Since the pioneering deployment of Canadarm in 1981,various on-orbit servicing (OOS) concepts have been flourishing and are continuously demonstrated by space agencies worldwide[1].This novel access to extend human cu...
Since the pioneering deployment of Canadarm in 1981,various on-orbit servicing (OOS) concepts have been flourishing and are continuously demonstrated by space agencies worldwide[1].This novel access to extend human curiosity in space is empowered by the remarkable OOS progress achieved by missions like the Robot technology Experiment (ROTEX),Engineering Test Satellite-Ⅶ(ETS-Ⅶ),
This paper designs a soft robot with a multi-chamber,multi-airbag mimicking soft biological structure,where the airbags of the same chamber are interconnected with each *** upper and lower chambers are separated by an...
详细信息
This paper designs a soft robot with a multi-chamber,multi-airbag mimicking soft biological structure,where the airbags of the same chamber are interconnected with each *** upper and lower chambers are separated by an intermediate layer(thin plate),which is extended and widened to achieve robot movement and *** applying pressure to the different chambers of the soft robot,it is possible to produce a variety of bionic movements of the inchworm and *** to the strong nonlinearity and infinite number of degrees of freedom properties of the material,it is impossible to obtain the analytical solution of the bending morphology and pressure of the soft robot ***,a method to establish a mathematical model of soft robot deformation based on the classical stacked plate theory is proposed,and a chain composite model of soft robot bending motion is established based on the large-deflection modeling *** paper proposes a method to generate a multi-mode soft robot motion control based on the Central Pattern Generator(CPG)using a single controller,which achieves the switching of sine wave-like patterns,half-wave-like patterns,and dragging patterns by adjust-ing frequency,amplitude and period of ***,a pneumatic control platform is built for the validation of the theoretical model and different experimental models of the motion of the *** comparation of the different motion modes of the soft robot under similar non-load and load conditions.
Photoacoustic detection has shown excellent performance in measuring thickness and detecting defects in metal ***,existing research on ultrafast lasers mainly focuses on using picosecond or nanosecond lasers for large...
详细信息
Photoacoustic detection has shown excellent performance in measuring thickness and detecting defects in metal ***,existing research on ultrafast lasers mainly focuses on using picosecond or nanosecond lasers for large-scale material processing and *** theoretical study of femtosecond laser sources for photoacoustic nondestructive testing(NDT)in nanoscale thin film materials receives much less emphasis,leading to a lack of a complete physical model that covers the entire process from excitation to *** this study,we developed a comprehensive physical model that combines the two-temperature model with the acoustic wave generation and detection *** the basis of the physical model,we established a simulation model to visualize the ultrafast lasermaterial interaction *** damage threshold of the laser source is determined,and the effect of key parameters(laser fluence,pulse duration,and wavelength)for AlCu nanofilms on the femtosecond photoacoustic NDT process is discussed using numerical results from the finite element *** numerical results under certain parameters show good agreement with the experimental results.
Hierarchical multi-granularity image classification is a challenging task that aims to tag each given image with multiple granularity labels *** methods tend to overlook that different image regions contribute differe...
详细信息
Hierarchical multi-granularity image classification is a challenging task that aims to tag each given image with multiple granularity labels *** methods tend to overlook that different image regions contribute differently to label prediction at different granularities,and also insufficiently consider relationships between the hierarchical multi-granularity *** introduce a sequence-to-sequence mechanism to overcome these two problems and propose a multi-granularity sequence generation(MGSG)approach for the hierarchical multi-granularity image classification ***,we introduce a transformer architecture to encode the image into visual representation ***,we traverse the taxonomic tree and organize the multi-granularity labels into sequences,and vectorize them and add positional *** proposed multi-granularity sequence generation method builds a decoder that takes visual representation sequences and semantic label embedding as inputs,and outputs the predicted multi-granularity label *** decoder models dependencies and correlations between multi-granularity labels through a masked multi-head self-attention mechanism,and relates visual information to the semantic label information through a crossmodality attention *** this way,the proposed method preserves the relationships between labels at different granularity levels and takes into account the influence of different image regions on labels with different *** on six public benchmarks qualitatively and quantitatively demonstrate the advantages of the proposed *** project is available at https://***/liuxindazz/mgs.
Reinforcement learning algorithms are central to the cognition and decision-making of embodied intelligent agents. A bilevel optimization(BO) modeling approach, along with a host of efficient BO algorithms, has been p...
详细信息
Reinforcement learning algorithms are central to the cognition and decision-making of embodied intelligent agents. A bilevel optimization(BO) modeling approach, along with a host of efficient BO algorithms, has been proven to be an effective means of addressing actor-critic(AC) policy optimization problems. In this work, based on a bilevelstructured AC problem model, an implicit zeroth-order stochastic algorithm is developed. A locally randomized spherical smoothing technique, which can be applied to nonsmooth nonconvex implicit AC formulations and avoid the closed-form lower-level mapping, is introduced. In the proposed zeroth-order scheme, the gradient of the implicit function can be approximated through inexact lower-level value estimations that are practically available. Under suitable assumptions,the algorithmic framework designed for the bilevel AC method is characterized by convergence guarantees under a fixed stepsize and smoothing parameter. Moreover, the proposed algorithm is equipped with the overall iteration complexity of O(n2L20 L20?-1). The convergence performance of the proposed algorithm is verified through numerical simulations.
Robustness walking and recovery ability in uneven terrains and unexpected collisions are crucial for the practical application of humanoid ***, existing methods struggle to effectively balance stability, motion safety...
详细信息
Dear Editor,This letter investigates the level curve tracking problem of unknown scalar fields using an unmanned aerial vehicle(UAV)and presents a robust reinforcement learning(RL)-guided model predictive control(MPC)...
详细信息
Dear Editor,This letter investigates the level curve tracking problem of unknown scalar fields using an unmanned aerial vehicle(UAV)and presents a robust reinforcement learning(RL)-guided model predictive control(MPC)scheme for the ***,we formulate the MPC trajectory tracking problem,wherein an RL-based trajectory planning algorithm provides the reference trajectory to guide the UAV towards the desired concentration.
When humanoid robots attempt to walk on terrain such as shaking platforms,time-varying disturbances are introduced to the support *** abrupt changes of inclination angle can cause the robot to lose balance upon landin...
详细信息
When humanoid robots attempt to walk on terrain such as shaking platforms,time-varying disturbances are introduced to the support *** abrupt changes of inclination angle can cause the robot to lose balance upon landing,presenting significant challenges for balance control *** address this issue,we propose a novel divergent component of motion(DCM)-based time-varying disturbance walking(DCM-TVDW)*** method allows the robot to walk on rugged surfaces and helps to maintain dynamic balance when subjected to large time-varying *** the DCM-TVDW control method,we first adjust the robot's center of mass and stride height to adapt to transitions between different terrain types via a variable height stabilization method,and hold these quantities constant as base *** then combine DCM with the N-step capturability *** combination allows for dynamic balance through multi-step adjustments from the initially unstable region,thereby extending the robots stability *** and experimental results demonstrate that the DCM-TVDW method enables the SJ-Bruce robot to traverse a dynamically shaking platform with an inclination angle of approximately 22°.
Versatile jumping is crucial for agile and adaptive behavior of humanoid robots in complex environments. However, humanoid robots have to manage impact forces while maintaining balance and stability throughout the tak...
详细信息
Discrete event system(DES)models promote system engineering,including system design,verification,and *** advancement in manufacturing technology has endowed us to fabricate complex industrial ***,the adoption of advan...
详细信息
Discrete event system(DES)models promote system engineering,including system design,verification,and *** advancement in manufacturing technology has endowed us to fabricate complex industrial ***,the adoption of advanced modeling methodologies adept at handling complexity and scalability is ***,industrial systems are no longer quiescent,thus the intelligent operations of the systems should be dynamically specified in the *** this paper,the composition of the subsystem behaviors is studied to generate the complexity and scalability of the global system model,and a Boolean semantic specifying algorithm is proposed for generating dynamic intelligent operations in the *** traditional modeling approaches,the change or addition of specifications always necessitates the complete resubmission of the system model,a resource-consuming and error-prone *** with traditional approaches,our approach has three remarkable advantages:(i)an established Boolean semantic can be fitful for all kinds of systems;(ii)there is no need to resubmit the system model whenever there is a change or addition of the operations;(iii)multiple specifying tasks can be easily achieved by continuously adding a new ***,this general modeling approach has wide potential for future complex and intelligent industrial systems.
暂无评论