Search

Micro air vehicles’ movements can be significantly affected by physical contact with unknown moving objects from a random direction. To address this problem, low-latency sensors (cameras or radars) and high-speed processors installed in six directions of aerial vehicles. However, this method is limited by hardware costs, load and energy. In this paper, a fast recovery strategy for micro air vehicles is proposed, which is inspired by the recovery manoeuvre of mosquitoes hit by raindrops. When hit by falling raindrops, mosquitoes will descent along the falling direction of raindrops to gradually offset the kinetic energy of the impact induced by the raindrops. During the descent, mosquitoes will gradually tilt to one side but move in the opposite direction, and finally deviate from the falling route of raindrops. To simulate the recovery strategy, a micro air robot called RobMos is designed. The RobMos is an overactuated system that can decouple the position control from the attitude control. Using the recovery strategy, the robot can quickly recover to a stable attitude after being impacted by unknown movingobjects. The onboard inertial measurement unit (IMU) is used to estimate the impact location. Finally, the prototype can bear the impact of objects two times heavier than its own weight. Moreover, the generalizability of this strategy is also discussed, and it is proved that the RobMos can even recover attitude after being hit by the unknown moving object in a lateral direction.

Differentially private stochastic gradient descent (DP-SGD) that perturbs the clipped gradients is a popular approach for private machine learning. Gaussian mechanism GM, combined with the moments accountant (MA), has demonstrated a much better privacy-utility tradeoff than using the advanced composition theorem. However, it is unclear whether the tradeoff can be further improved by other mechanisms with different noise distributions. To this end, we extend GM ( $p=2$ ) to the generalized $p$ -power exponential mechanism ( $p$ EM with $p>0$ ) family and show its privacy guarantee. Straightforwardly, we can enhance the privacy-utility tradeoff of GM by searching noise distribution in the wider mechanism space. To implement $p$ EM in practice, we design an effective sampling method and extend MA to $p$ EM for tightly estimating privacy loss. Besides, we formally prove the non-optimality of GM based on the variation method. Numerical experiments validate the properties of $p$ EM and illustrate a comprehensive comparison between $p$ EM and the other two state-of-the-art methods. Experimental results show that $p$ EM is preferred when the noise variance is relatively small to the signal and the dimension is not too high.

How to enable robotic compliant manipulation has become a critical problem in the robotics field. Inspired by a biomimetic adaptive control strategy, this work presents a novel representation model named human-like compliant movement primitives (Hl-CMPs) which could allow a robot to learn human-like compliant behaviours. The state-of-the-art approaches can hardly learn complete compliant profiles for a specific task. Comparatively, our model can encode task-specific parametric movement trajectories, correspondingly associated with dynamic trajectories including both impedance and feedforward force profiles. The compliant profiles are learned based on a biomimetic control strategy derived from the human motor learning in the muscle space, enabling the robot to simultaneously learn the impedance and the force while executing the movement trajectories obtained from human demonstration. Furthermore, both the kinematic and the dynamic profiles are learned in the parametric space, thus enabling the representation of a skill using corresponding parameters (i.e, task-specific parameters).

This paper presents a thorough evaluation of twenty-one state-of-the-art widely-used crop segmentation algorithms, motived by their significance in vision tasks for further analysis. An ideal crop segmentation algorithm can effectively extract crop information, thus providing an important precondition for the application of intelligent agriculture analytics. In order to enable researchers in this field to fully understand various crop segmentation methods, this paper proposes a new classification strategy of object segmentation by dividing the algorithms into pixel-based and region-based approaches at first, and then systematically evaluating various crop segmentation methods with a unified data benchmark and four common metrics. A new dataset which incorporates crop variety, environment condition and observation distance into consideration is constructed for demonstrating the experiments and comparisons. The effectiveness and robustness of these algorithms were evaluated by three sets of comparative experiments. Based on the quantitative results, we summarize the advantages and disadvantages of the evaluated algorithms from the segmentation performances with four metric indicators. Furthermore, the discussion and evaluation results will provide great support for precision agriculture analysis.

In this work, a novel process monitoring method in a block-wised partitioning manor is proposed for plant-wide processes which can be partitioned into several sub-blocks and monitored parallelly. The focus of this method is to reduce the high complexity of global plant-wide process, while to improve the efficiency of local feature extraction. In this method, considering that not all process knowledge is available in the block division process strategy, a novel community discovery (CD) algorithms, based on the similarity of neighbor node weighted Louvain, is introduced into the framework of the multi-block Bayesian inference and principal component analysis (PCA) based plant-wide process monitoring scheme. Firstly, the complex network (CN) theory is used to establish the network topology structure for the global variables of the plant-wide process. Secondly, by analyzing the graph characteristic structure, considering the connection strength between nodes, a more reasonable sub-block division is conducted according to the improved Louvain algorithm. Then, PCA method is used to establish process monitoring model for each sub-block to obtain sub-block monitoring statistics. Finally, the total joint statistics is obtained through Bayesian inference for fault detection. The feasibility and effectiveness, in terms of the detection performance, of this method are demonstrated in a simulated plant-wide process by compared with other state-of-the-art PCA based monitoring methods.

In a repetitive human-robot collaboration (HRC) task, robots typically are required to learn the intended motion of the human user to improve the collaboration efficiency. However, the human user's trajectory is of uncertainty when repeating the same task (e.g., human hand tremor and uncertain movement speed), which may directly deteriorate the learning performance. To address this issue, a path characterized by spatial correlation parameters, is of necessity to be learned by robots so that the aforementioned time-related uncertainty will be avoided. In this article, based on the path parameterization, a gradient-based iterative path learning (IPL) strategy is designed for the robot to learn the desired path of human. The proposed IPL strategy draws on the iterative learning methods with a properly designed performance index. Since the gradient of the performance index is hard to directly obtain in HRC, two learning methods with gradient search (GS) and gradient estimation (GE) are developed. The GS estimates the gradient online and has an advantage of easy implementation. By contrast, the advantage of GS is more obvious when the number of learned parameters is considerable due to its high learning efficiency. With these two methods, the unknown path parameters can be iteratively updated toward the desired values. To verify the effectiveness of the proposed IPL algorithm, experiments are carried out. In the experiment, a comparison between GS and GE methods is made to display their respective advantages. Besides, the proposed IPL is compared with an existing trajectory learning method subject to two different kinds of uncertainties and its better learning performance verifies its stability and capability in dealing with uncertainty.

In this paper, reflection, absorption and transmission characteristics of 1-50GHz frequency electromagnetic wave by the sub-dense heterogeneous plasma are studied by combining theoretical modeling and simulation with ground experimental measurement.

A swarm of robots is a typical self-organizing system. Parameters of self-organizing model can regulate the interaction among the individuals to achieve the desired collective motion. Thus, it makes sense to optimize the parameters for better performance in self-organization. The evolutionary algorithm (EA) is one promising heuristic optimization algorithm. However, quantitative evaluation of a self-organizing model is still an open scientific question. In this paper, we proposed three metrics for collective motion from the perspective of grouping, polarization and spatial distribution of group. Then, we designed the fitness function with the proposed metrics to bridge the gap between EA and parameter tuning of self-organizing model. In the experimental study, we optimized two kinds of self-organizing models using the differential evolution algorithm (DE) and validated the effectiveness in simulation. Moreover, we validated the optimized parameters on a swarm of physical robots. The experimental results show that the proposed metrics are effective to quantitatively evaluate the self-organizing model so that the DE performs well for the automatic parameter tuning of controlling swarm robots.

Industrial automation requires robot dexterity to automate many processes such as product assembling, packaging, and material handling. The existing robotic systems lack the capability to determining proper grasp strategies in the context of object affordances and task designations. In this paper, a framework of task-oriented dexterous grasping is proposed to learn grasp knowledge from human experience and to deploy the grasp strategies while adapting to grasp context. Grasp topology is defined and grasp strategies are learned from an established dataset for task-oriented dexterous manipulation. To adapt to various grasp context, a reinforcement-learning based grasping policy was implemented to deploy different task-oriented strategies. The performance of the system was evaluated in a simulated grasping environment by using an AR10 anthropomorphic hand installed in a Sawyer robotic arm. The proposed framework achieved a hit rate of 100% for grasp strategies and an overall top-3 match rate of 95.6%. The success rate of grasping was 85.6% during 2700 grasping experiments for manipulation tasks given in natural-language instructions.

This paper presents a novel method for learning and tracking of the desired path of the human partner in physical human-robot interaction. Combining the Adam optimization algorithm with iteration learning control (ILC), a path learning method is designed to generate and update reference waypoints according to the human partner’s desired path. This method firstly uses the Adam optimization algorithm to update the robot’s reference waypoints in an online manner. Then, an ILC is developed to further modify the waypoints and reduce the difference between the robot’s actual path and the human partner’s desired path in an iterative manner. Simulations and experiments on a 7-DOF Sawyer robot are carried out to show the effectiveness of our proposed method.

With the development of 5G millimeter wave and related technologies, working frequency of transmission lines becomes higher and higher, and the high-frequency electromagnetic radiation is a key factor affecting the working of integrated circuits. Therefore, the method to suppress electromagnetic radiation of high-frequency transmission line has become a research hotspot. Two shielding methods are proposed for a flexible segmented coplanar waveguide (SCPW) transmission line at 5G FR2 band (24.25-52.6 GHz) in this paper, adding copper film or adding copper film and metal vias. Through simulation and optimization of CST, results show that adding copper film can effectively suppress electromagnetic radiation of the SCPW transmission line in available frequency band, the shielding effectiveness of one-unit transmission line can reach 10-25 dB, while that of three-unit transmission line is 35-45 dB. On this basis, punching copper vias can further enhance the electromagnetic radiation suppression capability, the shielding effectiveness of one-unit transmission line can reach 50-60 dB, while that of three-unit transmission line is 50-65 dB. However, copper film will weaken the transmission performance, and metal vias will further affect that of SCPW transmission line. In addition, considering that metal vias will make the processing of transmission lines more difficult, if the shielding effectiveness of transmission lines shielded by copper film meets the requirements, it is not recommended to punch metal vias.

This paper proposes an approach to improve robot assisted physical training subject to human uncertainties. This approach is based on impedance control which is used to regulate the dynamic relationship between the robot’s position and contact force. Repetitive exercise is considered and impedance parameters are adapted in accordance with the human user to provide physical training as needed. Different from the existing approaches, the proposed one has the capacity to deal with time-length-varying cycles, which is a critical issue in physical training of human’s upper limbs. By theoretical analysis and experimental results, we show that the approach can effectively learn the required robot’s impedance parameters and improve the performance of physical training.

Transferring human stiffness regulation strategies to robots enables them to effectively and efficiently acquire adaptive impedance control policies to deal with uncertainties during the accomplishment of physical contact tasks in an unstructured environment. In this work, we develop such a physical human-robot interaction (pHRI) system which allows robots to learn variable impedance skills from human demonstrations. Specifically, the biological signals, i.e., surface electromyography (sEMG) are utilized for the extraction of human arm stiffness features during the task demonstration. The estimated human arm stiffness is then mapped into a robot impedance controller. The dynamics of both movement and stiffness are simultaneously modeled by using a model combining the hidden semi-Markov model (HSMM) and the Gaussian mixture regression (GMR). More importantly, the correlation between the movement information and the stiffness information is encoded in a systematic manner. This approach enables capturing uncertainties over time and space and allows the robot to satisfy both position and stiffness requirements in a task with modulation of the impedance controller. The experimental study validated the proposed approach.

As a typical cross-modal problem, visual question answering (VQA) has received increasing attention from the communities of computer vision and natural language processing. Reading and reasoning about texts and visual contents in the images is a burgeoning and important research topic in VQA, especially for the visually impaired assistance applications. Given an image, it aims to predict an answer to a provided natural language question closely related to its textual contents. In this paper, we propose a novel end-to-end textual content based VQA model, which grounds question answering both on the visual and textual information. After encoding the image, question and recognized text words, it uses multi-modal factorized high-order modules and the attention mechanism to fuse question-image and question-text features respectively. The complex correlations among different features can be captured efficiently. To ensure the model's extendibility, it embeds candidate answers and recognized texts in a semantic embedding space and adopts semantic embedding based classifier to perform answer prediction. Extensive experiments on the newly proposed benchmark TextVQA demonstrate that the proposed model can achieve promising results.

MicroRNAs (miRNAs) are small non-coding RNAs that achieve post-transcriptional regulation of RNA silencing and gene expression by targeting messenger RNAs (mRNAs). Rapid and effective detection of miRNAs target sites is a significantly important topic in bioinformatics. In this study, a deep learning approach based on fusion of stacked denoising autoencoders (SDA) and Convolutional denoising autoencoders (CAE) is developed for sequence and structure data respectively with the help of an existing duplex sequence model. Compared with four conventional machine learning methods, the proposed fusion model performs better in terms of the accuracy, precision, recall, AUC (Area under the curve) and Fl-score. A web system is also developed to identify and display the microRNA target sites effectively and Rapidly.

Fractal flexible interconnect structure is an important aspect of flexible printed circuit and wearable electronics. However, with the scale reduction of flexible chip and the increase of working frequency for 5G communication technology, high frequency transmission characteristics of interconnects under mechanical deformation become an important influence factor to flexible devices. In this paper, ⊔-shape fractal structure is adopted for interconnect lines. Fractal structures with different fractal orders, and with different unit numbers at same length are simulated respectively. Then, electromagnetic characteristics of fractal flexible interconnects under bending and twisting are analyzed from 0.01GHz to 12GHz. Different fractal flexible interconnects are manufactured on Polyimide. Comparing with the straight states, variety of transmission lines in bending and twisting states are tested and analyzed.

In order to acquire a time-optimal path in the presence of time-invariant but position-dependent currents, this paper develops a 3D path planning algorithm for AUV traversing from a starting location to the target location using optimal control theory. First, we consider the kinematics of the AUV for the design, which describes the relationship between AUV’s velocity in 3D space and the motion speed and its yaw and pitch angles. Then, under the constraints that the motion speed, initial and final positions are given, we obtain an optimal control law for the navigation yaw and pitch angles by using Pontryagin’s minimum principle of variational analysis in optimal control theory. Finally, by using the state-space theory of general linear time-invariant dynamical systems, the initial yaw and pitch angles are calculated, after which the optimal transformation time is also obtained. Compared with the existing iterative methods, such as graph search and evolutionary algorithms, this method generates a continuous time-optimal path underthe optimal navigation control law which is more computationally fast. What’s more, considering the position-dependent currents, the path planning of AUV in complex environment is realized. It is verified by simulation that this method can obtain the time-optimal path in the 3D plane in the presence of ocean currents.