No abstract

No abstract

We present an experimental study of the high-pressure, high-temperature behaviour of cerium up to ∼22 GPa and 820 K using angle-dispersive x-ray diffraction and external resistive heating. Studies above 820 K were prevented by chemical reactions between the samples and the diamond anvils of the pressure cells. We unambiguously measure the stability region of the orthorhombic oC4 phase and find it reaches its apex at 7.1 GPa and 650 K. We locate the α-cF4–oC4–tI2 triple point at 6.1 GPa and 640 K, 1 GPa below the location of the apex of the oC4 phase, and 1–2 GPa lower than previously reported. We find the α-cF4 → tI2 phase boundary to have a positive gradient of 280 K (GPa)−1, less steep than the 670 K (GPa)−1 reported previously, and find the oC4 → tI2 phase boundary to lie at higher temperatures than previously found. We also find variations as large as 2–3 GPa in the transition pressures at which the oC4 → tI2 transition takes place at a given temperature, the reasons for which remain unclear. Finally, we find no evidence that the α-cF4 → tI2 is not second order at all temperatures up to 820 K.

This paper considers the problem of how robots in long-term space operations can learn to choose appropriate sources of assistance to recover from failures. Current assistant selection methods for failure handling are based on manually specified static look up tables or policies, which are not responsive to dynamic environments or uncertainty in human performance. We describe a novel and highly flexible learningbased assistant selection framework that uses contextual multiarm bandit algorithms. The contextual bandits exploit information from observed environment and assistant performance variables to efficiently learn selection policies under a wide set of uncertain operating conditions and unknown/dynamically constrained assistant capabilities. Proof of concept simulations of long-term human-robot interactions for space exploration are used to compare the performance of the contextual bandit against other state of the art assistant selection approaches. The contextual bandit outperforms conventional static policies and non-contextual learning approaches, and also demonstrates favorable robustness and scaling properties.