scholarly journals Position Control of a Pneumatic Muscle Actuator Using RBF Neural Network Tuned PID Controller

2015 ◽  
Vol 2015 ◽  
pp. 1-16 ◽  
Author(s):  
Jie Zhao ◽  
Jun Zhong ◽  
Jizhuang Fan
Keyword(s):  
Neural Network ◽  
Experimental Data ◽  
Phenomenological Model ◽  
Pid Controller ◽  
Position Control ◽  
Rbf Neural Network ◽  
Pneumatic Muscle ◽  
Highly Nonlinear ◽  
Experimental Apparatus ◽  
Pneumatic Muscle Actuator

Pneumatic Muscle Actuator (PMA) has a broad application prospect in soft robotics. However, PMA has highly nonlinear and hysteretic properties among force, displacement, and pressure, which lead to difficulty in accurate position control. A phenomenological model is developed to portray the hysteretic behavior of PMA. This phenomenological model consists of linear component and hysteretic component force. The latter component is described by Duhem model. An experimental apparatus is built up and sets of experimental data are acquired. Based on the experimental data, parameters of the model are identified. Validation of the model is performed. Then a novel cascade position PID controller is devised for a 1-DOF manipulator actuated by PMA. The outer loop of the controller is to cope with position control whilst the inner loop deals with pressure dynamics within PMA. To enhance the adaptability of the PID algorithm to the high nonlinearities of the manipulator, PID parameters are tuned online using RBF Neural Network. Experiments are performed and comparison between position response of RBF Neural Network based PID controller and that of classic PID controller demonstrates the effectiveness of the novel adaptive controller on the manipulator.

Position Control of PWM-Actuated Pneumatic Muscle Actuator System

2011 ◽  
Author(s):  
Ville Jouppila ◽  
Asko Ellman
Keyword(s):  
Position Control ◽  
System Model ◽  
Modeling Error ◽  
Stick Slip ◽  
Pneumatic Muscle ◽  
Positioning Systems ◽  
Highly Nonlinear ◽  
Pneumatic Muscle Actuator ◽  
Non Linear System ◽  
Real Challenge

Pneumatic servo positioning systems have been in use for long time and subject to wide spectrum of studies due to their numerous advantages: inexpensive, clean, safe and high ratio of power to weight. However, the compressibility of air and the inherent non-linearity of these systems continue to make achieving accurate position control a real challenge. Conventional pneumatic servo systems are based on cylinder actuators that are difficult to control precisely due to the aforementioned nonlinearities as well as the nonlinear behavior of the air flow through the valve, the friction between the cylinder and the piston, and the stick slip effect at the low velocity of the system. In this paper, a position servo control system using a pneumatic muscle actuator is studied. Pneumatic muscle actuator is a novel type of actuator which has even higher force to weight ratio than the cylinder. In addition, muscle actuator introduces a stick slip free operation giving an interesting option for positioning systems. However, significant hysteresis and position dependant force result in a highly nonlinear system, a real challenge for good control performance. In this paper, pneumatic muscle actuator is controlled by a low-cost on/off valve with PWM-strategy instead of costly servo or proportional valve. The main processes of the system, including flow dynamics, pressure dynamics, force dynamics and load dynamics are derived to provide a full nonlinear model that captures all the major nonlinearities of the system. This model is used for analyzing and tuning the controller performances by simulations before implementing in the real system. In addition, a recently introduced method of using bipolynomial functions to model the valve flow rate is utilized to provide a continuous and invertible description of flow for controller designs. A proportional plus velocity plus acceleration controller with feed-forward component (PVA+FF) is designed based on the linearized system model. For a comparison, a sliding mode controller (SMC) based on linear as well as non-linear system model are designed. The performance of the designed controllers is studied by simulations. The stability and performance analysis includes the effects of friction modeling error and valve modeling error. The robustness of the controllers is tested by varying the payload mass of the system.

The Optimization of Adaptive PID Control Algorithm Based on RBF Neural Network

2014 ◽  
Vol 998-999 ◽  
pp. 943-946
Author(s):  
Jing Liu ◽  
Guo Xin Wang
Keyword(s):  
Neural Network ◽  
Pid Controller ◽  
Rbf Neural Network ◽  
High Reliability ◽  
Simple Algorithm ◽  
Poor Performance ◽  
Learning Ability ◽  
The Neural Network ◽  
Pid Controller Tuning ◽  
Deterministic Control

As the earliest practical controller, PID controller has more than 50 years of history, and it is still the most widely used and most common industrial controllers. PID controller is simple to understand and use, without a prerequisite for an accurate model of the physical system, thus become the most popular, the most common controller. The reason why PID controller is the first developed one is that its simple algorithm, robustness and high reliability. It is widely used in process control and motion control, especially for accurate mathematical model that can be established deterministic control system. But the conventional PID controller tuning parameters are often poor performance, poor adaptability to the operating environment. The neural network has a strong nonlinear mapping ability, competence, self-learning ability of associative memory, and has a viable quantities of information processing methods and good fault tolerance.

A RBF Neural Network Approach for Fitting Creep Curve of Sandstone

2010 ◽  
Vol 171-172 ◽  
pp. 274-277
Author(s):  
Yun Liang Tan ◽  
Ze Zhang
Keyword(s):  
Neural Network ◽  
Experimental Data ◽  
Creep Curve ◽  
Bp Neural Network ◽  
Rbf Neural Network ◽  
Reference Value ◽  
Network Approach ◽  
Neural Network Approach ◽  
Training Time ◽  
Set Up

In order to quest an effective approach for predicate the rheologic deformation of sandstone based on some experimental data, an improved approaching model of RBF neural network was set up. The results show, the training time of improved RBF neural network is only about 10 percent of that of the BP neural network; the improved RBF neural network has a high predicating accuracy, the average relative predication error is only 7.9%. It has a reference value for the similar rock mechanics problem.

Simulation and Experimental Study in Position Control of Servo Hydraulic System With Flexible Load Using Adaptive Neural Network and Differential Evolution Strategy

2006 ◽  
Author(s):  
Hassan Yousefi ◽  
Heikki Handroos
Keyword(s):  
Neural Network ◽  
Differential Evolution ◽  
Position Control ◽  
Differential Evolution Algorithm ◽  
System Stability ◽  
Minimum Problem ◽  
Nonlinear Behaviour ◽  
Neural Network Controller ◽  
Highly Nonlinear ◽  
Flexible Load

Hydraulic position servos with an asymmetrical cylinder are commonly used in industry. These kinds of systems are nonlinear in nature and generally difficult to control. Because of parameters changing during extending and retracting, using constant gain will cause overshoot, poor performance or even loss of system stability. The highly nonlinear behaviour of these devices makes them ideal subjects for applying different types of sophisticated controllers. This paper is concerned with a second order adaptive model reference and an artificial neural network controller to position tracking of a servo hydraulic with a flexible load. In present study, a neural network with two outputs is presented. One of the outputs of neural network is used for system’s dynamic compensator and another one for gain scheduling controller. To avoid the local minimum problem, Differential Evolution Algorithm (DEA) is used to find the weights and biases of neural network. The proposed controller is verified with a common used p-controller. The simulation and experimental results suggest that if the neural network is chosen and trained well, it improves all performance evaluation criteria such as stability, fast response, and accurate reference model tracking in servo hydraulic systems.

Adaptive Neural Network in Compensation the Dynamics and Position Control of a Servo-Hydraulic System With a Flexible Load

2005 ◽  
Author(s):  
Hassan Yousefi ◽  
Heikki Handroos
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Hydraulic System ◽  
Position Control ◽  
Reference Model ◽  
Differential Evolution Algorithm ◽  
Nonlinear Behavior ◽  
System Stability ◽  
Hydraulic Systems ◽  
Highly Nonlinear

Asymmetrical servo-hydraulic systems are commonly used in industry. These kinds of systems are nonlinear in nature and generally difficult to control. Because of changing system parameters, using the same gain will cause overshoot or even loss of system stability. The highly nonlinear behavior of these devises makes them idea subjects for applying different types of sophisticated controllers. This paper is concerned with using two artificial neural networks in compensation the dynamics and position tracking of a second order model reference in a flexible servo-hydraulic system. In present study, a neural network as an acceleration feedforward and another one as a gain scheduling of a proportional controller are proposed. Differential evolution algorithm is used to find the weights and biases to avoid the local minima. The proposed controller was verified with a commonly used p-controller. The results suggest that if the neural networks choose and train well, they improve all performance evaluation criteria such as stability, fast response, and accurate reference model tracking in servo-hydraulic systems.

Adaptive Integral-Type Neural Sliding Mode Control for Pneumatic Muscle Actuator

2014 ◽  
Vol 8 (6) ◽  
pp. 888-895 ◽  
Author(s):  
Dang Xuan Ba ◽  
◽  
Kyoung Kwan Ahn ◽  
Nguyen Trong Tai ◽  
Keyword(s):  
Neural Network ◽  
Working Conditions ◽  
Sliding Mode ◽  
System Stability ◽  
Integral Type ◽  
Real Time System ◽  
Pneumatic Muscle ◽  
Adaptive Sliding Mode Controller ◽  
Weight Factors ◽  
Pneumatic Muscle Actuator

This paper presents an integral-type adaptive sliding mode controller integrated into a neural network for position-tracking control of a pneumatic muscle actuator testing system. Stability of the closed-loop system is covered by the sliding mode algorithm while both control error and control energy are minimized by the neural network. With only four weight factors in the hidden layer and two weight factors in the output layer, the network provides a very high calculation speed. Then, the approach is successfully verified on a real-time system under different working conditions. By comparing it with a proportional-integraldifferential controller on the same system and under the same working conditions, the effectiveness of the designed controller is confirmed.

Sign in / Sign up

Export Citation Format

Share Document