Search

Biped robots usually adopt feet with a rigid structure that simplifies walking on flat grounds and yet hinders ground adaptation in unstructured environments, thus jeopardizing stability. We recently explored in the SoftFoot the idea of adapting a robotic foot to ground irregularities along the sagittal plane. Building on the previous results, we propose in this paper a novel robotic foot able to adapt both in the sagittal and frontal planes, similarly to the human foot. It features five parallel modules with intrinsic longitudinal adaptability that can be combined in many possible designs through optional rigid or elastic connections. By following a methodological design approach, we narrow down the design space to five candidate foot designs and implement them on a modular system. Prototypes are tested experimentally via controlled application of force, through a robotic arm, onto a sensorized plate endowed with different obstacles. Their performance is compared, using also a rigid foot and the previous SoftFoot as a baseline. Analysis of footprint stability shows that the introduction of the transverse arch, by elastically connecting the five parallel modules, is advantageous for obstacle negotiation, especially when obstacles are located under the forefoot. In addition to biped robots' locomotion, this finding might also benefit lower-limb prostheses design. [ABSTRACT FROM AUTHOR]

Knee-stretched walking is desirable for biped robot because it is a characteristic of human walking as well as lower energy consumption. In this paper, we proposed a new knee-stretched bipedal walking pattern generation approach named Cosine-law-based Spatially Quantized Gait (C-SQG). C-SQG includes two parts, the first one is the Cosine law-based kinematic joint angles generation for knee-stretched biped walking with the hip horizontal displacement; the second one is combining with the concept of Spatially Quantized Dynamics (SQD) and the optimization-based methods. Based on these two approaches, the balanced human-like walking patterns with fully extended knee joints for humanoid robots can be generated. The proposed approach was tested on the CoppeliaSim simulation platform, and the experimental results indicate that the knee-stretched walking patterns for different walking speeds and foot locations can be generated. Compared to the existing knee-stretched gait generation approach, the C-SQG is more robust and adaptive to different desired foot locations appropriately. [ABSTRACT FROM AUTHOR]

Biped robots usually adopt feet with a rigid structure that simplifies walking on flat grounds and yet hinders ground adaptation in unstructured environments, thus jeopardizing stability. We recently explored in the SoftFoot the idea of adapting a robotic foot to ground irregularities along the sagittal plane. Building on the previous results, we propose in this paper a novel robotic foot able to adapt both in the sagittal and frontal planes, similarly to the human foot. It features five parallel modules with intrinsic longitudinal adaptability that can be combined in many possible designs through optional rigid or elastic connections. By following a methodological design approach, we narrow down the design space to five candidate foot designs and implement them on a modular system. Prototypes are tested experimentally via controlled application of force, through a robotic arm, onto a sensorized plate endowed with different obstacles. Their performance is compared, using also a rigid foot and the previous SoftFoot as a baseline. Analysis of footprint stability shows that the introduction of the transverse arch, by elastically connecting the five parallel modules, is advantageous for obstacle negotiation, especially when obstacles are located under the forefoot. In addition to biped robots’ locomotion, this finding might also benefit lower-limb prostheses design.

A computational approach is presented to obtain the optimal path of the end-effector for the 10 DOF bipedal robot to increase its load carrying capacity for a given task from point to point. The synthesizing optimal trajectories problem of a robot is formulated as a problem of trajectory optimization. An Iterative Linear Programming method (ILP) is developed for finding a numerical solution for this nonlinear trajectory. This method is used for determining the maximum dynamic load carrying capacity of bipedal robot walking subjected to torque actuators, stability and jerk limits constraints. First, the Lagrangian dynamic equation should be written to be suitable for the load dynamics which together with kinematic equations are substantial for determining the optimal trajectory. After that, a representation of the state space of the dynamic equations is introduced also the linearized dynamic equations are needed to obtain the numerical solution of the trajectory optimization followed by formulation for the optimal trajectory problem with a maximum load. Finally, the method of ILP and the computational aspect is applied to solve the problem of trajectory synthesis and determine the dynamic load carrying capacity (DLCC) to the bipedal robot for each of the linear and circular path. By implementing on an experimental biped robot, the simulation results were validated.

Rehabilitation robots are gaining significant popularity for impaired gait rehabilitation. However, to make the recovering individual feel natural while walking and restore their original gait pattern, adapting the rehabilitation system according to the individual's need and walking characteristics becomes imperative. In this paper, we have compared four deep learning models for their ability to generate a personalized gait trajectory at different gait speeds. The first three models are primitive and are the basic implementations of long short term memory (LSTM), convolutional neural network (CNN) and gated recurrent unit (GRU). The fourth model is our proposed model, which is a sequential combination of LSTM and CNN. We considered hip, knee and ankle joints data as human gait is represented as the joint angle trajectories of these joints in the sagittal plane. We trained these models on a benchmark public human walking dataset consisting of treadmill walking data of 42 healthy individuals at eight different walking speeds. Anthropometric and demographic data along with gait speeds were given as input to the models. Our proposed LSTM-CNN sequential model is able to generate stable gait trajectories in the speed range of 0.49-1.76 m/s with a high correlation of 0.98 between the actual and the predicted trajectories, and an R2 Score of 0.94 is obtained. This work can be utilized for providing personalized gait reference trajectories for the rehabilitation of amputees and stroke patients using rehabilitation systems such as exoskeleton robots and prosthetic legs. Also, this work can be utilized for generating stable walking trajectories for bipedal robots. [ABSTRACT FROM AUTHOR]

In this research, a biped robot is programmed to learn to walk using genetic algorithms. The biped robot has the ability to rise after a fall. The system is made according to human genetics, where a child tries to learn to walk, takes a few steps, falls, and then stands up somehow after a fall. After this process, the system restarts and tries to walk again. The biped robot evolves after each generation, and after each generation it will have a more natural walking posture. The system is unpredictable, because the biped robot sometimes evolves faster and sometimes evolves slower, like in human genetics where some children are faster and other children are slower to walk. The biped robot is considered to have evolved enough after having a walking posture similar to that of a human and does not fall. [ABSTRACT FROM AUTHOR]

This work proposes a parameters auto-tuning strategy for biped locomotion in whole-body stabilization control (inverse kinematics based and inverse dynamics based) and active impedance control based on Bayesian optimization(BO). Using the domain knowledge, the parameter space is divided into three sub-spaces and optimized by decoupling BO and alternating BO algorithms. The effectiveness of the proposed method is demonstrated in simulation using a torque-controlled biped robot that we developed. The 32 control parameters are tuned in less than 400 evaluations. In addition, the auto-tuned parameters are robust to different top-level velocity inputs and show compliant behavior with balance in push recovery scenarios. To the best of our knowledge, this is the first work to automatically tune the parameters of the three controllers (inverse kinematics, inverse dynamics and active impedance control) jointly. [ABSTRACT FROM AUTHOR]

Virtual decomposition control (VDC) is a powerful approach to control/regulate the motion of high degrees-of-freedom (DOF) robots. It is a control strategy based on Newton-Euler (N‑E) formulation with recursive implementations. As a result, this work is focused on designing a fully decentralized control structure for an open-chain robotic mechanism. The dynamics of the target robot is formulated based on N-E formulation, while the robotic system is controlled using the function approximation technique (FAT)-based adaptive control. The key idea is to control each robot link subsystem individually such that the coupling interaction term is estimated using the FAT. The stability of each link subsystem and the whole system is proved using the Lyapunov theory. In this work, we assume that the target robot is equipped with direct-drive motors such that the joint dynamics is integrated with the link subsystem. To this end, a six-link biped robot is simulated using MATLAB/SIMULINK package with the proposed controller. In effect, to stabilize the locomotion of a biped mechanism a multi-level control strategy is required. However, the current controller is considered a low-level control for tracking the desired joint references. The results prove the precise tracking for the proposed control architecture. [ABSTRACT FROM AUTHOR]

Bipedal robots have gained increasing attention for their human-like mobility which allows them to work in various human-scale environments. However, their inherent instability makes it difficult to control their balance while they are physically interacting with the environment. This study proposes a novel balance controller for bipedal robots based on a behavior cloning model as one of the machine learning techniques. The behavior cloning model employs two deep neural networks (DNNs) trained on human-operated balancing data, so that the trained model can predict the desired wrench required to maintain the balance of the bipedal robot. Based on the prediction of the desired wrench, the joint torques for both legs are calculated using robot dynamics. The performance of the developed balance controller was validated with a bipedal lower-body robotic system through simulation and experimental tests by providing random perturbations in the frontal plane. The developed balance controller demonstrated superior performance with respect to resistance to balance loss compared to the conventional balance control method, while generating a smoother balancing movement for the robot. [ABSTRACT FROM AUTHOR]

In general, the stability control of the biped robot is based on Zero Moment Point (ZMP) which excludes the foot-ground contact stability characteristics. The concept of Contact Wrench Cones (CWC) is introduced in the stable gait generation of Humanoid robots for incorporating contact stability in recent research. The CWC confines the contact forces within limits as well as controls the Center of Pressure (CoP). This paper presents a CWC-based optimum gait generation for the dynamic biped locomotion. The kinematic model for a 12 Degree of Freedom (DoF) Biped robot is developed in the Screw theory framework. The dynamic equations of motion are derived using the recursive Newton–Euler (N–E) method. Stable dynamic gaits are generated for flat and step surfaces. Controlled Elitist Non-dominated Sorting Genetic Algorithm (NSGA-II) is used for the optimization of the dynamically stable biped gaits. The contact forces and the variation in friction coefficient are estimated for all the above cases. The peak values of the friction coefficient are used for assessing the conditions for kinetic friction and slip. [ABSTRACT FROM AUTHOR]

This paper considers dynamic stepping of biped robots when walking through a big obstacle, which is a great problem for these kinds of robots, using nonlinear model predictive control technique. No need for trajectory planning is one of the key advantages of this method. Hence, the gait length is not predefined and the controller determines it on the basis of the physical restraints and dynamic constancy conditions for approaching the obstacle and stepping over it. The stability and good performance for stepping over large obstacles are guaranteed by the definition of appropriate linear and nonlinear constraints and cost functions. Furthermore, to solve the uncertainty problem corresponding to the biped dynamic model, neural networks are used to identify the model of robot. For stability analysis, the Poincaré map with the fixed-point technique is utilized to guarantee the robot stability. Simulation outcomes indicate excellent efficiency of the suggested technique used to a five-link biped robot crossing over a 40x15 cm obstacle in sagittal plane when preserving a safety clearance from it with no need to a predefined trajectory. [ABSTRACT FROM AUTHOR]

This paper modifies the single rigid body (SRB) model, and considers the swinging leg as the disturbances to the centroid acceleration and rotational acceleration of the SRB model. This paper proposes deep reinforcement learning (DRL)-based model predictive control (MPC) to resist the disturbances of the swinging leg. The DRL predicts the swing leg disturbances, and then MPC gives the optimal ground reaction forces according to the predicted disturbances. We use the proximal policy optimization (PPO) algorithm among the DRL methods since it is a very stable and widely applicable algorithm. It is an on-policy algorithm based on the actor–critic framework. The simulation results show that the improved SRB model and the PPO-based MPC method can accurately predict the disturbances of the swinging leg to the SRB model and resist the disturbance, making the locomotion more robust. [ABSTRACT FROM AUTHOR]

In this research, a biped robot is programmed to learn to walk using genetic algorithms. The biped robot has the ability to rise after a fall. The system is made according to human genetics, where a child tries to learn to walk, takes a few steps, falls, and then stands up somehow after a fall. After this process, the system restarts and tries to walk again. The biped robot evolves after each generation, and after each generation it will have a more natural walking posture. The system is unpredictable, because the biped robot sometimes evolves faster and sometimes evolves slower, like in human genetics where some children are faster and other children are slower to walk. The biped robot is considered to have evolved enough after having a walking posture similar to that of a human and does not fall.

Recently, much research has focused on the design of biped robots with stable and smooth walking ability, identical to human beings, and thus, in the coming years, biped robots will accomplish rescue or exploration tasks in challenging environments. To achieve this goal, one of the important problems is to design a chip for real-time calculation of moving length and rotation angle of the biped robot. This paper presents an efficient and accurate coordinate rotation digital computer (CORDIC)-based efficient chip design to calculate the moving length and rotation angle for each step of the biped robot. In a previous work, the hardware cost of the accurate CORDIC-based algorithm of biped robots was primarily limited by the scale-factor architecture. To solve this problem, a binomial approximation was carefully employed for computing the scale-factor. In doing so, the CORDIC-based architecture can achieve similar accuracy but with fewer iterations, thus reducing hardware cost. Hence, incorporating CORDIC-based architecture with binomial approximation, pipelined architecture, and hardware sharing machines, this paper proposes a novel efficient and accurate CORDIC-based chip design by using an iterative pipelining architecture for biped robots. In this design, only low-complexity shift and add operators were used for realizing efficient hardware architecture and achieving the real-time computation of lengths and angles for biped robots. Compared with current designs, this work reduced hardware cost by 7.2%, decreased average errors by 94.5%, and improved average executing performance by 31.5%, when computing ten angles of biped robots. [ABSTRACT FROM AUTHOR]

This study investigates the effect of vertical movements of the pelvis on the joint of a biped robot. We evaluate experimentally two cases of walking on flat terrains and avoiding a toll bar obstacle. The experimental results show that the torque and velocity of the knee joint vary with the vertical movement of the pelvis. The knee joint velocity can be reduced when walking with the vertical movement compared to that when walking without the vertical movement of the pelvis. Based on the experimental results, the vertical movement of the pelvis can increase the stride lengths or relax the knee-actuator performance requirements. In addition, the trajectories of the center of mass (CoM) generated to avoid a toll bar obstacle using the preview control that includes the vertical motion of the CoM are compared to those using the preview control that neglects the vertical motion of the CoM. The results show that the trajectory generated by including the vertical motion of the CoM can reduce energy consumption and walk more stable. [ABSTRACT FROM AUTHOR]

This paper investigates the connection between non-prehensile manipulation, specifically juggling, and legged locomotion, focusing on biped robots. In this direction, the hybrid nature of juggler robot dynamics and biped robots, the zero moment point stability, and the non-prehensile dynamic grasping conditions are provided and analysed. The similarities between juggling actions and bipedal robot locomotion are discussed. The descriptive and mathematical analysis demonstrates many similitudes between juggler systems with cubic objects and flat-feet biped robots in throwing, catching, and stabilization phases, and between juggler actions with impact and point-feet legged robots in their switching (hybrid) dynamic behaviours. Finally, a common control framework based on the zero dynamic concept and integral sliding mode approach is proposed, and it applies to both classes of non-prehensile juggler systems and biped robots. To test the performance of the devised control system, a three-degrees-of-freedom juggler robot and a two-link walker are selected as case studies, whose simulation results demonstrate the effectiveness and feasibility of the proposed unified control framework for both systems. [ABSTRACT FROM AUTHOR]

This paper presents a passivity-based control strategy dealing with underactuated two-degree-of-freedom (2-DoF) mechanical systems. Such a methodology, which is based on the interconnection and damping assignment passivity-based control (IDA-PBC), rooted within the port-controlled Hamiltonian framework, can be applied to a very large class of underactuated 2-DoF mechanical systems. The main contribution, compared to the previous literature, is that the new methodology does not involve the resolution of any partial differential equation, since explicit solutions are given, while no singularities depending on generalised momenta are introduced by the controller. The proposed strategy is applied to two case studies: a) the stabilisation of a translational oscillator with a rotational actuator (TORA) system; b) the gait generation for an underactuated compass-like biped robot. The performances of the presented solution are evaluated through numerical simulations. [ABSTRACT FROM AUTHOR]

Biped robots are expected to be able to work in complex environments. However, these robots will inevitably fall at times and such falls may cause injury to the robot itself or to people nearby. Therefore, it is necessary to detect that the robot is falling to be able to warn the robot in sufficient time when it is about to fall and to switch its controller to protect the vulnerable parts of the robot. Modeling and analysis of the biped robot falling problem cannot be fully accurate and many current learning-based methods rely on large quantities of fall data that are difficult to use to train fragile robots. A machine learning-based fall detection method is therefore proposed in this paper. This method requires only a small amount of training data to obtain good fall detection, making the training process on the robot platform much safer. A support vector machine is used to determine the state of the robot and the decision boundary of the stable state is updated during motion to enable the classifier to match the motion capability of the robot. [ABSTRACT FROM AUTHOR]

Bipedal robots have gained increasing attention for their human-like mobility which allows them to work in various human-scale environments. However, their inherent instability makes it difficult to control their balance while they are physically interacting with the environment. This study proposes a novel balance controller for bipedal robots based on a behavior cloning model as one of the machine learning techniques. The behavior cloning model employs two deep neural networks (DNNs) trained on human-operated balancing data, so that the trained model can predict the desired wrench required to maintain the balance of the bipedal robot. Based on the prediction of the desired wrench, the joint torques for both legs are calculated using robot dynamics. The performance of the developed balance controller was validated with a bipedal lower-body robotic system through simulation and experimental tests by providing random perturbations in the frontal plane. The developed balance controller demonstrated superior performance with respect to resistance to balance loss compared to the conventional balance control method, while generating a smoother balancing movement for the robot.

Multimodal and controllable locomotion in complex terrain is of great importance for practical applications of insect-scale robots. Robust locomotion plays a particularly critical role. In this study, a locomotion mechanism for magnetic robots based on asymmetrical friction effect induced by magnetic torque is revealed and defined. The defined mechanism overcomes the design constraints imposed by both robot and substrate structures, enabling the realization of multimodal locomotion on complex terrains. Drawing inspiration from human walking and running locomotion, a biped robot based on the mechanism is proposed, which not only exhibits rapid locomotion across substrates with varying friction coefficients but also achieves precise locomotion along patterned trajectories through programmed controlling. Furthermore, apart from its exceptional locomotive capabilities, the biped robot demonstrates remarkable robustness in terms of load-carrying and weight-bearing performance. The presented locomotion and mechanism herein introduce a novel concept for designing magnetic robots while offering extensive possibilities for practical applications in insect-scale robotics., (© 2024 The Authors. Advanced Materials published by Wiley‐VCH GmbH.)

This paper modifies the single rigid body (SRB) model, and considers the swinging leg as the disturbances to the centroid acceleration and rotational acceleration of the SRB model. This paper proposes deep reinforcement learning (DRL)-based model predictive control (MPC) to resist the disturbances of the swinging leg. The DRL predicts the swing leg disturbances, and then MPC gives the optimal ground reaction forces according to the predicted disturbances. We use the proximal policy optimization (PPO) algorithm among the DRL methods since it is a very stable and widely applicable algorithm. It is an on-policy algorithm based on the actor–critic framework. The simulation results show that the improved SRB model and the PPO-based MPC method can accurately predict the disturbances of the swinging leg to the SRB model and resist the disturbance, making the locomotion more robust.

This paper presents a novel biped-wheeled-wearable machine, named HANDSHAKE, and obtained by an evolution of two robots presented in other works: one flexible-wheeled leg and one biped-flexible-wheeled robot. A critical design analysis of these two robots helped the author to propose a novel machine able to revolutionize the lower body exoskeletons' world. Conceptual and functional design, mechanical behavior (kinematics and dynamics), and multibody simulation of the biped-wheeled exoskeleton are presented in this paper, and a first reduced scale prototype is used to show the feasibility of the proposed solution. The simple control architecture used in this work underlines the enormous advantages to use the HANDSHAKE system for people with a complete absence of mobility, which are completely supported by this machine. This is possible thanks to the wheeled feet of the HANDSHAKE system which allow to support more weights respect to the classical exoskeletons, available on market and literature. The proposed machine increases stability, dynamic balance, autonomy, reducing power supply and complexity in comparison with classical exoskeleton systems because the wheeled feet are always in contact with the ground. These advantages, recognized in humanoid robots, may be used also in exoskeletons. [ABSTRACT FROM AUTHOR]

This paper deals with the problem of compliance in biped robots locomotion. After a first literature review, we designed and simulated a 3D virtual model of a biped robot conceived with the same link dimensions (and weights) of a standard man. In all simulation, the same input to the robot actuators are proposed, modifying only the compliance of the attached links. We first validated the model and then compared results increasing and reducing compliance on hip, knee and ankle joints. The very good results underline how the robot performances may be increased including the compliant element in the knee. In particular, we noted how including a proper value of the compliance in the knee, power and torque of the actuators may be reduced increasing robot speed.

In this paper, we propose a method for generating the center-of-mass (CoM) pattern for biped robots. One of the most common methods for CoM pattern generation is to approximate the robot's complex dynamics as alinear inverted pendulum model, applies preview controls, and generates CoM trajectories. However, the vertical motion of CoM is neglected during this approximation process. In this study, we formulate the preview control problem considering the dynamics of the CoM vertical motion and propose an algorithm to solve it. The results of numerical experiments to evaluate the proposed algorithm show it is more stable than the existing algorithm and can be sufficiently fast for online operation. [ABSTRACT FROM AUTHOR]

This paper presents an adaptive time-delay control method to handle the balance of a class of biped robots. Considering model uncertainties and unknown external disturbances, a high-dimensional integral Lyapunov function has been developed, then time-delay estimation technique (TDE) is utilized, and in order to compensate for the errors produced by TDE, an adaptive control law is designed. Based on the evolution of the error trajectories, the adaptive control law can automatically regulate the control gains without any threshold value. By the Lyapunov synthesis, the semiglobally uniformly ultimately boundness (UUB) of the closed-loop control system has been theoretically proved. We have conducted the experimental studies on biped robot to demonstrate the effectiveness of the proposed control method. [ABSTRACT FROM AUTHOR]

This paper provides a brief overview the need for humanoid robots specifically bipedal robots. The advantages and shortcomings of bipedal robots have been highlighted. The paper discusses in detail the various types of controllers that have been applied for the control of humanoid robots. Typical controllers such as slide mode control, active force control, computed torque control etc., require the boundaries of uncertainties to be known, which is very difficult considering the real world terrain. The application of artificial neural networks has thus become a priority in controller design for bipedal robots. Some of the neural network based on the controller design and their comparisons to typical controllers have been surveyed in this paper. The survey presented visibly shows that the application of neural networks results in an exponential increase in accuracy for bipedal robots in comparison to classic approaches. [ABSTRACT FROM AUTHOR]

Linear inverted pendulum model (LIPM) is an effective and widely used simplified model for biped robots. However, LIPM includes only the single support phase (SSP) and ignores the double support phase (DSP). In this situation, the acceleration of the center of mass (CoM) is discontinuous at the moment of leg exchange, leading to a negative impact on walking stability. If the DSP is added to the walking cycle, the acceleration of the CoM will be smoother and the walking stability of the biped will be improved. In this paper, a linear pendulum model (LPM) for the DSP is proposed, which is similar to LIPM for the SSP. LPM has similar characteristics to LIPM. The dynamic equation of LPM is also linear, and its analytical solution can be obtained. This study also proposes different trajectory-planning methods for different situations, such as periodic walking, adjusting walking speed, disturbed state recovery, and walking terrain-blind. These methods have less computation and can plan trajectory in real time. Simulation results verify the effectiveness of proposed methods and that the biped robot can walk stably and flexibly when combining LIPM and LPM.

Featured Application: The final product is an algorithm that allows a simulated robot to learn sequences of poses (single configurations of its joints) in order to learn to walk (and extendable to other tasks) by finding the combination of these poses. Programming robots for performing different activities requires calculating sequences of values of their joints by taking into account many factors, such as stability and efficiency, at the same time. Particularly for walking, state of the art techniques to approximate these sequences are based on reinforcement learning (RL). In this work we propose a multi-level system, where the same RL method is used first to learn the configuration of robot joints (poses) that allow it to stand with stability, and then in the second level, we find the sequence of poses that let it reach the furthest distance in the shortest time, while avoiding falling down and keeping a straight path. In order to evaluate this, we focus on measuring the time it takes for the robot to travel a certain distance. To our knowledge, this is the first work focusing both on speed and precision of the trajectory at the same time. We implement our model in a simulated environment using q-learning. We compare with the built-in walking modes of an NAO robot by improving normal-speed and enhancing robustness in fast-speed. The proposed model can be extended to other tasks and is independent of a particular robot model. [ABSTRACT FROM AUTHOR]

This paper aims to develop an active disturbance rejection controller design scheme for compass-like biped robots. In the previous study, with a special designed mechanical structure on compass-like biped, we have generated a high-efficient walking gait. The original controller applied linearization approximation based on the method transverse coordinate control, with which the ability of disturbance rejection is insufficient. We introduce the active disturbance rejection controller method into the control scheme, which only requires the information of angular momentum and has the ability to compensate the model error and the external disturbance. With the new control strategy, the linearization approximation method is replaced by an active disturbance rejection controller and the robustness is improved. In order to further apply this control method, a switching control strategy has been proposed and a series of numerical simulations show that active disturbance rejection controller has a good disturbance rejection effect. [ABSTRACT FROM AUTHOR]

To realize the dynamic balance optimization and control of biped robots under the perturbing external forces in the double-support phase, a systematic scheme is proposed in this paper. First, a constrained dynamic model of biped robots and a reduced order dynamical model for the double-support phase are formulated. Considering the dynamic external wrench applied on biped robots, we present a dynamic force distribution approach based on quadratic objective function for computing the optimal contact forces to equilibrate the dynamic external wrench. As a result, the sum of the normal force components is minimized for enhancing safety and energy saving. Then, one primary recurrent neural network (RNN) is adopted to solve the optimization problem subject to both equality and inequality constraints. For the derived optimized contact force and motion, hybrid motion/force control is proposed based on another RNN to approximate unknown dynamic functions. Adaptive learning algorithms for learning the parameters of the RNN are provided as well. The proposed control can deal with the uncertainties including approximation errors and external disturbances. Extensive simulations are presented to demonstrate the effectiveness of the proposed optimization and control approach. [ABSTRACT FROM AUTHOR]

Programming robots for performing different activities requires calculating sequences of values of their joints by taking into account many factors, such as stability and efficiency, at the same time. Particularly for walking, state of the art techniques to approximate these sequences are based on reinforcement learning (RL). In this work we propose a multi-level system, where the same RL method is used first to learn the configuration of robot joints (poses) that allow it to stand with stability, and then in the second level, we find the sequence of poses that let it reach the furthest distance in the shortest time, while avoiding falling down and keeping a straight path. In order to evaluate this, we focus on measuring the time it takes for the robot to travel a certain distance. To our knowledge, this is the first work focusing both on speed and precision of the trajectory at the same time. We implement our model in a simulated environment using q-learning. We compare with the built-in walking modes of an NAO robot by improving normal-speed and enhancing robustness in fast-speed. The proposed model can be extended to other tasks and is independent of a particular robot model.

In this paper, a novel method to determine the center of mass position of each link of human-like multibody biped robots is proposed. A first formulation to determine the total center of mass position has been tested in other works on a biped platform with human-like dimensions. In this paper, the formulation is optimized and extended, and it is able to give as output the center of mass positions of each link of the platform. The calculation can be applied to different types of robots. The optimized formulation is validated using a simulated biped robot in MATLAB. [ABSTRACT FROM AUTHOR]

In this paper, a method to find equivalences among dynamic models is presented. The models are given in different generalized coordinates for the same mechanical system. This novel method is valid for planar biped robots, i.e., those whose motions are executed only in a plane. We show that no matter which generalized coordinates are used to get the dynamic models, there is always an equivalence among them by using a particular input matrix. We exhibit some advantages of getting the dynamic model using absolute coordinates instead of relative coordinates, and we show how to calculate the input matrix for getting the equivalence between these models. Because of its simplicity, a compass-like biped robot model is used as example to explain in detail the novel procedure. In order to appreciate the benefits of the proposed procedure in biped robots of high degrees of freedom, we also present the equivalence between two dynamic models for the single support phase of a 5 degrees of freedom biped robot using absolute coordinates and relative ones. [ABSTRACT FROM AUTHOR]

This article focuses on inverse kinematic formulation and dynamic modeling of the Nao biped robot's lower body, accompanied by verification with the joints' angles as experimental data. Dynamic modeling in two different planes is discussed and joint angles for the given positions, nominal conditions, and trajectory computations are simulated and graphically illustrated. A new approach for development of the inverse dynamics on the aforementioned robot's lower body is proposed in this paper, analytically studied, and compared with MSC Adams for two various scenarios of fixed supporting leg and ground contact implementation. [ABSTRACT FROM AUTHOR]

In this paper, neural network control strategies based on radial basis functions are designed for biped robots, which includes balancing and posture control. To deal with system uncertainties, neural networks are used to approximate the unknown model of the robot. Both full state feedback control and output feedback control are considered in this paper. With the proposed control, the trajectories of the closed-loop system are semiglobally uniformly bounded which can be proved via Lyapunov stability theorem. Simulations are also carried out to illustrate the effectiveness of the proposed control. [ABSTRACT FROM PUBLISHER]