scholarly journals Online Control for Biped Robot with Incremental Learning Mechanism

2021 ◽  
Vol 11 (18) ◽  
pp. 8599
Author(s):  
Liang Yang ◽  
Guanyu Lai ◽  
Yong Chen ◽  
Zhihui Guo
Keyword(s):  
Incremental Learning ◽  
Technical Problem ◽  
Dynamic Environment ◽  
Biped Robot ◽  
Training Data ◽  
Dynamics Simulation ◽  
Iteration Algorithm ◽  
Biped Robots ◽  
Learning Mechanism ◽  
Interval Type

In this paper, we develop a new online walking controller for biped robots, which integrates a neural-network estimator and an incremental learning mechanism to improve the control performance in dynamic environment. With the aid of an iteration algorithm for updating, some newly incoming data can be used straightforwardly to update into the original well-trained model, in order to avoid a time-consuming retraining procedure. On the other hand, how to maintain the zero-moment-point stability and counteract the effect of yaw moment simultaneously is also a key technical problem to be addressed. To this end, an interval type-2 fuzzy weight identifier is newly developed, which assigns weight for each walking sample to deal with the imbalanced distribution problem of training data. The effectiveness of the proposed control scheme has been verified through a full-dynamics simulation and a practical robot experiment.

scholarly journals Effects of Torso Pitch Motion on Energy Efficiency of Biped Robot Walking

2021 ◽  
Vol 11 (5) ◽  
pp. 2342
Author(s):  
Long Li ◽  
Zhongqu Xie ◽  
Xiang Luo ◽  
Juanjuan Li
Keyword(s):  
Energy Efficiency ◽  
Energy Consumption ◽  
Biped Robot ◽  
Gait Pattern ◽  
Pattern Generation ◽  
Biped Robots ◽  
Pitch Motion ◽  
Double Support ◽  
Gait Parameters ◽  
Support Phase

Gait pattern generation has an important influence on the walking quality of biped robots. In most gait pattern generation methods, it is usually assumed that the torso keeps vertical during walking. It is very intuitive and simple. However, it may not be the most efficient. In this paper, we propose a gait pattern with torso pitch motion (TPM) during walking. We also present a gait pattern with torso keeping vertical (TKV) to study the effects of TPM on energy efficiency of biped robots. We define the cyclic gait of a five-link biped robot with several gait parameters. The gait parameters are determined by optimization. The optimization criterion is chosen to minimize the energy consumption per unit distance of the biped robot. Under this criterion, the optimal gait performances of TPM and TKV are compared over different step lengths and different gait periods. It is observed that (1) TPM saves more than 12% energy on average compared with TKV, and the main factor of energy-saving in TPM is the reduction of energy consumption of the swing knee in the double support phase and (2) the overall trend of torso motion is leaning forward in double support phase and leaning backward in single support phase, and the amplitude of the torso pitch motion increases as gait period or step length increases.

Sliding mode incremental learning algorithm for interval type-2 Takagi–Sugeno–Kang fuzzy neural networks

2012 ◽  
Vol 3 (3) ◽  
pp. 179-188 ◽  
Author(s):  
Sevil Ahmed ◽  
Nikola Shakev ◽  
Andon Topalov ◽  
Kostadin Shiev ◽  
Okyay Kaynak
Keyword(s):  
Neural Networks ◽  
Incremental Learning ◽  
Sliding Mode ◽  
Learning Algorithm ◽  
Fuzzy Neural Networks ◽  
Fuzzy Neural ◽  
Interval Type ◽  
Takagi Sugeno

scholarly journals Comparative stable walking gait optimization for small-sized biped robot using meta-heuristic optimization algorithms

2018 ◽  
Vol 40 (4) ◽  
pp. 407-424
Author(s):  
Tran Thien Huan ◽  
Ho Pham Huy Anh
Keyword(s):  
Humanoid Robot ◽  
Differential Evolution Algorithm ◽  
Biped Robot ◽  
The Novel ◽  
Biped Robots ◽  
Walking Gait ◽  
Gait Parameters ◽  
Lift Height ◽  
Evolution Algorithm ◽  
Gait Optimization

This paper proposes a new way to optimize the biped walking gait design for biped robots that permits stable and robust stepping with pre-set foot lifting magnitude. The new meta-heuristic CFO-Central Force Optimization algorithm is initiatively applied to optimize the biped gait parameters as to ensure to keep biped robot walking robustly and steadily. The efficiency of the proposed method is compared with the GA-Genetic Algorithm, PSO-Particle Swarm Optimization and Modified Differential Evolution algorithm (MDE). The simulated and experimental results carried on the prototype small-sized humanoid robot demonstrate that the novel meta-heuristic CFO algorithm offers an efficient and stable walking gait for biped robots with respect to a pre-set of foot-lift height value.

Incremental Learning of Object Detector with Limited Training Data

2021 ◽  
Author(s):  
Muhammad Abdullah Hafeez ◽  
Adnan Ul-Hasan ◽  
Faisal Shafait
Keyword(s):  
Incremental Learning ◽  
Training Data

System Identification Based on Dynamical Training for Recurrent Interval Type-2 Fuzzy Neural Network

2013 ◽  
pp. 188-205
Author(s):  
Tsung-Chih Lin ◽  
Yi-Ming Chang ◽  
Tun-Yuan Lee
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
System Identification ◽  
Nonlinear System ◽  
Fuzzy Neural Networks ◽  
Training Data ◽  
Fuzzy Neural ◽  
Interval Type ◽  
Recurrent Interval

This paper proposes a novel fuzzy modeling approach for identification of dynamic systems. A fuzzy model, recurrent interval type-2 fuzzy neural network (RIT2FNN), is constructed by using a recurrent neural network which recurrent weights, mean and standard deviation of the membership functions are updated. The complete back propagation (BP) algorithm tuning equations used to tune the antecedent and consequent parameters for the interval type-2 fuzzy neural networks (IT2FNNs) are developed to handle the training data corrupted by noise or rule uncertainties for nonlinear system identification involving external disturbances. Only by using the current inputs and most recent outputs of the input layers, the system can be completely identified based on RIT2FNNs. In order to show that the interval IT2FNNs can handle the measurement uncertainties, training data are corrupted by white Gaussian noise with signal-to-noise ratio (SNR) 20 dB. Simulation results are obtained for the identification of nonlinear system, which yield more improved performance than those using recurrent type-1 fuzzy neural networks (RT1FNNs).

Sign in / Sign up

Export Citation Format

Share Document