Investigation of fuzzy control based LCL resonant converter in RTOS environment

Transcription

1 Journal of Intelligent & Fuzzy Systems 26 (2014) DOI: /IFS IOS Press 913 Investigation of fuzzy control based LCL resonant converter in RTOS environment S. Selvaperumal a, and C. Christober Asir Rajan b a Department of EEE, Syed Ammal Engineering College, Ramanathapuram, Tamil Nadu, India b Department of EEE, Pondicherry Engineering College, Pondicherry, India Abstract. This paper presents a comparative evaluation of Fuzzy Logic (FLC) Controller and Open loop Controller for a modified LCL Resonant Converter has been simulated and the performance is analyzed. A three element LCL Resonant converter working under load independent operation is presented in this paper. In this work, the applicability of the ARM (Advanced RISC Machine) processor LPC 2148 is to be investigated as the controller for resonant converter. The simulation study indicates the superiority of Fuzzy Logic control over the conventional control method. The evaluation version of MATLAB was used to model the LCL topology for varied loads and LCL configurations. A LCL Resonant Inverter is proposed for applications in high frequency distributed AC power systems and Resonant Converter is proposed for applications in many space and radar power supplies. The advantages of the LCL topology are low total harmonic distortion (THD) high efficiency and the ability to handle varying loads. Keywords: ARM processor, dynamic response, fuzzy controller, resonant converter 1. Introduction In recent years the design and development of various DC-DC Resonant Converters (RC) have been focused for telecommunication and aerospace applications. It has been found that these converters experience high switching losses, reduced reliability, electromagnetic interference (EMI) and acoustic noise at high frequencies. The Series Parallel Resonant Converters (SPRC) are found to be suitable, due to various inherent advantages. The series and parallel Resonant Converter (SRC and PRC respectively) circuits are the basic resonant converter topologies with two reactive elements. The merits of SRC include better load efficiency and inherent dc blocking of the isolation transformer due to the series capacitor in the resonant network. However, the load regulation is poor and output-voltage regulation at Corresponding author. S. Selvaperumal, Associate Professor, Department of EEE, Syed Ammal Engineering College, Ramanathapuram, Tamil Nadu, India. perumal.om@gmail.com. no load is not possible by switching frequency variations. On the other hand, PRC offers no-load regulation but suffers from poor load efficiency and lack of dc blocking for the isolation transformer. It has been suggested to design Resonant Converter with three reactive components for better regulation. The LCL tank circuit based DC-DC Resonant Converter has been experimentally demonstrated and reported by many researchers [1 12]. M. Borage et al., has experimentally demonstrated with independent load when operated at resonant frequency, making it attractive for application as a constant voltage (CV) power supply. It has been found from the literature that the LCL tank circuit connected in series-parallel with the load and operated in above resonant frequency improves the load efficiency and independent operation [13]. Later, Mangesh Borage et al. [14], have demonstrated an LCL-T half bridge resonant converter with clamp diodes. The output current or voltage is sensed for every change in load because the output voltage or constant current increases linearly. The feedback control circuit has not been provided. LCL-T RC /14/$ IOS Press and the authors. All rights reserved

2 914 S. Selvaperumal and C.C.A. Rajan / Investigation of fuzzy control with constant current supply operated at resonant frequency is presented [15]. The parallel operation is simple without any complex control circuit which increases to ripple frequency. M. Qiu [16] has demonstrated different approaches which offer the fuzzy logic control (FLC). This control technique relies on the human capability to understand the system s behavior and is based on qualitative control rules. The FLC approach with same control rules can be applied to several dc dc converters. However, some scale factors must be tuned according to converter topology and parameters. The author utilized the proposed control technique for Buck- Boost converter and demonstrated. D. Tschirhart et al. [17] have demonstrated a DC/AC series resonant converter with fixed load value considering two control approaches. Later T.S. Sivakumaran et al. [18] have demonstrated a CLC SPRC using FLC for load regulation and line regulation. The performance of controller has been evaluated and found that the load independent operation may not be possible. The FLC based Zero Voltage Switching quasi-resonant converter has been demonstrated by Mangesh B. Borage et al. [19, 20]. The load independent operation was not realized and power handling capacity of the converter is found to be poor. It is clear from the above literatures that the output voltage regulation of the converter against load and supply voltage fluctuations have important role in designing high-density power supplies. LCL Resonant Converter is expected the speed of response, voltage regulation and better load independent operation. Keep the above facts in view, the LCL Resonant Converter has been module and analyses for estimating various responses. The closed loop state space modules have been derived and simulate using MAT LAB/Simulink for comparing the performance with existing converter. RTOS stands for Real Time Operating System It overcomes, to a certain extent, the drawback of the von Newman Architecture that is Fetch decode execute sequence. With RTOS a number of tasks can be carried out in a near simultaneous manner. In the existing system a micro controller is used to develop the two inverts required for driving the Switches of the two legs of the single phase inverter. In a closed loop system an ADC is required to monitor at least any one of the system parameters usually the terminal voltage on the output DC side. With a single micro controller it is not possible to carry out the two jobs of monitoring & controlling independently. With the advent of the Advanced RISC Microprocessors we can do many tasks in parallel. The operating system loaded onto the Micro controller is the basic facility that makes it possible to carry out multiple tasks. The LPC 2148 has two sets of ADCs ADC0 and ADC 1. The ADC0 has 8 channels while ADC1 can handle 6 channels. All these channels can work independently. Typically a Multiple Input Multiple Output system can therefore be controlled by the LPC2148. With the LPC 2148 the gap between the MIMO system design and implementation is reduced. 2. Proposed LCL resonant converter The block diagram of LCL RC with Fuzzy controller is shown in Fig. 1. The resonant tank consisting of three reactive energy storage elements (LCL) has overcome the conventional resonant converter that has only two elements. The first stage converts a dc voltage to a high frequency ac voltage. The second stage of the converter is to convert the ac power to dc power by suitable high frequency rectifier and filter circuit. Power from the resonant circuit is taken either through a transformer in series with the resonant circuit or series in the capacitor comprising the resonant circuit. In both cases the high frequency feature of the link allows the use of a high frequency transformer to provide voltage transformation and ohmic isolation between the dc source and the load. In LCL RC the load voltage can be controlled by varying the switching frequency or by varying the phase difference between the inverters. The phase domain control scheme is suitable for wide variation of load condition because the output voltage is independent of load. The dc current is absent in the primary side of the transformer, there is no possibility of current balancing. Another advantage of this circuit is that the device currents are proportional to load current. This increases the efficiency of the converter at light loads to some extent because the device losses also decrease with the load current. If the load gets short at this condition, very large current would flow through the circuit. This may damage the switching devices. To make the circuit short circuit proof, the operating frequency should be changed. A schematic diagram of full-bridge LCL-T SPRC is shown in Fig. 2. The resonant circuit consist of series inductance L 1, parallel capacitor C and series inductance L 2. S1-S4 are switching devices having base /gate turn-on and turn-off capability. D1 to D4 are anti-parallel diodes across these switching devices. The MOSFET (say S1) and its anti parallel diode (D1) act as a bidirectional switch. The gate pulses for S1 and S2 are in phase but 180 degree out of phase with the gate pulses for S3 and S4. The positive portion of switch current flows through the MOSFET and negative portion

3 S. Selvaperumal and C.C.A. Rajan / Investigation of fuzzy control 915 flows through the anti-parallel diode. The RLE load is connected across bridge rectifier via L 0 and C 0. The voltage across the point AB is rectified and fed to RLE load through L 0 and C 0. In the analysis that follows, it is assumed that the converter operates in the continuous conduction mode and the semiconductors have ideal characteristics. 3. Design of fuzzy logic controller The Fuzzy Logic Controller (FLC) provides an adaptive control for better system performance. Fuzzy logic Fig. 1. Block Diagram of LCL Resonant Converter. Fig. 2. LCL Resonant Converter. is aimed to provide solution for controlling non-linear processes and to handle ambiguous and uncertain situations. Fuzzy control is based on the fundamental of fuzzy sets. The fuzzy control for the chosen SPRC is developed using input membership functions for error e and change in error ce and the output membership function for D, the duty ratio of converter. The fuzzy control scheme is described using a dc to dc converter as a platform. The Fig. 3 shows the block diagram of the fuzzy logic control configuration for a dc to dc converter. The output of the fuzzy control algorithm is the change in duty cycle [$d(k)].the duty cycle d(k), at the kth sampling time, is determined by adding the previous

4 916 S. Selvaperumal and C.C.A. Rajan / Investigation of fuzzy control duty cycle [d(k 1)] to the calculated change in duty cycle. Fig. 3. Block Diagram of Fuzzy Control scheme for LCL. d(k) = d(k 1) + $ d(k) (1) 4. Simulation of the proposed system 4.1. Fuzzy logic control (FLC) Fuzzy control involves three stages: fuzzification, inference or rule evaluation and defuzzification as shown in Fig. 3. LCL is modeled by using MAT- LAB software. Fuzzy control is developed using the fuzzy toolbox. The fuzzy variables e, ce and D are described by triangular membership functions. Seven triangular membership functions (Fig. 4) are chosen and Table 1 shows the fuzzy rule base considered in this manuscript based on intuitive reasoning and experience. Fuzzy memberships NB, NM, NS, ZE, PS, PM, PB are defined as negative big, negative medium, negative small, zero and positive small, positive medium and positive big Rule table and inference engine Table 1 Fuzzy rule matrix E e NB NM NS ZE PS PM PB NB NB NB NB NB NM NS ZE NM NB NB NB NM NS ZE PS NS NB NB NM NS ZE PS PM ZE NB NM NS ZE PS PM PB PS NM NS ZE PS PM PB PB PM NS ZE PS PM PB PB PB PB ZE PS PM PB PB PB PB the Table 1. The Graphical representations for the fuzzy rules are shown in Fig. 5. Rule firing is the process of pointing out the existence. The rule firing contains the following steps. (1) From the membership functions of the input 1, the lingual under whose range, the actual values is present is found. Also it s decimal height DOB Degree of Belief is got. (2) From the membership functions of input 2, the corresponding lingual is also found along with the DOB. (3) For these two inputs, what is the output lingual? It is fetched from the rule matrix. The control rules that relate the fuzzy output to the fuzzy inputs are derived from general knowledge of the system behavior, perception and experience. The rule table for the designed fuzzy controller is given in 4.3. Development of the FLC At every sampling interval, the instantaneous RMS values of the sinusoidal reference voltage and load

5 Output 1 S. Selvaperumal and C.C.A. Rajan / Investigation of fuzzy control 917 voltage are used to calculate the error(e) and change in error(ce) signals that act as the input to the FLC. The stage of fuzzification, fuzzy inference and defuzzification are then perform program as generally described in the flowchart of Fig. 6. Fig. 4. Fuzzy memberships used for simulation for LCL. e=v r V L (2) C e = e p e (3) where Vr is the reference or the desired output voltage and V L is the actual output voltage. The subscript k denotes values at the beginning of kth sampling cycle. The duty ratio of the converter will be determined by the fuzzy inference. For instance, if the output voltage continues to increase gradually while the current is low during the charging process, the fuzzy controller will maintain the increase in voltage to reach the set point. A drop in the output voltage level triggers the fuzzy controller to increase the output voltage of the converter by modifying the duty cycle of the converter error rate Error Fig. 5. Graphic representation of Table The resolution of fuzzy logic control system relies on the fuzziness of the control variables while the fuzziness of the control variables depends on the fuzziness of their membership functions. The Closed loop simulation using FLC and PI controller for the SPRC is carried out using MATLAB/ Simulink software. Depending on error and the change

6 918 S. Selvaperumal and C.C.A. Rajan / Investigation of fuzzy control Fig. 6. Flow chart. in error, the value of change of switching frequency is calculated. Set parameter instruction and function blocks available in MATLAB are used to update the new switching frequency of the pulse generators. The closed loop Simulink diagram of LCL Resonant Converter using FLC is shown in Fig. 7. The entire system is simulated with a switching frequency of 50 KHz. The simulated converter output voltage V o and load current I o for a step change in load from 0.3 to 0.5 its applied at 10 milliseconds. It is observed that the FLC for LCL regulates the output voltage with a settling time of 0.06 millisecond.

7 S. Selvaperumal and C.C.A. Rajan / Investigation of fuzzy control 919 DC SOCKET JTAG Connecter Fig. 7. Closed loop Simulink diagram of LCL Resonant Converter using FLC. CAN1 CAN2 USB PORT FOR LPC 2148 UART1 4LED ARRAY RESET SW UART2(ISP) 4 INT SW Fig. 8. ARM Processor LPC RTOS based control ARM Processor LPC 2148 shown in Fig. 8. In this work the applicability of the Philips ARM processor LPC 2148 is investigated as the controller for the LCL resonant converter. The time sharing feature of the LPC2148 offers ample possibility for its use in the designed LCL RC which has a resonance frequency of 50 KHz. The RTOS output waveform is shown in Fig. 9.

8 920 S. Selvaperumal and C.C.A. Rajan / Investigation of fuzzy control Fig. 9. Software Output using RTOS. Fig. 10. Inverter Voltage and Current Waveforms. Fig. 11. LCL Resonant Converter Output Voltage. 5. Results and discussion The proposed model has been simulated using MAT- LAB/Simulink toolbox. The fuzzy controller and PI controller has been designed for LCLRC. The simulated wave forms of resonant voltage, resonant current, Output Voltage, and output Current are shown in Figs The fuzzy controller performance was also compared with the Open Loop controller performance for the converter.

9 S. Selvaperumal and C.C.A. Rajan / Investigation of fuzzy control 921 thd Fig. 13. THD of 100% Load (FLC) Open loop response Fig. 14. THD of 50% Load (FLC). The response for a reference voltage of 50 V and output voltage is 48 V, in the open loop response, the overshoot and the settling time are very high, and the response is oscillatory. The proposed control strategy is able to eliminate the peak overshoot and reduce the settling time. The resonant inverter voltage, resonant current and output voltage are shown in Fig. 10. The output voltage of the open loop LCL- RC are shown in Fig. 11. Here the settling time 0.6 for 50% of load and 0.9 for 100% of load, the steady state error for 50% of load is 0.06 and 100% of load is Fig. 12. LCL Converter Output Voltage and Current (FLC) FLC closed loop response The output voltage and current of the FLC based LCL-RC shown in Fig. 12. The response for a reference voltage of 50 V the output voltage is 48 V. In the closed loop response by using Fuzzy Controller, the overshoot and settling time is less compared to open loop controller, and the response is oscillatory. The Harmonic spectrum (shows Figs. 13 and 14) and AC component present in output voltage are very less compared to Open Loop controller. From the results, the settling time for 50% of load and 0.07 for 100% of load, the steady state error for 50% of load is and 100% of load is The result is justified that settling time of output voltage in Open Loop controller is more than that of the settling time in FLC. The output voltage response is flexible and sensitive Performance evaluation The open loop LCL and Closed loop RC have been estimated and provided in Tables 2 and 3. It is seen that the Fuzzy based closed loop controller provides better settling time. This ensures that the system can be controlled effectively. As far concerned to Tables 2 and 3. It is obvious that the rise time and settling time of open loop and Fuzzy controller has been compared and concluded that Fuzzy has got better performance. The percentage THD and Efficiency performance of both open loop and closed loop controller for 100%, 50%, 11%, 100% inductive and 100% capacitive load conditions are given in Tables 2 and 3. A 50 KHz, 133 W, 50 V prototype, shown in Fig. 15, is built to verify the proposed LCL resonant converter. Figure 16, shows cur-

10 922 S. Selvaperumal and C.C.A. Rajan / Investigation of fuzzy control Table 2 Summary of Performance Evalution for open loop control Load Parameters Rise Time Setlling Time Steady State THD % Efficiency % in ms in ms Error in ms Full Load Resistive % Load Resistive % Load Resistive Full load Inductive Full load Capacitive Table 3 Summary of Performance Evalution for FLC Closed loop control Load Parameters Rise Time Setlling Time Steady State THD % Efficiency % in ms in ms Error in ms Full Load Resistive % Load Resistive % Load Resistive Full load Inductive Full load Capacitive Fig. 15. Prototype model for Closed Loop LCL RC. Fig. 16. VAB and ILs at 50% Load Resistive Load. rent through the inductor ILs and voltage across the terminals A and B at 50% resistive load. The graph Fig. 17. for Load verus percentage THD for Open Loop and Closed Loop Controls has been plotted which depicts that the percentage THD increases for lower load and as the load increases the Fig. 17. Load verus % THD for Open Loop and Closed Loop Controls. percentage THD gradually decreases and remain constant at greater loads. Among the three curves FLC is well defined. The graph Fig. 18. for Load verus percentage Efficiency for Open Loop and Closed Loop Controls has been plotted which depicts that the percentage Efficiency increases for higher load and as the load increases the percentage Efficiency gradually increases and remain constant at greater loads. Among the three curves FLC is well defined. The above discussion the fuzzy Controller parameters are easy to determine. The proposed new control strategy the parametric and the load Sensitivity is much reduced. The results obtained indicate that the FLC is an

11 S. Selvaperumal and C.C.A. Rajan / Investigation of fuzzy control 923 A FLC based LCL RC circuit has been simulated in MAT LAB/Simulink and Experimentally done using ARM Processor. This converter with a voltage type load and current type load shows it provides load independent operation, output voltage regulation. So, the switching power losses are minimized. The effectiveness of FLC as compared with Open Loop Controller and open loop is verified by simulation studies. The LCL RC can be used for applications such as Space and radar high voltage power supplies with the appropriate turns ratio of HF transformer. The targeted design constraints, including efficiency, output voltage and output power, were met successfully. ARM (Advanced RISC Machine) processor LPC 2148 is to be investigated as the controller for both OLC and Fuzzy based resonant converter. An important factor in the test procedure was the self inductance of the Lab-Volt resistance boxes. This prohibited measurement of a purely resistive load and hindered efforts to accurately vary the power factor. Another important consideration to mention is the disproportionate variation of the inherent inductance with decreased load. As the load was decreased the inductance increased dramatically making 11% load results nearly meaningless as the power factor of the load was so low. Despite the aforementioned difficulties, high efficiencies were achieved. At full resistive (partially inductive) load an efficiency of 96.64% was obtained with less than 6.7% total harmonic distortion (THD). As the load was decreased, efficiencies remained high. At 11% load the efficiency of the circuit was measured as 80.09% with THD of 7.6%. The fact that the large inherent inductance did not significantly affect the result obtained at reduced loads helps to validate the operation of the LCL resonant tank. References Fig. 18. Load verus % Efficiency for Open Loop and Closed Loop Controls. effective approach for DC-DC converter output voltage regulation. 6. Conclusions [1] J.A.K.S. Bhat, Fixed frequency PWM series-parallel resonant converter, IEEE ransactions on Industrial Electronics 28(5) (1992), [2] J.A. Sabate, M.M. Jovanic, F.C. Lee and R.T. Gean, Analysis and design- optimization of lcc resonant inverter for highfrequency ac distributed power system, IEEE Transactions on Industrial Electronics 42(1) (1995), [3] A.K.S. Bhat and S.B. Dewan, A Generalized Approach for the Steady-State Analysis of Resonant Inverters, IEEE Transactions on Industrial Electronics 25(2) (1989), [4] C.Q. Lee, R. Liu and S. Sooksatra, Nonresonant and resonant coupled zero voltage switching converters, IEEE Transactions on Powerl Electronics 5(4) (1990), [5] Z. Ye, J.C.W. Lam, P.K. Jain and P.C. Sen, A robust onecycle controlled full-bridge series-parallel resonant inverter for a high-frequency AC (HFAC) distribution system, IEEE Transactions on Powerl Electronics 22(6) (2007), [6] T.-J. Liang, R.-Y. Chen and J.-F. Chen, Current-fed parallelresonant DC-AC inverter for cold-cathode fluorescent lamps with zero-current switching, IEEE Transactions on Power Electronics 23(4) (2008), [7] M.S. Agamy and P.K. Jain, A variable frequency phase-shift modulated three-level resonant single-stage power factor correction converter, IEEE Transactions on Power Electronics 23(5) (2008), [8] D. Fu, F.C. Lee, Y. Qiu and F. Wang, A novel highpowerdensity three-level LCC resonant converter with constantpower-factor-control for charging applications, IEEE Transactions on Power Electronics 23(5) (2008), [9] H. Pollock, Simple constant frequency constant current loadresonant power supply under variable load conditions, Inst Electr Eng Electron Lett 33(18) (1997), [10] H. Seidel, A high power factor tuned class D converter, in Proc IEEE PESC (1988), pp [11] H. Irie and H. Yamana, Immittance converter suitable for power electronics, Trans Inst Electr Eng Jpn, 117D(8) (1997), [12] M. Borage, S. Tiwari and S. Kotaiah, Analysis and design of LCL-T resonant converter as a constant-current power supply, IEEE Transactions on Industrial Electronics 52(6) (2005), [13] M. Borage, S. Tiwari and S. Kotaiah, LCL-T resonant converter with clamp diodes: A novel constant-current power supply with inherent constant-voltage limit, IEEE Transactions on Industrial Electronics 54(2) (2007), [14] M. Borage, S. Tiwari and S. Kotaiah, A constant-current, constantvoltagehalf-bridge resonant power supply for capacitor charging, Proc Inst Electr Eng Electr Power Appl 153(3) (2006), pp [15] M. Borage, K.V. Nagesh, M.S. Bhatia and S. Tiwari, Design of LCL-T resonant converter including the effect of transformer winding capacitance, IEEE Transactions on Industrial Electronics 56(5) (2009),

12 924 S. Selvaperumal and C.C.A. Rajan / Investigation of fuzzy control [16] M. Qiu, P.K. Jain and H. Zhang, An APWM resonant inverter topology for high frequency AC power distribution systems, IEEE Transactions on Power Electronics 19(1) (2004), [17] D. Tschirhart and P. Jain, A CLL resonant asymmetrical pulse width modulated converter with improved efficiency, IEEE Transactions on Power Electronics 55(1) (2008), [18] T.S. Sivakumaran and S.P. Natarajan, Development of Fuzzy Control of Series-ParallelLoaded Resonant converter- Simulation and Experimental Evaluation, Proceedings of India International Conference on Power Electronics (2006), pp [19] M. Borage, K.V. Nagesh, M.S. Bhatia and S. Tiwari, Design of LCL-Resonant converter including the effect of transformer winding capacitance, IEEE Transactions on Industrial Electronics 56(4) (2009), [20] B. Mangesh, K.V. Borage and M.S. Nagesh, Bhatia, Sunil Tiwari, Characteristics and Design of an Asymmetrical Duty-Cycle-Controlled LCL-T Resonant Converter, IEEE Transactions on Power Electronics 24(10) (2009),