We present an optimization based real-time walking controller for a full size humanoid robot. The consists of two levels optimization, high level trajectory optimizer that reasons about center mass and swing foot trajectories, low tracks those trajectories by solving floating base body inverse dynamics problem using Quadratic Programming. Our is capable on rough terrain, also achieves longer steps, faster speed, heel-strike toe push-off. Results are demonstrated with Boston Dynamics' Atlas...

We propose a framework to use full-body dynamics for humanoid state estimation. The main idea is decouple the full body vector into several independent vectors. Some decoupled vectors can be estimated very efficiently with steady Kalman Filter. In Filter, covariance computed only once during initialization. Furthermore, decoupling speeds up numerical linearization of dynamic model. demonstrate that these estimators are capable handling walking on flat ground and rough terrain.

High-bandwidth and long-range nanoprecision motion is crucially required by many advanced instruments robotic systems, which rely on the nanopositioners. However, traditional nanopositioners suffer from their fixed stiffness can only obtain performance, leading to narrow applications. Here, a novel nanopositioner called normal-stressed electromagnetic-driven stiffness-tunable (EDSTN) presented, consists of three parts: A electromagnetic actuator generate actuation force, high-stiffness...

Rate-dependent hysteresis seriously deteriorates the positioning accuracy of piezoelectric actuators, especially when tracking high-frequency signals. As a widely-used nonparametric Bayesian method, Gaussian process (GP) has proven its effectiveness in nonlinear modeling. In this paper, dimension input to GP model is extended consider more dynamic features signal so as improve rate-dependent modeling accuracy. contrast with traditional training set containing only position and speed...

Uneven terrain walking is one of the key challenges in bipedal walking. In this paper, we propose a motion pattern generator for slope 3D dynamics using preview control zero moment point (ZMP). method, future ZMP locations are selected with respect to known gradient. The trajectory center mass (CoM) robot generated by controller maintain at desired location. Two models walking, namely upslope and downslope, investigated. Continuous on slopes different gradients also studied enable robots...

Excellent tracking accuracy and remarkable disturbance rejection capability are quite important for fast tool servo (FTS) operations in ultraprecision manufacturing. For these purposes, this article proposes an enhanced real-time iterative compensation (ERIC) method with resonance suppression the FTS system. The dual-loop controller is first designed optimized to damp high-frequency increase control bandwidth of Then, a observer employed eliminate impact various unmodeled disturbances...

In this article, an intelligent feedforward prediction and compensation scheme to combine with a dual-loop high-bandwidth controller is proposed for high-speed high-precision tracking controls of nanopositioning stage. First, the consisting inner loop damping outer developed all parameters optimized simultaneously, which could provide control bandwidth over first resonant frequency Next, Gaussian process machine learning model employed capture dynamic characteristics error controlled plant....

With several microservice architectures comprising of thousands web services in total, used to serve 630 million customers, companies like Meituan face challenges the verification and validation their software. The use automated techniques, especially advanced AI-based ones, could bring significant benefits here. EvoMaster is an open-source test case generation tool for services, that exploits latest advances field Search-Based Software Testing research. This paper reports on our experience...

Using Lissajous scan trajectory is a promising way to get rid of issues resulting from the raster in several advanced applications, like atomic force microscopy and fast steering mirror. Although repetitive control (RC)-based schemes, which are tailed track periodic trajectories, have been successfully utilized improve tracking performance enabled limited due unavoidable dynamic cross-coupling effect. To address this issue, article proposes dual-notch-based RC (DNRC) scheme via generating...

Repetitive disturbances exist widely within the automation systems, which is one of major issues that hinder achievement precision operations. Dedicated to mitigating these disturbances, a generalized periodic-disturbance observer (PDOB) using spectrum-selection filtering scheme proposed in this paper with an application non-minimum phase system. The design process filter derives from comb-like notch for PDOB presented detail. variants are also distributed only odd- and even-harmonics. To...

Step planning is becoming an increasingly important research topic for humanoid robots. Most cost functions step in the literature are designed based on terrain information. The energy to perform each action usually ignored. In walking, consumption depends gait features such as length and width. this paper, we use three simple intuitive different lengths, widths, turning angle. These inspired by human walking analysis, function parameters tuned match computed costs optimal motions obtained...

For autonomous navigation system of intelligent vehicle, robust and stable control with accurate tracking ability is one the key requirements. In this paper, we present a systematic controller design approach for vehicle system. The proposed integrates dynamic model online updated path by quadratic programming (QP) cost function, which considers both error stability. A novel planner based on differential (DDP) consideration kinematic feasibility used. accuracy has been improved utilizing...

The repetitive control (RC) is a promising technique to achieve the high-speed periodic operation of nanopositioning stages. However, conventional RC fairly sensitive period-time uncertainty reference input and tends amplify tracking error at nonperiodic frequencies. To overcome these two limitations simultaneously, robust (RRC) proposed by integrating band-stop filter into spectrum-selection scheme in this article. With developed general structure filter, <inline-formula...