Treffer: Real-Time Performance Optimization of Uam Propulsion Sys-tem Using Ddpg Algorithm.

Com Propulsion system for Urban Air Mobility (UAM) vehicles and characterized by non-linear system dynamics, have traditionally relied on classical PID controllers for control Optimal tuning of PID gains for these nonlinear systems is commonly derived from empirical process, or for optimal control, commonly derived using methods such as the Riccati equation or Linear Quadratic Regulator (LQR), which linearize the sys-tem around an operating point. These approaches often lose optimality when the UAM propulsion system deviates significantly from the linearization point, which is frequent during various flight phases. Physical characteristics of the engine change over time due to aging and wear, necessitating manual retuning and incurring additional maintenance costs to sustain optimal performance. To address these critical challenges and maintain optimal control performance continuously for UAM propulsion, this paper proposes a reinforcement learning-based approach. Specifically, the Deep Deterministic Policy Gradient (DDPG) algorithm is applied to implement an adaptively optimized PID controller, enabling real-time performance optimization of the propulsion system. The proposed method is validated through comprehensive simulations conducted in Matlab Simulink, demonstrating its effectiveness in maintaining optimal and adaptive control under the highly nonlinear and time-varying operational conditions characteristic of UAM propulsion systems. [ABSTRACT FROM AUTHOR]

Copyright of Advances in Consumer Research is the property of Association for Consumer Research and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)