Search

This paper develops a novel adaptive $H_{\infty}$ control scheme for adaptive cruise control (ACC) system via the data-driven learning. In the proposed technique, a continuous time ACC system is first constructed with unknown system dynamics. To estimate the unknown system dynamics, an adaptive estimator is then formulated utilizing the vectorization and Kronecker's products operations, enabling the reconstruction of the unknown system dynamics through the detectable input/output information. An adaptive law is utilized to ensure the convergence of the estimated parameters. Moreover, a data-driven learning technique is employed to resolve the constructed Riccati equation in an online manner. To accomplish this, the Riccati equation is reformulated via the Kronecker's products by using another adaptive law, whose convergence can be also effectively guaranteed. Unlike existing neural network-based approximate dynamic programming (ADP) algorithms, the data-driven learning scheme proposed in this paper does not involve neural network approximation errors, so the solution is relatively more accurate. Finally, simulation and experimental verification results are provided to verify the effectiveness of the presented $H_{\infty}$ control and data-driven learning algorithm.

Although tracking control of servo mechanisms has been extensively studied, optimal tracking control of servo mechanisms via reinforcement learning (RL) is still a challenging problem, especially when there is inevitable friction in actual servo mechanisms. This article presents an optimal tracking control approach of servo mechanism by using a generalized policy learning (GPL) algorithm, which avoids actor neural network (ANN) and solving the Hamilton-Jacobi–Bellman equation (HJBE) (the classic RL-based optimal control method generally requires solving HJBE, which is essentially a partial differential equation, and thus, it is not easy to solve directly; note that the method designed in this article does not require solving HJBE, avoiding larger errors caused by solving HJBE). To achieve this objective, an adaptive steady-state controller is developed to ensure the desired tracking performance while designing an adaptive optimal controller to further adjust the tracking error dynamics by using an optimal manner. To this end, a critic neural network (CNN) is employed to modify the specified optimal value function, and this is conducive to learning the critic with Bellman scheme. Moreover, the stability of the system is ensured through the Lyapunov theory. Finally, the efficacy of the proposed control technique is tested through simulations and experiments.

Robust and efficient vehicle detection is an important task of environment perception of intelligent vehicles, which directly affects the behavior decision-making and motion planning of intelligent vehicles. Due to the rapid development of sensor and computer technology, the algorithm and technology of vehicle detection have been updated rapidly. But, there are few reviews on vehicle detection of intelligent vehicles, especially covering all kinds of sensors and algorithms in recent years. This article presents a comprehensive review of vehicle detection approaches and their applications in intelligent vehicle systems to analyze the development of vehicle detection, with a specific focus on sensor types and algorithm classification. First, more than 300 research contributions are summarized in this review, including all kinds of vehicle detection sensors (machine vision, millimeter-wave radar, lidar, and multisensor fusion), and the performance of the classic and latest algorithms was compared in detail. Then, the application scenarios of vehicle detection with different sensors and algorithms were analyzed according to their performance and applicability. Moreover, we also systematically summarized the methods of vehicle detection in adverse weather. Finally, the remaining challenges and future research trends were analyzed according to the development of intelligent vehicle sensors and algorithms.

The durability of proton exchange membrane fuel cells (PEMFC) is an important issue that restricts their large-scale application. To improve their reliability during use, this paper proposes a short-term performance degradation prediction model using particle swarm optimization (PSO) to optimize the gate recurrent unit (GRU). After training using only the data from the first 300 h, good prediction accuracy can be achieved. Compared with the traditional GRU algorithm, the proposed prediction method reduces the root mean square error (RMSE) and mean absolute error (MAE) of the prediction results by 44.8 % and 35.1 %, respectively. It avoids the problem of low accuracy in predicting performance during the temporary recovery phase in traditional GRU models, which is of great significance for the health management of PEMFC.