| POWER MANAGEMENT STRATEGY FOR THE 48 V MILD HYBRID ELECTRIC VEHICLE BASED ON THE CHARGE-SUSTAINING CONTROL |
| Jeongwon Sohn1, Myoungho Sunwoo2, Kyunghan Min2, Jaewook Shin2, Manbae Han3 |
1LG Electronics Inc.
2Hanyang University
3Keimyung University |
|
|
|
|
|
|
|
| ABSTRACT |
|
To enhance the 48 V mild hybrid electric vehicle performance using a smaller capacity and lower voltage
battery than the full hybrid electric vehicle, a novel power management strategy needs to be established that considers the
characteristics and limitations of the components. This paper proposes a charge-sustaining control strategy as a ground
principle of the 48 V hybrid electric vehicle control for managing the battery state-of-charge (SOC) to stay near the most
efficient regime. The base efficiency characteristics of the component models including engine, motor/generator, and battery
are determined in the form of efficiency maps using the powertrain analysis tool. Then the control strategy is formulated as
a nonlinear optimal regulation problem that meets two conflicting control objectives, such as fuel efficiency improvement and
state-of-charge maintenance. The optimal regulation problem implements a discrete-time Hamilton-Jacobi-Bellman approach.
The proposed strategy is evaluated by comparing with the reference strategy applying the Dynamic Programming (DP), i.e.
a global optimal result, under urban dynamometer driving schedule and worldwide harmonized light duty test cycle. Through
the evaluation, the fuel efficiency of the proposed strategy with three different electrical loads is slightly deteriorated at most
by 5.03 % from the DP results with staying within a desirable SOC. This suggests that the proposed strategy is operating very
closely to global optimal performances. |
| Key Words:
48 V mild hybrid electric vehicle, Charge-sustaining control, Hamilton-Jacobi-Bellman, Optimal regulation
problem, Power management strategy |
|
|
|
Corresponding Author.
Manbae Han , Email. 
Preview
Full text via DOI
Download Citation 
Print 
