ELSEVIER

Available online at www.sciencedirect.com

ScienceDirect

Procedía Technology 23 (2016) 240 - 247

Procedía

Technology

3rd International Conference on Innovations in Automation and Mechatronics Engineering,

ICIAME 2016

A Critical Evaluation of Advanced Multi-Carrier Modulation Scheme for 15-Level Inverter via PSO-PID Controller

*P. Hemachandu a, Dr. V. C. Veera Reddyb

aReseach Scholar, Department of EEE, SV University, Tirupathi, AndhraPradesh, India., bProfessor, Department of EEE, SV University,

Tirupathi, Andhra Pradesh, India, .,

Abstract

Particle swarm optimizing mechanism has been shown as a fruitful problem solving tool in a global optimization areas. So far, most of this algorithm is working properly by utilizing an imperative learning pattern, which process the all swarms utilizes in the unique strategy. The grid integrated system is interpretation of stand-by module to co-generation scheme; it combines a PV/FC arrangement with a high recognized power conditioning. Utilization of high voltage gain power conditioning units which plays a crucial role in a grid connected system with a single module. The module ranges from 100V to 300 V getting from 10V to 40 V input, then incorporated to micro-grid system via proposed asymmetrical inverter with PSO-PID controller. The formal tuning of the PID controller generates an unreasonable peak over-shoot; more error content will be counteracting by meta-heuristics approaches by a swarm optimizer. This controller forecasts the optimal modulation index & switching angles for attaining enhanced output voltages, in-limit THD resolution & averts the sudden variations. In this behaviour, the proposed scheme provides perfect sinusoidal grid voltages which are in-phase with the current, good harmonic limitation values by using PSO-PID controller. A simulink model is implemented to validate the appearance of proposed 15-level asymmetrical inverter topology with PSO-PID controller by advanced modulation schemes using Matlab/Simulink platform & results are conferred.

© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.Org/licenses/by-nc-nd/4.0/).

Peer-reviewunder responsibilityofthe organizingcommittee ofICIAME2016

Keywords: Adaptive Neuro-Fuzzy Inference System (ANFIS); Asymmetrical Modern Multi-level Inverter; Fuel Cell Stacks; PV Arrays; Total Harmonic Distortion (THD).

1. Introduction

At present control objective research area, the evolutionary computation tool performs an active tool over past decades. The global benchmark optimized concerns have overtake more complex from easy uni-modal functions,

* P. Hemachandu, E-mail address: chandusvuphd@gmail.com.

2212-0173 © 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Peer-review under responsibility of the organizing committee of ICIAME 2016 doi:10.1016/j.protcy.2016.03.023

finding the global optimum of a particular function has overtaken the more practical as well as impossible for many concerns. Moreover, this type of imperative algorithm structure is always needed for real world concerns with propitious results with best local minima value. The perception of renewable energy sources, which influences the favorable advantages [1] which are potential to paw the stand-by micro-grid system with merely naïve connectivity [2]. Attaining the safe, clean, eco-friendly specifications, the fuel cell (FC) & photo-voltaic (PV) stacks are virtually used as primary power generating CO-units. The CO or Combinational generating units extends the established flexibility & delightful scalability for the supremacy of incredible energy management theme, mostly preferred for remote area locations. Actually, this co-generation scheme is merely efficacious to partial shading, diminished energy system yields, system monitoring [1]-[4].

The AC module, which should be recommended to eradicate these problems, consists of several units of regular input/inverter sources. A very eminent power density is acquired by isolated formation of FC/PV have better dynamic stability, minimizing the fast load steps by the support of co-generation scheme [4]. These PV/FC systems are isolated to micro-grid system by using power conditioning units for floating the power with continuing the stable dispatch. There are two patterns such as; (i) high voltage gain DC/DC converter, (ii) interfaced DC-AC inverter module, high voltage DC/DC converters are strengthen the FC/PC stack voltage at DC bus & interfaced DC-AC inverter are behaves the barrier between the DC bus & micro-grid system.

fHOPOSKD 1S-I.LVEL MÜIJ1RN \ll II I I t\ hi INYEKTKK H£D [ MIC:KO GRID SYSTEM WITH OPTIMAL I'H) PARAMETERS TUNING L'SINC; SWARM OPTIMIZATION

II1GII STEP I1" IIOOST CONVERTER MODULES

Fig.1 Overall Design of Proposed CO-Generation Scheme with Effective PID-PSO controller for micro-grid system via

Reduced Switch Type-Modern Asymmetrical 15-Level Inverter

Amelioration of quality voltage & incredible efficiency is improved by counteract the utmost switching stress by incorporating the multi-level inverter structure [4] -[6]. The multi-level terminology has been surveyed by few years ago & attaining the wonderful advantages such as harmonic eradication, greater qualified voltage, low THD content, incredible efficiency, low EMI loss, low dv/dt stress coming from many number of voltage levels by administering the very advanced topologies. The general characteristics of classical multi-level inverters are neutral-point clamped type [7], balanced/flying capacitor type [8], cascaded H-bridge type [9], are vital regular topologies by so many surveys, these are useless because of uneven disadvantages. Symmetrical cascaded connected H-bridge topologies are more popular due to many switches/many sources, based on this concern most of concepts are related to asymmetrical topologies. Flying capacitor/diode clamped types requires more switching components & difficult to generate asymmetrical voltages. The main affection while preferring asymmetrical topology, it may provide the more number of levels with fewer switches by the consisted unequal DC sources, space requirement, and cost.

Multiple unique MLI structures have been implemented by many literatures in [5]-[9].

This paper proposes a new advanced asymmetrical multi-level inverter topology has been evaluated by using multi-carrier modulation schemes by optimal tuning of PID gains by PSO algorithm. Acquiring the optimum values of modulation index & switching angles for perfect operation of device & getting the good results, which are very realistic THD resolution, minimize the steady state error; eradication of the peak overshoot, incredible stability. The main distinction of proposed algorithm is a stochastic nature based on selection principle & natural search. There are many evidences of intelligence for well-posed domains on plants, normal living things & animals. The optimal parameters effectively control the proposed 15-level inverter to minimize the transient stability, acquiring qualified RMS voltage, stability index. Various comparisons are made for distinct control objectives with proposed intelligent over classical controllers. Finally, formalize the proposed module for micro-grid system with optimal tuning of PID gains by PSO algorithm is evaluated by Matlab/Simulink platform & results are conferred.

2. Proposed Co-Generation Scheme

Fig. 1 illustrates the overall design of proposed co-generation scheme with effective PID-PSO controller for micro-grid system via an advanced asymmetrical 15-level inverter. This co-generation scheme constitutes the merge of photo-voltaic (PV) & fuel cell stacks are required for irresistible power generation attitude which have incredible efficiency, greater reliability, tidiness. The activity of power conversion of PV/FC outcome power into self-reliant unique voltage sources with myriad relationship which are integrated to power inverter module through a high voltage gain DC-DC converter to regulate the DC link voltage as a constant. The turn's ratio of the coupled inductor is amplified by voltage gain & the secondary winding of the coupled inductor is series connection with respect to a switching capacitor for acquiring the enhanced voltage.

The zeta converter is aligned from the M1 coupled inductor with streamed active switch Sa. The primary winding N1 of a coupled inductor M1 is akin to the source inductor of the formal boost converter, barred from the D1 diode & C1 capacitor. The N2 secondary winding is appended with alternative pair of D2 diode & C2 capacitor; all these components are cascaded with N1. The rectified D3 diode is connected to its output C3 capacitor & DC bus. The distinction of Zeta converter is; low component stress, high voltage gain, recycling the coupled inductor leakage energy, restrained the voltage spikes at active switch, low cost, greater efficiency, the active switch isolates the input source during the non-working conditions, thus prevents the electric hazards to facilitate the human beings. The advanced asymmetrical multi-level inverter constitutes of three sources, seven switches, three balanced diodes, the input voltages are 100v, 200v, 400v, represent as a Vdc produces the 7-level DC voltage by presence of sub-multi-cell module.

The inclusive 15-level output voltage is V(t)=Vdc+2Vdc+4Vdc by switching the S1, S2, S3 switches. The basic unit of asymmetrical structure constitutes of 3 switches/diodes/sources (level generation), besides with a full bridge inverter module (polarity generation). Acquiring the respective voltage levels by acceptable switching between the switches & provides the positive, negative, zero level voltages. This full bridge converter converts the 7-level DC voltage into 15-level AC voltage with the help of polarity generation concept & is interfaced to micro-grid via filter inductors. In this process, the proposed scheme produces the pure sinusoidal outcome of both currents/voltages and these are in-phase with each other related to grid voltage then integrated to grid voltage. This converter requires only 7 switches for production of 15-levels, in order to amplify the conversions structure which has several advantages like low complex function, low switch stress, very low cost conversion, and greater efficiency & finally called as a smart inverter. In this scheme, fewer switches are casted-off to regulate the 15-levels & fabricate the user friendly module with low switch loss by minimum number of switches. Requirement of specific power conversion device & maximum outcome voltage levels for proposed module is contrived as;

Vomax = (2k -l)*Vdc (1)

n,cbt,s = (n + 4) (2)

NLevel,s = 2(k+1) - 1 _ _ (3)

Where the k represents the number of voltage sources used in proposed inverter module. The detailed switching scheme is illustrated in Table I & II. The grid voltage is recognized by the detection circuit & generates the unified amplitude signal by using phase locked loop (PLL) with a pure sine frame. The voltage at capacitor is recognized & compared to the reference voltage; the error values are acquired by specific process to drive the PID-PSO controller. The outcome of PLL sine frame & controlled output coming from controller is multiplied to generate the reference

signal, while the current is detected by a current recognized circuit. The reference signal and the carrier signals are compared to proceeds the multi-carrier PWM operation for generation of switching states to 15-level inverter topology which accords to the characteristics table.

3. Proposed Multi-Carrier Modulation Techniques

Evolution of advanced multi-carrier modulation techniques are established for multi-level inverters, the most recommended near to pure sinusoidal outcome voltage with low harmonic content, low switch stress, reduced loss component compared over the classical modulation techniques no need of any large size filter. Essential modulation techniques are evaluated in many applications can be classified as level & phase shifting multi-carrier modulation methods are explored by B.E.S. Mohammed et al in [10], [11], based on these two methods author implements a modish modulation techniques which have adequate results over formal methods. The classical multi-carrier modulation scheme is proposed by authors i.e., combination of both level shift & phase shift PWM techniques, to trounce the concerns regarding the switching action of shifting algorithm to overcome the phase imbalance condition. In this technique all the carriers have equal frequencies and the inequality of peak amplitude, which are disposed vertically by switching angle. The classical modulation scheme for 15-level proposed inverter as depicted

in Fig.2.

MI„ =

Vcr (q-1) 360° ' 4(n - 1)

Optimal Modulation Scheme

0 006 0 008 0.01 0.012 0.014

Time (sec)

Fig.2 Switching Strategy for Optimal Modulation Scheme

Time (sec)

Fig.3 Modulation Scheme of Variable Frequency Switching Strategy

For this modulation scheme, (n-1) carriers are required based on necessity with vertically disposed by measurable angle. Where Vref constitutes the reference signal coming from PID-PSO controller, Vcr constitutes the carrier signal;

MI„ constitutes the modulation index & 0sh constitutes the disposed phase angle. The proposed variable frequency switching technique mainly regards to definite carrier frequency ranges with respect to certain multiple factor. The carrier frequencies of distinct carriers are 7050Hz, 5050Hz, 3050Hz are interpret to the reference signal for production of optimal switching angle to inverter from PSO-PID controller is depicted in Fig.3. As, Table I depicts the switching states for proposed 15-level Asymmetrical multi-level inverter for generation of DC voltage levels & Table II shows the switching states for 15 level Asymmetrical multi-level inverter for generation of specific polarities, in that "L" specifies the switch is at OFF state & "H" specifies the switch is at ON state.

Table I. Switching Pattern for Level Generation Scheme for Proposed Inverter Module

Vo S1 S2 S3

7Vs H H H

6Vs L H H

5Vs H L H

4Vs L L H

3Vs H H L

2Vs L H L

Vs H L L

Table II. Switching Pattern Level Generation Scheme for Proposed Inverter Module

Vo SA SB SC Sd

Zero State L H L H

Positive State H H L L

Negative State L L H H

4. PSO-PID Controller

The traditional PID controller includes the proportional, integral, derivative operated gains in a closed loop feedback system. The transfer function of the plant for this controller is given as;

C(s) =Kp+!± + Kds (6)

Where, Kp-proportional gain, ^¿-integral gain, Kd- derivative gains of traditional PID controller, are tuned by trail & error method and also evaluated by extra means like tuning condition. Here proposes a contemporary tuning method of PID controller by using particle swarm optimization (PSO) technique. It is an optimized meta-heuristic algorithm implemented in the part of soft evolutionary computational method. The general PSO is evaluated from the research on several swarms like as a fish schooling & birds flocking [12] techniques. A new parameter called inertia weight is appended to original PSO form where the inertia weight is linearly decreased during the iteration process with the addition to other common technique of PSO is explored by Clerc [13]-[14]. Various particles are constitutes in an algorithm by potentiality solution of the problem; every particle normalize its flying capacity followed by its possessed flying experience & its associates. Every particle is handled as a D-dimensional space point. The íh particle is established as X1= (xi1, xi1,...., xiD). The best minimum fitness values of specific particle is appoint as P1= (pi1, pi1,...., piD), is called as thepbest. The index function of the best particle all over the particles in a stipulated population is initiated by symbol g, called as the gbest. The velocity for every particle of i is represented as V1=(vi1, vi1,...., viD). The acquired particles are renovated with respect to following equations:

v?d+1w. v?d + q. rand Q. (p?d - x?d) + c2. rand Q. (p^d - x£d) (7)

= + v?/1 . . (8)

Where the c1, c2 =1.2 are two attracted positive constants, n represents the several number of iterations, rand () constitutes the numbers extract by randomly in between 0 to 1. Calculation of each particle with new velocities (7) & followed by its descent velocity and the distance of its present position from its own best experience with the expectancy of group best. Then the particles are moved to a new current position followed (8). The emergence of each particle is recorded accords to a pre-organized fitness function, which is allied to the problem to be decoded. Inertia weight w is accompanied into the followed equation to regulate the global & local search capability [12]. It could be a positive linear/non-linear function of time. A highly assured convergence of PSO is proposed & set as approximate of 0.9. Swarm size of 10-60, here generally taken as 40. This impressed algorithm can be used for optimal tuning of PID gains to ensure best control objective at standard operating level. PSO is commissioned to

tune the gains of Kp, KL, Kd of PID controller by inline model. A certain set of PID gain parameters can tolerate by an acceptable system response & minimization the error index. The flow chart of proposed PSO scheme is depicted in Fig.4.

Fig.4. Flow Chart of Proposed PSO Scheme

5. Matlab/Simulink Results

Validate the performance analysis of proposed PV/FC co-generation scheme by using a Matlab simulation tool & compared with respect to several control objectives. The parameters of the proposed scheme are depicted in Table III; it is merely suitable for single-phase grid connected system with a 700V, 50Hz supply.

Table III Operating Parameters for Proposed Asymmetrical 15-Level Inverter for Grid System

S.No Operating Parameters Values

1 PV Source 10-40 V

2 Fuel Cell Stacks 10-20 V

3 Boost Inductor (L) 1 mH

4 Capacitors (C) C1/C2=47 ^F; C3=100 ^F

5 Filter Inductor (Lf) 1.5 mH

6 Switching Frequency 3050 Hz to 7050 Hz

7 PI Controller Kp-0.7, Ki-0.05

As below Fig.5 illustrates the various simulations of proposed 15-Level advanced inverter topology for grid integrated system via PID-PSO controller, in that (a) depicts the several DC link voltages over several controllers, proposed PID-PSO controller have low steady state error values under step changes; (b) depicts the stepped 15-level output voltage; (c) depicts the grid voltage/current are pure sinusoidal & both are in-phase each other represents the unity power factor; (d) depicts the harmonic distorted analysis (THD) of proposed 15-Level advanced inverter topology for grid integrated system via PID-PSO controller is 2.68%, it would be in IEEE standards limit without any extra filter units.

DC Link Voltage

Time (sec)

(a) DC Link Voltage

Output Voltage

Time (sec)

(b) 15-Level Output Voltage

Grid Voltage & Grid Current

0.15 Time (sec)

Frequency (Hz)

(d) THD of 15-Level Output Voltage Fig.5 Several Simulations of Proposed 15-Level MLI topology for Grid Integrated System via PID-PSO Controller

Table IV FFT Analysis of Output Voltage with Proposed Controller over Classical Controllers

Type of Controller Regular PWM Variable Frequency PWM Scheme Optimal PWM Scheme

PD POD APOD PD POD APOD PD POD APOD

PI Controller 7.75% 7.76% 7.76% 7.65% 7.67% 7.67% 7.68% 7.60% 7.71%

Fuzzy Controller 5.86% 5.96% 5.95% 4.99% 5.03% 5.03% 6.06% 6.12% 6.12%

ANFIS Controller 4.44% 4.56% 4.57% 3.68% 3.62% 3.62% 4.67% 4.69% 4.73%

PID-PSO Controller 3.09% 3.20% 3.24% 2.80% 2.68% 2.68% 3.33% 3.31% 3.40%

6. Conclusion

Effectuated simulated results, the excellent tuning of PID gains by optimized PSO tool have very low steady state response, reduced peak overshoot, better stability performance and low THD content over formal controllers. The formal tuning scheme is virtue for dispending the erupted the point of setting the PID gains. The major comfort of utilizing the smart optimization approach is constituted by a supplement solution to magnify the implementation of PID gains. The contemporary approach of explored co-generation scheme with PID-PSO controller includes regulating the PV/FC energy via asymmetrical zeta type power conditioning unit using an advanced multi-carrier modulation schemes. The several conclusions are updated from the evaluation of proposed 15-level inverter requires only 7 switches, requires low space requirement, high comfort, reduced cost. As number of levels goes to increases, then automatically THD approximate near to IEEE standards & common mode voltage also decreases, stress is low, minimizes the load side filter. THD comparison is takes place in Table IV, in that variable switching frequency modulation technique with PID-PSO controller has better THD % reduction over classical controllers.

References

[1] Mastromauro. R. A, Liserre. M, Dell. A, " Control issues in single stage PV systems: MPPT current & voltage control," IEEE Trans. Ind. Informat., vol. 8, no. 2, pp. 241-254, may 2012.

[2] P. Thounthong, S. Pierfederici, J. P. Martin, M. Hinaje, and B. Davat, "Modeling and control of fuel cell/super capacitor hybrid source based on differential flatness control," IEEE Trans. Veh. Technol., vol. 59, no. 6, pp. 2700-2710, Jul. 2010.

[3] M. Liserre, T. Sauter, J.Y. Hung, "Future energy systems: integrating renewable energy sources into the smart power grid through industrial electronics" IEEE Ind. Electron. Mag. 4(1), 18-37, 2010.

[4] S. Seme, G. Stumberger, J. Vorsic, "Operational properties of a photovoltaic system with three single phase inverters" In; International Conference on Renewable Energies and Power Quality, pp. 1-4, 2010.

[5] K. Hasegawa and H. Akagi, "Low-modulation index operation of a five-level diode-clamped pwm inverter with a dc-voltage-balancing circuit for a motor drive," IEEE Trans. Power Electron., vol. 27, no. 8, pp. 3495-3505, Aug. 2012.

[6] E. Pouresmaeil, D. Montesinos-Miracle, and O. Gomis Bellmunt, "Control scheme of three-level NPC inverter for integration of renewable energy resources into AC grid," IEEE Syst. J., vol. 6, no. 2, pp. 242-253, Jun. 2012.

[7] J. D. Barros, J. F. A. Silva, and E. G. A Jesus, "Fast-predictive optimal control of NPC multilevel converters," IEEE Trans. Ind. Electron., vol. 60, no. 2, pp. 619-627, Feb. 2013.

[8] S. Choi and M. Saeedifard, "Capacitor voltage balancing of flying capacitor multilevel converters by space vector PWM," IEEE Trans. Power Delivery, vol. 27, no. 3, pp. 1154-1161, Jul. 2012.

[9] L. Maharjan, T. Yamagishi, and H. Akagi, "Active-power control of individual converter cells for a battery energy storage system based on a multilevel cascade pwm converter," IEEE Trans. Power Electron., vol. 27, no. 3, pp. 1099-1107, Mar. 2012.

[10] S. M. E. Bahr, K. S. Rama Rao, "A new Multi Carrier Based PWM for Multilevel Converter" 2011 IEEE Applied Power Electronics Colloquium (IAPEC), IEEE 2011, pp. 63-68.

[11] X. Mao, R. Ayyanar, H. k. Krishnamurthy, "Optimal Variable Switching Frequency Scheme for Reducing Switching Loss in Single-Phase Inverters Based on Time-Domain Ripple Analysis," IEEE Transactions on Power Electronics, vol.24, no.4, pp.991-1001, April 2009.

[12] Y.H. Shi and R.C. Eberhart, "A modified particle swarm optimizer". IEEE International Conference on Evolutionary Computation, Anchorage, Alaska, 1998.

[13] Y.H. Shi and R.C. Eberhart "Comparing inertia weights and constriction factors in particle swarm optimization". Proc. The 2000 Congress on Evolutionary Computation. Vol.1, pp. 84-88.

[14] M. Clerc "The Swarm and the queen: towards a deterministic and adaptive particle swarm optimization. Proceedings of the Conference on Evolutionary Computation, pp. 1951-1957, 1999.