The Optimal Mars Entry Guidance with External Disturbance Using Neural Network Solution

In this paper, a novel control strategy is presented for reinforcement learning with disturbance compensation to solve the problem of quadrotor positioning under external disturbance. The proposed control scheme applies a trained neural-network-based reinforcement learning agent to control the quadrotor, and its output is directly mapped to four actuators in an end-to-end manner. The proposed control scheme constructs a disturbance observer to estimate the external forces exerted on the three axes of the quadrotor, such as wind gusts in an outdoor environment. By introducing an interference compensator into the neural network control agent, the tracking accuracy and robustness were significantly increased in indoor and outdoor experiments. The experimental results indicate that the proposed control strategy is highly robust to external disturbances. In the experiments, compensation improved control accuracy and reduced positioning error by 75%. To the best of our knowledge, this study is the first to achieve quadrotor positioning control through low-level reinforcement learning by using a global positioning system in an outdoor environment.

Download Full-text

Adaptive neural network control for maglev vehicle systems with time-varying mass and external disturbance

Neural Computing and Applications ◽

10.1007/s00521-021-05874-2 ◽

2021 ◽

Author(s):

Yougang Sun ◽

Junqi Xu ◽

Guobin Lin ◽

Ning Sun

Keyword(s):

Neural Network ◽

External Disturbance ◽

Network Control ◽

Time Varying ◽

Neural Network Control ◽

Adaptive Neural Network ◽

Maglev Vehicle ◽

Adaptive Neural Network Control ◽

Vehicle Systems ◽

Varying Mass

Download Full-text

Prediction of Fouling in Condenser Based on Fuzzy Stage Identification and Chebyshev Neural Network

Measurement Science Review ◽

10.2478/msr-2013-0017 ◽

2013 ◽

Vol 13 (2) ◽

pp. 94-99 ◽

Cited By ~ 3

Author(s):

Shaosheng Fan ◽

Qingchang Zhong

Keyword(s):

Neural Network ◽

Prediction Model ◽

Parameter Optimization ◽

Adaptive Algorithm ◽

External Disturbance ◽

Operational Parameters ◽

Chebyshev Neural Network ◽

Adaptive Parameter ◽

Novel Approach ◽

Variation Trends

The prediction of fouling in condenser is heavily influenced by the periodic fouling process and dynamics change of the operational parameters, to deal with this problem, a novel approach based on fuzzy stage identification and Chebyshev neural network is proposed. In the approach, the overall fouling is separated into hard fouling and soft fouling, the variation trends of these two kinds of fouling are approximated by using Chebyshev neural network, respectively, in order to make the prediction model more accurate and robust, a fuzzy stage identification method and adaptive algorithm considering external disturbance are introduced, based on the approach, a prediction model is constructed and experiment on an actual condenser is carried out, the results show the proposed approach is more effective than asymptotic fouling model and adaptive parameter optimization prediction model.

Download Full-text

Cerebellum-inspired neural network solution of the inverse kinematics problem

Biological Cybernetics ◽

10.1007/s00422-015-0661-7 ◽

2015 ◽

Vol 109 (6) ◽

pp. 561-574 ◽

Cited By ~ 8

Author(s):

Mitra Asadi-Eydivand ◽

Mohammad Mehdi Ebadzadeh ◽

Mehran Solati-Hashjin ◽

Christian Darlot ◽

Noor Azuan Abu Osman

Keyword(s):

Neural Network ◽

Inverse Kinematics ◽

Network Solution ◽

Inverse Kinematics Problem

Download Full-text

Increase of the Resistance to Noise in Data for Neural Network Solution of the Inverse Problem of Magnetotellurics with Group Determination of Parameters

Artificial Neural Networks and Machine Learning – ICANN 2016 - Lecture Notes in Computer Science ◽

10.1007/978-3-319-44778-0_59 ◽

2016 ◽

pp. 502-509 ◽

Cited By ~ 4

Author(s):

Igor Isaev ◽

Eugeny Obornev ◽

Ivan Obornev ◽

Mikhail Shimelevich ◽

Sergey Dolenko

Keyword(s):

Neural Network ◽

Inverse Problem ◽

Network Solution ◽

Determination Of Parameters

Download Full-text

Resource Efficient Arithmetic Effects on RBM Neural Network Solution Quality Using MNIST

2011 International Conference on Reconfigurable Computing and FPGAs ◽

10.1109/reconfig.2011.79 ◽

2011 ◽

Cited By ~ 7

Author(s):

Antony W. Savich ◽

Medhat Moussa

Keyword(s):

Neural Network ◽

Solution Quality ◽

Network Solution

Download Full-text

KnowMeme: A Knowledge-enriched Graph Neural Network Solution to Offensive Meme Detection

10.1109/escience51609.2021.00029 ◽

2021 ◽

Author(s):

Lanyu Shang ◽

Christina Youn ◽

Yuheng Zha ◽

Yang Zhang ◽

Dong Wang

Keyword(s):

Neural Network ◽

Network Solution

Download Full-text

Adaptive Integrated Control for Omnidirectional Mobile Manipulators Based on Neural-Network

Transdisciplinary Advancements in Cognitive Mechanisms and Human Information Processing ◽

10.4018/978-1-60960-553-7.ch020 ◽

2011 ◽

pp. 310-325

Author(s):

Xiang-min Tan ◽

Dongbin Zhao ◽

Jianqiang Yi ◽

Dong Xu

Keyword(s):

Neural Network ◽

Large Scale ◽

Integrated Control ◽

External Disturbance ◽

Torque Control ◽

Mobile Manipulator ◽

Mobile Manipulators ◽

Modeling And Control ◽

Control Approach ◽

Proposed Model

An omnidirectional mobile manipulator, due to its large-scale mobility and dexterous manipulability, has attracted lots of attention in the last decades. However, modeling and control of such systems are very challenging because of their complicated mechanism. In this paper, an unified dynamic model is developed by Lagrange Formalism. In terms of the proposed model, an adaptive integrated tracking controller, based on the computed torque control (CTC) method and the radial basis function neural-network (RBFNN), is presented subsequently. Although CTC is an effective motion control strategy for mobile manipulators, it requires precise models. To handle the unmodeled dynamics and the external disturbance, a RBFNN, serving as a compensator, is adopted. This proposed controller combines the advantages of CTC and RBFNN. Simulation results show the correctness of the proposed model and the effectiveness of the control approach.

Download Full-text