scholarly journals A New Approach to Modeling and Controlling a Pneumatic Muscle Actuator-Driven Setup Using Back Propagation Neural Networks

Complexity ◽  
2018 ◽  
Vol 2018 ◽  
pp. 1-9 ◽  
Author(s):  
Jun Zhong ◽  
Xu Zhou ◽  
Minzhou Luo
Keyword(s):  
Back Propagation ◽  
Weight Ratio ◽  
Adaptive Controller ◽  
Back Propagation Neural Network ◽  
Sampled Data ◽  
Nonlinear Characteristics ◽  
New Approach ◽  
Pneumatic Muscle ◽  
Muscle Actuators

Pneumatic muscle actuators (PMAs) own excellent compliance and a high power-to-weight ratio and have been widely used in bionic robots and rehabilitated robots. However, the high nonlinear characteristics of PMAs due to inherent construction and pneumatic driving principle bring great challenges in applications acquired accurately modeling and controlling. To tackle the tricky problem, a single PMA mass setup is constructed, and a back propagation neural network (BPNN) is employed to identify the dynamics of the setup. An offline model is built up using sampled data, and online modifications are performed to further improve the quality of the model. An adaptive controller based on BPNN is designed using gradient descent information of the built-up model. Experiments of identifying the PMA setup using BPNN and position tracking by adaptive BPNN controller are performed, and results demonstrate the good capacity in accurate controlling of the PMA setup.

scholarly journals Active Motion Control of a Knee Exoskeleton Driven by Antagonistic Pneumatic Muscle Actuators

Actuators ◽  
2020 ◽  
Vol 9 (4) ◽  
pp. 134
Author(s):  
Wei Zhao ◽  
Aiguo Song
Keyword(s):  
Sliding Mode ◽  
Weight Ratio ◽  
Practical Application ◽  
Active Motion ◽  
Pneumatic Muscle ◽  
The Past ◽  
Robotic Devices ◽  
Muscle Actuators ◽  
Further Development ◽  
Rehabilitation Robotic

The pneumatic muscle actuator (PMA) has been widely applied in the researches of rehabilitation robotic devices for its high power to weight ratio and intrinsic compliance in the past decade. However, the high nonlinearity and hysteresis behavior of PMA limit its practical application. Hence, the control strategy plays an important role in improving the performance of PMA for the effectiveness of rehabilitation devices. In this paper, a PMA-based knee exoskeleton based on ergonomics is proposed. Based on the designed knee exoskeleton, a novel proxy-based sliding mode control (PSMC) is introduced to obtain the accurate trajectory tracking. Compared with conventional control approaches, this new PSMC can obtain better performance for the designed PMA-based exoskeleton. Experimental results indicate good tracking performance of this controller, which provides a good foundation for the further development of assist-as-needed training strategies in gait rehabilitation.

Modeling and Enhancement of Mckibben Pneumatic Muscle Actuators

2012 ◽  
Vol 591-593 ◽  
pp. 793-796 ◽  
Author(s):  
He Zuo ◽  
Guo Liang Tao ◽  
Xiao Cong Zhu
Keyword(s):  
Dynamic Performance ◽  
System Model ◽  
Model Parameters ◽  
Mass Ratio ◽  
Slow Time ◽  
Nonlinear Characteristics ◽  
Pneumatic Muscle ◽  
Fast Switching ◽  
Dynamic Performances ◽  
Muscle Actuators

Mckibben pneumatic muscle actuators (PMA) have many advantages such as high power-mass ratio and low price, but their strong nonlinear characteristics makes modeling and controlling very hard, which limits their applications. This paper presents the modeling of PMA and methods to enhance their dynamic performances. Considering the incorporated models of fast-switching valves and PMAs, the entire system model is modified in some aspects and the dominant model parameters are determined through experimental results to estimate the dynamic characteristics precisely. Simulation and proper experiments reveal that the dynamic performance of PMA can be improved through filling the PMA with materials of high thermal conductivity. The slow time-varying disturbance caused by the temperature variation of inner gas can be reduced much, which lowers the difficulty of controlling.

LICENSE-PLATE RECOGNITION USING DWT AND NEURAL NETWORK

2006 ◽  
Vol 04 (04) ◽  
pp. 601-615 ◽  
Author(s):  
TIAN-DING CHEN
Keyword(s):  
Neural Network ◽  
Back Propagation ◽  
Back Propagation Neural Network ◽  
High Accuracy ◽  
Discrete Wavelet ◽  
License Plate ◽  
License Plate Recognition ◽  
Structure Modification ◽  
New Approach ◽  
Low Pass

This paper presents a new approach for license-plate recognition using Discrete Wavelet Transform (DWT) and Plastic Perception Neural Network (PPNN). It accomplishes the preliminary license-plate localization by applying low-pass wavelet coefficients. Since the amount of data reduces to 1/4, this approach saves a lot of running time, simplifies computational complexity, and economizes memory usage. It adopts the LL and HH sub-bands, which come from a two-dimensional Haar DWT to implement the localization and segmentation for license plates. The proposed methodology provides high accuracy for locating a license plate from an image, and has a high tolerance for license plate displacement in the images. Back-Propagation Neural Network (BPNN) has the advantage of anti-noise and anti-distortion, but the problems of traditional BPNN are a longer learning period, iterations are not prone to convergence, and local minimum. The proposed methods combine the parallel distributive process concept with the BPNN structure modification to solve the above problems. This paper also utilizes PPNN to solve taking position, scale and rotation of the license-plate recognition.

Experimental Comparisons of Sliding Mode Controlled Pneumatic Muscle and Cylinder Actuators

2014 ◽  
Vol 136 (4) ◽  
Author(s):  
Ville Jouppila ◽  
S. Andrew Gadsden ◽  
Asko Ellman
Keyword(s):  
Sliding Mode ◽  
Weight Ratio ◽  
Industrial Applications ◽  
Stick Slip ◽  
Servo Systems ◽  
Pneumatic Muscle ◽  
Positioning Systems ◽  
Actuation System ◽  
Highly Nonlinear ◽  
Muscle Actuators

Pneumatic muscle actuators offer a higher force-to-weight ratio compared to traditional cylinder actuators, and introduce stick-slip-free operation that offers an interesting option for positioning systems. Despite several advantages, pneumatic muscle actuators are commonly avoided in industrial applications, mainly due to rather different working principles. Due to the highly nonlinear characteristics of the muscle actuator and pneumatic system, a reliable control strategy is required. Although muscle actuators are widely studied, the literature lacks detailed studies where the performance for servo systems is compared with traditional pneumatic cylinders. In this paper, a pneumatic servo actuation system is compared with a traditional cylinder actuator. As the overall system dynamics are highly nonlinear and not well defined, a sliding mode control (SMC) strategy is chosen for the control action. In order to improve the tracking performance, an SMC strategy with an integral action (SMCI) is also implemented. The control algorithms are experimentally applied on the pneumatic muscle and the cylinder actuator, for the purposes of position tracking. The robustness of the systems are verified and compared by varying the applied loads.

scholarly journals LBP and GLCM Based Image Forgery Recognition

2018 ◽  
Vol 7 (4.6) ◽  
pp. 217
Author(s):  
D. Vaishnavi ◽  
T. S. Subashini ◽  
G. N. Balaji ◽  
D. Mahalakshmi
Keyword(s):  
Neural Network ◽  
Local Binary Pattern ◽  
Back Propagation ◽  
Back Propagation Neural Network ◽  
Support Vector ◽  
Gray Level ◽  
Image Splicing ◽  
Naked Eye ◽  
Occurrence Matrix

The forgery of digital images became very easy and it’s very difficult to ascertain the authenticity of such images by naked eye. Among the various kinds of image forgeries, image splicing is a frequent and widely used technique. Even though various methods are available to detect image splicing forgery, authors have attempted to provide a novel hybrid method which can yield greater accuracy, sensitivity and specificity. In this method, gray level co-occurrence matrix (GLCM) features are extracted using local binary pattern (LBP) operator on the image and the detection of the splicing forged images among the authentic images is done using the popular pattern recognition algorithms such as combined k-NN (Comb-KNN), back propagation neural network (BPNN) and support vector machine (SVM). The recorded results are also compared with the existing results of the previous studies to ascertain the quality of the results.  

scholarly journals Quality Monitoring and Analysis of Xinjiang ‘Korla’ Fragrant Pear in Cold Chain Logistics and Home Storage with Multi-Sensor Technology

2019 ◽  
Vol 9 (18) ◽  
pp. 3895 ◽  
Author(s):  
Jingjing Liu ◽  
Xu Zhang ◽  
Zhigang Li ◽  
Xiaoshuan Zhang ◽  
Tomislav Jemric ◽  
...  
Keyword(s):  
Back Propagation ◽  
Back Propagation Neural Network ◽  
Soluble Solids ◽  
Sensor Data ◽  
Cold Chain ◽  
Sensor Technology ◽  
Index Measurement ◽  
Carbon Dioxide Co2 ◽  
Continuous Sensor

Korla fragrant pear is prestigious for its special texture and unique flavor but suffers storage and supply chain difficulties for its deterioration-prone properties. In order to improve the storage quality of Korla fragrant pears during the whole cold chain from the orchard to the customers, the paper deeply researches multiple influencing factors of cold chain logistics and home storage of Korla fragrant pears with multi-sensor technology (MST), such as the temperature, relative humidity, concentrations of oxygen (O2), carbon dioxide (CO2) and ethylene (C2H4). Cold chain logistics are assessed by sensory evaluation and physiological index measurement, and home storage environments are classified by using back propagation neural network (BPNN) in both refrigerators and ordinary rooms. Experimental results show that the MST-based detectors can improve the accuracy of continuous sensor data acquisition, such that the preservation quality of Korla fragrant pears is effectively enhanced by data analysis on gas contents, firmness, pH, and total soluble solids. These results indicate that Korla fragrant pears stored in refrigerators have a higher acceptance for customers.

Evaluating Artistic Voice of Singing Objectivly Using BPNN

2012 ◽  
Vol 546-547 ◽  
pp. 1240-1244 ◽  
Author(s):  
Xiao Wu Li ◽  
Hong Bo Ouyang
Keyword(s):  
Neural Network ◽  
Subjective Evaluation ◽  
Back Propagation ◽  
Average Energy ◽  
Back Propagation Neural Network ◽  
Fuzzy Synthetic Evaluation ◽  
Bp Network ◽  
Artistic Voice ◽  
And Training

In this paper, an objective way to evaluate artistic voice of singing is discussed. The model transforms artistic voice evaluation indicators into qualified data as BP network input and takes fuzzy synthetic evaluation results as output. The authors take F1(the first Formant), F3(the third Formant), vocal range, perturbation of F1, perturbation of F3 and average energy as the evaluating parameters and assess the quality of singing voices with BPNN(back propagation neural network). The results are then compared with the subjective evaluation of experienced professionals. Experiments show that BP neural network is effective to evaluate the singing voices, thus to be helpful to scientific guidance of selecting and training the talent of artistic voice.

Prediction for Air-Bending Springback Radius of Sheet Metal Using Back Propagation Neural Network and Micro Genetic Algorithm

2011 ◽  
Vol 219-220 ◽  
pp. 1174-1177
Author(s):  
Ze Min Fu ◽  
Guang Ming Liu
Keyword(s):  
Neural Network ◽  
Genetic Algorithm ◽  
Sheet Metal ◽  
Back Propagation ◽  
Back Propagation Neural Network ◽  
Accurate Prediction ◽  
Air Bending ◽  
Micro Genetic Algorithm ◽  
Bpnn Model

Springback radius is a very important factor to influence the quality of sheet metal air-bending forming. Accurate prediction of springback radius is essential for the design of air-bending tools. In this paper, a three-layer back propagation neural network (BPNN), integrated with micro genetic algorithm (MGA), is proposed to solve the problem of springback radius. A micro genetic algorithm is used for minimizing the error between the predictive value and the experimental one. Based on air-bending experiment, the prediction model of springback radius is developed by using the integrated neural network. The results show that more accurate prediction of springback radius can be obtained with the MGA-BPNN model. It can be taken as a valuable tool for air-bending forming of sheet metal.

Application of Neural Network in Free Air Ball Forming for Wire Bonder

1999 ◽  
Author(s):  
Jau-Liang Chen ◽  
Yeh-Chao Lin ◽  
Chun-Hsien Liu ◽  
Wen-Chang Kuo ◽  
Tzung-Ching Lee
Keyword(s):  
Neural Network ◽  
Back Propagation ◽  
Tail Length ◽  
Back Propagation Neural Network ◽  
Ball Bond ◽  
Error Back Propagation ◽  
Fine Pitch ◽  
Free Air ◽  
Network Method

Abstract The shape and size of free-air-ball formation deeply affect the quality of wire bonding. It not only affects the bondability of first bond (ball bond), but also affects the possibility of processing low loop height bonding for thin form packages and high I/O fine pitch packages. Several parameters, such as tail length, spark gap, supplied voltage, current and time of electrical flame-off unit etc., will affect the free-air-ball formation. This paper represents a study of using error-back-propagation neural network method to analyze the effect of each parameter and to predict the final result of the ball forming. From the experiment, it is shown that neural network can not only be used to precisely predict the size of ball formation, but also saves sampling time.

Multiplexed Force Control of Pneumatic Muscles

2006 ◽  
Author(s):  
V. Jouppila ◽  
A. Ellman
Keyword(s):  
High Power ◽  
Force Control ◽  
Weight Ratio ◽  
Pressure Control ◽  
Pneumatic Actuators ◽  
Pneumatic Muscle ◽  
Servo Valve ◽  
Highly Nonlinear ◽  
Muscle Actuators ◽  
Pneumatic Cylinders

Pneumatic actuators are often used in applications that require high power-to-weight ratio, combined with low price and clean and fast operation. However, due to the compressibility of air and highly nonlinear behavior of seal friction, the position and force control of these actuators is difficult to manage. As a result, pneumatic cylinders have been used for many years solely in simple repetitive tasks requiring only a very limited amount of system control. Nonetheless, the pneumatic actuators have properties such as compactness, high power-to-weight ratio, and simplicity that are desirable features in advanced robotics. To overcome the shortcomings, a number of advanced pneumatic components have been developed, of which the most promising is the pneumatic muscle. Compared to a cylinder, a pneumatic muscle not only has a higher power-to-weight and power-to-volume ratio but it is also almost frictionless and has zero leakage. In spite of the muscle actuator's nonlinear force-to-contraction characteristics, many motion and force control methods have been successfully applied to it. The characteristics of the actuator enable it to be used in simple positioning systems and as a variable gas spring. The actuator's almost linear pressure-to-force ratio is extremely well-suited to a variety of gripping and pressing applications. Due to the muscle actuator's characteristics and recent developments in pneumatic valve technology, there is an opportunity to share a single pressure control servo valve among multiple muscle actuators. The multiplexed control of the actuators with only one servo valve reduces the system costs significantly. In this paper we investigate the feasibility of employing multiplexed force control of four pneumatic muscle actuators. In the system, pressure is controlled by a single proportional pressure valve. High-speed switching valves are used for activating the pressure control for each muscle actuator in the desired manner. Pneumatic cylinders are attached to the other ends of the muscles in order to cause controllable position disturbances. The displacement, force and pressure of each muscle are measured with appropriate sensors. The system behavior is investigated under position disturbances.

Sign in / Sign up

Export Citation Format

Share Document