Search

Active non-cooperative target tracking is an important ability for free-floating space manipulator (FFSM) systems. Aiming at the difficulty of pose and posture estimation of non-cooperative targets and the complexity of FFSM planning and control, the paper proposed an end-to-end target tracking algorithm using deep reinforcement learning (DRL) algorithm. The algorithm uses raw camera images as input and directly outputs the manipulator joint velocities. To deal with the partially observable Markov decision processes (POMDP) problem caused by single image input, we combine the soft actor–critic algorithm with recurrent neural networks. Therefore, the velocity information of the target is introduced to transform the task into a fully observable Markov decision-making process (MDP). Our approach is completely data-driven. It avoids the position and posture estimation of non-cooperative target and the complex modeling motion planning process of FFSM. We trained and tested the algorithm by building a simulation environment in CoppeliaSim. The results showed that our approach achieved good results in active non-cooperative target tracking of FFSM systems and demonstrated the great potential of DRL algorithms in solving space tasks. [ABSTRACT FROM AUTHOR]

BACKGROUND: Baculoviruses provide long‐lasting control of crop pests and are harmless to humans and non‐target animals, making them attractive bioinsecticides. Autographa californica multiple nucleopolyhedrovirus (AcMNPV) has a wide‐host range and is one such commercial bioinsecticide, but its low infectivity to older larvae and less‐sensitive species precludes its large‐scale application. We sought to improve the infectivity of AcMNPV. RESULTS: Two enhancing factors, the truncated enhancin from Agrotis segetum granulovirus and GP37 from Cydia pomonella granulovirus, were expressed in fusion with the N‐terminal and middle domain of the polyhedrin envelope protein of AcMNPV. Western blotting and immunoelectron microscopy analysis indicated that the enhancing factors were expressed on the occlusion bodies of the resulting AcMNPV variants. Bioassays showed that the median lethal doses of the recombinant viruses were 3.9‐fold to 7.4‐fold lower than those of the wild‐type virus against the second and fourth instar of Spodoptera exigua larvae. The yields of occlusion bodies from the two recombinants in S. exigua larvae were comparable with those of the wild‐type virus both in vitro and in vivo. Further bioassays showed that the AcMNPV variants fusing the enhancing factors were incapable of infecting the second instar larvae of S. litura, Helicoverpa armigera, and Pyrausta nubilalis, which were not sensitive to the wild‐type AcMNPV. CONCLUSION: These genetically modified AcMNPV variants exhibited an enhanced infectivity and may offer better baculovirus control of crop pests. © 2019 Society of Chemical Industry [ABSTRACT FROM AUTHOR]