Soheil Sadeghi Eshkevari

Xiaocheng Tang, Soheil Sadeghi Eshkevari, Haoyu Chen et al.

Transformers have enabled breakthroughs in NLP and computer vision, and have recently began to show promising performance in trajectory prediction for Autonomous Vehicle (AV). How to efficiently model the interactive relationships between the ego agent and other road and dynamic objects remains challenging for the standard attention module. In this work we propose a general Transformer-like architectural module MnM network equipped with novel masked goal conditioning training procedures for AV trajectory prediction. The resulted model, named golfer, achieves state-of-the-art performance, winning the 2nd place in the 2022 Waymo Open Dataset Motion Prediction Challenge and ranked 1st place according to minADE.

Soheil Sadeghi Eshkevari, Xiaocheng Tang, Zhiwei Qin et al.

In this study, a real-time dispatching algorithm based on reinforcement learning is proposed and for the first time, is deployed in large scale. Current dispatching methods in ridehailing platforms are dominantly based on myopic or rule-based non-myopic approaches. Reinforcement learning enables dispatching policies that are informed of historical data and able to employ the learned information to optimize returns of expected future trajectories. Previous studies in this field yielded promising results, yet have left room for further improvements in terms of performance gain, self-dependency, transferability, and scalable deployment mechanisms. The present study proposes a standalone RL-based dispatching solution that is equipped with multiple mechanisms to ensure robust and efficient on-policy learning and inference while being adaptable for full-scale deployment. A new form of value updating based on temporal difference is proposed that is more adapted to the inherent uncertainty of the problem. For the driver-order assignment, a customized utility function is proposed that when tuned based on the statistics of the market, results in remarkable performance improvement and interpretability. In addition, for reducing the risk of cancellation after drivers' assignment, an adaptive graph pruning strategy based on the multi-arm bandit problem is introduced. The method is evaluated using offline simulation with real data and yields notable performance improvement. In addition, the algorithm is deployed online in multiple cities under DiDi's operation for A/B testing and is launched in one of the major international markets as the primary mode of dispatch. The deployed algorithm shows over 1.3% improvement in total driver income from A/B testing. In addition, by causal inference analysis, as much as 5.3% improvement in major performance metrics is detected after full-scale deployment.

Soheila Sadeghi Eshkevari, Soheil Sadeghi Eshkevari, Debarshi Sen et al.

To maintain structural integrity and functionality during the designed life cycle of a structure, engineers are expected to accommodate for natural hazards as well as operational load levels. Active control systems are an efficient solution for structural response control when a structure is subjected to unexpected extreme loads. However, development of these systems through traditional means is limited by their model dependent nature. Recent advancements in adaptive learning methods, in particular, reinforcement learning (RL), for real-time decision making problems, along with rapid growth in high-performance computational resources, help structural engineers to transform the classic model-based active control problem to a purely data-driven one. In this paper, we present a novel RL-based approach for designing active controllers by introducing RL-Controller, a flexible and scalable simulation environment. The RL-Controller includes attributes and functionalities that are defined to model active structural control mechanisms in detail. We show that the proposed framework is easily trainable for a five story benchmark building with 65% reductions on average in inter story drifts (ISD) when subjected to strong ground motions. In a comparative study with LQG active control method, we demonstrate that the proposed model-free algorithm learns more optimal actuator forcing strategies that yield higher performance, e.g., 25% more ISD reductions on average with respect to LQG, without using prior information about the mechanical properties of the system.

Liam M. Cronin, Soheil Sadeghi Eshkevari, Debarshi Sen et al.

This study proposes a learning-based method with domain adaptability for input estimation of vehicle suspension systems. In a crowdsensing setting for bridge health monitoring, vehicles carry sensors to collect samples of the bridge's dynamic response. The primary challenge is in preprocessing; signals are highly contaminated from road profile roughness and vehicle suspension dynamics. Additionally, signals are collected from a diverse set of vehicles vitiating model-based approaches. In our data-driven approach, two autoencoders for the cabin signal and the tire-level signal are constrained to force the separation of the tire-level input from the suspension system in the latent state representation. From the extracted features, we estimate the tire-level signal and determine the vehicle class with high accuracy (98% classification accuracy). Compared to existing solutions for the vehicle suspension deconvolution problem, we show that the proposed methodology is robust to vehicle dynamic variations and suspension system nonlinearity.

Soheil Sadeghi Eshkevari, Martin Takáč, Shamim N. Pakzad et al.

Data-driven models for predicting dynamic responses of linear and nonlinear systems are of great importance due to their wide application from probabilistic analysis to inverse problems such as system identification and damage diagnosis. In this study, a physics-based recurrent neural network model is designed that is able to learn the dynamics of linear and nonlinear multiple degrees of freedom systems given a ground motion. The model is able to estimate a complete set of responses, including displacement, velocity, acceleration, and internal forces. Compared to the most advanced counterparts, this model requires a smaller number of trainable variables while the accuracy of predictions is higher for long trajectories. In addition, the architecture of the recurrent block is inspired by differential equation solver algorithms and it is expected that this approach yields more generalized solutions. In the training phase, we propose multiple novel techniques to dramatically accelerate the learning process using smaller datasets, such as hardsampling, utilization of trajectory loss function, and implementation of a trust-region approach. Numerical case studies are conducted to examine the strength of the network to learn different nonlinear behaviors. It is shown that the network is able to capture different nonlinear behaviors of dynamic systems with very high accuracy and with no need for prior information or very large datasets.