Parameters Selection of Flywheel Energy Storage System Controller on Wind-Diesel Hybrid Power System using Immune Algorithm

Transcription

1 American International Journal of Contemporary Research Vol. 6, No. 5; October 206 Parameters Selection of Flywheel Energy Storage System Controller on Win-Diesel Hybri Power System using Immune Algorithm Chang Hyun Kim Jeong Phil Lee Assistant Professor Subivision of New & Renewable Electricity Kyungnam College of Information & echnology 45 Jurye-ro, Sasang-gu, Busan, Korea Abstract In this paper, robust parameters selection problem of flywheel energy storage system (FESS) controller using immune algorithm (IA) is investigate to enhance ynamic characteristics of win-iesel hybri power system. he aim for optimal parameters selection of the controller using IA is to minimize the objective function. he objective function in IA is represente as affinity of antigen an antiboy. herefore the selecte elements for calculating affinity have a ecisive effect on ynamic performance of the power system. he win generator frequency, the iesel generator frequency, the control input an the H -norm are use for calculating affinity of IA in this paper. o verify control performance of the esigne FESS controller, ynamic simulations are performe uner various isturbances such as suen step change of win power an loa as well as the ranom change of win power an loa. he control characteristics with the esigne FESS controller using IA are compare with that of the pitch controller an SMES. he simulation results show that performance of the FESS controller esigne by IA is improve significantly. Keywors: flywheel energy storage system, immune algorithm, affinity, win-iesel hybri power system, objective function.. Introuction Interconnection with a new an renewable energy has been increase an various loas have been complicately connecte in power system. Irregular output of new an renewable energy an continuous loa change cause frequency change of power system (Lee & Kim, 203; Lee & Kim, 204). Win-iesel hybri power system has been consiere in an isolate site which is ifficult to receive the electric power from the main power system. However the irregular output of the win power source causes a fluctuation of frequency an voltage in the isolate power system (Lee & Kim 205). In orer to solve the problem, many researches about frequency control of win-iesel hybri power system have been carrie out using various control metho such as pitch control metho of the win system an iesel generation system (ripathy, Kalantar, & Balasubramanian, 99; ripathy, 997). he PI control (Nanar, 202) an variable structure control (VSC) (Das, Aitya & Kothari, 999), fuzzy control (Goutham Govin Raju & Mohame Ali, 202; Leclercq, Robyns, & Grave, 2003; hameem Ansari, & Velusami, 200), H control (Lee & Kim, 205; Singh, Mohant, Kishor, & Ray, 203) an control metho using the superconucting magnetic energy storage system (SMES) (Cuk Supriyai, Hashiguchi, Goa & umiran, 20; ripathy, Kalantar, & Balasubramanian, 99; ripathy, 997) were propose. In this paper, flywheel energy storage system (FESS) is applie to control frequency of win-iesel hybri power system. he FESS is an electric power storage system in which the electrical energy is store by converting it into mechanical rotational energy. he FESS is an environment-frienly energy storage system which can be use for uninterruptible power supply (UPS), power quality improvement, storage of istribute power sources such as solar power an win power an loa leveling (Lee, et al., 2009; Lee, Han, & Park, 20). It is possible to control the frequency quickly in spite of the suen loa change, because the active power output of an FESS is very fast (Lee & Kim, 203; Lee & Kim, 204; Lee & Kim, 205). 28

2 ISSN X (Print), X (Online) Center for Promoting Ieas, USA Several esign methos of the FESS controller for win-iesel hybri system have been propose such as fuzzy controller (Leclercq, Robyns, & Grave, 2003), controller using genetic algorithm (GA) (Lee & Kim, 204) an H controller (Lee & Kim, 205). Since the controller performance using GA is influence by the objective function selection metho, how to select the objective function is very important. Since the H control is able to esign controller incluing the system uncertainties in controller esign stage, the esigne H controller provies the robust control performance. However, an orer of controller become very high orer an structure of the controller is very complicate. In this paper, robust parameters selection problem of flywheel energy storage system (FESS) controller using immune algorithm (IA) is investigate to enhance ynamic characteristics of win-iesel hybri power system. Like the GA, the controller performance using IA is influence by the objective function selection metho, therefore the frequency, the control input an the H -norm are inclue in objective function to esign robust controller. o verify control performance of the esigne FESS controller, ynamic simulations are performe uner various isturbances such as suen the step change of win power an loa as well as the ranom change of win power an loa. he control characteristics with the esigne FESS controller using IA are compare with that of the pitch controller an SMES (ripathy, 997). 2. Win iesel hybri power system moel Fig. A win iesel hybri power generation system with FESS Fig. shows the system configuration for the win iesel hybri power generation system with flywheel (Lee & Kim, 205). Fig. 2 shows the block iagram for the win iesel power generation system (Cuk Supriyai, Hashiguchi, Goa, & umiran, 20; Lee & Kim, 205; ripathy, Kalantar, & Balasubramanian, 99) with the pitch controller an the flywheel. his block iagram moel consists of a win system moel, a iesel system moel, a blae pitch control an a generator moel. Fig. 2: A block iagram for win iesel power generation system with pitch controller an flywheel 29

3 American International Journal of Contemporary Research Vol. 6, No. 5; October Win System Moel he win ynamics moel incluing blae pitch control of the win turbine (Lee & Kim, 204; ripathy & Mishra, 996) is as following t t t t H H u () p2 p2 K p2p K p2p H K p2 H H u (2) p2 p2 P K 3 K H P (3) m w p pc K m K fc fc Pm w w2 Pw (4) 2Hw 2Hw 2Hw 2Hw he transfer function of the hyraulic pitch actuator is split into two blocks. H is the hyraulic pitch actuator variable an H is ummy variable. P m is the win power eviation. Pw is change in the win power input. w is win frequency eviation. p, p2 is time constant of the hyraulic pitch actuator, K p2 is the hyraulic pitch actuator gain, K pc is the blae characteristic gain, K p3 is the ata fit pitch response gain, K fc is the flui coupling gain, H is the inertia constant of the win turbine system. w 2.2 Diesel System Moel he iesel ynamics moel incluing the governor system (Lee & Kim, 204; ripathy & Mishra, 996) is as following t t w 2 w K fcw2 Pf Ploa (5) 2H 2H P K (6) Pf t f w 2 K w2 P f Pf he transfer function of the iesel governor is split into two blocks. Pf an P f is ummy variable. Ploa is change in loa. H is the inertial constant of the iesel engine, gain of iesel governor, is time constant of the iesel governor. (7) is the iesel governor output variable K is the 2.3 Flywheel System moel he FESS can hanle high power level an charge/ischarge spee of the FESS is very fast. he FESS can be moele by the first orer transfer function. herefore the output power of the FESS can be written as following equation (Lee & Kim, 204). t Where PFESS PFESS u2 (8) FESS FESS PFESS change of the FESS output is, FESS is time constant of the FESS. he linearize equation of the win iesel hybri power system in Fig. 2 incluing the win system, the iesel system, the pitch control an the FESS is as following, x Ax Bu p (9) Where, x, u an p are the state, control an isturbance vector respectively. A, B an are constant matrices which epen on system parameters an the operating point. he state, control an isturbance variables without controller are as following, x [ H, H, P m, w, w2, f 2, P f, Pf, PFESS ] 30

4 ISSN X (Print), X (Online) Center for Promoting Ieas, USA u u, u ], p P w, P ] [ 2 [ Loa 3. Design of FESS controller using IA 3. Summary of H control theory he proceure for esigning H controller is as following (Lee & Kim, 205). Select two shaping filters W S (S) an W (s). 2 Specify the control structure of Fig. 3 an erive the corresponing plant. 3 Compute a H controller K(s) for this plant. Fig. 3: Augmente system moel incluing weighte function Both isturbance attenuation an robust stability for the power system were treate simultaneously by using mixe sensitivity problem. he robust stability an the performance for uncertainties of power system in this paper were represente by the same frequency weighte transfer function of reference Lee & Kim (205) 3.2 Immune Algorithm Immune algorithm is has been applie to various optimization problems (Chun, Kim, & Jung, 997; Huang, 999) as an optimization algorithm that simulates the human immune system. As compare to GA, IA performs the optimization using the memory cell in orer to ensure the convergence of the optimum solution. It has affinity calculations for implementing iversity in real immune systems an performs a self-ajusting function of the immune system by an expecte value calculation for the antigen. herefore, IA may be resolve premature convergence problems by maintaining a memory mechanism an iversity of antiboy. Summary for optimization proceure of IA is as following, Recognition of antigen 2 Initial antiboy population formulation 3 Affinity calculation 4 Differentiation towar the memory cell 5 Boost or restriction of antiboy prouction 6 Crossover an mutation 7 New antiboy generation 8 3~7 repetition 9 Optimal antiboy selection accoring to termination criterion 3

5 American International Journal of Contemporary Research Vol. 6, No. 5; October Optimal parameters selection using IA Fig: 4 A block iagram for selecting controller parameters of the flywheel Fig.4 shows the block iagram for selecting parameters for the FESS controller using IA. he input of the FESS controller is a system frequency 2 in Fig. 4. he FESS output power limit of -0.0 P FESS 0.0 (pukw) is consiere. Parameters K f 0, f, f 2 to be optimize by IA become an antiboy. In orer to obtain an optimal antiboy, the affinity calculation of the antigen an antiboy is neee. For calculating the affinity, propose objective function J is as following equation. J tte t0 t t t u t P (0) 2 Where t an te is simulation time an simulation termination time, respectively.,,, are weighting values of each parameter. In the case of the minimum value search problem of the objective function, the affinity is as following equation. affinity J 4. Simulation Results he system parameters for the computer simulation (Lee & Kim, 205; ripathy, 997) are shown in able. able System Parameters H w 3. 5s, H 8. 5s, K fc 6.2Hz / pukw () K p 4.0, K p2. 25, p 0. 60s, p s K p3.4, K p 0.08 pukw/ eg., K FESS 0. Fig. 5 shows the simulation results for the frequency eviation of the iesel an win system with the conventional PI pitch control, SMES (ripathy, 997) an esigne FESS controller (IAFESS) respectively for a step loa change of 0.0 (p.u.kw). he frequency oscillations with IAFESS in Fig. 5(a) an (b) are significantly suppresse an settling time of the frequency response is very fast. On the other han, maximum eviation of frequency using IAFESS is smaller than that using conventional PI an SMES. 32

6 ISSN X (Print), X (Online) Center for Promoting Ieas, USA (a) Diesel system frequency eviation (b) Win system frequency eviation Fig. 5 Frequency responses for 0.0 p.u. step loa variation (a) Diesel system frequency eviation (b) Win system frequency eviation Fig. 6 Frequency responses for 0.0 p.u. step win power variation 33

7 American International Journal of Contemporary Research Vol. 6, No. 5; October 206 Fig. 6 shows the simulation results for the frequency eviation of iesel an win system with the conventional PI, SMES an IAFESS respectively for a step win power input change of 0.0 (p.u.kw). he frequency oscillations with the IAFESS in Fig. 6(a) an (b) are significantly suppresse an settling time of the frequency response is very fast. Fig. 7 Ranom loa variation (a) Diesel system frequency eviation (b) Win system frequency eviation Fig. 8: Frequency responses for ranom loa variation Fig. 7 shows ranom loa changes. Fig. 8 shows a comparison of the ynamic simulation results for frequency variation when a ranom loa changes like Fig. 7 are applie. he frequency oscillations with the IAFESS in Fig. 8 are significantly suppresse an settling time of the frequency response is very fast. he results showe that the IAFESS was more robust than that using the conventional PI an SMES. 34

8 ISSN X (Print), X (Online) Center for Promoting Ieas, USA Fig. 9 Ranom win power input change (a) Diesel system frequency eviation (a) Win system frequency eviation Fig. 0: Frequency responses for ranom win power input change Fig. 9 shows ranom win power input changes. Fig. 0 shows a comparison of the ynamic simulation results for frequency variation when a ranom win power changes like Fig. 9 are applie. he frequency oscillations with the IAFESS in Fig. 0 are significantly suppresse an settling time of the frequency response is very fast. he results showe that the IAFESS was more robust than that using the conventional PI an SMES. 5. Conclusion In this paper, the robust controller problem of FESS using IA is investigate to enhance ynamic characteristics of win-iesel hybri power system. he frequency, the control input an the H -norm are inclue in objective function to esign robust controller. 35

9 American International Journal of Contemporary Research Vol. 6, No. 5; October 206 o verify control performance of the esigne FESS controller, ynamic simulations are performe uner various isturbances such as suen the step change of win power an loa as well as the ranom change of win power an loa. he simulation results showe that the FESS controller using IA provie better ynamic responses in comparison with the conventional PI an SMES. References Chun, J. S., Kim, M. K. an Jung, H. K., (997). Shape Optimization of Electromagnetic Devices using Immune Algorithm, IEEE rans. On Magnetics, 33(2), pp Cuk Supriyai A.N, akuhei Hashiguchi, aahiro Goa an umiran, (20). Control Scheme of Hybri Win- Diesel Power Generation System, From urbine to Win Farms - echnical Requirements an Spin-Off Proucts, Dr. Gesche Krause (E.), ISBN: , Inech Das, D., Aitya, S. K. an Kothari, D. P., (999). Dynamics of Diesel an Win urbine Generator on an Isolate Power System, Electrical Power an Energy System, 2, pp Goutham Govin Raju, R an Mohame Ali, S., (202). Fuzzy Base Gain Scheule PI Controller for an Isolate Win-Diesel Hybri Power System, Bulletin of Electrical Engineering an Infomatics, (3), pp Huang, S. J., (999). Enhancement of thermal unit commitment using Immune Algorithms Base Optimization Approaches, Electrical Power an Energy System 2, pp Lee, J.P., Jeong, N. H., Ham. Y. H., Han. S. C., Jung, S. Y., Park, B. C. an Sung,. H., (2009). Assessment of the Energy loss for SFES with Rotational Core ype PMSM/G, IEEE rans. On App. Superconuctivity, 9(3) Lee, J.P., Han, S. C. an Park, B. C. (20). Experimental Estimation on Magnetic Friction Superconuctor Flywheel Energy Storage System. Journal of Magnetics, 6(2), pp , oi: /JMAG Lee, J. P. an Kim, H. G., (203) Loa Frequency Control of Power System using Energy Storage System, International Smart Gri Conference &Exhibition, Jeju, Korea, pp Lee, J. P. an Kim, H. G., (204). Application of Flywheel Energy Storage System in Hybri Distribute Generation System, he 20th International Conference on Electrical Engineering, Jun 5-9, jeju, Korea, pp Lee, J. P. an Kim, H. G., (205). Design of Robust Controller of Flywheel Energy Storage System for Hybri Distribute Generation System, International Smart Gri Conference, October 2-4, Gwangju, Korea. Pp Leclercq, L., Robyns, B. an Grave. J. M., (2003). Control Base on Fuzzy Logic of a Flywheel Energy Storage System Assocciate with Win an Diesel Generator, Mathematics an Computer in Simulation 63, pp Mori, K. sukiyama, M an Fukua,., (993). Immune Algorithm with Searching Diversity an Its Application to Resource Allocation Problem,. IEE Japan, 3-C(0), pp Nanar, C. S. A., (202). Robust PI Control of Smart Controllable Loa for Frequency Stabilization of Microgri Power System, Elsevier, Renewable Energy, pp. -8, Singh. V. P., Mohanty. S. R., Kishor, N. an Ray, P. K., (203). Robust H-infinity Loa Frequency Control in Hybri Distribute Generation System, Electrical Power an Energy Systems 46, pp hameem Ansari, M. M an Velusami, S., (200). Dual Moe Lingustic Hege Fuzzy Logic Controller for ans Isolate Win-Diesel Hybri Power System with Superconucting Magnetic Energy Storage Unit, Energy Conversion an Management 5, pp ripathy, S. C. Kalantar, M. an Balasubramanian, R., (99). Dynamics an stability of win an iesel turbine generators with superconucting magnetic energy storage unit on an isolate power system, IEEE rans. On Energy Conversion, 6(4), pp , Dec. ripathy, S. C. an Mishra, I. P., (996). Dynamic performance of win-iesel power system with capacitive energy storage, Energy Convers. Mgmt, 37(2), pp ripathy, S. C., (997). Dynamic simulation of hybri win-iesel power generation system with superconuctor magnetic energy storage, Energy Convers. Mgmt, 38( 9), pp