arXiv subjects

M. Egerstedt

Publications and source records attributed to M. Egerstedt.

3 recordsLinked to original sources

arXiv2017

Performance Regulation and Tracking via Lookahead Simulation: Preliminary Results and Validation

This paper presents an approach to target tracking that is based on a variable-gain integrator and the Newton-Raphson method for finding zeros of a function. Its underscoring idea is the determination of the feedback law by measurements of the system's output and estimation of its future state via lookahead simulation. The resulting feedback law is generally nonlinear. We first apply the proposed approach to tracking a constant reference by the output of nonlinear memoryless plants. Then we extend it in a number of directions, including the tracking of time-varying reference signals by dynamic, possibly unstable systems. The approach is new hence its analysis is preliminary, and theoretical results are derived for nonlinear memoryless plants and linear dynamic plants. However, the setting for the controller does not require the plant-system to be either linear or stable, and this is verified by simulation of an inverted pendulum tracking a time-varying signal. We also demonstrate results of laboratory experiments of controlling a platoon of mobile robots.

math.OC↗

arXiv2016

Approximately Truthful Multi-Agent Optimization Using Cloud-Enforced Joint Differential Privacy

Multi-agent coordination problems often require agents to exchange state information in order to reach some collective goal, such as agreement on a final state value. In some cases, it is feasible that opportunistic agents may deceptively report false state values for their own benefit, e.g., to claim a larger portion of shared resources. Motivated by such cases, this paper presents a multi-agent coordination framework which disincentivizes opportunistic misreporting of state information. This paper focuses on multi-agent coordination problems that can be stated as nonlinear programs, with non-separable constraints coupling the agents. In this setting, an opportunistic agent may be tempted to skew the problem's constraints in its favor to reduce its local cost, and this is exactly the behavior we seek to disincentivize. The framework presented uses a primal-dual approach wherein the agents compute primal updates and a centralized cloud computer computes dual updates. All computations performed by the cloud are carried out in a way that enforces joint differential privacy, which adds noise in order to dilute any agent's influence upon the value of its cost function in the problem. We show that this dilution deters agents from intentionally misreporting their states to the cloud, and present bounds on the possible cost reduction an agent can attain through misreporting its state. This work extends our earlier work on incorporating ordinary differential privacy into multi-agent optimization, and we show that this work can be modified to provide a disincentivize for misreporting states to the cloud. Numerical results are presented to demonstrate convergence of the optimization algorithm under joint differential privacy.

math.OC↗

arXiv2016

Hamiltonian-Based Algorithm for Optimal Control

This paper proposes an algorithmic technique for a class of optimal control problems where it is easy to compute a pointwise minimizer of the Hamiltonian associated with every applied control. The algorithm operates in the space of relaxed controls and projects the final result into the space of ordinary controls. It is based on the descent direction from a given relaxed control towards a pointwise minimizer of the Hamiltonian. This direction comprises a form of gradient projection and for some systems, is argued to have computational advantages over direct gradient directions. The algorithm is shown to be applicable to a class of hybrid optimal control problems. The theoretical results, concerning convergence of the algorithm, are corroborated by simulation examples on switched-mode hybrid systems as well as on a problem of balancing transmission- and motion energy in a mobile robotic system.

math.OC↗