Th Fuzzy Logic Based Space Vector Computer Science Essay

Published Date: 02 Nov 2017

1 Mr.Jarupula Somlal, 2 Dr.M.Venu Gopala Rao.,3 Maddu Anusha Priya

1 Professor, 2 Assistant Professor, 3M.Tech Student

1,2,3 EEE Department, KL University,Vijayawada,A.P,India-522502

1 [email protected],2 [email protected],3 [email protected]

Abstract: This paper investigates with a Hybrid Active Power Filter with fuzzy logic based space vector PWM controller for mitigating the harmonics, improving the power factor and increasing the distribution power of the three phase distribution system. This paper concluded that the power conditioning by fuzzy logic based SVPWM hybrid active power filter is superior than PI controller based SVPWM technique. In the proposed control filter three control circuits are used such as PI control unit, fuzzy unit SVPWM control unit. Fuzzy arithmetic’s are used for adjusting proportional–integral coefficients timely. The desired output voltage is generated based on generated reference voltage by fuzzy based SVPWM. A MATLAB code is developed to generate the SVPWM switching pulses fed to the two-level inverter topology. Simulations are carried out using MATLAB. It is found that the %THD has been improved from 2.67 to 1.57and power factor is improved to 0.9718. The simulation results shows that the effectiveness and feasibility of the proposed filter.

Key Words: Hybrid Active Power Filter, Fuzzy Logic Controller, IGBT Inverter, Space Vector PWM, Total Harmonic Distortion (THD).

I.INTRODUCTION

Power quality is the main problem that the industry is facing today. The quality of power has been deteriorating with the presence of various current and voltage harmonics, low power factor, voltage sags and swells, flicker and many other disturbances. Among the various disturbances, Harmonic distortion is one of the most serious power quality problems. Particularly, in the distribution systems, harmonics are the major concerned problem. The growing use of electronic equipments is one of the major causes to impute the harmonics, which led to distortion of voltage and current waveforms and increased reactive power demand in ac mains as they pass through the system impedance.

However, in the present situation various power quality improvement solutions are available; Isolate harmonic loads on separate circuits (with or without harmonic filters), Harmonic mitigating transformers, Phase shifting (zig-zag) transformers, Filter capacitor banks, Line Reactors, K-Rated / Drive Isolation Transformers, Harmonic Mitigating / Phase Shifting Transformers, Passive parallel / series tuned Filters and Active Filters.

Passive filtering is the simplest conventional solution to reduce the harmonics. But they have many demerits such as; a) the number of passive filters installed would depend on the number of harmonic component to be compensated, this demands for the information of harmonic content to be know in advance. b) These cannot function under the saturated conditions, c) At some frequencies, these filters may lead to resonance. All the above demerits of the filters are overcome by the use of active filters. But, for high-power applications, the Active filters are not cost effective due to their large rating and high switching-frequency requirement of the pulse width modulation inverter.

For harmonic current tracking controls, there are two schemes .One is the linear current control and the other is nonlinear current control. Hysteresis nonlinear control method is simple but leads to a widely varying switching frequency. This limitation has been improved with variable hysteresis band switching strategies but it requires a complex controller to achieve satisfactory performance. Predictive current control offers the best potential for precise current control, but the implementation of a practical system can be difficult and complex. In this paper, Fuzzy based SVPWM controller was proposed. The proposed controller filter shows shorter response time and higher control precision. The simulation and experimental results also show that the new control method is not only easy to be calculated and implemented, but also very effective in reducing harmonics.

II.PRINCIPAL OF OPERATION

In each switching cycle the controller samples are the supply current ia ,ib & ic are calculated as

-ia= ib + ic (1)

As the summation of three supply currents is zero. These three phase supply currents are measured & transformed into direct & quadrature axis components of two dimensional plane. The fundamental component of supply currents are transformed into d-q axis & supply current amplitude Is is generated. That Is is controlled by the fuzzy controller with Vdc & Vref (Reference value of DC bus voltage). The output of fuzzy controller is equivalent to reference voltage vector .By using Fourier magnitude block, voltage magnitude & angle is calculated. From the obtained signal, these values are fed to developed code & compare to the relative sequence. The generated switching actions are applied to & balancing of the filter takes place.

III. HYBRID ACTIVE POWER FILTER

The Fig 3.1shows the hybrid APF with non linear load consist of both active & Passive Filter. The Passive filter connected in shunt with the distribution system and is tuned to present low impedance at a particular harmonic current. The shunt active passive filter takes a three phase voltage source inverter as the main circuit & uses capacitor (C) as the Voltage storage element on the DC side to maintain the DC bus voltage Vdc constant. The hybrid active power filter is implemented with fuzzy based SVPWM current controlled voltage source inverter (VSI) and is connected at the point of common coupling for compensating the current harmonics and reactive power. This system is investigated and the performances of parameters are verified under different non-linear load conditions.

It can be assumed that the supply voltage and current is ideal and sinusoidal and the three-phase balanced parameters are shown as below:

Xsa=XsSin(wt) (2)

Xsb=XsSin(wt-2Ï€/3) (3)

Xsc=XsSin(wt+2Ï€/3) (4)

Where Xs represents the supply voltage or supply current, If equations (2),(3) and (4) are the three –phase voltages [ Xsa Xsb Xsc ] in a-b-c can be expressed as two-phase representation in d-q reference frame by Clark’s transformation and it is given by equation (5).

(5)

Above equation can be reduced as

[Xs] =2/3[Xsa a0+Xsb a1+Xsc a2] = Xd + jXq =Xs Ï´s

(6) Where a=ej2/3Ï€ ,Ï´s is angle of supply voltage or supply current.

Fig.3.1.Hybrid active power filter construction for non linear load

IV.COMPENSATION PRINCIPLE

The equivalent circuit and equivalent impedance circuit model of hybrid active power filter are shown in Fig.4.1 and Fig.4.2 respectively. Equation (7) and (8) shows source current and load current in the circuit.

Fig.4.1 Equivalent circuit of hybrid active power filter

Is = (7)

Il = - Iaf (8)

Fig. 4.2 Equivalent impedance circuit model

Total impedance of hybrid active power filter

Zeq= (9)

Total voltage from Fig.4.2 is calculated by

Vs-Is Zs=IL Zeq (10)

V.CONTROL STRATEGIES OF HYBRID ACTIVE POWER FILTER

a. Control Block Diagram of Fuzzy Based SVPWM Controller

Fig.5.1 shows a fuzzy based SVPWM controller to generate the required pulses for inverter operation. FLC is the best control method of reference frame. The three phase supply currents are measured & transformed into synchronous d-q axis then time constant t1, t2, to on time of v1, v2 and v0 are calculated and the generated switching actions are applied to active power filter and inverter of active power filter will generate the desired compensating harmonics. The Active Power Filter inject an equal but opposite distortion harmonics back into the power line and cancel with the original distorted harmonics on the line.

.

Fig.5.1 Control block diagram of Fuzzy based SVPWM

VI. FUZZY LOGIC CONTROLLER

Once the fuzzy controller were developed and incorporated into the simulated system, the simulation performances helped in the iteration of the controllers and best adaptive controller to the linear and non linear systems. Fuzzy controller main parts are evaluation and control rules from the rule base and data base is called fuzzifier and defuzzifier is takes highest MF component . Then the Fuzzy controller is fine tuned in several stages of that iteration.

Fig.6.1.Block diagram of Fuzzy Logic Controller

Recently, fuzzy logic controllers (FLC) [8] have received a great deal of attention for their application in active power filters (APFs). The advantages of FLC over conventional controllers are that they do not require an accurate mathematical model, can work with imprecise inputs, can handle non-linearity and are more robust than the conventional controllers. The Mamdani type of FLC is used for the control of an APF and it gives better results, but it has the drawback of a larger number of fuzzy rules. The FLC having different membership functions (M.Fs) to analyze the performance of instantaneous real active and reactive current (id–iq) control strategy [9] for extracting reference currents of SHAF under different source voltage conditions. PWM pattern generation based on carrier less hysteresis current control is used for quick response. In addition, the id–iq method is used for obtaining reference currents in the system, because in this strategy, angle ‘u’ is calculated directly from the main voltages and enables operation to be frequency independent; thereby, this technique avoids a large number of synchronization problems. The fuzzy inference system (FIS) in the Fuzzy

Fuzzy logic toolbox, it consists of FIS editor, M.F editor, Rule editor, Rule viewer, Surface viewer. Membership functions are symmetrical or asymmetrical .It is multi dimensional . Membership functions shows multi curves and forming hypersurface.In fig 6.2. andFig.6.3.M.F editor of kp and ki having two inputs (error, cerror)and single output curves (Kp and Ki ).we have different types of membership functions: They are Triangular ,Trapezoidal ,sigmoid Tripozidal,Sigmoid,Gaussian,delta .in these functions levels of information is not adequate and differ with interval –valued membership functions .They are not to relative frequencies ,as it is probability density function. The M.F editor is used to define the shapes of all the M.Fs

associated with each variable. The rule editor is used for editing the list of control rules that define the behavior of the system. In the present model 49 rules are developed in below. Fig .6.4.and fig.6.5.shows rule view of Kp and Ki. Fig.6.6 and fig.6.7. shows hyper surface view of Kp and Ki are simply come from control rules and also positive high when both e and ce are positive high.

Fig. 6.2.MF editor for kp Fig.6.3.MF editor for ki

Fig.6.4 Rule view for Kp Fig.6.5Rule view for ki

Fig .6.6.HyperSurface view of kp Fig.6.7.HyperSurface view of ki

The fuzzy control rule design involves defining rules that relates to the output model properties. For designing the control rule base for tuning ΔKp and ΔKi, the following important factors have been taken into account.

1) For large values of /e/, a large Δkp is required, and for small values of /e/,a small Δkp is required.

2) For e, ec >0, a large Δkp is required and for e, ec >0 a small Δkp is required.

3) For large values of /e/ and /ec/, ΔKp is set to zero, which can avoid control saturation.

4) For small values of /e/, ΔKp is effective ,and Δkp is larger when /e/ is smaller ,which is better to decrease the steady state error .so the tuning rule of ΔKp and ΔKi can be obtained

Table:6.1. Adjusting parameters of ΔKp Table:6.2. Adjusting parameters of ΔkI

ΔKP

ec

NB

NM

NS

O

PS

PM

PB

e

NB

PB

PB

NB

PM

PS

PS

0

NM

PB

PB

NM

PM

PS

0

0

NS

PM

PM

NS

PS

0

NS

NM

0

PM

PS

0

0

NS

NM

NM

PS

PS

PS

0

NS

NS

NM

NM

PM

0

0

NS

NM

NM

NM

NB

PB

0

NS

NS

NM

NM

NB

NB

ec

ΔKi

NB

NM

NS

0

PS

PM

PB

e

NB

0

0

NB

NM

NM

0

0

NM

0

0

NM

NM

NS

0

0

NS

0

0

NS

NS

0

0

0

0

0

0

NS

NM

PS

0

0

PS

0

0

0

PS

PS

0

0

PM

0

0

PS

PM

PM

0

0

PB

0

0

NS

PM

PB

0

0

VII. SIMULATION RESULTS

The proposed methods are implemented by simulation and simulation results are obtained by using MATLAB circuit design . simulation results of wave forms of source voltage ,load current and source currents Fig:7.1.shows wave form of PID controlled SVPWM Fig:7.2.shows wave form of fuzzy based

(a) (a)

(b) (b)

Fig.7.1.Wave Forms Of PID controlled SVPWM (a)Source Voltage, Fig. 7.2.Wave Forms oFuzzy based SVPWM (a)Source Voltage, (b)load currnet and (c)Source Current (b)Load Current and (c)Source Current

%THD values at load current and source current wave forms are shown in below ,they are Fig:7.3.load current harmonic spectrum for PID based SVPWM ,Fig: 7.4.source current harmonic spectrum for PIDbased SVPWM . Fig:7.5.load current harmonic spectrum for fuzzy based SVPWM ,Fig: 7.6.source current harmonic spectrum for fuzzy based SVPWM .

Fig.7.3. load current Harmonic representation for PID based SVPWM Fig.7.4. source current harmonic spectrum of PID based SVPWM

Fig.7.5. load current Harmonic representation for Fuzzy SVPWM Fig.7.6. source current harmonic spectrum of Fuzzy SVPWM

VII. CONCLUSION

In this paper , a control methodology for the APF using FUZZY LOGIC CONTROL based SVPWM proposed. The performance of APF with these methods done in MATLAB / Simulink. Comparison of %THD and POWER FACTOR of WITH SVPWM and FLC based SVPWM.

Table.7.1.

% THD

Power factor

WITH SVPWM

2.67

0.9678

FLC BASED SVPWM

1.57

0.9718