Keywords

1 Introduction

A UPQC derives from the UPFC [1] that is present at the level of distribution. It is composed of joined shunt and series converters for the consequent adjustment of current and voltage unbalance available in a supply feeder [2]. It restricts the reactive, unbalances, and harmonic power that is being demanded by the load. The best solution to attain a healthy power distribution system is to use the MC-UPQC [3].

The power electronics-oriented devices draw the reactive and harmonic power from the supply owing to their inherent non-linearity. In the case of three-phase systems, they create unbalance and produce a vast amount of neutral currents [4]. The excessive neutral currents, unbalance, reactive power burden, and injected harmonics create worst power factor and less system efficiency. Additionally, the power system is given to several transients such as flickers, swells, and voltage sags. These transients damage the voltage that is present at the distribution levels [5]. The excessive reactive power in the loads enhances the transmission losses of the lines and also enhances the producing capacity of generating stations. Therefore, it becomes necessary to supply the reactive power at the load ends [6].

Power quality enhancement by means of Artificial Intelligence (AI)-controlled strategies is the recent research field in the area of custom power devices that are related to the electrical distribution systems. The intelligent control strategies, which employ AI approaches like artificial bee colony algorithm, fuzzy logic, and Artificial Neural Networks (anns) are the best substitutes for the traditional control strategies in identifying the power quality enhancement in the traditional scenarios [7]. The power quality identification is composed of performing load balancing, eradicating the harmonics that occur because of the non-linear loads, and handling the voltage quality with respect to the constant frequency and load imperfections [3, 8, 9]. Nowadays, due to the enhancement in the non-linear power electronic equipment usage, power quality seems to be the main conflict, and hybrid AI approaches are in the area of research for the UPQCS.

The major improvement of the paper is as follows.

  • To manage the power quality in the distribution system by means of a novel controller-oriented MC-UPQC.

  • To implement the control schemes of MC-UPQC by means of the fuzzy theory with SRF, thereby reducing the THD.

  • To evaluate the effectiveness of the developed fuzzy controlled-MC-UPQC by comparing it with the existing models.

The organization of the paper is shown as Sect. 1 provides the introduction of the MC-UPQC. The literature related works of the MC-UPQC are described in Sect. 2. The enhanced MC-UPQC for power quality improvement is explained in Sect. 3. Section 4 provides the fuzzy controller-oriented MC-UPQC for power quality improvement. The results and discussions are explained in Sect. 5. Section 6 ends the paper.

2 Literature Survey

2.1 Related Works

In 2018, Nagireddy et al. [10] have proposed a model of the hybrid fuzzy backpropagation control strategy. The reference currents were described with the help of the back-propagation algorithms that contain the load and source currents as the input control parameters. This investigation was done in a multilevel UPQC. The outcomes of the mitigation of the voltage sag, load balancing, dynamic performance, and total harmonic distortion were examined by means of the MATLAB/Simulink.

In 2015, Boddepalli and Sangameswara [11] have addressed the control and design operation of MC-UPQC. A series VSC was included in the nearby feeder. The device was joined among multiple feeders that arise from distinct substations. The combined system was used to reduce the voltage and the current fluctuations. The control schemes used here were the p-q theory and d-q method that were composed of extended mathematical modeling. The artificial intelligence techniques were used to reduce the harmonics by means of the MATLAB/SIMULINK Software.

In 2013, Boddepalli and Raju [12] have developed a new structure for a three-phase four-wire (3P4W) distribution system that employed the MC-UPQC. It also proposed a novel control scheme for handling the unbalanced load currents. The simulation outcomes revealed the efficiency of the MC-UPQC-oriented 3P4W distribution system. The control schemes were based on Neuro-Fuzzy controller, FUZZY, and PI of the MC-UPQC.

In 2015, Mahanty [13] has introduced a new MC-UPQC for continuous adjustment of current and voltage in multifeeder/multibus distribution systems. A MC-UPQC joined multiple apfs and one shunt APF. MC-UPQC compensated load current unbalances and supply voltage on the main feeder and complete adjustment of supply voltage unbalances on the remaining feeders. Here, a two-feeder MC-UPQC was employed that was composed of two series apfs and one shunt APF. The entire apfs shared a common DC-link capacitor. A new hybrid fuzzy-PI controller was modeled that was non-linear and robust to parameter variations and was composed of a quick dynamic response together with the superior steady-state response. The outcomes showed that the MC-UPQC compensated load current and supply voltage imperfections on feeder 1 and completely protected the critical/sensitive load on feeder 2 against the interruption, swell/sag, and distortion.

In 2015, Gaikwad et al. [14] have proposed a MC-UPQC that has the capability of consequent compensation for the current and voltage in multifeeder/multibus systems. It consisted of multiple VSCS and one shunt VSC. The power compensated for the interruption and swell/sag. It was simulated in MATLAB Simulink and power transfer was compared among the two nearby feeders for the interruption compensation. It returned minimum Total Harmonic Distortion (THD). The major problem solved was the voltage dip.

2.2 Review

The MC-UPQC offers high current and voltage, minimizes the cost since series transformer is not necessary, easier capacity expansion, redundancy, and working at maximum rating, etc. But, it lacks from high conduction loss, centralized approach, difficult capacity expansion, etc. These challenges must be handled quickly. Few features with the challenges are shown in Table 1. Hybrid fuzzy-backpropagation control [10] regulates the DC voltage without any undershoot or overshoot beneath abnormal conditions and also estimates the gating signals for the shunt as well as the series VSCS of the multilevel UPQCS. But, within distinct levels, voltage unbalance occurs. NFC [11] achieves better compensation functions and also reduces the THD with the current and the voltage. Still, it cannot be utilized in multilevel operations. Power quality theory [12] frees the load currents from the distortion and can be executed with the help of simple analog hardware. Yet, it returns less efficiency near the nominal operating point. Hybrid Fuzzy-PI control [13] can be extended to multifeeder/multibus distribution systems with the addition of several series apfs and is robust to parameter variations, non-linear, and is composed of quick dynamic response together with steady-state performance. But, high DC-link losses are resulted. Novel multifeeder distribution system approach [14] shares the power compensation capabilities among the two nearby feeders and also handles the voltage dip problem. Still, heavy and bulky dc inductor is produced in some cases. Hence, these challenges are acted as a motivation in developing a fuzzy controlled MC-UPQC for power quality improvement.

Table 1 Features and challenges of state-of-the-art MC-UPQC methods
Full size table

3 Power Quality Improvement Using Improved MC-UPQC

3.1 MC-UPQC

The reactive current coming from the source is likely to be in-phase with the currently present feeder voltages. The feeder 1 is linked to a non-linear load and feeder 2 is linked to a linear load, hence both these feeders are joined in the MC-UPQC. This is used to handle the power quality related problems. As non-linear load is present in feeder 1, it gets affected by the unbalanced currents/voltage, harmonic distortions, and the feeder sources. Feeder 2 is composed of linear load, and so it is not composed of any interruption, current/voltage unbalance, harmonic distortion, swell, and sag. Thus, both the feeders do not contain any effects. The two series VSCs are linked by means of a series transformer. The switching harmonics get rejected by power RC high pass filter that contains a communication reactor \(L\) with the complete VSCs. It is done by verifying the single strategy of the MC-UPQC with the FLC. The diagrammatic model of the MC-UPQC is portrayed in Fig. 1.

Fig. 1
figure 1

Diagrammatic illustration of MC-UPQC

Full size image

3.2 Control Strategies

The suggested method generates the reference signals for the shunt as well as series voltage source converter. The current as well as the voltage unbalance, voltage sags, load currents of feeders, reactive and harmonic component, swell and harmonics, and source voltage distributions are extracted by the control technique.

The control algorithm that is associated with the shunt VSC block is explained here. The shunt VSC is composed of the MSRF theory and Fuzzy Logic Classifier (FLC) technique. Synchronization is achieved by a PLL with the help of the supply voltage. At every phase, it calculates the 120° phase displacement. Depending on the unit vector template, the shunt VSC follows the MSRF theory approach. The phase angle is extracted using three independent two-phase system that is denoted by lag or \({\Pi \mathord{\left/ {\vphantom {\Pi 2}} \right. \kern-\nulldelimiterspace} 2}\). The theory exists independently for the three-phase of the entire phase system. It is displayed in Eqs. (1) and (2).

$$ \left[ {\begin{array}{*{20}c} {i_{n - d} } \\ {i_{n - q} } \\ {i_{n - 0} } \\ \end{array} } \right] = \left[ {\begin{array}{*{20}c} {I_{d} } \\ {I_{q} } \\ {I_{0} } \\ \end{array} } \right]\;\;\left[ {\begin{array}{*{20}c} {i_{n - a} } \\ {i_{n - b} } \\ {i_{n - c} } \\ \end{array} } \right] $$
(1)
$$ \left[ {\begin{array}{*{20}c} {I_{d} } \\ {I_{q} } \\ {I_{0} } \\ \end{array} } \right] = \frac{2}{3}\left[ {\begin{array}{*{20}c} {\sin \;\omega z} & {\sin \left( {\omega z - \frac{2\pi }{3}} \right)} & {\sin \left( {\omega z + \frac{2\pi }{3}} \right)} \\ {\cos \;\omega z} & {\cos \left( {\omega z - \frac{2\pi }{3}} \right)} & {\cos \left( {\omega z + \frac{2\pi }{3}} \right)} \\ \frac{1}{2} & \frac{1}{2} & \frac{1}{2} \\ \end{array} } \right]\;\left[ {\begin{array}{*{20}c} {I_{a} } \\ {I_{b} } \\ {I_{c} } \\ \end{array} } \right] $$
(2)

The primary direct axis component current is given to the DC as portrayed in Eqs. (3) and (4).

$$ i_{g - d}^{rce} = \overline{i}_{nd} + \Delta I_{dc} $$
(3)
$$ i_{g - q}^{rce} = i_{n - q} $$
(4)

The power released from the DC-link capacitor minimizes the average value that is present in the DC bus voltage. These are given to the FLC that is used to lessen the error between the measured as well as the desired capacitor voltage. The controlling signal output is used as input to the shunt VSC’s current control system. The power is returned from the source and the DC capacitor voltage is stabilized with the help of the shunt VSC. Hence, feeder 1 does not contain harmonic as well as reactive component. This behavior is shown in Eq. (5).

$$ \left[ {\begin{array}{*{20}c} {i_{g - a}^{rce} } \\ {i_{g - b}^{rce} } \\ {i_{g - c}^{rce} } \\ \end{array} } \right] = \left[ {\begin{array}{*{20}c} {I_{a} } \\ {I_{b} } \\ {I_{c} } \\ \end{array} } \right]\;\;\left[ {\begin{array}{*{20}c} {i_{g - d}^{rce} } \\ {i_{g - q}^{rce} } \\ {i_{g - 0}^{rce} } \\ \end{array} } \right] $$
(5)

The addition of shunt currents to the \(abc\) reference currents takes place. The controlling of currents occurs by the sensation of the reference frame currents. The current in the shunt VSC is given to the controller part. The series VSC is composed of the MSRF theory as well as the enhanced PWM generator. The series VSC block is operated by the MSRF theory. This performance is depicted in Eqs. (6) and (7).

$$ \left[ {\begin{array}{*{20}c} {v_{n - d} } \\ {v_{n - q} } \\ {v_{n - 0} } \\ \end{array} } \right] = \left[ {\begin{array}{*{20}c} {v_{d} } \\ {v_{q} } \\ {v_{0} } \\ \end{array} } \right]\;\;\left[ {\begin{array}{*{20}c} {v_{n - a} } \\ {v_{n - b} } \\ {v_{n - c} } \\ \end{array} } \right] $$
(6)
$$ \left[ {\begin{array}{*{20}c} {v_{d} } \\ {v_{q} } \\ {v_{0} } \\ \end{array} } \right] = \frac{2}{3}\left[ {\begin{array}{*{20}c} {\sin \;\omega z} & {\sin \left( {\omega z - \frac{2\pi }{3}} \right)} & {\sin \left( {\omega z + \frac{2\pi }{3}} \right)} \\ {\cos \;\omega z} & {\cos \left( {\omega z - \frac{2\pi }{3}} \right)} & {\cos \left( {\omega z + \frac{2\pi }{3}} \right)} \\ \frac{1}{2} & \frac{1}{2} & \frac{1}{2} \\ \end{array} } \right]\;\left[ {\begin{array}{*{20}c} {v_{a} } \\ {v_{b} } \\ {v_{c} } \\ \end{array} } \right] $$
(7)

The load voltage is kept sinusoidal with the stable amplitude. Equation (8) shows the subtraction of the forecasted load synchronous reference \(dq0\) voltages from the \(V_{m - dq0}\). This is depicted in Eq. (9).

$$ \left[ {\begin{array}{*{20}c} {v_{g - d}^{rce} } \\ {v_{g - q}^{rce} } \\ {v_{g - 0}^{rce} } \\ \end{array} } \right] = \left[ {\begin{array}{*{20}c} {v_{n - d} } \\ {v_{n - q} } \\ {v_{n - 0} } \\ \end{array} } \right]\;\;\left[ {\begin{array}{*{20}c} {v_{g - d}^{\exp o} } \\ {v_{g - q}^{\exp o} } \\ {v_{g - 0}^{\exp o} } \\ \end{array} } \right] $$
(8)
$$ \left[ {\begin{array}{*{20}c} {v_{tg - a}^{rce} } \\ {v_{tg - b}^{rce} } \\ {v_{tg - c}^{rce} } \\ \end{array} } \right] = \left[ {\begin{array}{*{20}c} {v_{d} } \\ {v_{q} } \\ {v_{0} } \\ \end{array} } \right]^{ - 1} \;\left[ {\begin{array}{*{20}c} {v_{g - d}^{rce} } \\ {v_{g - q}^{rce} } \\ {v_{g - 0}^{rce} } \\ \end{array} } \right] $$
(9)

4 Fuzzy Controller-Based MC-UPQC for Power Quality Improvement

4.1 Proposed Fuzzy Controller

The main concept of the developed technique is to design a MC-UPQC that can perform the power quality issues present in the distribution system. This technique works on the principle of the SRF theory. Normally, the MC-UPQC consists of two series VSC, in which the power transformation from other feeders eliminates the “voltage sag, swell, interruption, and transient response of the system”. The Fuzzy logic controllers are very cheaper to develop. It is very useful for Natural language processing. It is also used for the modern control systems. Here, the power quality is managed by the novel controller-oriented MC-UPQC. The control strategies related to the MC-UPQC are implemented by the fuzzy theory that contains the SRF. The major aim of the fuzzy controlled-MC-UPQC is to reduce the THD. The effectiveness is measured by comparing it over the existing models. The block diagram for fuzzy controller is shown in Fig. 2.

Fig. 2
figure 2

Block diagram for fuzzy controller

Full size image

FLC [15, 16] models a system by means of the MC-UPQC with the concept of fuzzy theory. The logical variables are examined as input values that consider the continuous variables from 0 to 1. The controllers offer the linguistic schemes by means of the expert knowledge. FLC was introduced by Professor Lotfia Zadeh at California University in the year 1965. The inaccurate data is processed here. The applications cannot be revealed without the computers as well as the controllers. The controller operation works on the principle of the fuzzy rules that are made by the fuzzy set theory. Fuzzy controller compensates the power quality problem. The phases available in the fuzzy controller are the “decision-making, defuzzification, and fuzzification”. In fuzzification, the crisp value is altered to the fuzzy value. It is got by several types of fuzzifier. The fuzzy set shapes exist as triangular, trapezoidal, etc. The fuzzified process output is produced by the rule creation. The FLC model returns the dynamic performance of the small and the large signal, but it is impossible with the linear control approach. The FLC input represents the alteration of error as well as the voltage error. The membership functions are assumed to be triangular. The fuzzy approach is assumed to be the area center.

The knowledge base rules and the database output create the inference relation \(CS\) as displayed in Eq. (10). The variables are assumed to consist of fuzzy set description. The fuzzy memberships are shown in Eq. (10).

$$ \begin{gathered} CS^{{\left( {QO} \right)}} = IF\;U_{1} \;is\;GH_{1} \;AND\;U_{2} \;is\;GH_{2} \cdots U_{oi} \;is\;GH_{oi} \hfill \\ \quad \quad \quad \quad \quad \quad \quad then\;V\;is\;DS^{{\left( {QO} \right)}} \hfill \\ \end{gathered} $$
(10)

In the above equation, the term \(QO = 1,2,3 \cdots OI\), where \(OI\) denotes the rule count, \(GH_{1} ,GH_{2} , \cdots GH_{oi}\) defines the fuzzy sets, \(oi\) defines the fuzzy variable count, \(U_{1} ,U_{2} , \cdots U_{oi}\) defines the input variable vector, and \(V\) defines the output variable that is also called as control variable. The fuzzy controller computes the input signals to define the efficient control action. The FLC is implemented by the two input state variables as portrayed in Eq. (11).

$$ v_{err} \;fuzz = v_{dc} - v_{dc}^{rce} $$
(11)

The fuzzy control rules are used for \(V_{dc}\) and \(\Delta V_{dc}\). It regulates the voltage using less real loss quantity that is being taken as the output of FLC. The MSRF-oriented currents are provided to the relay and a sensing takes place in the shunt VSC control circuit.

5 Results and Discussions

5.1 Simulation Setup

The proposed fuzzy controller MC-UPQC was performed in MATLAB 2019a and the tests were carried out. The transmission of power from feeder 1 to feeder 2 was done to eliminate the “voltage sag, swell, interruption and transient response of the system”. The proposed method was compared with distinct models like “MC-UPQC, FLC-MC-UPQC, and NN-MC-UPQC” for describing that the THD is reduced in the developed method.

5.2 MC-UPQC Performance in Connection with Feeder 1 with Different Controllers

The proposed method was compared with the traditional models that are connected with feeder 1 as portrayed in Fig. 3. The outcomes revealed the “bus 1 voltage, series compensation voltage and load 1 voltage”. The source voltage is provided with sag from 0.1 to 0.2 s and it tends to swell from 0.2 to 0.3 s. If fault happens in feeder 1, then there occurs an effect in feeder 1 with voltage across the linear load having interruptions, swell, and sag. This problem is handled by connecting an FLC to the shunt VSC. This is efficiently handled by the FLC than the remaining controllers.

Fig. 3
figure 3

Analysis of proposed and existing MC-UPQC a General MC-UPQC system, b NN-MC-UPQC, and c FLC-MC-UPQC for “Bus 1, series compensation, and load 1 voltages in feeder 1”

Full size image

5.3 MC-UPQC Performance in Connection with Feeder 2 with Different Controllers

The proposed method is compared over the state-of-the-art models that are linked with the feeder 2 as in Fig. 4. It clearly portrays the “bus 1 voltage, series compensation voltage and load 1 voltage”. The disturbances and voltage sag swell are reduced by joining the system with the two-feeder system. The source voltage is applied that contains the sag from 0.1 to 0.2 s and swelling from 0.2 to 0.3 s. If fault occurs in feeder 2, then the effect of voltage occurs along the linear load that contains the interruptions, sag, and swell. This problem is solved by linking the FLC with the shunt VSC. It is effectively portrayed in the load voltage that is solved by the proposed method in an efficient manner than the various controllers.

Fig. 4
figure 4

Analysis of proposed and existing MC-UPQC a General MC-UPQC system, b NN-MC-UPQC, and c FLC-MC-UPQC for “Bus 1, series compensation, and load 1 voltages in feeder 2”

Full size image

5.4 Harmonic Analysis

The harmonic analysis of the proposed method over the traditional models that are linked with the feeder 1 and feeder 2 is portrayed in Fig. 5, respectively. It portrays the THD of the general MC-UPQC, FLC-MC-UPQC, and NN-MC-UPQC at distinct seconds. Here, the THD of the developed method is minimized over the time sequence when compared with the existing models. At 0.15 s, the THD of the proposed method is 33.33 and 20% better than general MC-UPQC and NN-MC-UPQC system. Therefore, MC-UPQC reduced the THD in the developed method than the traditional models when it is joined with the feeder 1 and feeder 2.

Fig. 5
figure 5

Harmonic Analysis of proposed and existing MC-UPQC a General MC-UPQC system, b NN-MC-UPQC, and c FLC-MC-UPQC for feeder 1 and feeder 2

Full size image

6 Conclusion

This paper has managed the power quality in distribution system by means of a new controller-oriented MC-UPQC. The fuzzy theory containing the SRF implemented the control schemes of MC-UPQC. As a major objective, the fuzzy controlled-MC-UPQC minimized the THD. MC-UPQC is used to mitigate voltage and current harmonics and for improving the voltage regulation and to compensate reactive power. The effectiveness was described by comparing it with traditional models. From the analysis, at 0.15 s, the THD of the proposed method was 33.33% and 20% better than general MC-UPQC and NN-MC-UPQC system. Hence, MC-UPQC reduced the THD in the proposed method than the traditional models when it was combined with the feeder 1 and feeder 2.