As the basic conditions for laser inertial confinement fusion (ICF) research, the targets are required to be well specified and elaborately fabricated. Because of the characteristics of the targets, the research and fabrication process is a systematically tough task, which needs fundamental and deep insights into film deposition, mechanical machining, precise measurement and assembly, etc. As a result, knowledge of material science, physics, mechanical as well as electronics is a necessity for target researchers. In this paper, we give introductions to the state of art on target fabrication for ICF research at Research Center of Laser Fusion (RCLF) in China.

For assembly of trans-scale micro-device capsule fill tube assemblies (CFTA) for inertial confinement fusion (ICF) targets, a high-precision space assembly approach based on micro-vision is proposed in this paper. The approach consists of three modules: (i) a posture alignment module based on a multi-vision monitoring model that is designed to align two trans-scale micro-parts in 5DOF while one micro-part is in ten microns and the other one is in hundreds of microns; (ii) an insertion depth control module based on a proposed local deformation detection method to control micro-part insertion depth; (iii) a glue mass control module based on simulation research that is designed to control glue mass quantitatively and to bond micro-parts together. A series of experiments were conducted and experimental results reveal that attitude alignment control error is less than ±0.3°, position alignment control error is less than ±5 µm, and insertion depth control error is less than ±5 μm. Deviation of glue spot diameter is controlled at less than 15 μm. A CFTA was assembled based on the proposed approach, the position error in 3D space measured by computerized tomography (CT) is less than 5 μm, and glue spot diameter at the joint is 56 μm. Through volume calculation by the cone calculation formula, the glue mass is about 23 PL when the cone height is half the diameter.

For the dual peg-in-hole compliance assembly task of upper and lower double-hole structural micro-devices, a skill-learning method is proposed. This method combines offline training in a simulation space and online training in a realistic space. In this paper, a dual peg-in-hole model is built according to the results of a force analysis, and contact-point searching methods are provided for calculating the contact force. Then, a skill-learning framework is built based on deep reinforcement learning. Both expert action and incremental action are used in training, and a reward system considers both efficiency and safety; additionally, a dynamic exploration method is provided to improve the training efficiency. In addition, based on experimental data, an online training method is used to optimize the skill-learning model continuously so that the error caused by the deviation in the offline training data from reality can be reduced. The final experiments demonstrate that the method can effectively reduce the contact force while assembling, improve the efficiency and reduce the impact of the change in position and orientation.