CHAPTER 6 OPTIMIZING SWITCHING ANGLES OF SRM

Transcription

1 111 CHAPTER 6 OPTIMIZING SWITCHING ANGLES OF SRM 6.1 INTRODUCTION SRM drives suffer from the disadvantage of having a low power factor. This is caused by the special and salient structure, and operational characteristics of SRM drives. The low power factor will obviously result in unduly high losses in the power system. Hence, algorithms for correcting or improving the power factor of SRM drives become very challenging objects of research study. Kwon et al (1997) proposes a Switching Power Converter with Power Factor Correction (SPC-PFC), DC/DC converter, and the converter for the motor drive. The SPC-PFC is designed to improve the power factor. Jurgen Reinert and Stefan Schroder (2002) present a topology, which consists of a pair of boost-buck power converters and a motor converter, where the boost converter is employed to correct the power factor. Barnes and Pollock (1998) change the existing power electronics converter of the SRM drive by providing active filtering and correct the power factor. The method proposed by Rim et al (1994) with a boost PFC circuit is introduced into the converter circuit of the SRM drive to implement PFC. Sharma et al (1997) presents the converter system, from which the SRM drive, is fed. The current controlled converter is utilized to correct the power factor. Venkatesan et al (2006) proposes a new PFC circuit based on Buck Boost converter for P.F. improvement.

2 112 It is seen from the above brief review that the common strategy from previous studies is to introduce the PFC circuits into the control topologies of the SRM drives. These PFC circuits generally include some inductors and capacitors, and consequently those methods are applicable to small ratings of SRM drives and they are generally not suitable for large power ratings because of the capacity, size, and cost of these devices. Furthermore, these PFC circuits will result in more complicated topologies and controls for SRM drives. The switching angles (the turn-on and turn-off angles) are flexible control parameters for SRM drives. The studies by Lang and Torrey (1991) and Sozer et al (2003) show that the switching angles have a great effect on the efficiency of the SRM drives and the efficiency can be improved by adjusting the turn-on angle and the turn-off angle. The simulation and experimental results given in Xue et al (2002) indicate that power factor in the SRM drive is dependent upon the switching angles. Therefore, this study attempts to improve the power factor, based on the further work of Xue et al (2004) by adjusting the switching angles rather than using hardware circuits. The novel strategy to improve the power factor in SRM drives is summarized as follows: the power factor is improved by changing the turn-on and turn-off angles. Hence, the present study attempts to improve the power factor in SRM drives by varying the intrinsic control parameters. It is clear that any additional external hardware circuit is not needed for the strategy proposed in this research study. Contrary to the previous methods, the proposed strategy is simpler in topology and control, incurs a lower cost, and is suitable for not only small, but also large ratings of SRM drives. The contribution of this study is described briefly as follows. A new control strategy to improve the power factor is presented. fuzzy and adaptive neuro fuzzy controllers are used to improve the power factor by optimizing the switching angles on a prototype of the 6/4 SRM drive. Simulation algorithms to search for the optimal turn-on and turn-off

3 113 angles are given. The simulation results are found to validate the proposed novel control strategy. Furthermore, the proposed scheme to improve the power factor is developed by optimizing the switching angles. 6.2 NOVEL STRATEGY TO IMPROVE POWER FACTOR IN SRM DRIVES The basic analysis of the effect of the switching angle upon the performance of SRM drives is summarized as follows: (i) (ii) (iii) (iv) Speed has a great effect upon the power factor of the SRM drives. In general, the power factor decreases with an increase in motor speed under single pulse operation. The torque generated by the SRM drive and the input AC voltage or the DC link voltage has only a rather weak, virtually negligible effect upon the power factor of SRM drives. In SRM drives with voltage PWM control, the duty cycle of the PWM affects the power factor in the SRM drives considerably. Under both single-pulse and the PWM operations, the turnon and turn-off angles influence the power factor of the SRM drives significantly. An interesting result is that, there is always a specific turn-on angle and a specific turn-off angle, regardless of the motor speed in SRM drives, which would give rise to a maximum power factor. From the above conclusions, it is seen that the power factor in SRM drives is affected by some specific parameters, which are, the turn-on angle, turn-off angle, speed, duty cycle of PWM, voltage, and the torque.

4 114 ω ref + ω - Voltage or Current Control signal The fixed Turn-on and turnoff angles Converter SRM Load Optimizing the turn-on angles turn-off angles AC supply Figure 6.1 Schematic diagram of the novel strategy to improve power factor Furthermore, the turn on angle, turn-off angle, and the speed are the key parameters affecting the power factor of SRM drives (Sozer and Torrey 2007). Thus, the novel strategy for improving the power factor of SRM drives is fulfilled by varying the turn-on and turn-off angles, together with adjusting the average voltage applied to the phase winding (or the winding current) or DC link voltage to compensate for the speed changes produced by variations in the turn-on and turnoff angles. Figure 6.1 illustrates the proposed novel strategy. It can be shown that by varying the switching angles, one can improve the power factor of SRM drives. However, such a control strategy gives rise to speed variation simultaneously. For the speed to remain at specified values when optimizing the switching angles, the following approaches can be utilized. For constant power operation under the single-pulse control, a) DC converter control; or b) changing the switching angles both for improving the power factor and controlling the speed. For constant torque operation, a) voltage PWM control; or b) current hysteresis control. The proposed strategy herein, is therefore based on the above-mentioned characteristics.

5 ADAPTIVE SWITCHING ANGLE CONTROLLER The neuro-fuzzy model in this research study uses the (ANFIS) techniques, which provide a method for the fuzzy modelling procedure to earn information about a data set, in order to compute the membership function parameters that best allow the associates fuzzy inference system to track the given input/output data. This learning method works similar to that of neural networks. In SRM drives, both the average torque and torque ripple are affected by the turn-on and turn-off angles and by the current waveforms in the motor phases. These characteristics change as a function of the motor speed. In many applications, for instance electric vehicle drives, it is highly desirable to have the highest torque/ampere ratio and lowest torque ripple, over the widest speed range possible. The SRM torque characteristic can be optimized by applying appropriate pre-calculated turn-on and turn-off angles according to the motor current and speed (Blanque et al 2005) Power Factor Improvement The turn on angle of the SRM is calculated by means of the following expression, which was first proposed by Bose et al (1985): L na * Nm θ on = θm 6 I (6.1) V dc The turn off angle is computed from the turn off angle theory proposed by Gribble et al (1999) using the following formula: ( ) 2 24* I * 1 x * N m θ off = θa θm α + α + R ua * V dc * θm (6.2)

6 116 where inductances R a and R u are reciprocals of the aligned (L a ) and unaligned (L na ) R ua =R u -R a (6.3) R R a α = (6.4) ua N m speed (rpm) V dc, DC link voltage (V) x is a constant (usually between I is phase current (A) θ m is rotor position middle angle ( o ) 1 2 and ) 2 3 The parameters mentioned above, influence the power factor in an SRM drive. It gives a basic analysis of the effect of the switching angle upon the performance of SRM drives. These influences are summarized as follows: (i) (ii) (iii) (iv) Speed has a great effect upon the power factor of SRM drives. In general, the power factor decreases with an increase in motor speed under single pulse operation. The torque generated by the SRM drive and the input AC voltage or the DC link voltage has only a weak, virtually negligible effect upon the power factor of SRM drives. In SRM drives with voltage PWM control, the duty cycle of the PWM affects the power factor considerably. Under both single-pulse and PWM operations, the turn-on and turnoff angles influence the power factor of the SRM drives significantly. An interesting result is that there is always a specific turn-on angle and a specific turn-off angle, regardless of the motor speed in the SRM drives which would give rise to a maximum power factor.

7 Power factor Computation The power factor in SRM drives can be computed by using the following expressions: Power factor = P S a a + Pb + S b + Pc + S c (6.5) P m = T 1 U m ( t) * I m ( t) dt, (6.6) m = a, b, c S m = U m rms. I m rms (6.7) where P a, P b, and P c are the phase average power; S a, S b, and S c are the phase apparent power, T denotes the period of AC sinusoidal voltage and is equal to 0.02 s, and U m (t) denotes the instantaneous value of the AC phase voltage. I m (t) denotes the instantaneous value of the AC phase current; U m, rms denotes the rms value of the AC phase voltage; and I m, rms denotes the rms value of the AC phase current. The above power factor is a non-linear function of the turn-on and turn-off angles, speed, torque, voltage, and the duty cycle of PWM.For these parameters, the turn-on and, turn-off angles, and speed are found to influence the power factor strongly. Therefore, the power factor in SRM drives can be regarded as a non-linear function of the turn-on and, turn-off angles and speed, and the function is assumed to be independent of the other parameters. Power factor = f (θ on, θ off ) (6.8)

8 118 where θ on represents the turn-on angle,θ off represents the turn-off angle and ω r represents the motor speed. For an SRM drive, the turn-on angle θ on and the turn-off angle θ off cannot be changed arbitrarily due to constraints in control method and construction topology (Ramani and Eshari 1994). Thus, the variables in Equation (6.8) should satisfy the proper constrains which are given by Equations (6.9 to 6.11). θ θ θ ons offs cons θ θ (6.9) on off one θ θ (6.10) con offe θ θ (6.11) cone where θ ons, θ offs, and θ cons are the minimum values of θ on, θ off, and the conduction angle, θ con, is equal to the difference between θ off and θ on, respectively. θ one, θ offe, and θ cone are the maximum values of θ on, θ off, and θ con, respectively. They depend on the number of phases, construction of the topology circuit, and control strategy of SRM drives. 6.4 RESULTS AND DISCUSSION Using the 6/4 SRM model, the proposed approaches, fuzzy and ANFIS-based switching angles optimization have been tested. The performances of the simulated model using the proposed fuzzy approach are shown in Figures and those using the neuro fuzzy approach are shown in Figure 6.6 to 6.13.

9 Optimization based on Fuzzy Logic Controller To control the phase current for optimal performance, the change of inductance should be known. But, it is difficult to express it as a mathematical equation because of its non-linear characteristics. A high efficiency drive by optimizing the switching angle with a precise speed control scheme is simulated and tested. This system has excellent dynamic torque control characteristics. The fuzzy based optimized turn-on and turn-off angle controller Surface Viewer are shown in Figures 6.2 and 6.3 respectively. The 3-D plot represents the input-output relationship of the designed controller. The surface view shows optimal mapping control behavior between the inputs and the output control action to be given to turn-off and turn-on angle tuning. Figure 6.2 Surface Viewer of Fuzzy based turn-on angle controller

10 120 Figure 6.3 Surface Viewer of Fuzzy based turn-off angle controller Figure 6.4 shows the simulated waveform of the power factor by the fuzzy logic controller. The power factor has improved to an average value of Fuzzy 0.9 power factor Time (sec) Figure 6.4 Power Factor of SRM by Fuzzy Controller

11 Optimization Based on Neuro-Fuzzy Controller Generally, steps in developed neuro-fuzzy-based models are: (i) Generate training/checking data from measured or simulation data (ii) (iii) (iv) (v) Use ANFIS to load training data and use the subtractive clustering or grid partition algorithm to generate the FIS. Train the initial FIS created. Validate the model using the original loaded data. Evaluate the model create. Figure 6.5 shows the MATLAB/ Simulink block diagram of the simulation. The sampling period at the input terminals of the ANFIS position estimator is selected as 75 µs. Figure 6.5 Simulink diagram of Switching angle Optimization using ANFIS

12 122 Figure 6.6 ANFIS model structure of Switching angle Optimization The ANFIS will try to formulate an optimized mapping structure from the input space to the output space, based on the training data. The training data sets are arranged in matrix form with the first and second columns as the input data and the last column as the output data, corresponding to the inputs. With the training data established, subtractive clustering is used to generate an initial FIS. The designed ANFIS model for the switching angle optimization of the 6/4 SRM is shown in Figure 6.6. The torque developed in the SRM is shown in Figure 6.7.

13 Torque (N.m.) Torque (N.m) Time (s) Time (s) Figure 6.7 Torque developed in the SRM The optimization of the turn-off and turn-on angle of the SRM using the ANFIS based approach is shown in Figures 6.8 and Turn off angle ( o ) Turn off angle (deg) Time (sec) Time (s) Figure 6.8 Turn-off angle Optimization

14 Turn on angle ( o ) Turn on angle (deg) Time (sec) Figure 6.9 Turn-on angle Optimization Figures 6.10 and 6.11 show the surface view plot of the proposed neuro-fuzzy controller. The 3-D plot represents the input-output relationship of the designed controller. The surface view shows the optimal mapping control behavior between the inputs and the output control action to be given to turn-off and turn-on angle tuning. The proposed neuro-fuzzy controller gives an efficient controller structure and optimally provides the input-output mapping. Time (s) Figure 6.10 Surface Viewer of ANFIS based turn-on angle controller

15 125 Figure 6.11 Surface Viewer of ANFIS based turn-off angle controller Neuro-Fuzzy 0.9 power factor Time (sec) Figure 6.12 Power factor variation of SRM by Optimizing the Switching angle

16 Speed (rpm) Speed (rpm) Time (sec) Time (s) Figure 6.13 Speed performance curve of SRM The power factor variations and their improvement by optimizing the switching angle are shown in Figure The speed performance of the SRM using the proposed approach is shown in Figure Power factor Controllers PID Controller Fuzzy Controller ANFIS Controller Figure 6.14 Switching angle Optimization Results Comparison

17 127 Figure 6.14 shows the comparison of the power factor values of the 6/4 SRM by optimizing the switching angles through various methods. From the comparison, it can be seen that the power factor of the SRM drive is very much improved by optimizing the turn on and turns off angle using the adaptive neuro fuzzy controller. 6.5 CONCLUSION Optimizing the switching angles in SRM drives, using an Adaptive Neuro Fuzzy controller, has been proposed and successfully simulated. The principles of operation, design considerations and simulation results have been presented. The turn on and turn off angles are optimized in order to obtain a high power factor in SRM drives and to maintain the specified speed. Improving the power factor of SRM drives is fulfilled by varying the turn-on and turn-off angles together with adjusting the average voltage applied to the phase winding. The fixed turn-on and turn-off angles are easily implemented using the Adaptive neuro fuzzy controller. The proposed strategy is suitable for both small ratings and large ratings of SRM drives. It is an effective strategy to improve the power factor in SRM drives.