Non-linear friction compensation using backstepping control and robust friction state observer with recurrent fuzzy neural networks

2009 ◽  
Vol 223 (7) ◽  
pp. 973-988 ◽  
Author(s):  
B S Kim ◽  
S I Han
Keyword(s):  
State Observer ◽  
Fuzzy Neural Networks ◽  
Friction Model ◽  
Backstepping Control ◽  
Ball Screw ◽  
Control Scheme ◽  
Non Linear ◽  
Linear Friction ◽  
Fuzzy Neural ◽  
Lugre Friction

A robust precise tracking control for a servomechanical system with non-linear dynamic friction is presented. The LuGre friction that is adopted as a non-linear friction model contains both a directly immeasurable friction state variable and the uncertainty caused by incomplete parameter identification and change of the condition of contact surface. To provide an efficient solution to these problems, a composite robust control scheme is proposed, which consists of a robust friction state observer, a recurrent fuzzy neural network (RFNN) approximator, and an adaptive reconstructed error compensator with backstepping control. A robust friction state observer is designed to estimate the unknown internal state of the LuGre friction model. Next, a proposed RFNN scheme approximates the lumped friction torque uncertainty. Finally, an adaptive error compensator eliminates a reconstructed error arising from RFNN approximation. Some simulations and experiments for a ball-screw servosystem are carried out to demonstrate the performance of a proposed control scheme.

Linear friction tester design and validation

2021 ◽  
pp. 135065012110454
Author(s):  
Mehran Shams Kondori ◽  
Saied Taheri
Keyword(s):  
Wear Rate ◽  
Normal Load ◽  
Lateral Force ◽  
Friction Model ◽  
Ball Screw ◽  
Dynamic Friction ◽  
Test Bed ◽  
Friction Tester ◽  
Linear Friction ◽  
Novel Design

Due to the complexity of friction phenomena, empirical analysis is the best way to predict the friction coefficient. To accomplish this, laboratory test rigs are needed. Although a rotary dynamic friction test bed was available to the authors, it had its limitations, such as low speed, inducement of lateral force, and the limitation of testing samples with different shapes. This paper will explain the process of designing and manufacturing a novel test setup for measuring friction and wear of the tire. The newly designed test rig can apply dynamic loading during the tests, and it can automatically measure the wear rate and temperature between cycles. In addition, it can be used for measuring the wear rate of rubber samples sliding on different types of surfaces. Therefore, experiments can be run under more controlled conditions. The designed linear friction tester can slide flat and round rubber samples approximately three meters across a large flat surface. The frictional force of rubber samples can be measured for various normal loads, velocities, and surface conditions. The new setup can automatically control the applied normal load on the sample using proportional–integral–derivative controller control. The important difference between this novel design and the existing testers used by other researchers is implementing the ball screw technology and the servo motor with high accuracy encoder to achieve highly accurate test results. In this design, the new mechanism for the ball screw is designed to increase the speed limit and eliminating vibrations while keeping the precision. In addition, in this design, the sample's mass can be measured automatically after each test cycle, thus providing a measure of the rate of wear of the rubber. In this study, the data collected by the linear friction tester is validated by comparing the data to the data collected by the dynamic friction tester (a validated rotary friction tester that exists in CenTiRe Lab). The data collected by the new setup was later used to benchmark the Persson analytical friction model.

Application of interval type-2 fuzzy neural networks in non-linear identification and time series prediction

2013 ◽  
Vol 18 (6) ◽  
pp. 1213-1224 ◽  
Author(s):  
Oscar Castillo ◽  
Juan R. Castro ◽  
Patricia Melin ◽  
Antonio Rodriguez-Diaz
Keyword(s):  
Neural Networks ◽  
Time Series ◽  
Time Series Prediction ◽  
Fuzzy Neural Networks ◽  
Non Linear ◽  
Fuzzy Neural ◽  
Interval Type
Sign in / Sign up

Export Citation Format

Share Document