Mitigation of supply current harmonics in fuzzy-logic based 3-phase induction motor

ABSTRACT


INTRODUCTION
Induction motor drives are in use for domestic, agriculture, and industrial applications owing to the many specifications related to cost, size, portability, operational conditions, and high performance they provide to the tip users. The high-performance AC drive system has been employed with the modern power electronic converter. Power electronic converter fed induction motor (IM) drive system may introduce additional electrical noises and disturbances through harmonics injected into the power system. The cause for power quality glitches in power systems is the failure of the electrical components and the evolution of these harmonics. A diode rectifier fed motor for power factor correction circuits employing 1-ϕ active power filter fashioned out of power switches and reactive components are presented [1]. One cycle control [2] had achieved an active power filter with reduced switching loss. Harmonic mitigations have been done with fuzzy controller tuned proportional derivative controller (PID), with d-q-0 reference frame theory for three-phase power shunt active filter (SAF) enhancing power quality has been reported [3]− [5]. A similar result of the total harmonic distortion (THD) values using proportional integral PI and fuzzy logic controller (FLC) controller with RL and RC loads was presented by [6]. The FLC-based speed control of the induction motor, ANN-controlled active filter is discussed in [7] and [8]. Underbalanced and unbalanced load conditions with three-phase SAF using proportional integral (PI) and fuzzy logic controller (FLC) suggested reduction of harmonics [9]. Various ways like condenser banks, passive filters, active filters, and hybrid filters, are recommended and adopted for reducing the harmonics caused by power electronic converters and nonlinear loads. Recent research has been reported in [10]− [20] to reduce the harmonic effects in the induction motor drive with SAF integrated PI, PID, fuzzy, and artificial neural network (ANN) controllers. The above literature does not confirm the curtailment of harmonics using SAF at the supply side of a vector-controlled IM drive. The overall performance of the proposed induction motor (IM) with parallel SAF has been found out the usage of PI, PID, and fuzzy logic controller. To understand the reduction in the current harmonic analysis, the SAF is connected to the supply-side to diminish the current harmonics in power systems. In the present work, a fuzzy logic (FL) controller is implemented to improve the IM drive performance and to limit the supply current harmonics was proposed.

PROPOSED SYSTEM
Nowadays, nonlinear loads are predominantly used in many applications such as uninterruptible power supply (UPS), switched mode power supplies (SMPS), electric furnaces, fluorescent lamps, various power electronic converters and electric drives. These nonlinear loads draw non-sinusoidal currents from the supply, i.e., harmonic currents, and create severe power quality problems such as heating losses, resonance, electromagnetic interference (EMI) disturbances in a communication network, de-rating of the motor. The current SAF method has been used to reduce harmonics and improve power factors using reactive power compensation techniques are addressed [21]− [27]. The adjustable speed drive system used with a front 3ϕ diode bridge rectifier or controlled rectifier acts as a front converter. The general block diagram in Figure 1(a) illustrates the vertical speed indicator voltage source inverter (VSI) fed IM drive without an active shunt power filter and Figure 1(b) presents the general block diagram of the existing induction motor drive system with an influence of shunt active filter (SAF). In this system, a 1-ϕ input power SAF was introduced parallel at the supply side of PI, PID, and FLC-based VSI fed induction motor drive system. The in-depth simulation studies and analysis have been carried out with PI, PID, and FLC based IM drive with SAF.  Figure 2 illustrates the circuit elements within the projected system with a shunt active power filter. For DC-AC conversion, a 1-ϕ diode rectifier is utilized in the front-end converter. A DC-link capacitance is exercised for load-balancing energy storage between the 1-ϕ diode bridge rectifier and three-phase voltage supply inverter. Vector control will control the motor speed by exploiting the stator voltage.

CONTROL TECHNIQUE
Various control methods are adopted to extract harmonic compensating currents in both the frequency domain and time-domain approaches. Harmonic compensating currents are compared with the generated reference current to switch pulses to the parallel SAF and inverter. In this proposed work, 1-ϕ instantaneous power (PQ) theory is applied to control active shunt power filters and to extract harmonic compensating current. In this system, PI, PID, and Fuzzy logic controllers are adapted for generating switching signals applied to SAF and the inverter to maintain source current harmonics and boost the IM performance, were reported [28]− [31]. SAF is introduced here to yield the compensating harmonic current, which is opposite to the current drawn by non-linear loads, and it reduces the source current harmonics. The active and the reactive powers have been computed using (1).
Both the real and reactive powers consist of harmonic and fundamental components, and from (1), the compensating harmonic reference current is given by (2).
The THD of the source current can be calculated by using the (3).
Where I1 characterizes the fundamental component of source current, and Ih indicates h th harmonic current of non-linear loads contributes to the harmonic currents. Equations involved in the proposed system are given in (4) to (11).  (7) and (8).
By taking Laplace transform on the differential of PID controller mentioned in (12), the transfer function is derived, and revealed in (13).
The PID controller parameters are tuned by means of Ziegler-Nichols method in many applications. Ziegler-Nichols method are step response and frequency response methods. The PID controller parameters are tuned by Ziegler-Nichols self-oscillatory method. This method is traditionally easier and is based on measurement of ultimate or critical gain Ku and ultimate or critical period Tu from the measured quantity of closed-loop system. The required PID controller gains such as kc = 0.56, ki = 0.45 and kd = 0.38 are used in this simulation analysis.
The many control algorithms are enforced to FLC. Figure 3 shows the elementary architecture of an FLC. The FLC has fuzzification, rule evaluator, and de-fuzzification blocks. Input variables are depicted in a linguistic manner, the logical thinking mechanism takes acceptable control action supported rule base block, and the output control signals have been converted into real-time signal with a de-fuzzification block. The fuzzy controller may be a content expressed in terms of fuzzy inference rules and a fuzzy inference engine with two input variables; error voltage () and changes in error voltage (), to come up with the required signal to satisfy the real-time system and Figure 3 shows the schematic representation block of the current reference generation. In the vector control method, 3-ϕ induction motor is reflected as DC shunt motor (separately-excited), which distinctly controls the decoupled torque and flux component of the induction motor stator currents. The error signal is detected by comparing DC link voltage with actual DC bus voltage. This error () and change in error voltage () are fed to PID/FL controllers to generate the reference currents. Hysteresis controllers are employed here to generate gate pulses of shunt active filter and SVM inverter. The control block for switching pulse generation using hysteresis current controller and FLC-PID is shown in Figure 4.
Comparing the reference and actual speed of the IM drive, an error signal (e) is generated. The specified gate pulses are generated with the influence of an error signal in the course of PI, PID, and FL controller. The provision of current harmonics has been regulated through the parallel SAF. Still, the FL controller provides higher performance with reduced supply current THDs. Within the system, the Mamdani kind, logical thinking engine has been adopted as a result of its just like human input. Error () and change in error () signals are delineated through seven fuzzy linguistic states negative-big (NB), negative-medium (NM), negative-small (NS), zero (ZE), positive-small (PS), positive-medium (PM), positive-big (PB)). Fuzzy variables are characterized by triangular membership functions within the fuzzification method, and therefore the PLL management approach is estimated by MATLAB/Simulink power tools. Fuzzy inference system (FIS) plays a vital role in the control system, i.e., in the fuzzy logic controller, which is defined in Table 1.  The control of the asynchronous motor drive based on the PI and PID controller without an active filter is simulated with the MATLAB/Simulink software. Here, the PI controller regulates the motor speed. The frequency FFT analysis of the asynchronous motor based on a PI controller without an active filter as shown in Figure 5(a), and the source current harmonics (THD) is found at 20.29%. Figure 5(b). depicts the THD of the drive, employing a PID controller, with no SAF, and the THD is 13.60%. The simulation parameters for the suggested system are given in Table 2.  Closed-loop control of the IM drive employing PI and PID controllers with SAF implementation, is laid out in Figure 6. The THD of the supply current, employing PI and PID controllers with SAF based IM drive, is revealed in Figure 6, and the THD of the supply current is found as 16.51% in Figure 6(a). From the frequency spectrum, Figure 6(b) depicts the THD of the induction motor drive with PID-SAF is found as 3.86%.
(a) (b) Figure 6. THD of the induction motor with SAF (a) PI controller and (b) PID controller A fuzzy controller-based induction motor with a parallel shunt active filter (SAF) is laid out in Figure 7. The speed and torque responses of IM drive control by means of the FL controller with SAF are as shown in Figure 8(a) and Figure 8(b) respectively. The speed and torque responses of the FLC-IM by means of SAF is exhibited in Figure 9(a) and Figure 9(b). From the FFT analysis, the THD of the FLC-based IM drive without SAF is found at 7.52% in Figure 10(a). From the FFT analysis, the THD of the FLC-SAF controlled induction motor drive revealed in Figure 10(b) was found as 2.66%. The FLC is considerably enhancing the speed and torque response of the IM drive when compared with PI and PID controllers. Additionally, FLC reduces the source current harmonics %THD is from 3.86 % to 2.66%.  Table 3. Comparison of source current harmonics (%THD) of the PI, PID & FL controller-based IM, with and without parallel SAF is depicted in Figure 11. Comparative analysis of supply current harmonics (%THDs) of PI, PID, and FL controller-based induction motor drive implementing with & without parallel SAF is tabulated in Table 4. The FLC is considerably enhancing the speed and torque waveform response of the induction motor drive when compared with PI and PID controllers. Additionally, FLC reduces the source current harmonics %THD is from 3.86 % to 2.66%. It was reported that the %THD is 2.66% for using Fuzzy logic controller. Therefore, in relation to the PID controller, the Fuzzy logic controller improves the dynamic response of the induction motor and the reduction of the source current THD.

CONCLUSION
In this work, the supply current harmonics are minimized by using parallel-connected SAF at the end. The system has a three-phase IM drive in a closed loop with PI, PID, and Fuzzy logic controllers, which are analyzed and simulated using MATLAB/ Simulink software. Simulation results indicate the dynamic response of the Fuzzy logic-controlled induction motor drive system could be advanced to PI and PIDcontrolled induction motor drive systems. The main reason is, the settling time of the system was reduced to 0.4sec by including FLC. The total harmonic distortion of supply current (%THD), employing FLC with shunt active filter for induction motor drive is 2.66%, that's beneath the 5% restriction of the global standards (IEC 61000-3-2 and IEEE 519). Therefore, the Fuzzy logic-controlled IM system is superior to the PIDcontrolled system.