Date of Award


Publication Type

Master Thesis

Degree Name



Electrical and Computer Engineering

First Advisor

Xiang Chen


autonomous operation, extremum seeking control, LiDAR




LiDAR detection is susceptible to ambient interference. Therefore, it is important to maintain LiDAR detection performance when it operates autonomously in varying environments. In this paper, an optimization approach is proposed to automatically regulate LiDAR detection range through a model-guided extremum seeking control (ESC) against the variation of ambient conditions. A neural network model is trained with experimental LiDAR data off-line to simulate the impact of ambient conditions, and an Environmental Index (EI) is proposed to classify the ambient conditions. In order to obtain the optimal LiDAR detection range for each classified ambient condition, a designed cost function is used to obtain off-line solutions for each ambient condition. In order to deal with modelling uncertainties, an on-line optimization algorithm, ESC, is employed with initial conditions originating in the results of off-line optimization. The effectiveness of this model-guided ESC mechanism is then validated with experiments involving a real LiDAR on a mobile carrier.