Design of Active Filter Controller Based on DSP Processor

Publisher:chwwdchLatest update time:2010-01-28 Source: 电网技术 Reading articles on mobile phones Scan QR code
Read articles on your mobile phone anytime, anywhere

1 Introduction

The rapid development of power electronics technology has led to the increasing application of various power electronic devices in industry, transportation and households, and the harmonic hazards caused by these nonlinear loads are becoming increasingly serious. Harmonics cause harmonic losses in components in the power grid, reduce equipment efficiency and power factor, and even damage power equipment such as capacitors; harmonics affect precision instruments and nearby communication systems, making them unable to work normally [1].

The number and size of harmonics in the power system change with the system load conditions. The traditional LC static filter cannot meet the requirements. The use of power active filters can dynamically compensate for harmonics with varying sizes and frequencies and varying reactive power. Therefore, the research and application of active filters are receiving more and more attention. The basic principle of active filters is: first detect the harmonic current from the compensation object, and then the compensation device generates a compensation current that is equal to the harmonic current in magnitude but opposite in phase. The two offset each other so that the grid current contains only the fundamental component. The controller is the core component of the active filter. It controls the behavior of the active filter by generating and controlling the pulses that drive the switching device, completing the function of dynamically compensating harmonics and reactive power.

2 Structure and basic functions of control system

The main circuit of the active filter uses a three-phase single-phase bridge voltage source converter, which is coupled to the system through a transformer. Its structure is shown in Figure 1. The three-phase bridge structure is adopted because the control of the three-phase bridges can be decoupled from each other, and it can also be used to eliminate the zero-sequence component and its harmonic current to achieve asymmetric control.

Control system structure

The control system of the active filter adopts a dual DSP structure, in which one DSP processor is used to complete data processing, control and high-level protection functions; the other DSP processor is used to generate high-precision PWM pulses. The controller mainly has the following functions:

(1) Controlling the behavior of active filters

The output voltage of the inverter is controlled according to the harmonic and reactive current components of the detected load current, so that the compensation current output by the active filter and the sum of the load harmonic current and reactive current cancel each other out, thereby making the system current a fundamental positive sequence active current.

(2) Generate trigger pulse

The drive circuit controls the on and off of the IGBT to generate PWM trigger pulses, so that the active filter can output the correct harmonic compensation current.

(3) Pulse synchronization

According to the synchronization pulse retrieved from the power grid, a pulse signal synchronized with the power grid voltage is generated, so that the voltage output by the active filter is synchronized with the power grid voltage.

(4) Self-fault tolerance

Once some components of the controller itself have errors (such as voltage transformer (PT) disconnection, etc.), the controller can immediately detect the error and alarm without causing the device to exit operation. The device can be easily restored after the fault is repaired.

(5) Protection function

When the active filter is operating in an overload or other abnormal state, and the current does not exceed the set value of the protection action, the controller can use the protection function to restore the active filter to a normal working state, avoiding its underlying protection action, thereby ensuring that the active filter can continue to operate normally [2].

3 Implementation of Active Filter Controller

The principle block diagram of the active filter controller is shown in Figure 2.

Active filter controller block diagram

The controller samples and A/D converts the load current, the compensation current output by the device and the system voltage at a sampling frequency of 60×50Hz (or higher). The sampled current is decomposed using harmonic separation algorithms such as dq decomposition or ab decomposition and other methods to filter out the fundamental active component and retain the harmonic current required for compensation. Then the control algorithm is used to calculate the harmonic voltage that the inverter should generate based on the circuit parameters. The instantaneous value of the harmonic voltage is sent to the DSP pulse generator, which uses the SPWM algorithm to determine the action of the inverter switch element based on the instantaneous value of the harmonic voltage. The pulse generator performs SPWM pulse calculation based on the instantaneous value of the voltage to generate a drive pulse.

The following is an introduction to the various functions of the controller.

3.1 Control Algorithm

The control algorithm of this system consists of two parts: detection of harmonic and reactive current and current tracking control. The calculation of harmonic and reactive current is based on the instantaneous reactive power theory of three-phase circuit [3], as shown in Figure 3.

Compensation current detection

formula

Since the active filter in this paper needs to control the voltage on the DC side, a certain fundamental active component needs to be included in the command current so that the DC side and the AC side of the active filter can exchange energy and adjust the capacitor voltage to a given value.

Figure 4 shows the current tracking control block diagram. The current tracking control of this system adopts PI control. The output control quantity is sent to the pulse generator through the dual-port RAM. The pulse generator generates a trigger pulse based on the waveform information obtained. After isolation and shaping, the pulse drives the IGBT of the main circuit to make the inverter output the corresponding voltage. The compensation current is generated by the difference between the output voltage of the inverter and the AC side power supply voltage acting on the inductor.

Current tracking control block diagram

Figure 5 is a digital simulation result of phase A using the active filter to compensate for the harmonics generated by the three-phase 6-pulse rectifier load. The simulation software uses PSCAD. It can be seen from the figure that the system current after compensation is in phase with the system voltage, and the current waveform is greatly improved. However, by comparing the load current and the system current, it can be seen that the compensation effect is poor at the moment when the load current changes rapidly (corresponding to the phase change of the rectifier bridge). This is because the APF is required to generate a very high harmonic voltage to compensate for the rapidly changing current. On the one hand, the active filter is required to have a very fast response speed, and on the other hand, it is required to generate high voltage on the DC side, which is difficult to achieve in actual devices. Therefore, when the load current changes very quickly, the compensation ability of the APF is poor. The impact of system asymmetry on the APF and its compensation for zero-sequence current are still under further study. Figure 6 is a harmonic analysis of the A-phase system current. The total harmonic distortion rate THD of the load current is 20.1%, and the total distortion rate of the system current after compensation is 9.4%. The content of the 5th, 7th, 9th, and 11th harmonic currents is less than 5%.

A-phase digital simulation results of using the active filter to compensate for the harmonics generated by the three-phase 6-pulse rectifier load

Harmonic Analysis of Phase A System Current

3.2 Data sampling and processing

The DSP processor synchronously samples the three-phase current and voltage signals on the load side and the current signal output by the active filter, and then performs data processing. The instantaneous active and reactive power are calculated based on the current and voltage values ​​on the load side, and then the reference value of the compensation current is calculated through the harmonic detection and separation algorithm. The difference between this value and the actual compensation current of the active filter is used to obtain the corresponding control signal through the PI control link.

3.3 High-level protection and reset functions of the controller

Once the active filter is overcurrent or overvoltage, the protection device will be activated to lock the IGBT and put the active filter in a locked state. At this time, the controller will make a judgment based on the system status and the status of the active filter itself. If both return to normal, the controller will choose an appropriate time to reset the active filter to restore it to normal operation [2].

4 High-precision pulse generator

In the past, the pulse width modulation implementation scheme based on single chip microcomputer was difficult to ensure the pulse accuracy due to the long execution time of the processor instruction, and was also significantly affected by phase jitter [4]. The fast computing power of digital signal processor makes it possible to use microprocessor structure to realize high-precision pulse generator. This method can generate various types of PWM pulses by modifying the program of the pulse generation part. It is simple, flexible and has good versatility [5].

4.1 Relationship between converter pulse signals

Figures 7(a) and 7(b) are the structural diagram and working principle diagram of the single-phase bridge voltage inverter based on IGBT. Assuming that the semiconductor switch in the figure is an ideal switch, the conduction and shutdown of the two switches in the same bridge arm are complementary (because the two switches in the same bridge arm cannot be turned on at the same time, otherwise the DC power supply will be short-circuited due to the bridge arm being directly connected). Assuming that the upper switch (SL and SR in Figure (a)) is turned on and the lower switch (SL′ and SR′ in Figure (a)) is turned off, the switch state is 1, otherwise it is 0. If both tubes are turned on at any time, the possible combinations of the single-phase bridge IGBT switch states are only 10 and 01, and the output voltages correspond to +Ed and -Ed respectively.

Structural diagram and working principle diagram of single-phase bridge voltage inverter based on IGBT

In this way, a 6-bit status word can be used to represent the output voltage of the three single-phase full-bridge inverter. For example, 100110B means that the output voltage at this time is A phase +Ed, B phase -Ed, and C phase +Ed.

4.2 Pulse generator software and hardware architecture and implementation

This system uses SPWM to compare the amplitude of the carrier wave with the reference wave, and determines the state of the output switch according to the comparison result. The design goal of this active filter system is to eliminate harmonics below the 25th order (1.25 kHz), that is, the highest frequency of the reference wave is 1.25kHz. According to the sampling theorem, the sampling frequency must be greater than or equal to twice the original signal frequency to maintain all the information of the original signal. Therefore, the minimum frequency of the carrier wave (triangle wave) in this system should be 2.5 kHz. Considering that increasing the frequency of the modulation wave increases the switching frequency of the power component and increases the loss, the frequency of the triangle wave in this system is 2.5kHz. Since the digital discretization method is used to compare the carrier wave and the reference wave, the higher the sampling frequency of the two signals, the smaller the error. Considering the real-time processing capability of the digital signal processor, this system adopts a method of comparing every 0.3°, that is, the sampling frequency is 60 kHz. Since the frequency of the periodic triangle wave is 2.5kHz, only 24 points of amplitude information are needed to meet the requirements. In actual application, two tables are constructed in the program, one is a 24-point modulated triangle wave amplitude table, and the other is a reference wave amplitude table, that is, a reference wave amplitude table with a total of 1200 points at intervals of 0.3° between 0° and 360°. The reference wave amplitude is provided by another control chip, and the digital interface of this system is provided through a dual-port RAM.

The hardware structure of the pulse generator is shown in Figure 8. The controller in the figure is implemented by another DSP chip (TMS320C31), and the output control variable is the reference value of the inverter output voltage. The two DSP chips exchange data through the dual-port RAM. The synchronous signal generation circuit completes the filtering and shaping of the grid voltage signal, and generates a narrow pulse to apply for external interrupt to the DSP at each negative zero crossing of the sine signal. The carrier value table is stored in the on-chip RAM. Each interrupt cycle refreshes the value in the index register to update the position of the data in the current table, so as to compare with the amplitude of the reference wave in the dual-port RAM. Timer 0 is triggered by the external synchronous pulse and converts the angle information value into the corresponding clock cycle number and loads it into the cycle counter of timer 1 and the serial port counter, which is used to trigger counter 1 and the serial port interrupt program.

Hardware structure of pulse generator

The software structure corresponding to the hardware structure is shown in Figure 9. System initialization includes writing control words, variable assignment, determining storage addresses, etc. Start timer O in the external interrupt service program, that is, execute the main program of the system. The frequency of the synchronization signal is calculated with the time interval between two consecutive negative zero crossings as the period, and converted into the corresponding number of clock cycles. The timer interrupt program is mainly used to maintain the synchronization of the trigger pulse and initialize the address register for table lookup, and save the last angle information. The serial port interrupt program is used to compare the amplitude of the reference wave and the modulation wave. The reference wave used for comparison each time is the three-phase amplitude. The corresponding bit of the status word sent is determined to be 1 or 0 based on the comparison result. Since the main circuit adopts a three-phase bridge structure and requires 6 trigger pulses, the status word is 6 bits. The status word is refreshed in real time according to the comparison result. The status word forms a continuous pulse after being latched by the output latch.

Software structure corresponding to hardware structure

4.3 Test results

FIG10 is a harmonic analysis diagram of the A-phase PWM pulse measured by a FLUKE 41B harmonic analyzer when the modulation wave is a fundamental sine wave superimposed with 11th and 13th harmonics, wherein the amplitudes of the 11th and 13th harmonics are both 1/4 of the amplitude of the fundamental wave.

Harmonic analysis diagram of phase A PWM pulse measured by FLUKE 41B harmonic analyzer when the modulation wave is the fundamental sine wave superimposed with 11th and 13th harmonics

5 Conclusion

This paper takes advantage of the digital signal processor's fast computing speed, high calculation accuracy, and accurate timing to design a pulse generator and controller based on the TMS320C31 DSP. The characteristics, structure, control method, and main functions of the active filter controller are introduced in detail, as well as the design, hardware and software structure, and field test results of the pulse generator. The field test and digital simulation results show that the pulse generator has high accuracy and good stability, and the performance of the controller meets the design requirements.

Reference address:Design of Active Filter Controller Based on DSP Processor

Previous article:Choosing the Right MOSFET Driver for Your Application
Next article:Research on Shot Noise Test Method of Short Channel MOSFET

Latest Analog Electronics Articles
Change More Related Popular Components

EEWorld
subscription
account

EEWorld
service
account

Automotive
development
circle

About Us Customer Service Contact Information Datasheet Sitemap LatestNews


Room 1530, 15th Floor, Building B, No.18 Zhongguancun Street, Haidian District, Beijing, Postal Code: 100190 China Telephone: 008610 8235 0740

Copyright © 2005-2024 EEWORLD.com.cn, Inc. All rights reserved 京ICP证060456号 京ICP备10001474号-1 电信业务审批[2006]字第258号函 京公网安备 11010802033920号