Method for Kalman filtering compensation of input end power grid of single-phase Vienna converter

Abstract

The invention discloses a low-cost and high-quality method for Kalman filtering compensation of an input end power grid of a single-phase Vienna converter. The method comprises the following steps: S1, initializing an initial state and parameters of a Kalman filtering algorithm; S2, acquiring a real-time measurement value zk from a voltage sensor in a Vienna converter in a single-phase mode; s3, predicting a state variable at a next moment according to the system model: predicting by using a state equation and a prediction error covariance matrix; s4, measurement updating: using Kalman gain to update state estimation: calculating the Kalman gain according to the measured value and the predicted value, and updating the state estimation; s5, compensation voltage calculation: calculating a compensation voltage value according to the updated state estimation; s6, the modulation wave Vm is adjusted according to the compensation voltage value, wherein the modulation wave Vm is adjusted in real time, and the detuning phenomenon is reduced; and S7, repeating the steps and circularly executing the steps until the system is stable or an expected effect is achieved. The invention is applied to the technical field of electronic power.

Description

Technical Field

[0001] The present invention relates to a method for compensating for Kalman filtering of an input power grid of a single-phase Vienna converter, and in particular to a method for compensating for a sinusoidal wave detuning phenomenon caused by asymmetric positive and negative characteristics of a sensor and a sampling circuit in a single-phase Vienna converter by using Kalman filtering technology. Background Art

[0002] In power electronic systems, Vienna converter is a highly efficient single-phase / three-phase rectification device and is widely used in industrial applications.

[0003] In power electronic equipment, sensors are used to detect grid voltage and current signals. However, due to various reasons such as sensor design, manufacturing, and use environment, the sensor may show an imbalance in positive and negative characteristics. Specifically, the sensor may have different response characteristics when measuring in the forward and reverse directions, and there may also be deviations in the positive and negative voltage characteristics of the sampling circuit. These reasons will cause positive and negative deviations in the measured values, thus affecting the stability and efficiency of the system.

[0004] The existing single-phase Vienna converter input sensor positive and negative characteristic imbalance problem is mainly solved by hardware filters and software algorithms, but these methods have certain limitations. For example, adding hardware filters or hardware positive and negative compensation circuits to compensate for the positive and negative detuning phenomenon of the sine wave, but this method increases the complexity of the system and the additional hardware cost, and will increase the volume and reduce the power density. Software algorithms are a more affordable solution, but traditional software algorithms such as PID control and sliding mode control do not work well under more complex working conditions. Summary of the invention

[0005] The technical problem to be solved by the present invention is to overcome the deficiencies of the prior art and provide a low-cost and high-quality method for input-end power grid Kalman filter compensation of a single-phase Vienna converter.

[0006] The technical solution adopted by the present invention is: the present invention comprises the following steps:

[0007] S1. Initialization - Initialize the initial state and parameters of the Kalman filter algorithm in the Vienna converter in single-phase mode: Choose a suitable initial state estimate and the initial error covariance matrix P0;

[0008] S2. Measurement value input - Get sensor measurement value: Get real-time measurement value from voltage sensor in Vienna converter in single-phase mode k ;

[0009] S3. State prediction - predict the state variables at the next moment based on the system model: use the state equation and the prediction error covariance matrix for prediction;

[0010] S4. Measurement update - Update state estimate using Kalman gain: Calculate Kalman gain based on measured and predicted values ​​and update state estimate;

[0011] S5. Compensation voltage calculation - Calculate the compensation voltage value based on the updated state estimate: Calculate the compensation voltage value using the compensation voltage formula;

[0012] S6. Real-time adjustment - adjust the modulation wave V according to the compensation voltage value m :Real-time adjustment of modulation wave V m , reduce detuning phenomenon;

[0013] S7. Repeat steps - cyclically execute the above steps: until the system is stable or the desired effect is achieved.

[0014] Furthermore, the Vienna converter in the single-phase mode disconnects one branch of the original three-phase Vienna converter from the circuit through a contactor or directly operates the switch tube of the branch to disconnect it, and connects the input ends of the remaining two branches to the positive and negative ends of the single-phase input respectively, and the Vienna converter in the single-phase mode also includes the voltage sensor, input voltage sampling circuit and controller electrically connected in sequence, the voltage sensor is connected to the branch of the branch inductor, the input voltage sampling circuit is used to process the analog quantity obtained by the voltage sensor, and the controller is a microprocessor or a DSP controller, which is used to convert the conditioned analog quantity output by the input voltage sampling circuit into a digital quantity for subsequent data processing and control.

[0015] Furthermore, in step S1, the positive and negative difference of the input terminal voltage is set to xk, and the formula is:

[0016] x k =U peak+ -U peak-

[0017] Among them U peak+ is the positive voltage value of the input terminal, where U peak- is the negative voltage value of the input voltage.

[0018] Furthermore, in step S2, z k It is an observed variable, which represents the positive and negative difference of the input voltage measured by the voltage sensor.

[0019] Furthermore, the state equation in step S3 is used to describe the change of the system state over time, and its formula is:

[0020]

[0021] in, is the predicted state at time k, A is the state transfer matrix, is the state estimate at time k-1, B is the control input matrix, u k is the control input vector;

[0022] The prediction error covariance matrix is ​​used to describe the uncertainty of the prediction state, and its formula is:

[0023] P k|k-1 =AP k-1|k-1 A T +Q

[0024] Among them, P k|k-1 is the prediction error covariance matrix at time k, Q is the process noise covariance matrix,

[0025] Further, the measurement update in step S4 includes the following steps:

[0026] S41. Measurement equation: describes the relationship between the measurement value and the system state. The formula is:

[0027]

[0028] Among them, z k is the positive and negative difference of the input voltage measured at time k, H is the observation matrix, v k is the measurement noise;

[0029] S42. Kalman gain: used to calculate the optimal estimate, the formula is:

[0030] K k =P k|k-1 H T (HP k|k-1 H T +R) -1

[0031] Among them, K k is the Kalman gain at time k, R is the measurement noise covariance matrix;

[0032] S43. State update: Update the state estimate based on the measurement value and the Kalman gain, and the formula is:

[0033]

[0034] S44. Error covariance matrix update: describes the uncertainty of the updated state estimate, and its formula is:

[0035] P k =(IK k H)Pk|k-1

[0036] Furthermore, in the calculation of the compensation voltage in step S5, it is assumed that half of the absolute value of the positive and negative difference of the input voltage needs to be compensated to the modulation wave V m The formula is:

[0037]

[0038] Among them, |x k | represents the absolute value of the state estimate, V comp is the compensation voltage.

[0039] Furthermore, the modulation wave V in step S5 m Compensation is to compensate the voltage V comp Add to the modulation wave V m The formula is:

[0040] V m ′=V m +V comp

[0041] Among them, V m ′ is the modulated wave after compensation.

[0042] The beneficial effects of the present invention are as follows: the present invention solves the positive and negative detuning problem of the sine wave caused by the positive and negative deviation characteristics of the sensor and the sampling circuit in the traditional way through the Kalman filter algorithm, so that the present invention has the following advantages: 1. Real-time compensation: it can detect and compensate for the positive and negative detuning phenomenon in real time during the operation of the system; 2. High robustness: it can adapt to changes in different working environments and improve the robustness of the system; 3. Low-cost implementation: it reduces costs by using software algorithms rather than adding additional hardware; 4. High efficiency: it uses the efficient computing power of the Kalman filter to ensure the response speed of the system in real-time applications; 5. It significantly improves the performance of the single-phase Vienna converter and improves the stability and power factor of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 is a schematic diagram of a Vienna converter in single-phase mode;

[0044] Figure 2 is a schematic diagram of the sampling bias and Kalman compensation process;

[0045] Figure 3 is a flow chart of the steps of the present invention;

[0046] Figure 4 This is a comparison diagram of the Kalman filter's prediction and tracking of the positive and negative peak differences and the actual positive and negative peak differences;

[0047] Figure 5This is a comparison chart demonstrating the compensation effect of sine waves through Kalman filtering. DETAILED DESCRIPTION

[0048] like Figure 1 As shown, in this embodiment, the technical solution for the transformation of the rectifier circuit scheme compatible with single / three-phase input of the present invention is briefly introduced here. Originally, the traditional Vienna converter has three branches, which are usually used in three-phase three-wire systems. Then, each branch is connected to the three phase lines A / B / C respectively. The technical solution for the transformation of the rectifier circuit scheme compatible with single / three-phase input of the present invention is to disconnect one of the branches from the circuit through a contactor (or directly operate the switch tube of the branch to disconnect). Then, the input ends of the remaining two branches are respectively connected to the positive and negative ends of the single-phase input. This is the single-phase mode of the Vienna converter compatible with single / three-phase input to which the present invention is applicable.

[0049] Figure 1 The basic circuit topology, simple connection relationship between voltage sensor, sampling circuit and controller are shown in Figure 1. Figure 1 In, V in is a single-phase input AC voltage source; L1 is a branch inductor; D1-D4 are diodes; S s1 -S s4 is a MOS tube; C1 is the voltage-dividing capacitor at the positive end; C2 is the voltage-dividing capacitor at the negative end; R O Represents the output load; U si is the output DC voltage.

[0050] The voltage sensor is connected to the branch of the branch inductor, and the specific form is not limited (it can be a Hall sensor to measure the current of a terminal of a converter to obtain an input voltage value, or it can be an external transformer to directly measure the voltage value).

[0051] The input voltage sampling circuit will perform some processing on the analog quantity obtained by the voltage sensor. This is a common circuit in the power electronics industry, which conditions the analog quantity obtained by the sensor to optimize the signal quality. It is usually a proportional operational amplifier circuit, and sometimes other filters are added due to certain EMC requirements. Sometimes, some automotive-grade power electronic products will have an additional peak holder circuit and some triggers to build a fault warning circuit. Because these things have little impact on the present invention, they are only briefly introduced here without too much explanation.

[0052] The controller is usually a microprocessor or a specialized DSP controller. They usually integrate ADC as a peripheral. They can convert the conditioned analog output of the input voltage sampling circuit into a digital quantity for later data processing and control.

[0053] In order to make the text more logical, here is a brief introduction to the common power factor correction methods of PFC such as Vienna converter. Most PFC switching signal modulation processes require a modulation wave Vm. Modulation waves are usually used to generate pulse width modulation (PWM) signals that control these switching elements (such as MOSFET or IGBT). These PWM signals determine when the switching elements are turned on or off, thereby controlling the flow of energy in the circuit. The power factor correction process requires that the input current be controlled to be at the same frequency and phase as the input voltage. This modulation wave Vm is usually a relatively regular sine wave, and it is at the same frequency and phase as the input voltage. It is used as a reference for controlling the input current in PFC technology. This modulation wave helps to improve the quality of the AC input current, making it closer to a sine wave, thereby improving the power factor.

[0054] There are two common ways to use Vm. In traditional PFC, the digital quantity converted from the analog quantity of the measured input voltage is directly used as the modulation wave Vm. In addition, some PFCs use a certain method (such as an adaptive notch filter ANF) to obtain the fundamental wave component with the same frequency and phase as the modulation wave Vm.

[0055] Regardless of the method, if the analog quantity of the input voltage is distorted during the measurement and conditioning process, it will affect the quality of the final power factor correction. In practical applications, due to various reasons, the sensor or sampling conditioning circuit may have asymmetric positive and negative characteristics, which will cause the sine wave of the input signal to be detuned. This asymmetric positive and negative characteristics is usually caused by the following factors:

[0056] 1. Sensor nonlinearity: The sensor may have a nonlinear response when measuring positive and negative signals, resulting in inconsistent gain or offset of positive and negative signals.

[0057] 2. Circuit defects: Defects in the design of the sampling and conditioning circuit or in the components, such as poor resistor matching, power supply ripple interference, temperature drift, etc., may lead to asymmetric processing of positive and negative signals.

[0058] 3. Environmental influences: External environmental conditions, such as temperature changes, electromagnetic interference, etc., may also cause changes in sensor or circuit performance, thereby causing asymmetry.

[0059] Therefore, a method is needed to compensate for the asymmetry of the positive and negative characteristics. Based on this, the present invention proposes a method for predicting the positive and negative difference of the input voltage through Kalman filtering and adjusting the modulation wave Vm according to the compensation voltage value. Figure 2 Schematic diagram demonstrating this sampling bias and Kalman compensation process. Figure 2In a healthy grid voltage, the voltage waveform is close to an ideal sine wave. After passing through the voltage sensor and sampling circuit, due to the above reasons, it becomes a positive and negative detuned sine wave, and then the Kalman compensation process of the present invention is used to recalculate the compensation voltage value to adjust the modulation wave V m Finally, the controller can get a modulation wave V that is close to the ideal sine wave. m .

[0060] like Figure 3 As shown, in this embodiment, the present invention includes the following steps:

[0061] S1. Initialization - Initialize the initial state and parameters of the Kalman filter algorithm in the Vienna converter in single-phase mode: Choose a suitable initial state estimate and the initial error covariance matrix P0;

[0062] S2. Measurement value input - Get sensor measurement value: Get real-time measurement value from voltage sensor in Vienna converter in single-phase mode k ;

[0063] S3. State prediction - predict the state variables at the next moment based on the system model: use the state equation and the prediction error covariance matrix for prediction;

[0064] S4. Measurement update - Update state estimate using Kalman gain: Calculate Kalman gain based on measured and predicted values ​​and update state estimate;

[0065] S5. Compensation voltage calculation - Calculate the compensation voltage value based on the updated state estimate: Calculate the compensation voltage value using the compensation voltage formula;

[0066] S6. Real-time adjustment - adjust the modulation wave V according to the compensation voltage value m :Real-time adjustment of modulation wave V m , reduce detuning phenomenon;

[0067] S7. Repeat steps - cyclically execute the above steps: until the system is stable or the desired effect is achieved.

[0068] In this embodiment, the positive and negative difference of the input terminal voltage is set to xk in step S1, and the formula is:

[0069] x k =U peak+ -U peak-

[0070] Among them U peak+ is the positive voltage value of the input terminal, where U peak- is the negative voltage value of the input voltage.

[0071] In this embodiment, z in step S2 k It is an observed variable, which represents the positive and negative difference of the input voltage measured by the voltage sensor.

[0072] In this embodiment, the state equation in step S3 is used to describe the change of the system state over time, and its formula is:

[0073]

[0074] in, is the predicted state at time k, A is the state transfer matrix, is the state estimate at time k-1, B is the control input matrix, u k is the control input vector;

[0075] The prediction error covariance matrix is ​​used to describe the uncertainty of the prediction state, and its formula is:

[0076] P k|k-1 =AP k-1|k-1 A T +Q

[0077] Among them, P k|k-1 is the prediction error covariance matrix at time k, Q is the process noise covariance matrix, P k-1|k-1 is the prediction error covariance matrix at time k-1, and A is the state transfer matrix. T is the transpose of the state transfer matrix, T is the transpose symbol of the matrix. When a matrix is ​​transposed, the positions of its rows and columns are swapped, and the new matrix is ​​recorded with a superscript T.

[0078] In this embodiment, the step S4 measurement update includes the following steps:

[0079] S41. Measurement equation: describes the relationship between the measurement value and the system state. The formula is:

[0080]

[0081] Among them, z k is the positive and negative difference of the input voltage measured at time k, H is the observation matrix, v k is the measurement noise;

[0082] S42. Kalman gain: used to calculate the optimal estimate, the formula is:

[0083] K k =P k|k-1 H T (HP k|k-1 H T +R) -1

[0084] Among them, K k is the Kalman gain at time k, R is the measurement noise covariance matrix;

[0085] S43. State update: Update the state estimate based on the measurement value and the Kalman gain, and the formula is:

[0086]

[0087] S44. Error covariance matrix update: describes the uncertainty of the updated state estimate, and its formula is:

[0088] P k =(IK k H)P k|k-1

[0089] I is the identity matrix of the same order. The identity matrix is ​​a special square matrix whose elements on the main diagonal are all 1 and all other elements are 0.

[0090] In this embodiment, in the step S5, in the compensation voltage calculation, it is assumed that half of the absolute value of the positive and negative difference of the input voltage needs to be compensated to the modulation wave V m The formula is:

[0091]

[0092] Among them, |x k | represents the absolute value of the state estimate, V comp is the compensation voltage.

[0093] In this embodiment, the modulation wave V in step S5 m Compensation is to compensate the voltage V comp Add to the modulation wave V m The formula is:

[0094] V m ′=V m +V comp

[0095] Among them, V m ′ is the modulated wave after compensation.

[0096] Here we simply use MATLAB to demonstrate the process of predicting the difference between positive and negative peak values ​​through Kalman filtering and calculating the compensation signal. The code can be implemented in other high-level languages ​​or other forms:

[0097]

[0098]

[0099] In order to demonstrate the process of Kalman filtering the input voltage in a single-phase Vienna converter to predict sampling deviation and compensate for it, we reproduced this process through MATLAB simulation.

[0100] like Figure 4 As shown. The first row of graphs above shows how the actual positive and negative peak differences change over time. The second row of graphs shows how the estimated positive and negative peak differences change over time. The third row of graphs shows how the value of the compensation signal changes over time.

[0101] It can be seen that the Kalman filter has a good effect on predicting and tracking the difference between positive and negative peak values. It can estimate the change value of the positive and negative peak values ​​relatively perfectly and calculate the compensation signal at the same time.

[0102] Figure 5 This is a demonstration of the compensation effect of sine wave through Kalman filtering in MATLAB simulation. Figure 5 The first row of graphs is the original sine wave of the actual input grid voltage. It can be seen that it is still a relatively ideal sine wave at this time. The second row of graphs is the sine wave after the sensor or sampling and conditioning circuit has been detuned. The third row of graphs is the sine wave compensated by the Kalman filter. Finally, this sine wave signal compensated by the Kalman filter can be used as the modulation wave Vm.

[0103] In summary, the present invention has the following advantages:

[0104] 1. Real-time and adaptability: The Kalman filter algorithm can process the input voltage signal in real time and respond quickly in a dynamically changing environment, ensuring that the system can compensate in time under various working conditions. Moreover, this method can dynamically adjust the compensation parameters according to the actual state of the system, improving the system's adaptability and robustness, so that the system can better cope with different working environments.

[0105] 2. High-precision compensation: The present invention uses the Kalman filter algorithm to accurately predict and compensate the input voltage signal, eliminating the sinusoidal wave detuning phenomenon caused by the asymmetric positive and negative characteristics of the sensor and the sampling circuit, and significantly improving the measurement accuracy and control accuracy of the system.

[0106] 3. Improve system stability and power factor: Through the effective compensation of the Kalman filter algorithm, the negative impact caused by the asymmetry of the positive and negative characteristics of the input voltage signal is significantly reduced, the stability and power factor of the system are improved, and the overall performance of the power electronic converter is optimized.

[0107] 4. Low-cost implementation: Compared with traditional hardware filters and compensation circuits, the present invention implements the compensation function through a software algorithm, avoiding the cost of adding additional hardware, reducing the complexity and volume of the system, and improving the power density.

[0108] 5. Dynamic adjustment capability: By real-time monitoring of sensor measurement values ​​and dynamically adjusting the compensation voltage, it ensures that the modulation wave is accurately compensated at any time, thereby improving the system's dynamic response capability and anti-interference ability.

[0109] The invention is applied to the technical field of power electronics.

[0110] Although the embodiments of the present invention are described with practical solutions, they do not constitute limitations on the meaning of the present invention. For those skilled in the art, it is obvious to modify the implementation scheme and combine it with other solutions based on this description.

Claims

1. A method for Kalman filter compensation of the input end grid of a single-phase Vienna converter, characterized in that: It includes the following steps: S1. Initialization - Initialize the initial state and parameters of the Kalman filter algorithm in the Vienna converter in single-phase mode: Choose a suitable initial state estimate and the initial error covariance matrix P0; S2. Measurement value input - Get sensor measurement value: Get real-time measurement value from voltage sensor in Vienna converter in single-phase mode k ; S3. State prediction - predict the state variables at the next moment based on the system model: use the state equation and the prediction error covariance matrix for prediction; S4. Measurement update - Update state estimate using Kalman gain: Calculate Kalman gain based on measured and predicted values ​​and update state estimate; S5. Compensation voltage calculation - Calculate the compensation voltage value based on the updated state estimate: Calculate the compensation voltage value using the compensation voltage formula; S6. Real-time adjustment - adjust the modulation wave V according to the compensation voltage value m :Real-time adjustment of modulation wave V m , reduce detuning phenomenon; S7. Repeat steps - cyclically execute the above steps: until the system is stable or the desired effect is achieved.

2. The method for input-end grid Kalman filter compensation of a single-phase Vienna converter according to claim 1, characterized in that: The Vienna converter in single-phase mode is to disconnect one branch of the original three-phase Vienna converter from the circuit through a contactor or directly operate the switch tube of the branch to disconnect it, and connect the input ends of the remaining two branches to the positive end and the negative end of the single-phase input respectively, and the Vienna converter in single-phase mode also includes the voltage sensor, input voltage sampling circuit and controller electrically connected in sequence, the voltage sensor is connected to the branch of the branch inductor, the input voltage sampling circuit is used to process the analog quantity obtained by the voltage sensor, and the controller is a microprocessor or a DSP controller, which is used to convert the conditioned analog quantity output by the input voltage sampling circuit into a digital quantity for subsequent data processing and control.

3. The method for input-end grid Kalman filter compensation of a single-phase Vienna converter according to claim 1, characterized in that: In step S1, the positive and negative difference of the input voltage is set to xk, and the formula is: x k =U peak+ -U peak- Among them U peak+ is the positive voltage value of the input terminal, where U peak- is the negative voltage value of the input voltage.

4. The method for input-end Kalman filter compensation of a single-phase Vienna converter according to claim 3, characterized in that: The z in step S2 k It is an observed variable, which represents the positive and negative difference of the input voltage measured by the voltage sensor.

5. The method for input-end Kalman filter compensation of a single-phase Vienna converter according to claim 4, characterized in that: The state equation in step S3 is used to describe the change of the system state over time, and its formula is: in, is the predicted state at time k, A is the state transfer matrix, is the state estimate at time k-1, B is the control input matrix, u k is the control input vector; The prediction error covariance matrix is ​​used to describe the uncertainty of the prediction state, and its formula is: P k|k-1 =AP k-1|k-1 From T +Q Among them, P k|k-1 is the prediction error covariance matrix at time k, and Q is the process noise covariance matrix.

6. The method for input-end Kalman filter compensation of a single-phase Vienna converter according to claim 5, characterized in that: The step S4 measurement update comprises the following steps: S41. Measurement equation: describes the relationship between the measurement value and the system state. The formula is: Among them, z k is the positive and negative difference of the input voltage measured at time k, H is the observation matrix, v k is the measurement noise; S42. Kalman gain: used to calculate the optimal estimate, the formula is: K k =P k|k-1 H T (HP k|k-1 H T +R) -1 Among them, K k is the Kalman gain at time k, R is the measurement noise covariance matrix; S43. State update: Update the state estimate based on the measurement value and the Kalman gain, and the formula is: S44. Error covariance matrix update: describes the uncertainty of the updated state estimate, and its formula is: P k =(I-K k H)P k|k-1 7. The method for input-end grid Kalman filter compensation for a single-phase Vienna converter according to claim 6, characterized in that: In the step S5, the compensation voltage is calculated, assuming that half of the absolute value of the positive and negative difference of the input voltage needs to be compensated to the modulation wave V m The formula is: Among them, |x k | represents the absolute value of the state estimate, V comp is the compensation voltage.

8. The method for input-end grid Kalman filter compensation of a single-phase Vienna converter according to claim 7, characterized in that: The modulation wave V in step S5 m Compensation is to compensate the voltage V comp Add to the modulation wave V m The formula is: V′ m =V m +V comp Among them, V′ m is the modulated wave after compensation.

Images