With the growing demand for in-situ exploration of complex extreme terrains such as polar craters in lunar exploration missions, quadruped robots have demonstrated unique advantages owing to their efficient jumping and locomotion capabilities in weak gravitational fields. However, lunar gravity is only one-sixth of Earth's, which substantially prolongs the flight phase of underactuated jumping. Incorrect takeoff actions on rugged terrain generate rotational torques that cause the robot to continuously rotate in mid-air, with the angular momentum accumulating over time and leading to catastrophic consequences, thus imposing significant uncertainty on landing control. Conventional optimal control methods rely on offline preplanned trajectories and are unable to cope with the variability of unknown rugged terrains; existing deep reinforcement learning models lack effective utilization of temporal causal relationships, relying solely on current states for decision-making and struggling to predict the long-term consequences of actions. Therefore, how to leverage historical sequence information to model temporal dependencies between actions, enabling the robot to possess forward-looking prediction and adaptive adjustment capabilities during the jumping process, has become a key technical challenge for jumping control on complex lunar terrains.

With the continuous growth of global communication demands, fiber-optic networks are undergoing rapid expansion, and the most widely used silica-based optical fibers are now approaching their intrinsic loss limits. Heavy-metal fluoride ZBLAN fibers theoretically exhibit losses in the mid-infrared band that can be approximately 20 times lower than those of silica fibers. However, this theoretical advantage has long been constrained by the inevitable crystallization problem during ground-based fabrication—ZBLAN glass possesses a critical cooling rate far higher than that of silica, rendering it highly susceptible to crystallization during the fiber drawing process on Earth, which leads to a significant increase in scattering loss. Although microgravity environments may reduce crystallization rates by suppressing buoyancy-driven convection, and theoretical models support this mechanism, the lack of space-qualified automated fiber drawing equipment has previously prevented the acquisition of fiber samples of sufficient length for systematic characterization. Therefore, validating the practical feasibility and performance enhancement effects of ZBLAN fiber drawing under microgravity conditions has become a key scientific issue in overcoming the bottleneck to realizing the ultra-low-loss potential of this material.

The largest nationwide Polish study on migraine and sleep, involving 5,311 adults, found a strong association between migraine and poorer sleep quality. Nearly two-thirds (64.1%) of people with migraine reported poor sleep, compared with 38.7% of participants without headaches. People with migraine experienced problems across all assessed aspects of sleep, including falling asleep, sleep duration and efficiency, nighttime awakenings, and daytime functioning.

Poorer sleep was also associated with a greater migraine burden, including more frequent attacks, greater disability, higher stress and anxiety, and lower quality of life. The study additionally found that people with migraine were more likely to have a later chronotype, although the relationship between chronotype and migraine severity was weak.

The findings support a bidirectional relationship between migraine and sleep disturbances and suggest that sleep assessment should be routinely included in migraine care. However, because the study was cross-sectional and relied on self-reported data, it cannot establish causality. Further longitudinal studies using objective sleep measurements are needed.