Traditional myoelectric prostheses that employ a static pattern recognition model to identify human movement intention from surface electromyography (sEMG) signals hardly adapt the changes in sEMG characteristics caused by interferences daily activities, which hinders clinical applications of such prostheses. In this paper, we focus on methods reduce or eliminate impacts three types (MPR), i.e., outlier motion, muscle fatigue, and electrode doffing/donning. We constructed an adaptive...

To generate local trajectory between initial states and target for autonomous vehicles, a feasible generation algorithm based on quartic Bézier curve is proposed. The problem of firstly separated into generating continuous bounded curvature profile to shape the linear velocity execute trajectory. further converted an optimization with only 3 parameters owing specific properties curve. Sequential quadratic programming employed find optimal solution respect objective function. avoid sideslip...

Abstract Robotics have important applications in the field of disaster medical rescue. The deployment urban rescue robots at earthquake site can help shorten response time, improve efficiency and keep personnel away from danger. This discussion introduces performance some actual scenarios, focuses on current research status that provide assistance, analyzes merits shortcomings each system. Based existing studies, limitations development directions are also discussed.

Abstract Global path planning of mobile robot aims to provide a safe and smooth for navigation. Traditional A * algorithm is the more turns, not smooth. Moreover, u-shaped terrain, stick obstacles. Aiming at shortcomings A* algorithm, cosine distance selected as heuristic function, direction information added normalized. The 36-order neighborhood search matrix solve problem fitting u-bend. In addition, post-processing method based on Bessel curve proposed. Make planned curvature change...

Introduction Shape memory alloy (SMA) actuators are attractive options for robotic applications due to their salient features. So far, achieving precise control of SMA and applying them human-robot interaction scenarios remains a challenge. Methods This paper proposes novel approach deal with the problem actuator. Departing from conventional mechanism models, we attempt describe this nonlinear plant using gray-box model, in which only input current output displacement measured. The scheme...

Surface electromyography (sEMG), a human-machine interface for gesture recognition, has shown promising potential decoding motor intentions, but variety of nonideal factors restrict its practical application in assistive robots. In this paper, we summarized the current mainstream recognition strategies and proposed method based on multimodal canonical correlation analysis feature fusion classification (MCAFC) condition that occurs daily life, i.e., posture variations. The deep features sEMG...

Autonomous decision-making is a hallmark of intelligent mobile robots and an essential element autonomous navigation. The challenge to enable complete navigation tasks in environments with mapless or low-precision maps, relying solely on sensors. To address this, we have proposed innovative algorithm called PEEMEF-DARC. This consists three parts: Double Actors Regularized Critics (DARC), priority-based excellence experience data collection mechanism, multi-source fusion strategy mechanism....

High-quality and complete human motion 4D reconstruction is of great significance for immersive VR even operation. However, it has inevitable self-scanning constraints, tracking under monocular settings also strict restrictions. In this paper, we propose a capture system combined with priors performance that only uses single RGB-D sensor. To break the constraint, generated mesh using front view input to initialize geometric capture. order construct correct warping field, most previous...

Abstract To overcome the shortcomings of whale optimization algorithm (WOA) such as slow convergence rate, and low accuracy, being easy to fall into local optimum, an improved based on tent chaotic mapping nonlinear factor (TNWOA)is proposed. Firstly, In this algorithm, mapping, which enhances diversity initialization population, is introduced therefore, search space can be searched more thoroughly; Secondly, trigonometric function beta distribution are in ‘a’, balance global ability well...

This work addresses a method to map the estimated 2D pose from color image 3D and obtain more accurate keypoints. In this paper, RGBD camera is used simultaneously acquire images depth images. AlphaPose estimate keypoints of target, are mapped corresponding maps keypoint. We adopt parametric human body model, SMPL, as prior body. The SMPL model data point cloud aligned by optimizing global orientation translation well error vertices experimental results show that our can effectively improve...

The control strategies of reconfiguration and obstacles negotiation are the intensive topic in field leg-wheel hybrid robot. Based on analysis structure characters robot disadvantage early method, this paper proposes a new method rough terrain an auto-control to get over channels, ridges etc. These experimented six-leg-wheel which is designed by Shenyang Institute Automation, results show that these can be effective used terrain.

Reliable mapping and hazard detection are prerequisites for autonomous navigation unmanned ground vehicles. Because of the uncertainty vibration induced by high-speed rugged terrain, problem off-road has not been completely solved yet. A relative probabilistic (RPM) algorithm is introduced to address problem. Firstly, map updated Kalman filter Gaussian Mixture based on exteroceptive measurements model. Then, terrain traversability evaluated identify obstacles in map. Experiments vehicle,...

Online loop-closure detection serves as an essential task for Unmanned Aerial Vehicles (UAVs) equipped with laser scanners. Due to the inherent errors in UAVs' pose estimation, a 3-D reconstruction algorithm is adopted perform map building, which establishes probabilistic models of system according assumption errors. To meet demand online using sequential 2-D data, robust ISW-NDT (incremental sliding-window-based NDT) approach proposed, compares appearance similarity between two scans by...

When disasters such as earthquakes and collapses occur, rescuers must go deep into dangerous regions rush to transport the wounded. Long-term rescue operations in environments challenge rescuers' physical strength threaten their lives. We propose a human-body reconstruction method for autonomous with wounded robot. This constructs human body model that can support robot's given limited scans of injured. Most casualties lie on ground supine positions, so this scanning limit is typical....

Embedded systems have stepped deeper into Integrated Avionics systems, and security is becoming an important concern. Most embedded present a number of software vulnerabilities, such as buffer overflows. Furthermore, the rapid growth pervasive use makes it easier for sophisticated attacker to gain physical access launch attacks on insecure off-chip main memory. This paper presents novel compiler/hardware assisted application code data protection architecture (CHCDP) monitor execution...

The Arctic and Antarctic, as the harshest environments on earth, are of great importance to nation. extreme can be harmful, even fatal, human beings mobile robots. To execute missions Antarctic in place human, State Key Laboratory Robotics Shenyang Institute Automation, Chinese Academy Sciences has designed three generations rovers. hardware software design implementation third generation rover, which been tested during 28th National Research Expedition 2011, will presented this paper....

When the training data is inadequate, it difficult to train a deep Convolutional Neural Network (CNN) from scratch with randomized initial weights. Instead, common source CNN model on very large set beforehand, and then use learned as an initialization target model. In learning realm, this procedure called fine-tuning CNN. This paper presents experimental study how combine collection of incrementally fine-tuned models for cross-domain multi-class object category recognition tasks. A group...

Aimed at structure of humanoid robot and requirements for control performance, this paper designs realizes joint controller based on CAN bus, structures an effective credible system by connecting all controllers, force sensors, task management computer upper limbs together; the transmission audio video information wireless LAN; constitutes whole robot. Characteristics are simple structure, high reliability, good real-time, friendly human-machine interface. This includes mainly: overall...

This paper provides the design and implementation of navigation control software for an aerospace-level lunar rover validation system. Based on analysing structure characteristics system, real-time involving interrupt service routine, tasks, task-task communication is designed implemented within embedded operation system VxWorks. Although current research our project still way, this here already successfully applied in some physical experiments platform. The practical employment indicates...

The rapid advances in embedded microprocessor technologies provide opportunities to promote digital avionic systems significantly. With the complexity and frequency increase, power consumption has quickly become a key design constraint designs. processors must utilize energy efficiently, as their payload is restricted by battery factor weight constraints aircrafts. This paper proposed new approaching using dynamic time slice turning with way prediction technology for achieving high...

Aiming at Humanoid Robot structure and requirement for control performance, this paper designs realizes joint controller based on CAN, structures an effective credible system by connecting controllers, force sensors harmonious level computer together. This includes mainly: overall of the system, design implement hardware software, topology bring forward some imaginations to enhance performance system.

Direct estimation of body weight through noncontact methods is crucial for applications such as health monitoring, surveillance, and robot-assisted casualty rescue. Existing from images are often affected by various factors, camera distance, human orientation, posture, etc., which ignore the fact that bodies typically inhabit 3D space. To address problem, we propose an approach estimating based on virtual anthropometric measurements deep features instead reasoning pixels. Specifically,...