Wahyu Caesarendra

Finger movement pattern recognition method using artificial neural network based on electromyography (EMG) sensor

In this study, the EMG signals are processed using 16 time-domain features extraction to classify the finger movement such as thumb, index, middle, ring, and little. The pattern recognition of 16 extracted features are classified using artificial neural network (ANN) with two layer feed forward network. The network utilizes a log-sigmoid transfer function in hidden layer and a hyperbolic tangent sigmoid transfer function in the output layer. The ANN uses 10 neurons in hidden layer and 5 neurons in output layer. The training of ANN pattern recognition uses Levenberg-Marquardt training algorithm and the performance utilizes mean square error (MSE). At about 22 epochs the MSE of training, test, and validation get stabilized and MSE is very low. There is no miss classification during training process. Based on the resulted overall confusion matrix, the accuracy of thumb, middle, ring, and little is 100%. The confusion of index is 16.7%. The overall confusion matrix shows that the error is 3.3% and overall performance is 96.7%.

An application of nonlinear feature extraction - A case study for low speed slewing bearing condition monitoring and prognosis

This paper presents the application of four nonlinear methods of feature extraction in slewing bearing condition monitoring and prognosis: these are largest Lyapunov exponent, fractal dimension, correlation dimension, and approximate entropy methods. Although correlation dimension and approximate entropy methods have been used previously, the largest Lyapunov exponent and fractal dimension methods have not been used in vibration condition monitoring to date. The vibration data of the laboratory slewing bearing test-rig run at 1 rpm was acquired daily from February to August 2007 (138 days). As time progressed, a more accurate observation of the alteration of bearing condition from normal to faulty was obtained using nonlinear features extraction. These findings suggest that these methods provide superior descriptive information about bearing condition than time-domain features extraction, such as root mean square (RMS), variance, skewness and kurtosis.

Pattern recognition methods for multi stage classification of parkinson's disease utilizing voice features

A number of papers has presented a pattern recognition method for Parkinsons Disease (PD) detection. However, the literatures only able to classify subjects as either healthy of suffering from PD. This paper presents a pattern recognition method for multi stage classification of PD utilizing voice features. 22 features are obtained from University of California-Irvine (UCI) data repository. These features are extracted using Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA). It is found that PCA is better than LDA in terms of extracting significant features. Some classifiers such as Support Vector Machine (SVM), Adaptive Boosting (AdaBoost), K-Nearest Neighbor (KNN) and Adaptive Resonance Theory-Kohonen Neural Network (ART-KNN) are then used and compared. These methods are applied in multi stage classification. The classification results show that SVM has better testing accuracy than the other methods.

Degradation Trend Estimation and Prognosis of Large Low Speed Slewing Bearing Lifetime

In many applications, degradation of bearing conditions is usually monitored by changes in time-domain features. However, in low speed (< 10 rpm) slewing bearing, these changes are not easily detected because of the low energy and low frequency of the vibration. To overcome this problem, a combined low pass filter (LPF) and adaptive line enhancer (ALE) signal pre-conditioning method is used. Time-domain features such as root mean square (RMS), skewness and kurtosis are extracted from the output signal of the combined LPF and ALE method. The extracted features show accurate information about the incipient of fault as compared to extracted features from the original vibration signal. This information then triggers the prognostic algorithm to predict the remaining lifetime of the bearing. The algorithm used to determine the trend of the non-stationary data is auto-regressive integrated moving average (ARIMA).

Development of robotic hand integrated with SimMechanics 3D animation

In this study, a robotic hand is developed using low cost material to meet the demand for prosthetic hand in Indonesia. It has five actuators utilizing five micro servo motors. The robot consists of three joints for index, middle, ring and little respectively and two joints for the thumb. The robotic hand movement is triggered using electromyography (EMG) sensor. The sensors read the information of muscle activities and send them to the microcontroller to drive the robotic hand. By using the same signal for opening and closing fingers, the three pattern grip is embedded in the robotic grip pattern. The patterns are power grip, hook and OK sign. The 3D CAD model of robotic hand is exported into SimMechanics First Generation. The 3D animation in SimMechanics First Generation is augmented to visualize the 3D virtual hand. Based on the experimental result, the robotic hand successfully shows three patterns grip motion driven by EMG sensor. The present roboctic hand is the initial model of the prosthetic hand.

Development of a low cost anthropomorphic robotic hand driven by modified glove sensor and integrated with 3D animation

In this paper, a low cost anthropomorphic robotic hand is developed using low cost materials. The robotic hand has 6 joints and 6 actuators. User or operator gives the hand movement command by a modified glove sensor. The glove consists of six flex sensors placed on the fingers and wrist join that detect the bend of the fingers into a joint angle in each finger. 3D CAD model of robotic hand is exported into SimMechanics model using SimMechanics link to generate SimMechanics block diagram that can run in MATLAB/ Simulink environment. The model in SimMechanics is utilized as 3D animation hand. The relationship of the servo motor rotation angle among metacarpal phalangeal (MCP), proximal inter phalangeal (PIP) and distal inter phalangeal (DIP) joints will be presented. Finally, the performance of robotic hand is tested to grasp various objects and to perform specific motion augmented with 3D animation. The experiment results show the successful development of a low cost anthropomorphic robotic hand that can perform activities of daily living (ADLs).

Frequency Domain Analysis of Sensor Data for Event Classification in Real-Time Robot Assisted Deburring

Process monitoring using indirect methods relies on the usage of sensors. Using sensors to acquire vital process related information also presents itself with the problem of big data management and analysis. Due to uncertainty in the frequency of events occurring, a higher sampling rate is often used in real-time monitoring applications to increase the chances of capturing and understanding all possible events related to the process. Advanced signal processing methods are used to further decipher meaningful information from the acquired data. In this research work, power spectrum density (PSD) of sensor data acquired at sampling rates between 40–51.2 kHz was calculated and the corelation between PSD and completed number of cycles/passes is presented. Here, the progress in number of cycles/passes is the event this research work intends to classify and the algorithm used to compute PSD is Welch’s estimate method. A comparison between Welch’s estimate method and statistical methods is also discussed. A clear co-relation was observed using Welch’s estimate to classify the number of cycles/passes. The paper also succeeds in classifying vibration signal generated by the spindle from the vibration signal acquired during finishing process.

Electromyography (EMG) signal recognition using combined discrete wavelet transform based on Artificial Neural Network (ANN)

Rapid disability patients increasing over time and need a solution in the future. Hand amputation is one form of disability that common in Indonesian society. A possible solution would be necessary at the moment is the development of prosthetic hand that has the ability as a human hand. The development of neuroscience has now reached the stage of the bodys ability to use the signal as an input signal to operate a system. One of the applications of the science development is the use of electromyography (EMG) signals as an input to the control system to operate the prosthetic hand. This study is divided into two stages: a preliminary study and further research. Initial research focus in the process of EMG signal pattern recognition and advanced research focus in the development of a prototype prosthetic hand that is integrated with the controller system. Preliminary research indicates that the results of pattern recognition EMG signal using wavelet transform and Artificial Neural Network (ANN) classification has an accuracy rate of about 77.5 %. Based on these results, it can be concluded that the study results could be used as a signal input to program control of the prosthetic hand that will be developed in phase two.

A pattern recognition method for stage classification of Parkinson's disease utilizing voice features

This paper presents a pattern recognition method for multi-class classification of Parkinsons disease based on PCA, LDA and SVM. 22 voice features which are extracted and reduced using PCA and LDA. SVM is then used during the classification step. The classification accuracy between single features and PCA and LDA features are presented and the results show that the PCA features have greater accuracy than LDA features and the single features.

Speech control of robotic hand augmented with 3D animation using neural network

In this paper, speech control of robotic hand augmented with 3D animation system has been proposed and presented. Artificial neural network is employed for speech recognition method with tansig and softmax transfer function in hidden layer and output layer. Stream processing method is incorporated for processing the input signal in real time. Thirteen features in frequency domain and time domain that are commonly used in the EMG analysis are utilized in this system. To reduce the influence of noise, voice in noisy environment in the room is recorded as training data set. From the experimental results in offline speech recognition, ANN can recognize the voice command with very high accuracy. In the online real time speech recognition incorporating stream processing, the recognition accuracy decreased about 10%. The proposed speech control of robotic hand augmented with 3D animation is reliable enough with noisy environment.