EVS28 KINTEX, Korea, May 3-6, 2015 Parameter design of regenerative braking strategy and battery range of use of electric vehicle using the Optimization Technique Kiyoung Kim 1, Seungwan Son 1, Sukwon Cha 1 1 School of Mechanical & Aerospace Engineering, Seoul National University, Gwanak-ro 1, Gwanak-gu Seoul 151-744, Republic of Korea, kims1988@snu.ac.kr
Introduction I. Target electric vehicle Mortor Inverter Mechanical connection forward direction Electrical connection Target vehicle Battery Rear drive Compact car Component Vehicle weight Value 1260kg Radius of tire 0.35m Reduction ratio 3.6 Motor Battery capacity 50kW 16.4kWh 2
Optimization Technique I. Optimal parameter using Taguchi method 1. Statistical method developed by Genichi Taguchi Optimization using steepest gradient method known J= J(x1,., xn) Optimization using Taguchi method unknown SN= SN(x1,., xn) SN: signal-to-noise ratio 3
Objective and function parameter I. Objective 1. Maximization of efficiency of electric vehicle regardless of various using condition II. Function parameter 1. efficiency of electric vehicle (km/kwh) max y = y 1 y 1,ref y 1 y 1,ref : efficiency : efficiency reference 4
Forward simulator I. Analysis of efficiency by simulation program based on MATLAB/Simulink Forward simulation program considering power train dynamics Simulation using component data map Determine traction power comparing vehicle speed with target speed 5
Design parameter I. Initial SOC II. SOC range of use III. Ratio of front/rear hydraulic pressure based on vehicle deceleration Parameter combination in correlation Initial SOC & SOC range of use 6
Design parameter I. Initial SOC 1. Different OCV and internal resistance for SOC Different battery efficiency for SOC To find optimal initial SOC SOC : State of Charge = current charge amount Maximum charge amount 7
Design parameter II. SOC range of use 1. Different OCV and internal resistance for SOC Different battery efficiency for SOC To find optimal range of use SOC : State of Charge = current charge amount Maximum charge amount 8
Design parameter I. Ratio of front/rear hydraulic pressure based on vehicle deceleration 1. Total brake force = regenerative brake(motor) + mechanical brake(hybraulic) 2. Ideal brake distribution rate based on vehicle deceleration(blue line) 3. Real brake distribution rate only by mechanical brake (red line) 4. Regenerative braking makes up for deficient rear braking force 5. Meeting point ideal/real line as design parameter (green point) Rear brake force(n) 0.2g 0.4g 0.6g 0.8g Ideal brake force distribution Real brake force distribution Rear brake force(n) 0.2g 0.4g 0.6g 0.8g Ideal brake force distribution Real brake force distribution Front brake force(n) Front brake force(n) 9
Product using condition I. Driving propensity of driver 1. Indexing based on Aggressiveness factor 2. Aggressiveness factor? 1) Based on required traction power 2) Effect on efficiency according to driving pattern can be quantified Agg = a v + dt vdt <Fuel economy of conventional car for aggresiveness> <various driving cycle> 10
Level of design parameter and using condition I. Level of design parameter on primary Design parameter Description on design parameter Level 1 Level 2 Level 3 A Initial SOC(0~1) 0.75 0.95 - B SOC range of use(0~1) 0.55 0.60 0.65 C Meeting point ideal/real line 0.4g 0.6g 0.8g II. Level of using condition on primary Using condition aggressiveness Level 1 [N 1 ] mildest 0.0697 (HWFET cycle) Level2 [N 2 ] harshest 0.1646 (UDDS cycle) 11
Result of primary process I. L 18 (2 1 ⅹ3 2 ) Array Design parameter Using condition Com bina tion Function parameter S/N ratio 12
S/N ratio S/N ratio S/N ratio Analysis on sensitivity I. Table and graph of S/N ratio for level of design parameter Level Max. difference Level of A Level of B Level of C Design parameter A does not influence on function parameter due to low sesitivity Design parameter B and C influence on function parameter due to high sesitivity 13
Analysis on parameter correlation I. Table and graph of Correlation A with B slope of design parameter A&B for level is similar Low correlation between A and B 14
Estimation of S/N ratio I. Estimation of S/N ratio and comparison with result 1. Design parameter B&C influence on S/N ratio 2. Design parameter A and A-B do not influence on S/N ratio Design parameter Using condition Com bina tion Function parameter S/N ratio Level2 of A is selected because S/N ratio is lager than that of level 1 of A 15
Level of design parameter and using condition I. Level of design parameter on second Design parameter Description on design parameter Level 1 Level 2 Level 3 A Initial SOC(0~1) - 0.95 - B SOC range of use(0~1) 0.65 0.70 0.75 C Meeting point ideal/real line 0.8g 0.85g 0.9g II. Level of using condition on second Using condition aggressiveness Level 1 [N 1 ] mildest 0.0697 (HWFET cycle) Level2 [N 2 ] harshest 0.1646 (UDDS cycle) 16
Result of second process I. L 9 (3 2 ) Array Design parameter Using condition Com bina tion Function parameter S/N ratio 17
S/N ratio S/N ratio Analysis on sensitivity I. Table and graph of S/N ratio for level of design parameter Level B C Max. difference Level of B Level of C Design parameter B&C do not influence on function parameter No need to third process 18
Estimation of S/N ratio I. Estimation of S/N ratio and comparison with result 1. Design parameter B&C do not influence on S/N ratio 2. Combination number 6 is optimal combination Design parameter Using condition Com bina tion Function parameter S/N ratio S/N ratio more bigger than primary process 19
Conclusion I. Determine Optimal parameter of EV using Taguchi method II. Design parameter 1. Initial SOC 2. SOC range of use 3. Ratio of front/rear hydraulic pressure based on vehicle deceleration III. Product using condition 1. Aggressiveness factor IV. Optimal design parameter combination 1. Select parameter combination making S/N ratio maximized Design parameter Description on design parameter Optimu m A Initial SOC(0~1) 0.95 B SOC range of use(0~1) 0.7 C Meeting point ideal/real line 0.9g 20
Acknowledgement & Reference This work was supported by the Hyundai Motor Company and the National Research Foundation of Korea(NRF) grant funded by the Ministry of Science, ICT & Future Planning (MSIP) (No. 2009-0083495). Daeheung Lee et. Al., System Efficiency Analysis for Next Generation Eco-Friendly Vehicles with Aggressiveness of Real-World Driving Schedules, Transactions of KSAE, 2010.11, pp. 3178-3183 Mehrdad Ehsani et. Al., Modern Electric, Hybrid Electric, and Fuel Cell Vehicles, 2nd edition, CRC Press, 2010 김종원, 공학설계 : 창의적신제품개발방법론, 서울 : 문운당, 2008 Ho Gi Kim, Suppression Control of the Drivetrain-Oscillations of an Electric Vehicle using Taguchi method., Transaction of KSME, 2009. 5 Vol.33 No.5 pp.463-468 Chunhua Zheng, A study on Battery SOC Estimation by Regenerative Braking in Electric Vehicle Transaction of KSAE, Vol. 20 No. 1, pp.119-123 Yongsun Bak, Development of regenerative braking co operative control algorithm for electric vehicle equipped with booster brake Transactions of KSAE, 2013.5, pp. 1800-1804 21