Finite set model prediction direct power control method for single-phase three-level rectifier

Abstract

The invention discloses a finite set model prediction direct power control method for a single-phase three-level rectifier. The method comprises the following steps: carrying out SOGI conversion on alternating current side voltage and current of the single-phase three-level rectifier and output voltage of the rectifier to respectively obtain two-phase orthogonal rotating vectors; calculating to obtain active power and reactive power according to the rotation vector; performing finite control set model prediction control through the calculated active power and reactive power and the rotating vector output by the SOGI, circularly traversing the switching state vector, and predicting the active power and reactive power rate at the next moment and the voltage difference between the upper capacitor and the lower capacitor at the direct current side; and finding out an optimal switching state through cost function calculation to control the three-level rectifier. According to the invention, faster dynamic response can be realized, and the method is especially suitable for a system with transient load disturbance; the FCS-MPC is easy to add system constraints, so that the robustness to power grid fluctuation and load fluctuation is higher; and the influence caused by uncertain factors such as external interference and model errors is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of power electronics, and particularly relates to a finite set model predictive direct power control method for a single-phase three-level rectifier. Background Art

[0002] Compared with a single-phase two-level rectifier, a three-level rectifier can better provide high-power output, improve the withstand voltage level, and has high system reliability. Therefore, three-level rectifiers are widely used in power electronic systems and renewable energy systems.

[0003] In the industrial field, there are various three-level rectifier topologies widely used, including Neutral Point Clamped (NPC), Flying Capacitor (FC), and hybrid three-level rectifiers, etc. However, the number of components used in these topologies is relatively large, and the utilization rate of the DC bus voltage is relatively low. Due to the addition of more clamping diodes and capacitors, the conduction times of the two switching tubes on the inner and outer sides of each arm of the NPC type single-phase three-level rectifier are different, which will cause the switching losses of the two inner tubes to be greater, and the heat generation of different switching tubes on the same arm is uneven; although the FC type single-phase three-level rectifier omits the use of diodes, it adds a considerable number of capacitors, increasing the volume and cost, and generally difficult to meet the volume requirements in the traction system.

[0004] PI control (Proportional-Integral Control) is a classic control strategy widely used in various control systems. PI control combines proportional control and integral control. The proportional control part generates a control signal quickly according to the deviation between the actual output value of the system and the preset reference value in a proportional relationship. The greater the deviation, the stronger the control signal, so as to drive the system output to approach the reference value quickly and effectively improve the response speed of the system to the deviation. The integral control part performs an integral operation on the deviation in terms of time. As long as the deviation exists, the integral term will continue to accumulate, and its generated control effect will continuously increase with time until the steady-state error of the system is completely eliminated, ensuring that the system output can be accurately and stably maintained at the reference value. However, if the PI control parameters are unreasonable, overshoot is likely to occur, and even oscillation may be caused, resulting in system instability and difficulty in achieving the desired effect. In addition, when dealing with complex systems with large lag characteristics or non-linear changes, the control performance of PI control is greatly reduced and it is difficult to meet high-precision requirements. Finite set model predictive control (FCS-MPC) has obvious advantages in inverter control, such as excellent anti-interference ability, fast dynamic response, and flexible ability to handle multi-objective optimization.

[0005] FCS-MPC first establishes a discrete mathematical model of the system, predicts future states from a finite set of states at a fixed step size, calculates the cost function of the controlled object and the reference object under finite switching states, and then determines the optimal switching state applied to the rectifier in the next control period through online optimization. FCS-MPC performs optimization calculations on all switching states in each sampling period, predicts the power changes at multiple future moments in advance, so FCS-MPC can achieve a faster dynamic response.

[0006] Disadvantages of the prior art: The traditional single-phase three-level rectifier direct power control method controls the active and reactive powers of the rectifier through PI controllers respectively to obtain the voltage control command at the AC side input terminal, and this voltage command is used as the modulation signal of the pulse width modulation module to realize the on and off control of each switching device of the rectifier through the pulse width modulation technology. Although this achieves the expected effect, the control method is relatively complex, requires feedforward decoupling operation, and the response speed is not fast. Summary of the Invention

[0007] The object of the present invention is achieved through the following technical solutions.

[0008] Specifically, the present invention provides a finite control set model predictive direct power control method for a single-phase three-level rectifier, including:

[0009] Performing SOGI transformation on the AC side voltage and current of the single-phase three-level rectifier and the rectifier output voltage respectively to obtain two-phase orthogonal rotating vectors;

[0010] Calculating the active power and reactive power according to the rotating vectors;

[0011] Performing finite control set model predictive control through the calculated active power, reactive power and the rotating vectors output by SOGI, traversing the switching state vector in a loop, and predicting the active power, reactive power and the voltage difference between the upper and lower DC side capacitors at the next moment;

[0012] Finding the optimal switching state through the cost function to control the three-level rectifier.

[0013] Further, the single-phase three-level rectifier includes two symmetrical bridge arms composed of eight power switching tubes, and each bridge arm includes four series-connected power switching tubes;

[0014] In each bridge arm, the two outer power switching tubes are main switching tubes, which are directly connected to the positive and negative DC buses; the two inner power switching tubes are auxiliary switching tubes, which are connected in parallel with two diodes to form a midpoint clamping circuit.

[0015] Furthermore, the SOGI transformation of the AC side voltage and current of the single-phase three-level rectifier and the rectifier output voltage to obtain two-phase orthogonal rotating vectors includes:

[0016] In the stationary coordinate system, the voltage and current constructed by the second-order generalized integration module are expressed as:

[0017]

[0018] where v s For AC measurement of AC voltage, v sα and v sβ are two orthogonal outputs of SOGI, v sm is the peak voltage on the AC side, ω is the angular frequency, is the angle between the voltage and current on the AC side, i L is the AC side inductor current, i Lα and i Lβ are two orthogonal outputs of SOGI, i Lm is the peak value of the inductor current on the AC side.

[0019] Furthermore, the calculating of active power and reactive power according to the rotating vector includes:

[0020] The instantaneous active power and reactive power output by the rectifier are

[0021]

[0022] Where V m and I m are the fundamental voltage and current peak values respectively.

[0023] Furthermore, the active power and reactive power rates at the next moment are predicted, including:

[0024] Assume that the signal sampling period is T s , the instantaneous active power and reactive power prediction values at time k+1 are as follows

[0025]

[0026] Where R is the AC side resistance, L is the AC side inductance, ν abα and ν abβ Indicates ν ab Output from SOGI; P(k+1) represents the instantaneous active power prediction value at time k+1, Q(k+1) represents the instantaneous reactive power prediction value at time k+1, ω is the angular frequency, ν abα (k) and ν abβ (k) represents the AC side output voltage ν at the kth moment ab The value after SOGI transformation, νsα (k) and ν sβ (k) represents v at the k-th moment s The value after SOGI transformation.

[0027] Furthermore, predicting the voltage difference between the upper and lower DC-side capacitors at the next moment includes:

[0028]

[0029] where Δu C (k + 1) and Δu C (k) are the voltage differences between the upper and lower DC-side capacitors at the (k + 1)-th and k-th moments respectively, i0(k) represents the current flowing through the midpoint O at the k-th moment, and C1 is the upper DC-side capacitor.

[0030] Furthermore, the cost function G is

[0031] G = |Q ref (k) - Q(k + 1)| 2 + |P ref (k) - P(k + 1)| 2 + λΔu C (k + 1)

[0032] where λ is the weight coefficient, Q ref (k) is the reference reactive power value, and P ref (k) is the reference active power value.

[0033] The advantages of the present invention are as follows:

[0034] 1. In single-phase rectification control, FCS-MPC control is used to optimize and calculate a limited number of switching states in each control cycle, and predict the power and midpoint potential changes at the next moment in advance, which can achieve faster dynamic response, especially suitable for systems with transient load disturbances.

[0035] 2. Compared with traditional PI control, FCS-MPC is easy to add system constraints, such as current constraints, voltage constraints, etc., so it has stronger robustness to grid fluctuations and load fluctuations.

[0036] 3. FCS-MPC optimizes the system performance index in real time at each sampling moment and realizes local forward rolling in the time domain, which can greatly reduce the influence of uncertain factors such as external interference and model error, and obtain a more ideal control effect. Brief Description of the Drawings

[0037] By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present invention. Also, throughout the drawings, the same reference symbols are used to represent the same components. In the drawings:

[0038] Figure 1 Shows the NPC-type single-phase three-level rectifier circuit topology according to an embodiment of the present invention.

[0039] Figure 2 Shows the basic schematic diagram of SOGI according to an embodiment of the present invention.

[0040] Figure 3 Shows the single-phase three-level finite set model predictive direct power control block diagram according to an embodiment of the present invention.

[0041] Figure 4 Shows the midpoint potential fluctuation and AC side voltage and current waveform diagrams of the power feedforward decoupled direct power control according to an embodiment of the present invention.

[0042] Figure 5 Shows the midpoint potential fluctuation and AC side voltage and current waveform diagrams of the model predictive direct power control according to an embodiment of the present invention.

[0043] Figure 6 Shows the rectifier output voltage diagrams under two control modes of PI control and finite set model predictive control according to an embodiment of the present invention.

[0044] Figure 7 Shows the output power diagrams under two control modes of PI control and finite set model predictive control according to an embodiment of the present invention. Detailed Embodiments

[0045] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully communicated to those skilled in the art.

[0046] Term Explanation:

[0047] Single-phase diode-clamped three-level rectifier: Compared with traditional single-phase rectification, single-phase three-level rectification has more significant advantages. The three-level rectifier can output three levels: zero, positive, and negative, reducing the du / dt in the switching operating mode of power devices. Each power device in the converter only bears half of the DC bus voltage, improving the voltage withstand level of the rectifier and making it suitable for high-power applications. At the same time, compared with two-level rectification, its AC-side current is also closer to a sine wave and has less harmonic content. Therefore, three-level rectifiers are widely used in power electronics systems and renewable energy systems.

[0048] The second-order generalized integrator (SOGI) is a second-order dynamic system with specific frequency response characteristics. It is essentially a system that performs an integration operation on the input signal, and its integration characteristics exhibit special responses at specific frequencies. It can be regarded as a second-order filter that can achieve infinite gain at a specific frequency point while having a certain filtering effect on signals of other frequencies. SOGI takes the input signal and the reference frequency as inputs. When the input frequency is equal to the reference signal frequency, it generates two orthogonal sine wave signals.

[0049] Finite Control Set Model Predictive Control (FCS-MPC): FCS-MPC first establishes the discrete mathematical model of the system, then constructs a suitable cost function according to the control command and constraint conditions, calculates the cost function values under finite switching states within a control period, and then determines the optimal switching state applied to the inverter at the next moment through iterative optimization. FCS-MPC fully considers the nonlinearity of the system and has great advantages in the process of multi-objective optimization and multi-objective constraints.

[0050] Embodiment

[0051] A. Model construction

[0052] The topology of the diode-clamped (NPC) single-phase three-level rectifier is as Figure 1 shown, consisting of 8 power switches forming two symmetric bridge arms. Each bridge arm uses half of the power switching devices and is paralleled with two diodes. The two outer ones are the main switching tubes, directly connected to the positive and negative DC buses, and the two inner ones are the auxiliary switching tubes, which together with the two diodes form a midpoint clamping circuit. On the AC side, u s and i L are the AC-side voltage and current respectively, L and R are the AC-side inductance and resistance respectively. On the DC side, u1 and u2 are the voltage withstands of the two support capacitors C1 and C2 respectively, u C1 and uC2 are the voltages of the two capacitors on the DC side, and R L is the load resistance. a and b are the two output terminals on the AC side, and point O is the connection point of the two support capacitors on the DC side.

[0053] According to Figure 1 define the switching function S i (i = a, b) as:

[0054]

[0055] It can be seen from Equation (1) that each bridge arm has three states: 1, 0, and -1. There are a total of 9 switching combinations for the two bridge arms, so there are 9 operating modes. The corresponding numerical relationships for each operating mode are as shown in Formula (2) and

[0056] Table I.

[0057]

[0058] Table I Different Switching Device Combinations and Corresponding Output Voltages

[0059]

[0060] Since the transient process of the single-phase power control system is directly affected by the response speed of the power calculation module, it must be fast and accurate. The average power algorithm is often used for calculation. However, with the development of power electronic systems, the performance requirements for rectifiers have increased, and transient control is often completed within several cycles. The traditional method can no longer meet the requirements. Also, since single-phase rectification cannot form a rotating vector like a three-phase system, it is impossible to combine the instantaneous power calculation theory to obtain the corresponding active and reactive powers. Therefore, the present invention uses a second-order generalized integrator to implement an orthogonal rotating vector.

[0061] Ideally, the active power P and reactive power Q of a single-phase rectifier can be expressed according to the grid voltage and current formulas as follows:

[0062]

[0063] where V m and I m are the peak values of the fundamental voltage and current respectively, is the power factor angle displacement. Compared with the three-phase system using the Prak transformation, single-phase rectification can improve the steady-state characteristics of the system by filtering the signal through the SOGI, and the response speed is relatively fast. Its transfer function is as follows:

[0064]

[0065] where v s (s), v sα (s) and vsβ (s) are two orthogonal quantities of the input AC voltage and the corresponding output in the complex frequency domain, i L (s), i Lα (s) and i Lβ (s) are two orthogonal quantities of the input AC current and the corresponding output in the complex frequency domain, ω is the angular frequency, s is the complex frequency domain variable, and k is the damping ratio. The schematic diagram of SOGI is as Figure 2 shown.

[0066] The AC-side voltage and current of the single-phase PWM rectifier can be equivalent to sinusoidal signals. The difference is that the current contains high-order harmonics. From Equation (3), the transfer function of SOGI can be approximately regarded as a combination of a low-pass filter and a band-pass filter, which suppresses the high-order harmonics in the current. Therefore, the AC-side voltage and current in the steady state can be approximately regarded as sinusoidal signals. That is, in the stationary coordinate system, the voltages and currents constructed by the second-order generalized integration module can be expressed as:

[0067]

[0068] where v s is the AC voltage on the AC side, v sα and v sβ are the two orthogonal outputs of SOGI, v sm is the peak value of the AC-side voltage, is the angle between the AC-side voltage and the current, i L is the inductor current on the AC side, i Lα and i Lβ are the two orthogonal outputs of SOGI, i Lm is the peak value of the inductor current on the AC side.

[0069] Through Equations (2), (4), and (5), the instantaneous active power and instantaneous reactive power of the rectifier output can be obtained as

[0070]

[0071] Taking the derivative of the above P and Q with respect to time, we can get

[0072]

[0073] where the derivative of i Lα and i Lβ with respect to time can be expressed as

[0074]

[0075] where R is the AC-side resistance, L is the AC-side inductance, ν abα and ν abβ represent the AC-side output voltage ν abOutput through SOGI. Assume that the signal sampling period is T s , the above formula is discretized by equations (2), (7), (8) and the first-order Euler equation to obtain the instantaneous active power and reactive power prediction values at time k+1 as follows:

[0076]

[0077] Where P(k+1) represents the instantaneous active power prediction value at time k+1, Q(k+1) represents the instantaneous reactive power prediction value at time k+1, ω is the angular frequency, ν abα (k) and ν abβ (k) represents the AC side output voltage ν at the kth moment ab The value after SOGI transformation, ν sα (k) and ν sβ (k) represents the kth moment v s The value after SOGI transformation.

[0078] B. DC side midpoint potential model

[0079] Assume that the capacitance value C1=C2, u C1 and u C2 are the voltages of the two capacitors on the DC side, i C1 and i C2 They are

[0080] The current flowing through the two capacitors is C1 and i C2 It can be expressed as:

[0081]

[0082] Assume that the signal sampling period is T s , discretize the above formula using the forward Euler method and get:

[0083]

[0084] where u C1 (k) and u C2 (k) represents the voltage value of the two capacitors on the DC side at the kth moment, i C1 (k) and i C2 (k) represents the current value of the two capacitors on the DC side at the kth moment, u C1 (k+1) and u C2 (k+1) represents the predicted voltage values of the two capacitors on the DC side at time k+1.

[0085] According to Kirchhoff's law, we can get

[0086] i0=i C1 -i C2(13)

[0087] where \(i_0\) is the current flowing through the midpoint \(O\). According to equations (11) and (12), it can be deduced that

[0088]

[0089] where \(\Delta u\) C (k + 1) and \(\Delta u\) C (k) are the voltage differences across the upper and lower DC-side capacitors at times \(k + 1\) and \(k\) respectively, and \(i_0(k)\) represents the current flowing through the midpoint \(O\) at time \(k\).

[0090] C. Construction of the cost function

[0091] To enable the finite-set model predictive control to accurately control the active and reactive powers of the rectifier and the midpoint potential fluctuation, the cost function \(G\) is constructed as

[0092] \(G=|Q\) ref (k)-Q(k + 1)| 2 +|P\) ref (k)-P(k + 1)| 2 +\(\lambda\Delta u\) C (k + 1)\ (15)

[0093] where \(\lambda\) is the weight coefficient, \(Q\) ref (k) is the reference reactive power value, which can be set to 0, and \(P\) ref (k) is the reference active power value, which is obtained by multiplying the reference value of the rectifier voltage output by the voltage closed-loop output. The specific control block diagram is as Figure 3 shown. The AC-side voltage and current and the rectifier output voltage are subjected to the SOGI transformation to obtain two-phase orthogonal rotating vectors \(\nu\) sα (k), \(\nu\) sβ (k), \(\nu\) abα , \(\nu\) abβ , \(i\) Lα and \(i\) Lβ . Using formula (7), the active power \(P\) and the reactive power \(Q\) are obtained. Through the calculated \(P\), \(Q\) and the \(\nu\) sα (k), \(\nu\) sβ (k), \(\nu\) abα and \(\nu\) abβ output by the SOGI, the finite control set model predictive control is performed. The 9 finite switching state vectors are cycled through, and the active power \(P(k + 1)\), the reactive power \(Q(k + 1)\) and \(\Delta u\) C (k + 1) at the next moment are predicted using formulas (10) and (14). Finally, the optimal switching state is found using formula (15) to control the three-level rectifier.

[0094] D. Simulation results and analysis

[0095] In the present invention, two methods, namely power feedforward decoupling direct power control based on PI control and finite set model predictive direct power control, are built on the Simulink platform. The system starts from no-load and adds a load once at 0.4 s and 0.8 s respectively. Figure 4 It represents the midpoint potential fluctuation and the AC side voltage and current waveforms of the power feedforward decoupling direct power control. Figure 5 It represents the midpoint potential fluctuation and the AC side voltage and current waveforms of the model predictive direct power control. Figure 6 It represents the output voltage of the two control methods during load fluctuation. Figure 7 It represents the output power of the two control methods.

[0096] From Figures 4-7 it can be analyzed that the two control methods have good control effects on the midpoint potential and power control effects, etc. However, compared with the traditional control algorithm, the finite set model predictive direct power control is faster in the response speed and has a lower distortion rate.

[0097] As described above, it is only the preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claimed rights.

Claims

1. A finite set model predictive direct power control method for a single-phase three-level rectifier, characterized in that, Including: Performing SOGI transformation on the AC-side voltage, current, and rectifier output voltage of the single-phase three-level rectifier to obtain two-phase orthogonal rotating vectors respectively; Calculating the active power and reactive power based on the rotating vectors; Performing finite control set model predictive control on the calculated active power, reactive power, and the rotating vectors output by SOGI, cycling through the switch state vectors, and predicting the active power, reactive power, and voltage difference between the upper and lower DC capacitors at the next moment; Finding the optimal switch state through the cost function to control the three-level rectifier.

2. The finite set model predictive direct power control method for a single-phase three-level rectifier according to claim 1, characterized in that The single-phase three-level rectifier includes two symmetric bridge arms composed of eight power switch tubes, and each bridge arm includes four serially connected power switch tubes; In each bridge arm, the two outer power switch tubes are main switch tubes, directly connected to the positive and negative DC buses; the two inner power switch tubes are auxiliary switch tubes, paralleled with two diodes to form a neutral point clamped circuit.

3. The finite set model predictive direct power control method for a single-phase three-level rectifier according to claim 1 or 2, characterized in that The performing SOGI transformation on the AC-side voltage and current of the single-phase three-level rectifier and the rectifier output voltage to obtain two-phase orthogonal rotating vectors respectively includes: In the stationary coordinate system, the voltages and currents constructed by the second-order generalized integral module are respectively expressed as: where v s is the AC-side AC voltage, v sα and v sβ are the two orthogonal output voltages of SOGI, v sm is the peak value of the AC-side voltage, ω is the angular frequency, is the angle between the AC-side voltage and current, i L is the AC-side inductor current, i Lα and i Lβ are the two orthogonal output currents of SOGI, i Lm is the peak value of the AC-side inductor current, t represents the time variable.

4. The finite set model predictive direct power control method for a single-phase three-level rectifier according to claim 3, characterized in that The calculating the active power and reactive power based on the rotating vectors includes: The instantaneous active power P and instantaneous reactive power Q of the rectifier output are Where V m and I m are the fundamental voltage and current peak values respectively.

5. The finite set model predictive direct power control method for a single-phase three-level rectifier according to claim 3 or 4, characterized in that The predicting the active power and reactive power at the next moment includes: Let the signal sampling period be T s , and the predicted values of the instantaneous active power and reactive power at the (k + 1)-th moment are as follows Where R is the AC-side resistance, L is the AC-side inductance, P(k + 1) represents the predicted value of the instantaneous active power at the (k + 1)-th moment, Q(k + 1) represents the predicted value of the instantaneous reactive power at the (k + 1)-th moment, ω is the angular frequency, ν abα (k) and ν abβ (k) represent the value of the AC-side output voltage ν ab after the SOGI transformation, ν sα (k) and ν sβ (k) represent the value of v s after the SOGI transformation.

6. The finite set model predictive direct power control method for a single-phase three-level rectifier according to claim 5, characterized in that The predicting the voltage difference between the upper and lower DC capacitors at the next moment includes: where Δu C (k + 1) and Δu C (k) are the voltage differences of the upper and lower capacitors on the DC side at the (k + 1)-th and k-th moments respectively, i0(k) represents the current flowing through the midpoint O at the k-th moment, and C1 is the upper capacitor on the DC side.

7. The finite set model predictive direct power control method for a single-phase three-level rectifier according to claim 6, characterized in that The cost function G is G = |Q ref (k) - Q(k + 1)| 2 + |P ref (k) - P(k + 1)| 2 + λΔu C (k + 1) where λ is the weight coefficient, and Q ref (k) is the reference reactive power value, and P ref (k) is the reference active power value.

Images