Deep reinforcement learning navigation via decision transformer in autonomous driving.

In real-world scenarios, making navigation decisions for autonomous driving involves a sequential set of steps. These judgments are made based on partial observations of the environment, while the underlying model of the environment remains unknown. A prevalent method for resolving such issues is reinforcement learning, in which the agent acquires knowledge through a succession of rewards in addition to fragmentary and noisy observations. This study introduces an algorithm named deep reinforcement learning navigation via decision transformer (DRLNDT) to address the challenge of enhancing the decision-making capabilities of autonomous vehicles operating in partially observable urban environments. The DRLNDT framework is built around the Soft Actor-Critic (SAC) algorithm. DRLNDT utilizes Transformer neural networks to effectively model the temporal dependencies in observations and actions. This approach aids in mitigating judgment errors that may arise due to sensor noise or occlusion within a given state. The process of extracting latent vectors from high-quality images involves the utilization of a variational autoencoder (VAE). This technique effectively reduces the dimensionality of the state space, resulting in enhanced training efficiency. The multimodal state space consists of vector states, including velocity and position, which the vehicle's intrinsic sensors can readily obtain. Additionally, latent vectors derived from high-quality images are incorporated to facilitate the Agent's assessment of the present trajectory. Experiments demonstrate that DRLNDT may achieve a superior optimal policy without prior knowledge of the environment, detailed maps, or routing assistance, surpassing the baseline technique and other policy methods that lack historical data.

Preparation and Characterization of Diene Rubbers/Silica Composites via Reactions of Hydroxyl Groups and Blocked Polyisocyanates.

To improve the curing reaction rate and efficiency of sulfur-cured diene-based rubbers, the introduction of some chemical compounds as activators and accelerants is inevitably required, causing potential harm to humans and ecological systems. Moreover, silica is usually employed as a green filling material for rubber reinforcement, and a silane coupling agent is always required to improve its dispersion. Herein, we reported an effective method to cure hydroxyl-functionalized rubbers/silica composites with blocked polyisocyanates, avoiding the use of any other additives. The enhanced dispersion of silica by interaction with hydroxyl groups on molecular chains endowed the composites with high-mechanical performance. The mechanical properties and crosslinking kinetics of the resultant silica composites can be regulated by adjusting the content of hydroxyl groups in the rubber, as well as the amount of the blocked polyisocyanates. The dynamic heat build-up was related to the distance between crosslinking points. A SBROH/B-TDI/silica composite prepared with blocked toluene diisocyanatem (TDI) exhibited comparable tanδ (0.21 at 0 °C and 0.11 at 60 °C) to that of silica composites cured by sulfur with the help of a silane coupling agent (SBR/S/Si69/silica, 0.18 and 0.10), suggesting great applicable potential for new tire rubber compounds.

Study on photoionization in a rubidium diode-pumped alkali laser gain medium with the optogalvanic method.

We use the optogalvanic method to calculate the concentration of rubidium ions produced by photoionization in a Rb diode-pumped alkali laser gain medium. With bias voltage added across the electrodes of a rubidium hollow cathode lamp, the measured optogalvanic current is 2.3×10(-7) A. Further study shows that the rubidium ion concentration is proportional to the pump intensity, and the drift velocity of rubidium ions is proportional to the bias voltage. When the photoionization process reaches dynamic equilibrium, the rubidium ion concentration will not increase with growing rubidium atom density. The calculated rubidium ion concentration is 1.5×10(5)-10(6) according to the experiment, and the ionization degree is less than 2.4×10(-7).