Shenyang Institute of Automation,Chinese Academy Of Sciences
Not a member yet
23582 research outputs found
Sort by
A positioning method for maize seed laser-cutting slice using linear discriminant analysis based on isometric distance measurement
No full textIt is necessary that vision system should aid laser-cutting manipulator to position the specified part of each maize seed for getting the slice breeding genotype analysis with high throughput. Each of trivial maize seeds should be recognized and positioned in a certain posture. Correlation area ratio (CAR) is defined as the metric of pixel attribute. A large template of round mask is adopted for seed morphological detection to measure the CAR values. We get the feature points extracted from the seed image through the isometric mapping operation. Iterative processes of linear discriminant analysis search the morphological data space to learn non-linear transformations to the space where data are linearly separable. Linear discriminant analysis utilizes the data directional distribution to position the major axis and distinguish different parts of maize seed. The labeling partition operation is applied for picking out the scattered pieces to be finely clustered. Without denoising process, the feature region could be recognized with accuracies by the synthetical methods. Extensive experiments on a large amount of seeds demonstrate the effectiveness of proposed methods. © 2021 China Agricultural University</p
Lifelong robotic visual-tactile perception learning
No full textLifelong machine learning can learn a sequence of consecutive robotic perception tasks via transferring previous experiences. However, 1) most existing lifelong learning based perception methods only take advantage of visual information for robotic tasks, while neglecting another important tactile sensing modality to capture discriminative material properties; 2) Meanwhile, they cannot explore the intrinsic relationships across different modalities and the common characterization among different tasks of each modality, due to the distinct divergence between heterogeneous feature distributions. To address above challenges, we propose a new Lifelong Visual-Tactile Learning (LVTL) model for continuous robotic visual-tactile perception tasks, which fully explores the latent correlations in both intra-modality and cross-modality aspects. Specifically, a modality-specific knowledge library is developed for each modality to explore common intra-modality representations across different tasks, while narrowing intra-modality mapping divergence between semantic and feature spaces via an auto-encoder mechanism. Moreover, a sparse constraint based modality-invariant space is constructed to capture underlying cross-modality correlations and identify the contributions of each modality for new coming visual-tactile tasks. We further propose a modality consistency regularizer to efficiently align the heterogeneous visual and tactile samples, which ensures the semantic consistency between different modality-specific knowledge libraries. After deriving an efficient model optimization strategy, we conduct extensive experiments on several representative datasets to demonstrate the superiority of our LVTL model. Evaluation experiments show that our proposed model significantly outperforms existing state-of-the-art methods with about 1.16%∼15.36% improvement under different lifelong visual-tactile perception scenarios.</p
SIASAIL-I solar sail: From system design to on-orbit demonstration mission
No full textSolar sails, characterized by passive propulsion, have attracted an increasing attention of researchers, which has important value in the field of deep space exploration. SIASAIL-I is an on-orbit verification mission for solar sails based on a 6U cubesat, designed and produced by Shenyang Institute of Automation, Chinese Academy of Sciences (SIA, CAS). SIASAIL-I deployment mechanism is composed by two-stage deployment system, which can achieve direction transformation, stretching and membrane deploying in the narrow space of the cubesat. SIASAIL-I only takes up within 0.5U (48 mm × 90 mm × 95 mm) space and the total mass is within 1kg. SIASAIL-I was successfully launched from the Xiaoxiang-1 07 satellite on August 31, 2019, and successfully deployed a 0.78 m × 0.78 m sail in low earth orbit in December. This paper introduces the design process of SIASAIL-I system, including deployment mechanism, a series of ground experiments, and an de-orbit mission analysis, which aims to provide some engineering experience for the development of solar sails.</p
Monocular 3d object detection based on uncertainty prediction of keypoints
No full textThree-dimensional (3D) object detection is an important task in the field of machine vision, in which the detection of 3D objects using monocular vision is even more challenging. We observe that most of the existing monocular methods focus on the design of the feature extraction framework or embedded geometric constraints, but ignore the possible errors in the intermediate process of the detection pipeline. These errors may be further amplified in the subsequent processes. After exploring the existing detection framework of keypoints, we find that the accuracy of keypoints prediction will seriously affect the solution of 3D object position. Therefore, we propose a novel keypoints uncertainty prediction network (KUP-Net) for monocular 3D object detection. In this work, we design an uncertainty prediction module to characterize the uncertainty that exists in keypoint prediction. Then, the uncertainty is used for joint optimization with object position. In addition, we adopt position-encoding to assist the uncertainty prediction, and use a timing coefficient to optimize the learning process. The experiments on our detector are conducted on the KI TTI benchmark. For the two levels of easy and moderate, we achieve accuracy of 17.26 and 11.78 in AP3D, and achieve accuracy of 23.59 and 16.63 in APBEV, which are higher than the latest method KM3D.</p
Characterization of interconnectivity of gelatin methacrylate hydrogels using photoacoustic imaging
No full textHydrogels can provide a three-dimensional microenvironment for cells and thus serve as an extracellular matrix in a biofabrication process. The properties of hydrogels, such as their porosity and mechanical properties, significantly influence the cell growth. However, there is still a lack of effective methods for characterizing the hydrogel structure noninvasively. Herein, a photoacoustic (PA) imaging-based method is proposed for the characterization of gelatin methacrylate (GelMA) hydrogels. Owing to their high PA contrast, red blood cells (RBCs) are included as mediators in the GelMA hydrogel to analyze its pore distribution. The interconnectivity of the pores is further analyzed through the lysis of RBCs. The diffusion of the RBC lysis buffer in the GelMA is consistent with the trend observed in simulations. The analyzed vitality of HEK293 cells in different GelMA hydrogels reveals that understanding the diffusion of solutes (i.e., nutrients) is a potential strategy to optimize the hydrogel parameters during biofabrication.</p
Event-triggered finite-time control for a class of switched nonlinear systems
No full textThis work addresses the event-triggered finite-time control problem for a class of uncertain switched nonlinear systems with arbitrary switchings, whose powers are positive odd rational numbers. The key difference from the results of similar problems is that the systems considered in this article are more general, which contains a special case when the powers are equal to 1, and the powers have switching signals. It is well known that such nonlinear systems have challenges because of the uncontrollability in the linearization process and the backstepping technique that successfully developed for low-order systems fail to work. To tackle this issue, combining the backstepping method, event-triggered strategy with adding one power integrator technique, an event-triggered control scheme is developed to make the controlled systems be globally finite-time stable. Besides, the Zeno-free behavior is proved to verify the feasibility of the proposed event-triggered mechanism. Finally, simulation results are given to validate the effectiveness of the developed control strategy
Theoretical investigation and implementation of nonlinear material removal depth strategy for robot automatic grinding aviation blade
No full textDue to the requirements of manufacturing accuracy and surface quality consistency, which will influence the dynamic properties and service cycle of aircraft engines, the high-precision material removal depth model is urgently needed for robot automatic grinding aviation blade system. This research developed a novel material removal depth (MRD) model to ensure quantitative grinding depth point by point for robot automatic grinding aviation blade. Firstly, the relationship between contact stress and contact force is developed based on simulation and theoretical derivation, and the contact contour is further investigated by the experiments. Then, the nonlinear material removal depth model is established to predict grinding depth according to Preston equation. Meanwhile, multiple linear regression method is introduced to analyze parameters of material removal depth model. The previous researches show that grinding parameters have strong correlation with MRD, especially for the non-negligible contact force consideration. Therefore, the non-uniform material removal with variable contact force strategy is developed according to the proposed model, and the proposed strategy is applied to the robot automatic grinding aviation blade system to achieve the quantitative grinding depth point by point on the aviation blade surface. The practice experiments further prove that the developed strategy is feasible and effective. Comparing the predicted and experimental MRD, the maximum errors and average relative errors are respectively 9.21% and 4.68%, when the aviation blade is ground by the proposed material removal strategy. Simultaneously, the experimental results tend to produce much better surface roughness with uniform surface texture which satisfies grinding requirements
Swarm of AUVs fission control algorithm combining topological interaction and information coupling degree
No full text在无协商、指派及中心控制的自主水下航行器(AUV)集群控制中,AUV群集在进行避碰、多目标追踪等行为时群集分裂的现象不可避免,如果分裂不可控,易出现AUV丢失或分裂后的子群规模差距过大等问题。为了面对外部刺激时群集能进行可控分群,提出一种融合拓扑交互和信息耦合度的分群控制算法,该算法使用信息耦合度作为评判指标衡量个体间的影响作用,并根据AUV在水下通信难、感知难的特点,引入了拓扑交互机制,减少群集组群和分群所需要的信息,提升AUV组群和分群的效果。仿真实验验证该算法的可行性和有效性。</p
A humanoid robotic hand capable of internal assembly and measurement in spacesuit gloves
No full textPurpose To identify the dexterity of spacesuit gloves, they need to undergo bending tests in the development process. The ideal way is to place a humanoid robotic hand into the spacesuit glove, mimicking the motions of a human hand and measuring the bending angle/force of the spacesuit glove. However, traditional robotic hands are too large to enter the narrow inner space of the spacesuit glove and perform measurements. This paper aims to design a humanoid robot hand that can wear spacesuit gloves and perform measurements. Design/methodology/approach The proposed humanoid robotic hand is composed of five modular fingers and a parallel wrist driven by electrical linear motors. The fingers and wrist can be delivered into the spacesuit glove separately and then assembled inside. A mathematical model of the robotic hand is formulated by using the geometric constraints and principle of virtual work to analyze the kinematics and statics of the robotic hand. This model allows for estimating the bending angle and output force/torque of the robotic hand through the displacement and force of the linear motors. Findings A prototype of the robotic hand, as well as its testing benches, was constructed to validate the presented methods. The experimental results show that the whole robotic hand can be transported to and assembled in a spacesuit glove to measure the motion characteristics of the glove. Originality/value The proposed humanoid robotic hand provides a new method for wearing and measuring the spacesuit glove. It can also be used to other gloves for special protective suits that have highly restricted internal space.</p
Engineering Biological Tissues from the Bottom-Up: Recent Advances and Future Prospects
No full textTissue engineering provides a powerful solution for current organ shortages, and researchers have cultured blood vessels, heart tissues, and bone tissues in vitro. However, traditional top-down tissue engineering has suffered two challenges: vascularization and reconfigurability of functional units. With the continuous development of micro-nano technology and biomaterial technology, bottom-up tissue engineering as a promising approach for organ and tissue modular reconstruction has gradually developed. In this article, relevant advances in living blocks fabrication and assembly techniques for creation of higher-order bioarchitectures are described. After a critical overview of this technology, a discussion of practical challenges is provided, and future development prospects are proposed