arXiv subjects

Amith Manoharan

Publications and source records attributed to Amith Manoharan.

6 recordsLinked to original sources

arXiv2026

Momentum Based Reward Design for Low Emission Traffic Signal Control

Urban traffic congestion is a growing global issue contributing significantly to long commute times and environmental pollution. Traditional traffic signal control systems often fail to adapt to dynamic traffic conditions. Adaptive traffic signal control can improve urban traffic without changing road infrastructure. Deep Reinforcement Learning (DRL) has shown strong performance for this task, but existing delay and queue-based rewards often produce short-sighted or unstable policies. This paper proposes a Momentum-Based Reward Function (MBRF) that encourages vehicles to keep moving rather than penalizing congestion alone. The method is evaluated in SUMO (Simulation of Urban MObility) using standard traffic metrics such as waiting time, queue length, throughput, and CO2 emissions. Results show that the proposed reward produces better throughput-emission trade-offs and more stable learning behavior than delay or queue-based rewards, as well as classical controllers such as Max Pressure and LQF.

cs.LG↗

arXiv2025

$\pi$-MPPI: A Projection-based Model Predictive Path Integral Scheme for Smooth Optimal Control of Fixed-Wing Aerial Vehicles

Model Predictive Path Integral (MPPI) is a popular sampling-based Model Predictive Control (MPC) algorithm for nonlinear systems. It optimizes trajectories by sampling control sequences and averaging them. However, a key issue with MPPI is the non-smoothness of the optimal control sequence, leading to oscillations in systems like fixed-wing aerial vehicles (FWVs). Existing solutions use post-hoc smoothing, which fails to bound control derivatives. This paper introduces a new approach: we add a projection filter $\pi$ to minimally correct control samples, ensuring bounds on control magnitude and higher-order derivatives. The filtered samples are then averaged using MPPI, leading to our $\pi$-MPPI approach. We minimize computational overhead by using a neural accelerated custom optimizer for the projection filter. $\pi$-MPPI offers a simple way to achieve arbitrary smoothness in control sequences. While we focus on FWVs, this projection filter can be integrated into any MPPI pipeline. Applied to FWVs, $\pi$-MPPI is easier to tune than the baseline, resulting in smoother, more robust performance.

cs.RO↗

arXiv2024

Bi-level Trajectory Optimization on Uneven Terrains with Differentiable Wheel-Terrain Interaction Model

Navigation of wheeled vehicles on uneven terrain necessitates going beyond the 2D approaches for trajectory planning. Specifically, it is essential to incorporate the full 6dof variation of vehicle pose and its associated stability cost in the planning process. To this end, most recent works aim to learn a neural network model to predict the vehicle evolution. However, such approaches are data-intensive and fraught with generalization issues. In this paper, we present a purely model-based approach that just requires the digital elevation information of the terrain. Specifically, we express the wheel-terrain interaction and 6dof pose prediction as a non-linear least squares (NLS) problem. As a result, trajectory planning can be viewed as a bi-level optimization. The inner optimization layer predicts the pose on the terrain along a given trajectory, while the outer layer deforms the trajectory itself to reduce the stability and kinematic costs of the pose. We improve the state-of-the-art in the following respects. First, we show that our NLS based pose prediction closely matches the output from a high-fidelity physics engine. This result coupled with the fact that we can query gradients of the NLS solver, makes our pose predictor, a differentiable wheel-terrain interaction model. We further leverage this differentiability to efficiently solve the proposed bi-level trajectory optimization problem. Finally, we perform extensive experiments, and comparison with a baseline to showcase the effectiveness of our approach in obtaining smooth, stable trajectories.

cs.RO↗

arXiv2022

Nonlinear Model Predictive Control Framework For Cooperative Three-Agent Target Defense Game

This paper presents cooperative target defense guidance strategies using nonlinear model predictive control (NMPC) framework for a target-attacker-defender (TAD) game. The TAD game consists of an attacker and a cooperative target-defender pair. The attacker's objective is to capture the target, whereas the target-defender team acts together such that the defender can intercept the attacker and ensure target survival. We assume that the cooperative target-defender pair do not have perfect knowledge of the attacker states, and hence the states are estimated using an Extended Kalman Filter (EKF). The capture analysis based on the Apollonius circles is performed to identify the target survival regions. The efficacy of the NMPC-based solution is evaluated through extensive numerical simulations. The results show that the NMPC-based solution offers robustness to the different unknown attacker models and has better performance than CLOS and A-CLOS based strategies.

eess.SY↗

arXiv2022

Multi-AAV Cooperative Path Planning using Nonlinear Model Predictive Control with Localization Constraints

In this paper, we solve a joint cooperative localization and path planning problem for a group of Autonomous Aerial Vehicles (AAVs) in GPS-denied areas using nonlinear model predictive control (NMPC). A moving horizon estimator (MHE) is used to estimate the vehicle states with the help of relative bearing information to known landmarks and other vehicles. The goal of the NMPC is to devise optimal paths for each vehicle between a given source and destination while maintaining desired localization accuracy. Estimating localization covariance in the NMPC is computationally intensive, hence we develop an approximate analytical closed form expression based on the relationship between covariance and path lengths to landmarks. Using this expression while computing NMPC commands reduces the computational complexity significantly. We present numerical simulations to validate the proposed approach for different numbers of vehicles and landmark configurations. We also compare the results with EKF-based estimation to show the superiority of the proposed closed form approach.

cs.RO↗

arXiv2021

NMPC-Based Cooperative Strategy For A Target Pair To Lure Two Attackers Into Collision

This paper presents a cooperative target defense strategy using nonlinear model-predictive control (NMPC) framework for a two--targets two--attackers (2T2A) game. The 2T2A game consists of two attackers and two targets. Each attacker needs to capture a designated target individually. However, the two targets cooperate to lure the attackers into a collision. We assume that the cooperative target pair do not have perfect knowledge of the attacker states, and hence they estimate the attacker states using an extended Kalman filter (EKF). The NMPC scheme computes closed- loop optimal control commands for the targets while respecting imposed state and control constraints. Theoretical analysis is carried out to determine regions that will lead to the targets' survival, given the initial positions of the attacker and target agents. Numerical simulations are carried out to evaluate the performance of the proposed NMPC- based strategy for different scenarios.

eess.SY↗