The robotic manipulation of deformable linear objects (DLOs) has broad application prospects in many fields. However, a key issue is to obtain the exact deformation models (i.e., how robot motion ...affects DLO deformation), which are hard to theoretically calculate and vary among different DLOs. Thus, the shape control of DLOs is challenging, especially for large deformation control that requires global and more accurate models. In this article, we propose a coupled offline and online data-driven method for efficiently learning a global deformation model, allowing for both accurate modeling through offline learning and further updating for new DLOs via online adaptation. Specifically, the model approximated by a neural network is first trained offline on random data, then seamlessly migrated to the online phase, and further updated online during actual manipulation. Several strategies are introduced to improve the model's efficiency and generalization ability. We propose a convex-optimization-based controller and analyze the system's stability using the Lyapunov method. Detailed simulations and real-world experiments demonstrate that our method can efficiently and precisely estimate the deformation model and achieve the large deformation control of untrained DLOs in 2-D and 3-D dual-arm manipulation tasks better than the existing methods. It accomplishes all 24 tasks with different desired shapes on different DLOs in the real world, using only simulation data for the offline learning.
The shape control of deformable linear objects (DLOs) is challenging, since it is difficult to obtain the deformation models. Previous studies often approximate the models in purely offline or online ...ways. In this paper, we propose a scheme for the shape control of DLOs, where the unknown model is estimated with both offline and online learning. The model is formulated in a local linear format, and approximated by a neural network (NN). First, the NN is trained offline to provide a good initial estimation of the model, which can directly migrate to the online phase. Then, an adaptive controller is proposed to achieve the shape control tasks, in which the NN is further updated online to compensate for any errors in the offline model caused by insufficient training or changes of DLO properties. The simulation and real-world experiments show that the proposed method can precisely and efficiently accomplish the DLO shape control tasks, and adapt well to new and untrained DLOs.
The robot soccer game has been recognized as an excellent scenario to test the gaming algorithm of multi-agent systems. This paper develops a new simulation platform for the robot soccer game, and it ...has the advantage of open architecture, such that the formation control scheme, the path planning strategy for multiple robots, and many other algorithms can be implemented and tested. Specifically, both a Deep Reinforcement Learning (DRL) scheme and a role-assignment-based method have been successfully realized in this platform to drive multiple robots to play the soccer game, including 2V2,3V3,4V4, and so on. It is believed that the developed simulation environment can be used for data collection and transfer learning (TL), hence bridging the gap of Sim2Real technique in actual implementations.
Many industrial and domestic applications involve the manipulation of deformable objects, such as wire, soft tissue, and food material. Compared to rigid objects, they are more challenging to ...manipulate. One reason being that it is difficult to obtain the exact deformation model, which relates to the external forces and the state of the object. This paper considers the problem of modeling and the force control of deformable linear objects (DLOs), where the force control input exerted on the DLO is related to the change of features along it. An online adaptation law is proposed to update the unknown model parameters, by exploring the property of linear parameterization; Then, a model-based force control scheme is applied to actively shape the DLO, in the absence of vision feedback. The proposed method has the advantages of data efficiency due to the prior physics model and of low computational complexity (and hence better real-time performance). The convergence of estimated parameters to the actual value is rigorously proved using Lyapunov methods. A series of comparative studies are presented to validate the efficient and effective performance of the proposed modeling method. Experimental results on model learning and shape control are also given to illustrate the proposed method.
The shape control of deformable linear objects (DLOs) is challenging, since it is difficult to obtain the deformation models. Previous studies often approximate the models in purely offline or online ...ways. In this paper, we propose a scheme for the shape control of DLOs, where the unknown model is estimated with both offline and online learning. The model is formulated in a local linear format, and approximated by a neural network (NN). First, the NN is trained offline to provide a good initial estimation of the model, which can directly migrate to the online phase. Then, an adaptive controller is proposed to achieve the shape control tasks, in which the NN is further updated online to compensate for any errors in the offline model caused by insufficient training or changes of DLO properties. The simulation and real-world experiments show that the proposed method can precisely and efficiently accomplish the DLO shape control tasks, and adapt well to new and untrained DLOs.
Robotic manipulation of deformable linear objects (DLOs) has broad application prospects in many fields. However, a key issue is to obtain the exact deformation models (i.e., how robot motion affects ...DLO deformation), which are hard to theoretically calculate and vary among different DLOs. Thus, shape control of DLOs is challenging, especially for large deformation control which requires global and more accurate models. In this paper, we propose a coupled offline and online data-driven method for efficiently learning a global deformation model, allowing for both accurate modeling through offline learning and further updating for new DLOs via online adaptation. Specifically, the model approximated by a neural network is first trained offline on random data, then seamlessly migrated to the online phase, and further updated online during actual manipulation. Several strategies are introduced to improve the model's efficiency and generalization ability. We propose a convex-optimization-based controller, and analyze the system's stability using the Lyapunov method. Detailed simulations and real-world experiments demonstrate that our method can efficiently and precisely estimate the deformation model, and achieve large deformation control of untrained DLOs in 2D and 3D dual-arm manipulation tasks better than the existing methods. It accomplishes all 24 tasks with different desired shapes on different DLOs in the real world, using only simulation data for the offline learning.
The deformable linear objects (DLOs) are common in both industrial and domestic applications, such as wires, cables, ropes. Because of its highly deformable nature, it is difficult for the robot to ...reproduce human's dexterous skills on DLOs. In this paper, the unknown deformation model is estimated in both the offline and online manners. The offline learning aims to provide a good approximation prior to the manipulation task, while the online learning aims to compensate the errors due to insufficient training (e.g. limited datasets) in the offline phase. The offline module works by constructing a series of supervised neural networks (NNs), then the online module receives the learning results directly and further updates them with the technique of adaptive NNs. A new adaptive controller is also proposed to allow the robot to perform manipulation tasks concurrently in the online phase. The stability of the closed-loop system and the convergence of task errors are rigorously proved with Lyapunov method. Simulation studies are presented to illustrate the performance of the proposed method.
The lithium–sulfur batteries show the great potential to be the most promising candidate for high energy applications. However, the shuttling of soluble polysulfides deteriorates the battery ...performance tremendously. To suppress the diffusion of soluble polysulfides, diatomite that has abundant natural three-dimensional ordered pores is incorporated into the cathode to trap polysulfides. The composite cathode material(S-DM-AB for short), including sulfur(S), diatomite(DM), and acetylene black(AB) is prepared by an impregnation method. For comparison, another composite cathode material(S-AB for short) including sulfur and acetylene black is also prepared by the same method. The battery with S-DMAB composite cathode material delivers a discharge capacity of 531.4 m Ah/g after 300 cycles at 2 C with a capacity retention of 51.6% at room temperature. By contrast, the battery with S-AB composite cathode material delivered a capacity of only 196.9 m Ah/g with a much lower capacity retention of 18.6% under the same condition. The addition of diatomite in the cathode is proved to be a cheap and effective way to improve the life time of the lithium sulfur batteries.
•Free radicals obviously suppressed ARG conjugative transfer between E. coli strains.•Radicals exposure decreased antioxidant stress and SOS responses in donor strains.•Decreased intercellular ...contact contributed to depressed ARGs conjugative transfer.•Energy driving force for ARGs conjugative transfer decreased by radicals exposure.•Transcriptome data and gene expressions were performed to reveal the mechanisms.
Spread of antibiotic-resistant genes (ARGs) is a global public safety issue and inhibition their transfer is imperative. In this study, a novel strategy using environmental free radical exposure was developed to inhibit conjugative transfer of ARGs (RP4 plasmid) in aqueous solutions. Long-time free radical (·OH, 1O2, and O2·−) exposure significantly suppressed the conjugative transfer frequency of ARGs between Escherichia coli (E. coli) strains, and ·OH was more likely to attack ARG, thereby inhibiting the conjugate transfer frequency, compared to 1O2 and O2·−. Compared with the control, the conjugative transfer frequency significantly decreased from 4.08 × 10−5 to 1.2 × 10−8 after 10 min free radical exposure, confirming that the transfer and proliferation of ARGs were well inhibited. Correspondingly, the number of transconjugant significantly decreased by 61.7% after 10 min free radical exposure. Significant reductions in reactive oxygen species levels (ROS content and enzyme levels) and DNA damage-induced responses in the donor strains were observed after 10 min free radical exposure. Concurrently, intercellular contact was also weakened via inhibiting the synthesis of polysaccharides in extracellular polymeric substances. Moreover, the expressions of plasmid transfer genes were down-regulated after 10 min exposure due to the shortage of adenosine-triphosphate supply. This study firstly disclosed the underneath mechanisms for depressing ARGs transfer and dissemination via environmental free radical exposure.
Display omitted
