

[0071] Example 1:

[0072] The traditional FCS-MPC method has the following problems: (1) the switching frequency is not fixed, making it difficult to design a filter to remove high-frequency harmonics based on the switching frequency; (2) it does not consider the DC side oscillation problem, and when there is a large oscillation on the DC side, it will affect the voltage quality on the AC side; (3) the weighting factor is a fixed value, the design process is complicated, the system does not have autonomous adjustment capability, and the anti-interference capability is not strong. In view of this, this embodiment proposes a microgrid oscillation suppression control method based on an adaptive static-dynamic model. This embodiment first proposes a fixed-frequency processing method with zero voltage vector injection, then adds a DC side control target on the basis of model predictive control, designs an adaptive objective function, evaluates the objective function to obtain the optimal voltage vector, and finally uses space vector pulse width modulation to obtain the corresponding switch state.

[0073] like Figure 1 As shown, the method proposed in this embodiment includes the following steps:

[0074] Step 100: Using a pre-built prediction model, based on the sampled DC-side capacitor voltage value, DC-side inductor current value, AC-side three-phase capacitor voltage value, AC-side three-phase inductor current value, and AC-side three-phase output current value at the current moment, predict the DC-side capacitor voltage value, AC-side three-phase capacitor voltage value, and AC-side three-phase inductor current value at the next moment.

[0075] Step 200: Using the predicted DC-side capacitor voltage value, AC-side three-phase capacitor voltage value, and AC-side three-phase inductor current value and their expected values ​​as control targets, and combining the weights of each control target calculated according to the adaptive weight function, a control target function is established.

[0076] Step 300: The zero-voltage vector injection modulation method is used to modulate the voltage vector output by the control objective function to achieve fixed-frequency control.

[0077] Furthermore, traditional model predictive control is based on eight voltage vectors (such as...) Figure 2 As shown, sequential evaluation is performed. After determining the optimal voltage vector, this optimal voltage vector will continue to act throughout the entire control cycle, equivalent to the full duty cycle state in PWM modulation technology. Therefore, when using the traditional model prediction method, each switch only has two states in one control cycle: zero duty cycle and full duty cycle. For example, in the process of changing from [0 0 1] to [1 1 0], all three bridge arms have switching actions (defined as S respectively). a S b S c The process of changing from [0 0 1] to [1 0 1] involves only one bridge arm switching, leading to a problem of inconsistent frequency. Table 1 shows the relationship between the switching state and the output voltage vector in traditional model predictive control.

[0078] Table 1 Relationship between switching state and output voltage vector

[0079]

[0080] Unlike traditional model prediction, SVPWM modulation includes processes such as sector determination, vector action time calculation, seven-segment processing, and modulation wave generation. It inserts two zero vector switching states, [0 0 0] and [1 1 1], within the control cycle, so that each bridge arm switches only once in a control cycle. Therefore, this embodiment uses the SVPWM modulation method to modulate the voltage vector, which can achieve fixed frequency control.

[0081] In order to allow the addition of a zero vector within a single control cycle, the original eight voltage vectors are reduced in size:

[0082] (1)

[0083] Among them, V i Let be the (i+1)th voltage vector, where i = 0, 1, 2, ..., 7; and m be the voltage vector reduction factor. Reducing the voltage vector is equivalent to inserting a fixed-time zero vector within each modulation cycle, thus increasing the duration of the zero voltage vector. Since the voltage change rate generated by the zero voltage vector is small, while that of the non-zero vector is large, increasing the duration of the zero voltage vector leads to a decrease in output voltage ripple and an increase in steady-state error. To reduce steady-state error, the value of m should be chosen to be close to 1.

[0084] like Figure 3 As shown, after reducing the voltage vector in Table 1, when the optimal vector is a non-zero vector, V within one modulation period i (i=1,2,…6), all three voltage vectors, V0 and V7, are active; when the optimal vector is zero, V0 and V7 are active. Therefore, the evaluation of the switch state also changes.

[0085] Let S a * S b * S c * To evaluate the equivalent switching state after voltage vector reduction, its magnitude is related to the conduction time of the transistors on the three bridge arms, expressed in terms of S. a * For example, when the upper pipe is activated for time T S At that time, S a * =1; the pipe opening time is T. S At that time, S a *=0; To achieve fixed-frequency processing, the evaluation voltage vector was reduced, and a zero voltage vector was inserted into the modulation period. Since m is slightly less than 1, the duration of V7 can be ignored (except when the optimal vector is zero); therefore, the conduction time of each bridge arm's upper transistor is:

[0086] (2)

[0087] Among them, T ai T bi T ci The optimal vector is V. i The conduction time of the upper tubes of the three bridge arms, T S The sampling period is specified; note that when the optimal vector is the zero vector, V0 and V7 each have a specific effect, T. S / 2, therefore we have:

[0088] (3)

[0089] According to equations (2) and (3), the relationship between the equivalent switching state and the output voltage vector after the evaluation voltage vector reduction process is shown in Table 2.

[0090] Table 2 Relationship between equivalent switching states and output voltage vector

[0091]

[0092] Based on the voltage vector action time relationship, its SVPWM modulation process (taking the modulation process of V1 as an example) is as follows: Figure 4 As shown. Traditional model predictive control has three possible actions for each switch within a control cycle: remain in its original state, turn on, or turn off. This results in a non-fixed switching frequency in traditional model predictive control; for example, by... Figure 4 As can be seen, the fixed-frequency processing method used in this embodiment inserts a zero vector in each modulation cycle, and each switching transistor turns on and off once in one modulation cycle. Therefore, its switching frequency is determined by the frequency of the modulation wave, thus achieving fixed-frequency processing.

[0093] Furthermore, the prediction model construction process specifically includes the following steps:

[0094] First, establish the mathematical model of the DC-side LC filter. The differential equation of the DC-side LC filter is:

[0095] (4)

[0096] Among them, L dc C dc These are the DC-side inductance and DC-side capacitance values, respectively. s v dc i poli dc These are the DC-side input voltage, DC-side capacitor voltage, inverter input current, and DC-side inductor current, respectively.

[0097] The formula for the inverter input current is:

[0098] (5)

[0099] Where T is Clark's constant-amplitude transformation matrix, S a * S b * S c * To evaluate the equivalent switching state after the switch vector is reduced, i Lα Let i be the α component of the three-phase inductor current. Lβ Let β be the β component of the three-phase inductor current. For the DC side, when the DC inductance is relatively large, the DC current can be considered constant. Therefore, the discrete-time model for the DC side is:

[0100] (6)

[0101] Among them, i pol_i and i pol_f T represents the inverter input current at time k and the inverter input current at time k+1, respectively. s The sampling period is denoted as . Due to the presence of a large inductance on the DC side, the current change between two sampling periods can be considered as a disturbance, which can be regarded as a constant value. This constant value can be considered as the average value of the inverter input current at the two moments.

[0102] Then, the mathematical model of the AC-side LC filter is established. The differential equation of the AC-side LC filter is:

[0103] (7)

[0104] Where L and C are the AC side inductance and AC side capacitance values, respectively, and v i,αβ v C,αβ i L,αβ and i o,αβ These represent the inverter's three-phase output voltage, three-phase capacitor voltage, three-phase inductor current, and three-phase output current in the α-β two-phase coordinate system, respectively.

[0105] According to equation (6), the spatial state equation of the system can be obtained. By using the zero-order preserved discretization method, the discrete-time model of the AC side can be obtained as follows:

[0106] (8)

[0107] in,

[0108] , , , .

[0109] Furthermore, the process of establishing the objective function for predictive control specifically includes:

[0110] For inverters in a DC microgrid, the control objective is to ensure AC side voltage quality while reducing DC microgrid oscillations. Based on this control objective, the established control objective function is:

[0111] (9)

[0112] in,

[0113] (10)

[0114] Where, λ dc , λ i and λ v Each control objective item (DC item) AC voltage term and alternating current term The weighting factor of ) Indicates the AC side capacitance value. Indicates the reference angular frequency. , , These represent the reference values ​​for the DC-side capacitor voltage, the α component of the AC-side capacitor voltage, and the β component of the AC-side capacitor voltage, respectively. , These represent the α and β components of the AC-side capacitor voltage, respectively. , These represent the α and β components of the AC-side inductor current, respectively.

[0115] Furthermore, since the weighting factor is set to a fixed value, the design process is relatively complex and lacks adaptive capability. Therefore, this embodiment uses an adaptive weighting function instead of a fixed weighting factor, enabling the system to stabilize the DC microgrid while ensuring voltage quality on the AC side.

[0116] Specifically, the first consideration is the allowable error range. It's important to note that the priority of AC-side control objectives is higher than that of DC-side control objectives. When the system is stable, the errors of each control objective are within the allowable error range. At this time, the DC-side voltage is stable, and a smaller value is selected for the DC-side weighting factor to ensure AC-side voltage quality. However, when the system is subjected to disturbances, significant oscillations occur, and the errors of each control objective exceed the allowable error range. In this case, a larger value is selected for the DC-side weighting factor to quickly stabilize the DC microgrid and improve the system's dynamic response speed.

[0117] The adaptive weighting function increases with the increase of error, thus giving the system a certain degree of adaptability. Furthermore, to improve the AC side voltage quality when the system is stable, the DC weighting factor is 0 when the DC side error is 0; while when the AC side error is 0, the AC side adaptive weighting factor maintains a certain initial value, increasing the AC weight ratio when the system is stable, thereby improving the system output voltage quality.

[0118] The error values ​​of each control target at time k are expressed as follows:

[0119] (11)

[0120] The adaptive weighting function for the DC-side capacitor voltage is:

[0121] (12)

[0122] The adaptive weighting function for the AC side capacitor voltage is:

[0123] (13)

[0124] The adaptive weighting function for the AC side inductor current is:

[0125] (14)

[0126] in, Indicates the AC side capacitance value. Indicates the reference angular frequency. , , These represent the reference values ​​for the DC-side capacitor voltage, the α component of the AC-side capacitor voltage, and the β component of the AC-side capacitor voltage, respectively. , , These represent the actual error values ​​of the DC-side capacitor voltage, AC-side capacitor voltage, and AC-side inductor current, respectively. , , This indicates their allowable error value. , These represent the α and β components of the AC-side capacitor voltage, respectively. , Indicates the reference phase voltage. , These represent the α and β components of the AC-side inductor current, respectively.

[0127] This embodiment employs a zero-voltage vector injection fixed-frequency processing method and adds a DC-side control objective to the traditional model predictive control. A static-dynamic adaptive objective function is designed to improve the system's adaptive capability. Under large oscillations, it can stabilize the DC microgrid in a short time and maintain good voltage quality on the AC side, thereby improving the system's stability and dynamic performance.

[0128] Based on the same technical concept described above, this embodiment also proposes a microgrid oscillation suppression control system based on an adaptive static-dynamic model, such as... Figure 5 As shown, the control system proposed in this embodiment includes:

[0129] The prediction module uses a pre-built prediction model to predict the DC-side capacitor voltage, DC-side inductor current, AC-side three-phase capacitor voltage, AC-side three-phase inductor current, and AC-side three-phase output current values ​​at the current moment.

[0130] An adaptive weight calculation module is used to calculate the weights of each control target based on an adaptive weight function.

[0131] In addition, there is an objective function evaluation module, which uses the predicted DC-side capacitor voltage value, AC-side three-phase capacitor voltage value, and AC-side three-phase inductor current value, combined with the weights of each control objective, to establish a control objective function.

[0132] Furthermore, the control system proposed in this embodiment also includes:

[0133] The vector control module uses a zero-voltage vector injection modulation method to modulate the voltage vector output by the control objective function, thereby achieving fixed-frequency control.

[0134] Example 2:

[0135] To verify the effectiveness of the control method proposed in the above embodiments, this embodiment compares its performance with that of traditional model predictive control methods. A model is built in Simulink for a series of simulation verifications. The experimental parameters are set as shown in Table 3:

[0136] Table 3 Simulation Parameters

[0137]

[0138] First, the effectiveness of the zero-voltage vector injection fixed-frequency processing method proposed in the above embodiments is verified. Based on the fixed-frequency processing, an objective function containing a fixed weighting factor is used to evaluate each voltage vector, and the results are compared with traditional model predictive control. A simulation time of 0.2s and a load parameter of 30Ω are selected. To ensure that a zero vector can be inserted within one modulation cycle, a fixed-frequency reduction factor m of 0.95 is selected. The load jumps from 30Ω to 20Ω at 0.1s, and the dynamic response speed and steady-state performance under large oscillations are observed.

[0139] Design a switching count integration module. Within a fixed period of 0.02s, the switching count is accumulated when the rising and falling edges of the switching transistor's drive waveform are detected. That is, the count is performed when the transistor turns on or off, and the accumulated value is reset to zero at the end of the period. The slope of the switching count integration curve can be used to determine whether the frequency is constant. Figure 6 As shown in (a), after frequency fixing, the slope of the integral curve of the switch is constant, and the frequency is a fixed value; as Figure 6 As shown in (b), the slope of the integral curve of the number of switching operations under traditional MPC control is not fixed and the frequency changes continuously.

[0140] Figure 7 (a) The AC-side output waveform and DC-side waveform are shown when the fixed-frequency stabilization control method (fixed weight factor) proposed in this paper is used. Figure 7 (b) shows the AC and DC output waveforms under traditional MPC control. After the load jump, Figure 7 (a) The DC-side voltage overshoot is 11.03V and the response time is 6.28ms; Figure 7 (b) The DC-side voltage overshoot is 10.23V and the response time is 5.823ms; this shows that after frequency fixing, the DC voltage and AC-side capacitor voltage can be tracked after load changes in a short time and the system remains stable.

[0141] Figure 8 (a) is the FFT analysis diagram of the output three-phase AC voltage when the fixed-frequency stabilization control method (fixed weight factor) proposed in this paper is used; Figure 8 (b) is the FFT analysis diagram of the output three-phase AC voltage under traditional MPC control. For example... Figure 8As shown, when using the zero-vector injection fixed-frequency method, the system's output voltage amplitude is lower than that of the traditional MPC, and the total harmonic distortion (THD) is also lower. This is because the proposed fixed-frequency processing method inserts a zero vector for a fixed time in each modulation cycle, thus increasing the zero vector's duration. Since the voltage change rate generated by the zero vector is small, while that generated by the non-zero vector is large, the increased proportion of the zero vector leads to a reduction in system output voltage ripple and an increase in output voltage steady-state error. The above comparison demonstrates that the system can still achieve stable control when using the zero-vector injection fixed-frequency method proposed in the above embodiment.

[0142] Secondly, it was verified that the fixed-frequency stabilization control method proposed in the above embodiments can maintain the original steady-state and transient performance of the system. Next, an adaptive weighting function is used instead of the fixed weighting function to further improve the system performance. To achieve the control objectives of stabilizing DC microgrid oscillations and ensuring AC side voltage quality, the adaptive system can stabilize DC-side oscillations by increasing the DC weight ratio when DC-side oscillations are large; and increase the AC weight ratio to ensure AC-side voltage quality when large fluctuations occur on the AC side. Note that the priority of the AC-side control objective is higher than that of the DC-side control objective; taking the DC side as an example, when λ... dc Observe the changes in the DC and AC side performance of the system when the values ​​are 0, 0.3, 1 and 5 respectively.

[0143] like Figure 9 As shown, Figure 9 (a), (b), (c), and (d) are respectively λ dc The waveforms of the DC-side voltage and AC-side capacitor voltage are shown for values ​​of 0, 0.3, 1, and 5. Observing the waveforms, one can intuitively see that: Figure 9 As shown in (a), without a DC-side control target, large oscillations occur on the DC side. Excessive oscillations can affect the AC-side voltage quality, ultimately leading to system instability. Figure 9 Comparing (b), (c), and (d), as the DC weighting factor increases, the stabilization speed of the DC-side voltage improves, and the dynamic performance is enhanced. However, the steady-state error of the AC-side output voltage increases, the THD increases, and the steady-state performance deteriorates. Therefore, when the system is stable, a smaller DC weighting factor is used to improve the system's steady-state performance; when the system is disturbed, a larger DC weighting factor is used to improve the system's dynamic performance. The adaptive weighting function proposed in this paper can effectively accomplish this control task.

[0144] The DC-side adaptive weighting function proposed in the above embodiments sets a maximum allowable error. When the error exceeds the allowable error range, the DC weighting factor takes a larger value; within the allowable error range, the DC weighting factor increases with the increase of the error. The AC-side adaptive objective function is designed in the same way as the DC-side weighting factor. Since the AC-side control objective has a higher priority than the DC-side control objective, as shown in formulas (13) and (14), a certain initial value is set for the AC weighting factor. That is, when the AC error is 0, the AC weight still maintains a large proportion; at the instant the load jumps, the DC-side voltage error is greater than the allowable error value, and the DC-side stabilization speed at this time depends on the proportion of the DC weight at this time.

[0145] (15)

[0146] Where, ψ dc and ψ ac These represent the DC weighting percentage and the AC weighting percentage, respectively; set v allow_dc For 5, v allow_ac For 2, i allow_ac Set the value to 3 and observe the changes in the proportion of communication weight during the simulation process; for example... Figure 10 As shown, the AC ratio remains above 90% during stable conditions, ensuring the steady-state performance of the AC side of the system; during load jumps, the AC ratio decreases and the DC ratio increases, improving the DC side response speed and enhancing the dynamic performance of the system.

[0147] Further verification is needed to demonstrate the improvement of the system's steady-state and dynamic performance by the adaptive weighting function proposed in the above embodiments. For example... Figure 11 As shown, Figure 11 (a) The AC-side output waveform and DC-side waveform are shown when the fixed-frequency stabilization control method (adaptive weighting function) proposed in this paper is used. Figure 11 (b) The AC and DC output waveforms are shown when the proposed fixed-frequency stabilization control method (fixed weighting factor) is used. After the load jump, Figure 11 (a) The DC side voltage overshoot is 6.96V and the response time is 4.557ms; Figure 11 (b) The DC-side voltage overshoot is 11.03V, and the response time is 6.28ms. After replacing the fixed weighting factor with an adaptive weighting function, the system's dynamic performance is significantly improved. Compared with traditional MPC, Figure 7 (b) The DC overshoot is 10.23V and the response time is 5.283ms, which shows that the adaptive fixed-frequency model predictive control method proposed in this paper can effectively improve the dynamic performance of the system.

[0148] Next, the steady-state performance of the system is compared, such as... Figure 12 As shown, Figure 12In the figure, (a), (b), (c), and (d) are FFT analysis graphs of the AC side capacitor voltage before and after load jump when using the fixed-frequency stabilization control method (adaptive weighting function) proposed in this paper, and before and after load jump when using the fixed-frequency stabilization control method (fixed weighting factor) proposed in this paper. Figure 12 (a) Same Figure 12 (c) In comparison, after replacing the fixed weighting factor with an adaptive weighting function, the AC side capacitor voltage amplitude increased from 109.1V to 109.6V, while the total harmonic distortion (THD) value did not change significantly; Figure 12 (b) Same Figure 12 (d) Comparing the two systems, after a load step change, the system oscillation increases. When using the adaptive weighting function, the system self-adjusts according to the error, the AC side capacitor voltage amplitude remains at 109.6V, and the THD value decreases slightly. However, when using a fixed weighting factor, due to the increased oscillation and the fixed weighting factor, the system cannot self-adjust, the AC side capacitor voltage amplitude decreases to 108.7V, and the THD value increases. This demonstrates that replacing the fixed weighting factor with an adaptive weighting function significantly improves the system's steady-state performance.

[0149] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0150] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0151] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0152] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0153] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of this application. It should be understood that the above description is only a specific embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A microgrid oscillation suppression control method based on an adaptive static-dynamic model, characterized in that, The microgrid oscillation suppression and control methods include: Using a pre-built prediction model, based on the sampled current DC-side capacitor voltage, DC-side inductor current, AC-side three-phase capacitor voltage, AC-side three-phase inductor current, and AC-side three-phase output current, the DC-side capacitor voltage, AC-side three-phase capacitor voltage, and AC-side three-phase inductor current values ​​at the next moment are predicted. The deviations of the predicted DC-side capacitor voltage, AC-side three-phase capacitor voltage, and AC-side three-phase inductor current values ​​and their expected values ​​are used as control targets. The control target function is established by combining the weights of each control target calculated according to the adaptive weight function. A zero-voltage vector injection modulation method is used to modulate the voltage vector output by the control objective function to achieve fixed-frequency control; The established control objective function is expressed as follows: ； in, ； in, , and DC terms AC voltage term and alternating current term Weighting factors Indicates the AC side capacitance value. Indicates the reference angular frequency. , , These represent the reference values ​​for the DC-side capacitor voltage, the α component of the AC-side capacitor voltage, and the β component of the AC-side capacitor voltage, respectively. , These represent the α and β components of the AC-side capacitor voltage, respectively. , Let v represent the α and β components of the AC-side inductor current, respectively. dc Here, k represents the DC-side capacitor voltage value; The adaptive weighting function includes: The adaptive weighting function for the DC-side capacitor voltage is: ； The adaptive weighting function for the AC side capacitor voltage is: ； The adaptive weighting function for the AC side inductor current is: ； in, ； , , These represent the allowable voltage error on the DC side, the allowable voltage error on the AC side, and the allowable current error on the AC side, respectively. , , DC terms AC voltage term and alternating current term The error value, , Indicates the reference phase voltage; The zero-voltage vector injection modulation method specifically includes: The original eight voltage vectors are reduced in size: ； in, Let m be the (i+1)th voltage vector, where i = 0, 1, 2, ..., 7; m is the voltage vector reduction factor. A zero-voltage vector of a fixed duration is inserted within each modulation cycle, thereby increasing the duration of the zero-voltage vector's action.

2. The microgrid oscillation suppression control method based on an adaptive static-dynamic model according to claim 1, characterized in that, The prediction model construction process specifically includes: The mathematical model of the DC-side LC filter is established as follows: ； Among them, L dc C dc These are the DC-side inductance and DC-side capacitance values, respectively; v s v dc i pol i dc These are the DC-side input voltage, DC-side capacitor voltage, inverter input current, and DC-side inductor current, respectively. The discrete-time model for the DC side is as follows: ； Among them, i pol_i and i pol_f T represents the inverter input current at time k and the inverter input current at time k+1, respectively. s v is the sampling period dc (k+1) and v dc (k) represents the DC-side capacitor voltage value at the next moment and the DC-side capacitor voltage value at the current moment, respectively.

3. The microgrid oscillation suppression control method based on an adaptive static-dynamic model according to claim 2, characterized in that, The prediction model construction process also includes: The mathematical model of the AC-side LC filter is established as follows: ； Where L and C are the AC side inductance and AC side capacitance values, respectively, and v i,αβ v C,αβ i L,αβ and i o,αβ These represent the inverter's three-phase output voltage, three-phase capacitor voltage, three-phase inductor current, and three-phase output current in the α-β two-phase coordinate system, respectively. Using the zero-order preserved discretization method, the discrete-time model of the AC side is obtained as follows: ； in, ， ， ， ； Among them, i L,αβ (k+1), i L,αβ (k) represents the AC three-phase inductor current value at the next moment and the AC three-phase inductor current value at the current moment, respectively; v C,αβ (k+1), v C,αβ (k) represents the AC three-phase capacitor voltage value at the next moment and the AC three-phase capacitor voltage value at the current moment, respectively.

4. A microgrid oscillation suppression control system based on an adaptive static-dynamic model, characterized in that, The microgrid oscillation suppression and control system includes: The prediction module uses a pre-built prediction model to predict the DC capacitor voltage, DC inductor current, AC three-phase capacitor voltage, AC three-phase inductor current and AC three-phase output current values ​​at the current moment based on the sampled DC capacitor voltage, DC inductor current, AC three-phase capacitor voltage, AC three-phase inductor current and AC three-phase output current values ​​at the next moment. An adaptive weight calculation module is used to calculate the control target weights according to an adaptive weight function. The system also includes an objective function evaluation module, which uses the deviations of the predicted DC-side capacitor voltage, AC-side three-phase capacitor voltage, and AC-side three-phase inductor current values ​​from their expected values ​​as control objectives, and establishes a control objective function by combining the weights of each control objective. The microgrid oscillation suppression control system further includes: The vector control module employs a zero-voltage vector injection modulation method to modulate the voltage vector output by the control objective function, thereby achieving fixed-frequency control. The established control objective function is expressed as follows: ； in, ； in, , and DC terms AC voltage term and alternating current term Weighting factors Indicates the AC side capacitance value. Indicates the reference angular frequency. , , These represent the reference values ​​for the DC-side capacitor voltage, the α component of the AC-side capacitor voltage, and the β component of the AC-side capacitor voltage, respectively. , These represent the α and β components of the AC-side capacitor voltage, respectively. , Let v represent the α and β components of the AC-side inductor current, respectively. dc Here, k represents the DC-side capacitor voltage value; The adaptive weighting function includes: The adaptive weighting function for the DC-side capacitor voltage is: ； The adaptive weighting function for the AC side capacitor voltage is: ； The adaptive weighting function for the AC side inductor current is: ； in, ； , , These represent the allowable voltage error on the DC side, the allowable voltage error on the AC side, and the allowable current error on the AC side, respectively. , , DC terms AC voltage term and alternating current term The error value, , Indicates the reference phase voltage; The zero-voltage vector injection modulation method specifically includes: The original eight voltage vectors are reduced in size: ； in, Let m be the (i+1)th voltage vector, where i = 0, 1, 2, ..., 7; m is the voltage vector reduction factor. A zero-voltage vector of a fixed duration is inserted within each modulation cycle, thereby increasing the duration of the zero-voltage vector's action.

5. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1-3.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-3.

Images