A Maximum Power Point Tracking Technique Based on Ripple Correlation Control for Single-Phase Single-Stage Grid Connected Photovoltaic System

Transcription

1 A Maximum Power Point Tracking Technique Based on Ripple Correlation Control for Single-Phase Single-Stage Grid Connected Photovoltaic System Satish R, Ch L S Srinivas, and Sreeraj E S Department of Electrical and Electronics National Institute of Technology Goa, India. satish20ramaiah@gmail.com, chlssrinivas@gmail.com, sreeraj@nitgoa.ac.in Abstract This paper presents a novel maximum power point tracking (MPPT) technique for single-phase, single-stage gridconnected photovoltaic (PV) systems. Single-stage topology reduces the cost and complexity of the system compared to multistage topology and such systems are reliable because of the reduced component count. In single-phase single-stage system, PV array voltage is subjected to 100 Hz ripple (double the grid frequency) introduced by the instantaneous power oscillations. The method which makes use of such ripple and correlates this with switching function to control the operating point of PV array is called ripple correlation control (RCC). In this paper, a modified RCC method has been proposed to implement the MPPT technique with reduced component count which makes the implementation easier compared to existing techniques presented in literature. Compared to existing methods which uses only sign of the error signal generated by the MPPT block, this method uses both magnitude and sign of the error signal which makes the system respond faster. The effectiveness of the proposed method has been verified by performing numerical simulation studies using MATLAB/SIMULINK. The real time simulation results of the proposed scheme are presented in this paper using Real time digital simulator (RTDS). Index terms Photovoltaic (PV), maximum power point tracking (MPPT), ripple correlation control (RCC), real time digital simulator (RTDS). I. INTRODUCTION The concept of distributed generation has enabled any individual single-phase consumer to generate power and also sell the excess power to the utility grid [1]. Because of this, small single phase photovoltaic (PV) generating units are becoming more and more popular. Such distributed units should be operated reliably without much maintenance and should be cost effective. A typical single phase grid connected PV system has more than one stage of energy conversion. The first stage is usually a dc-dc converter, which boosts the dc-link voltage level such that it can draw peak available power from PV panels. The second stage is an inverter, which ensures that whatever the energy extracted from PV array is fed to the utility grid. The cost and complexity of the system can be reduced by employing a single-stage topology. Moreover, such a system can be more reliable, because of the reduced component count Fig. 1. Single-phase, single-stage PV system. [1], [2]. The schematic of a typical single-stage PV system is shown in Fig. 1 in which the output of PV array is connected to the dc-link of the inverter, and the output of the inverter is fed to the grid. The dc-ac inverter present in single-stage systems performs the function of extracting the peak available power as well as dumping the extracted power to the grid. Since PV cells shows nonlinear v-i characteristic which is dependent on solar irradiation and temperature [3], there is a need to track the maximum power point (MPP). Various maximum power point tracking (MPPT) techniques such as Perturb and Observe, Hill-climbing, incremental conductance have been reported in the literature [1] [4]. All these MPPT techniques makes sure that the PV array is operating around the MPP, and in order to achieve this, a small perturbation is given to the PV array voltage. However perturb and observe and incremantal conductance methods have their own disadvantages such as (i) poor tracking performance during sudden change in solar irradiance, (ii) need of choosing an effective perturbation size [3]. Usually, these methods use fixed step size in which larger step size gives faster dynamics whereas it results in more steady state oscillations around operating point. And smaller step size gives slow dynamics with reduced steady state oscillations [4]. Therefore PV system compromises between dynamics and steady state oscillations for satisfactory performance of MPPT. When these methods

2 are applied in single-phase single-stage systems where the PV array voltage is inevitably subjected to 100 Hz ripple (double the grid frequency) because of the oscillating nature of the power fed to the grid [5] [10], the dc-link voltage perturbs around the MPP according to the perturbation size given. And the 100 Hz ripple is superimposed on these perturbing operating points. Since the 100 Hz ripple is inevitable in single-phase singlestage systems, this can be effectively used to determine the MPP of the PV array, and such techniques are reported in the literature by the name ripple correlation control (RCC) [5] [10]. Since the RCC uses inherently available ripples, no artificial perturbation is needed to track MPP which improves the tracking performance during sudden change in solar irradiance [11]. RCC is parameter insensitive and do not induce perturbations to the PV array voltage making MPPT faster [11]. However, the schemes given in [5] [10] requires the service of a phase locked loop (PLL), whose implementation is computationally intensive. Some pulse width modulation techniques used for grid connected inverters do not require the service of PLL to synchronize with the grid [12] [14], and the MPPT schemes given in [5] [10] cannot take advantages of such schemes. And for the implementation of ripple correlation control based MPPT algorithm given in [5], service of four filters, two high pass filters and two low pass filters (LPF) are required. This slows down the response of the system, and also makes implementation computationally intensive. Moreover, only the sign of the power derivative signal, and not its magnitude is used to determine the MPP, thereby making the system respond slowly if the operating point is far away from MPP. In this paper, a novel MPPT method based on modified RCC has been proposed. In this method the drawbacks of perturb and observe, incremental conductance method and the RCC based MPPT methods presented in [5] [10] are addressed. This method uses the product of power ripple and voltage ripple available in PV array as error signal to track the MPP. The MPPT block uses this error to generate a control signal whose magnitude vary according to distance of operating point from the MPP. The magnitude of control signal will be more if the operating point is far from the MPP and will be less if the operating point is near to MPP. Therefore this method gives faster dynamics and the steady state oscillations around the MPP will be limited to ripple content of the PV array voltage which is unavoidable. Moreover this scheme is implemented using only two LPFs, as compared to [5] thereby avoiding the use of two HPFs, which reduces the computational complexity and helps the system to respond faster. Extensive numerical simulations are carried out to prove the efficacy of the proposed MPPT scheme and verified using Real-Time Digital Simulator (RTDS) hardware. This paper is organized as follows: the working principle and implementation of the proposed algorithm is discussed in Section II. Simulation results of a PV generation system are presented in Section III. Detailed real time simulation results of the proposed scheme are presented in Section IV Fig. 2. Current and power of the PV panels versus voltage. Fig. 3. Dc-link voltage, power ripple and product of voltage and power ripple for different operating points. and concluding remarks are presented in Section V. II. OPERATING PRINCIPLE AND IMPLEMENTATION OF MPPT ALGORITHM For a single-phase system, the value of instantaneous power p g (t) injected into the grid pulsates at twice the grid frequency. This causes the dc-link voltage to oscillate at 100 Hz. For the single-stage system given in Fig. 1, the terminals of the PV array is directly connected across the dc link capacitor, and therefore the output voltage of the PV array, v(t) also oscillates at 100 Hz. The PV array current, i(t) and the power fed by the PV array, p(t) also contains a ripple. The ripple content of a general time varying quantity, x(t) can be expressed as x(t) = x(t) x(t) (1) where x(t) represents the ripple content, and x(t) represents the moving average component. The general quantity, x(t), can be PV array voltage, v(t); current, i(t); or power, p(t).

3 Fig. 4. Block diagram showing implementation of the proposed MPPT algorithm. The power can be obtained by finding the product of voltage and current as expressed in (2). p(t) = v(t)i(t) (2) Expressing v(t) and i(t) as in (1) and substituting them in (2), we get the PV array power as p(t) = v(t)i(t) + i(t)ṽ(t) + v(t)ĩ(t) + ṽ(t)ĩ(t) (3) Hence, the power ripple can be written as p(t) = i(t)ṽ(t) + v(t)ĩ(t) + ṽ(t)ĩ(t) (4) The product of ṽ(t) and p(t) can be expressed as [ p(t)ṽ(t) = ṽ 2 (t) i(t) + v(t) ĩ(t) ] + ṽ 2 (t)ĩ(t) (5) ṽ(t) Also, the PV power derivative can be expressed as dp(t) di(t) = i(t) + v(t) dv(t) dv(t) Linearization of v-i curve shown in Fig. 2 at a point (v o, i o ) yields ( ) di(t) = ĩ(t) (7) dv(t) v o ṽ(t) Using (6) and (7), p(t)ṽ(t) can be expressed as [ ] dp(t) p(t)ṽ(t) = ṽ 2 (t) + ṽ 2 (t)ĩ(t) (8) dv(t) As the average value of ṽ 2 (t)ĩ(t) over a cycle is zero and ṽ 2 (t) is always positive, the magnitude of the average value of error signal p(t)ṽ(t), henceforth called as e(t), is directly related to magnitude of dp/dv. The e(t) is a representation of the distance of the operating point from MPP. This can be further explained with the help of Fig. 3, where dc-link voltage, p(t),and e(t) is shown for various operating points. When the operating point is far left of MPP (Region 1 in Fig. 3), the average value of e(t) is positive. When the operating point is in the left vicinity of MPP (Region 2 in Fig. 3), the average value of e(t) is positive with smaller magnitude and when the operating point is at MPP (Region 3 in Fig. 3), average of the e(t) is zero. In Region 4, the operating (6) TABLE I. Main parameters of PV generation system Rating of PV array 1.55 kwp DC-link capacitance, C dc 1000 µf PWM carrier frequency, f sw AC-link inductor Time constant of LPF s PI controller (inside MPPT block) PI controller (outside MPPT block) Short circuit current (1000 W/m 2, 25 o C) Open circuit voltage (1000 W/m 2, 25 o C) 3.5 khz L = 15.5 mh t = 2.5 ms K P = 0.01 K I = 30 K P = 5e-7 K I = A 550 V Fig. 5. Simulated power versus voltage characteristics of the PV array for varying irradiance conditions point is on right of MPP and e(t) is negative. As the average value of error signal indicates the distance of the operating point from MPP, the operating point can be controlled by passing the average error signal through a PI controller. The implementation of the proposed MPPT algorithm is shown in Fig. 4. The ripples ṽ(t) and p(t) can be obtained by subtracting the average values from the respective signals using LPFs. The

4 (a) (b) Fig. 6. Simulated system performance during step change in solar irradiance. (a) PV power, (b) Reference voltage and dc-link voltage. Fig. 8. Harmonic spectrum of grid current. (a) With 1000 W/m 2 irradiance. (b) With 600 W/m 2 irradiance. product of these ripples is used as input to a PI controller. The output of the PI controller is considered as reference signal, V (t) to control the dc-link voltage. The reference signal thus obtained is compared with PV array voltage and the error obtained is passed through another PI controller to obtain the load angle, δ. This angle is used to generate the control signals to operate the inverter switches. Fig. 7. Simulated steady state performance of the system under irradiance of 1000 W/m 2 : Grid voltage (100 V/div.) and current (20 A/div.). III. SIMULATION RESULTS The proposed MPPT algorithm has been numerically simulated using Matlab/Simulink. The details of the PV system considered is summarized in Table I. The performance of the system connected to the photovoltaic array has been evaluated both in steady state and transient conditions by varying solar irradiance. Fig. 5 shows the simulated power versus voltage characteristics for the PV system considered in Table I. Point A shows the MPP for irradiance 1000 W/m 2. After 4 s, a step change in solar irradiance is applied from 1000 W/m 2 to 600 W/m 2 and operating point shifts to B whose corresponding voltage is same as previous MPP as dc-link voltage cannot change abruptly. Later the operating point moves towards new MPP

5 (a) Fig. 9. RTDS hardware setup. (b) (a) Fig. 11. Real time simulated system performance during step change in solar irradiance. (a) Reference voltage and actual DC-Link voltage. (b) DC-Link voltage (65 V/div) and PV panel output power (670 W/div). (b) Fig. 10. Real time simulated steady state results. (a) Grid voltage (100 V/div) and current (5 A/div) with 1000 W/m 2. (b) Grid voltage (100 V/div) and current (5 A/div) with 600 W/m 2. corresponding to 600 W/m 2. The blue colored portion in Fig. 5 indicates the PV array voltage ripple around the MPP. The simulated performance of the system under varying solar irradiance conditions are shown in Fig. 6. When the irradiation level is 1000 W/m 2, the values of PV array voltage and power is 445 V and 1.55 kw respectively which corresponds to MPP for irradiation level of 1000 W/m 2. The irradiation level is varied from 1000 W/m 2 to 600 W/m 2 at 4 s. The PV array voltage and power is settled at 434 V and 900 W which corresponds to MPP for irradiation level of 600 W/m 2. It can be verified from both Fig. 5 and Fig. 6 that the operating point is tracking MPP under varying irradiance condition. Fig. 7 shows the grid side voltage and current at steady state for irradiance of 1000 W/m 2. The total harmonic distortion (THD) in the gird current has been analyzed and it is observed that for irradiance of 1000 W/m 2 condition, THD is 2.94% with peak value of fundamental component as A and for iraadiance of 600 W/m 2, THD is 5.06% with peak value of fundamental component as A as shown in fig. 8(a) and 8(b) respectively. The satisfactory operation of system is verified by simulation under varying irradiation levels. IV. REAL TIME SIMULATION RESULTS In order to further verify the proposed method, the generated switching pulses are given to a photovoltaic inverter system simulated using real time digital simulator (RTDS). The RTDS hardware which consists of host computer and FPGA based target is used to implement the aforementioned system [15]. Fig. 9 shows the photograph of RTDS OP4500 test set up, where OP4500 is the compact real time power grid digital simulator from OPAL-RT technologies. The grid connected PV system through inverter system has been implemented using RTDS hardware with a real time simulation time step of 10 µs. The details of the PV generation system are given in Table I. It has been observed that the

6 real time simulation results are matching with the simulation results presented in section III. Fig. 10(a) gives the details of the grid voltage and grid current for 1000 W/m 2, grid voltage is 230 V and current is 5.34 A. The irradiance level is changed to 600 W/m 2 at 4 second and the grid voltage and currents are 230 V, 2.9 A respectively as shown in Fig. 10(b). Fig. 11(a) gives the details of reference voltage and dc link voltage and it can be observed that the operating point is tracking the desired MPP represented by reference voltage, V (t). Fig. 11(b) gives the details of dc link voltage and PV power, for 1000 W/m 2 irradiance condition, the values of PV array voltage and power is 445 V and 1.55 kw respectively which corresponds to MPP for irradiation level of 1000 W/m 2. The irradiation level is varied from 1000 W/m 2 to 600 W/m 2 after 4 s. The PV array voltage and power is settled at 434 V and 900 W which corresponds to MPP for irradiation level of 600 W/m 2. It has been verified that from both numerical simulation and real time simulation results, the proposed method is able to track the desired MPP under varying irradiation levels effectively. [9] C. Barth and R. Pilawa-Podgurski, Dithering digital ripple correlation control for photovoltaic maximum power point tracking, IEEE Trans. Power Electron., vol. 30, no. 8, pp , Aug [10] A. Bazzi and P. Krein, Ripple correlation control: An extremum seeking control perspective for real-time optimization, IEEE Trans. Power Electron., vol. 29, no. 2, pp , Feb [11] T. Esram, J. Kimball, P. Krein, P. Chapman, and P. Midya, Dynamic maximum power point tracking of photovoltaic arrays using ripple correlation control, IEEE Trans. Power Electron., vol. 21, no. 5, pp , Sept [12] Q.-C. Zhong, P.-L. Nguyen, Z. Ma, and W. Sheng, Self-synchronized synchronverters: Inverters without a dedicated synchronization unit, IEEE Trans. Power Electron., vol. 29, no. 2, pp , Feb [13] P. Xiao, K. Corzine, and G. Venayagamoorthy, Cancellation predictive control for three-phase pwm rectifiers under harmonic and unbalanced input conditions, in 32nd Annu. Conf. IEEE Ind. Electron., IECON, Nov 2006, pp [14] D. Roiu, R. Bojoi, L. Limongi, and A. Tenconi, New stationary frame control scheme for three-phase pwm rectifiers under unbalanced voltage dips conditions, IEEE Trans. Ind. Applicat., vol. 46, no. 1, pp , Jan [15] S. Mikkili, A. Panda, and J. Prattipati, Review of real-time simulator and the steps involved for implementation of a model from matlab/simulink to real-time, Journal of The Institution of Engineers (India): Series B, vol. 96, no. 2, pp , V. CONCLUSION A novel MPPT technique based on ripple correlation control is proposed for single-phase single-stage grid connected PV systems. The MPPT block uses the product of voltage and power ripple to drive the operating point towards MPP. The proposed algorithm is working satisfactorily under dynamic irradiance conditions. The efficacy of the proposed algorithm has been proved through extensive numerical simulations and real time simulation results using RTDS hardware. REFERENCES [1] Y. Chen and K. Smedley, A cost-effective single-stage inverter with maximum power point tracking, IEEE Trans. Power Electron., vol. 19, no. 5, pp , Sept [2] E. S. Sreeraj, K. Chatterjee, and S. Bandyopadhyay, One-cyclecontrolled single-stage single-phase voltage-sensorless grid-connected pv system, IEEE Trans. Ind. Electron., vol. 60, no. 3, pp , Mar [3] B. Subudhi and R. Pradhan, A comparative study on maximum power point tracking techniques for photovoltaic power systems, IEEE Trans. Sustainable Energy, IEEE Trans., vol. 4, no. 1, pp , Jan [4] F. Liu, S. Duan, F. Liu, B. Liu, and Y. Kang, A variable step size inc mppt method for pv systems, Industrial Electronics, IEEE Transactions on, vol. 55, no. 7, pp , July [5] D. Casadei, G. Grandi, and C. Rossi, Single-phase single-stage photovoltaic generation system based on a ripple correlation control maximum power point tracking, IEEE Trans. Energy Conv., vol. 21, no. 2, pp , June [6] R. Stala, K. Koska, and L. Stawiarski, Realization of modified ripplebased mppt in a single-phase single-stage grid-connected photovoltaic system, in IEEE Int. Symp. Industrial Electronics (ISIE), June 2011, pp [7] C. Boonmee and Y. Kumsuwan, Modified maximum power point tracking based-on ripple correlation control application for single-phase vsi grid-connected pv systems, in 10th Int. Conf. Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON)., May. 2013, pp [8] P. Krein, Ripple correlation control, with some applications, in Proc. IEEE Int. Symp. Circuits and Systems, ISCAS 99., vol. 5, 1999, pp vol.5.