Sangjae Bae

Sangjae Bae

Auto-generated profile, based on publicly available information

Sangjae Bae

Department of Civil and Environmental Engineering, University of California at Berkeley, Berkeley, CA, USA

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
18 March 2021 in IEEE Transactions on Smart Grid
Verified
Journal Article
Closed access
TZTeng Zeng
Teng Zeng
SBSangjae Bae
Sangjae Bae
BTBertrand Travacca
Bertrand Travacca
SMScott Moura
Scott Moura
Plug-in electric vehicle (PEV) charging station service capability is physically limited by the charger availability and local transformer capacity. However, the station operation performance has become an increasingly important factor for enhancing charging service accessibility. In this work, we propose an innovative station-level optimization framework to operate charging station with optimal pricing policy and charge scheduling. This model incorporates human behaviors to explicitly and effectively capture drivers' charging decision process. We propose a menu of price-differentiated charging services, which differ in both the per-unit prices and the energy delivery schedule. Involving human in the loop, the operation model also exploits the capability to alleviate the overstay issue that occurs when a PEV's charging session has completed. We then propose a muli-block convex transformation methodology to reformulate the resulted non-convex problem via Fenchel-Young Inequality; the Block Coordinate Descent algorithm is applied to solve the overall problem and the efficiency enables real-time implementation. As a result, our simulation demonstrates that the proposed control polices can realize benefits in three aspects: (i) net profits gains, (ii) overstay reduction, and (iii) quality-of-service increase.

3 August 2022 in IEEE Transactions on Control Systems Technology
Verified
Journal Article
Closed access
SBSangjae Bae
Sangjae Bae
DIDavid Isele
David Isele
ANAlireza Nakhaei
Alireza Nakhaei
PXPeng Xu
Peng Xu
AAAlexandre Miranda Añon
Alexandre Miranda Añon
SMScott Moura
Scott Moura
This article presents an online smooth-path lane-change control framework. We focus on dense traffic where intervehicle space gaps are narrow, and cooperation with surrounding drivers is essential to achieve the lane-change maneuver. We propose a two-stage control framework that harmonizes model predictive control (MPC) with generative adversarial networks (GANs) by utilizing driving intentions to generate smooth lane-change maneuvers. To improve performance in practice, the system is augmented with an adaptive safety boundary and a Kalman filter to mitigate sensor noise. Simulation studies are investigated at different levels of traffic density and cooperativeness of other drivers. The simulation results support the effectiveness, driving comfort, and safety of the proposed method.

20 December 2023 in The International Journal of Robotics Research
Verified
Journal Article
Closed access
HHHaimin Hu
Haimin Hu
DIDavid Isele
David Isele
SBSangjae Bae
Sangjae Bae
JFJaime F Fisac
Jaime F Fisac
The ability to accurately predict others’ behavior is central to the safety and efficiency of robotic systems in interactive settings, such as human–robot interaction and multi-robot teaming tasks. Unfortunately, robots often lack access to key information on which these predictions may hinge, such as other agents’ goals, attention, and willingness to cooperate. Dual control theory addresses this challenge by treating unknown parameters of a predictive model as stochastic hidden states and inferring their values at runtime using information gathered during system operation. While able to optimally and automatically trade off exploration and exploitation, dual control is computationally intractable for general interactive motion planning, mainly due to the fundamental coupling between the robot’s trajectory plan and its prediction of other agents’ intent. In this paper, we present a novel algorithmic approach to enable active uncertainty reduction for interactive motion planning based on the implicit dual control paradigm. Our approach relies on sampling-based approximation of stochastic dynamic programming, leading to a model predictive control problem that can be readily solved by real-time gradient-based optimization methods. The resulting policy is shown to preserve the dual control effect for a broad class of predictive models with both continuous and categorical uncertainty. To ensure the safe operation of the interacting agents, we use a runtime safety filter (also referred to as a “shielding” scheme), which overrides the robot’s dual control policy with a safety fallback strategy when a safety-critical event is imminent. We then augment the dual control framework with an improved variant of the recently proposed shielding-aware robust planning scheme, which proactively balances the nominal planning performance with the risk of high-cost emergency maneuvers triggered by low-probability agent behaviors. We demonstrate the efficacy of our approach with both simulated driving studies and hardware experiments using 1/10 scale autonomous vehicles.

27 January 2022 in IEEE Robotics and Automation Letters
Verified
Journal Article
Closed access
KLKeuntaek Lee
Keuntaek Lee
DIDavid Isele
David Isele
ETEvangelos Theodorou
Evangelos Theodorou
SBSangjae Bae
Sangjae Bae
It can be difficult to autonomously produce driver behavior so that it appears natural to other traffic participants. Through Inverse Reinforcement Learning (IRL), we can automate this process by learning the underlying reward function from human demonstrations. We propose a new IRL algorithm that learns a goal-conditioned spatio-temporal reward function. The resulting costmap is used by Model Predictive Controllers (MPCs) to perform a task without any hand-designing or hand-tuning of the cost function. We evaluate our proposed Goal-conditioned SpatioTemporal Zeroing Maximum Entropy Deep IRL (GSTZ)-MEDIRL framework together with MPC in the CARLA simulator for autonomous driving, lane keeping, and lane changing tasks in a challenging dense traffic highway scenario. Our proposed methods show higher success rates compared to other baseline methods including behavior cloning, state-of-the-art RL policies, and MPC with a learning-based behavior prediction model.

1 May 2021 in Applied Energy
Verified
Journal Article
Closed access
SWSoomin Woo
Soomin Woo
SBSangjae Bae
Sangjae Bae
SMScott J. Moura
Scott J. Moura

1 January 2022 in Journal of Dynamic Systems, Measurement, and Control
Verified
Journal Article
Closed access
PDF file
SBSangjae Bae
Sangjae Bae
YKYeojun Kim
Yeojun Kim
YCYongkeun Choi
Yongkeun Choi
JGJacopo Guanetti
Jacopo Guanetti
PGPreet Gill
Preet Gill
SMScott J. Moura
Scott J. Moura
This paper examines both mathematical formulation and practical implementation of an ecological adaptive cruise controller (ECO-ACC) with connected infrastructure. Human errors are typical sources of accidents in urban driving, which can be remedied by rigorous control theories. Designing an ECO-ACC is, therefore, a classical research problem to improve safety and energy efficiency. We add two main contributions to the literature. First, we propose a mathematical framework of an online ECO-ACC for plug-in hybrid electric vehicle (PHEV). Second, we demonstrate ECO-ACC in a real world, which includes other human drivers and uncertain traffic signals on a 2.6 (km) length of the corridor with eight signalized intersections in Southern California. The demonstration results show, on average, 30.98% of energy efficiency improvement and 8.51% additional travel time.