An adaptive hybrid data fusion based identification of skeletal muscle force with ANFIS and smoothing spline curve fitting (original) (raw)
Related papers
A hybrid adaptive data fusion with linear and nonlinear models for skeletal muscle force estimation
2010 5th Cairo International Biomedical Engineering Conference, 2010
Skeletal muscle force can be estimated using surface electromyographic (sEMG) signals. Usually, the sEMG location for the sensors is near the respective muscle motor unit points. Skeletal muscles generate a temporal and spatial distributed EMG signal, which causes cross talk between different sEMG signal sensors. In this paper, an array of three sEMG sensors is used to capture the information of muscle dynamics in terms of sEMG signals and generated muscle force. The recorded sEMG signals are filtered utilizing optimized nonlinear Half-Gaussian Bayesian filter, and a Chebyshev type-II filter prepares the muscle force signal. The filter optimization is accomplished using Genetic Algorithm (GA). Multi nonlinear Auto Regressive eXogenous (ARX) and Wiener-Hammerstein models with different nonlinearity estimators/classes are obtained using system identification (SI) for three sets of sensor data. The outputs of these models are fused with a probabilistic Kullback Information Criterion (KIC) for model selection and an adaptive probability of KIC. First, the outputs are fused for the same sensor and for different models and then the final outputs from each sensor. The final fusion based output of three sensors provides good skeletal muscle force estimates.
2013 6th International Symposium on Resilient Control Systems (ISRCS), 2013
In this work, an array of three surface Electrography (sEMG) sensors are used to acquired muscle extension and contraction signals for 18 healthy test subjects. The skeletal muscle force is estimated using the acquired sEMG signals and a Non-linear Wiener Hammerstein model, relating the two signals in a dynamic fashion. The model is obtained from using System Identification (SI) algorithm. The obtained force models for each sensor are fused using a proposed fuzzy logic concept with the intent to improve the force estimation accuracy and resilience to sensor failure or misalignment. For the fuzzy logic inference system, the sEMG entropy, the relative error, and the correlation of the force signals are considered for defining the membership functions. The proposed fusion algorithm yields an average of 92.49% correlation between the actual force and the overall estimated force output. In addition, the proposed fusionbased approach is implemented on a test platform. Experiments indicate an improvement in finger/hand force estimation.
Adaptive multi sensor based nonlinear identification of skeletal muscle force
2010
Skeletal muscle force and surface electromyographic (sEMG) signals are closely related. Hence, the later can be used for the force estimation. Usually, the location for the sEMG sensors is near the respective muscle motor unit points. EMG signals generated by skeletal muscles are temporal and spatially distributed which results in cross talk that is recorded by different sEMG sensors. This research focuses on modeling muscle dynamics in terms of sEMG signals and the generated muscle force. Here, an array of three sEMG sensors is used to capture the information of the muscle dynamics in terms of sEMG signals and generated muscle force. Optimized nonlinear Half-Gaussian Bayesian filters and a Chebyshev type-II filter are used for the filtration of the sEMG signals and the muscle force signal, respectively. A Genetic Algorithm is used for the optimization of the filter parameters. sEMG and skeletal muscle force is modeled using multi nonlinear Auto Regressive eXogenous (ARX) and Wiener-Hammerstein models with different nonlinearity estimators/classes using System Identification (SI) for three sets of sensor data. An adaptive probabilistic Kullback Information Criterion (KIC) for model selection is applied to obtain the fusion based skeletal muscle force for each sensor first and then for the final outputs from each sensor. The approach yields good skeletal muscle force estimates.
2010
Skeletal muscle force can be estimated using surface electromyographic (sEMG) signals. Usually, the sEMG location for the sensors is near the respective muscle motor unit points. Skeletal muscles generate a temporal and spatial distributed EMG signal, which causes cross talk between different sEMG signal sensors. In this paper, an array of three sEMG sensors is used to capture the information of muscle dynamics in terms of sEMG signals and generated muscle force. The recorded sEMG signals are filtered utilizing optimized nonlinear Half-Gaussian Bayesian filter, and a Chebyshev type-II filter prepares the muscle force signal. The filter optimization is accomplished using Genetic Algorithm (GA). Multi nonlinear Auto Regressive eXogenous (ARX) and Wiener-Hammerstein models with different nonlinearity estimators/classes are obtained using system identification (SI) for three sets of sensor data. The outputs of these models are fused with a probabilistic Kullback Information Criterion (KIC) for model selection and an adaptive probability of KIC. First, the outputs are fused for the same sensor and for different models and then the final outputs from each sensor. The final fusion based output of three sensors provides good skeletal muscle force estimates.
Adaptive finger angle estimation from sEMG data with multiple linear and nonlinear model data fusion
2011
This paper presents a novel approach to control the motion of a smart prosthesis using surface electromyographic (sEMG) signals. Currently, all sEMG based prosthetic hands are controlled based on preprogrammed motion sets, which are initiated when some threshold value of the measured sEMG signal is reached. In this paper, we present an approach that utilizes System Identification (SI) in order to obtain a dynamic finger angle model. Such a model allows for instantaneous control for the finger motions. The algorithm presented relays on an array of nine sEMG sensors. The sEMG and angle data is filtered using a nonlinear Teager-Kaiser Energy (TKE) operator based nonlinear spatial filter and a Chebyshev type-II filter respectively. The filtered signals are smoothed using a smoothing spline curve fitting. The smoothed sEMG data is used as input and the respective smoothed finger angle data is used as output for a system identification routine to obtain multiple linear and nonlinear models. To achieve better estimates of the finger angles, an adaptive probabilistic Kullback Information Criterion (KIC) for model selection based data fusion algorithm is applied to the linear and nonlinear model's outputs. Final fusion based output of this approach results in improved estimates of finger angles.
Precise and Accurate Multifunctional Prosthesis Control Based on Fuzzy Logic Techniques
2011
This paper proposes a fuzzy approach to classify single-channel surface electromyography (SEMG) signals for multifunctional prosthesis control. In this approach three variables root mean square, standard deviation & variance were selected for the analysis. These three parameters gave best results for discriminating hand movements from time domain analysis using SEMG. As the frequency range of SEMG signal is specified to be from 0 Hz to 400 Hz, therefore the analysis was divided into three sections of frequencies as: low(<60Hz), high pass (>250Hz) and band pass (70-250Hz). This is done to establish the frequency change in the three regions for specified selected output. The three parameters were used as input variables to fuzzy logic controller for discrimination of the hand movements i.e. whether the hand is closed or open. SEMG signal was taken from below elbow position using bipolar electrodes as this part was found to be more active during expansion and contraction of muscles. Out of the three parameters standard deviation gave best results for discriminating hand movements (opening and closing). Out of three filters, low pass filter gave the best results for discriminating opening and closing movements. Further, to develop precise control of the grip prosthetic, the work is extended to multilevel SEMG control. A gripexerciser was used to calibrate the executed force and SEMG signal. The result was a fuzzy logic controller for the accurate grip force.
Bio Systems, 2011
The aim of this paper is to create a model for mapping the surface electromyogram (EMG) signals to the force that generated by human arm muscles. Because the parameters of each person's muscle are individual, the model of the muscle must have two characteristics: (1) The model must be adjustable for each subject. (2) The relationship between the input and output of model must be affected by the force-length and the force-velocity behaviors are proven through Hill's experiments. Hill's model is a kinematic mechanistic model with three elements, i.e. one contractile component and two nonlinear spring elements. In this research, fuzzy systems are applied to improve the muscle model. The advantages of using fuzzy system are as follows: they are robust to noise, they prove an adjustable nonlinear mapping, and are able to model the uncertainties of the muscle. Three fuzzy coefficients have been added to the relationships of force-length (active and passive) and force-velocity existing in Hill's model. Then, a genetic algorithm (GA) has been used as a biological search method that can adjust the parameters of the model in order to achieve the optimal possible fit. Finally, the accuracy of the fuzzy genetic implementation Hill-based muscle model (FGIHM) is invested as following: the FGIHM results have 12.4% RMS error (in worse case) in comparison to the experimental data recorded from three healthy male subjects. Moreover, the FGIHM active force-length relationship which is the key characteristics of muscles has been compared to virtual muscle (VM) and Zajac muscle model. The sensitivity of the FGIHM has been evaluated by adding a white noise with zero mean to the input and FGIHM has proved to have lower sensitivity to input noise than the traditional Hill's muscle model.
Fuzzy EMG classification for prosthesis control
… IEEE Transactions on, 2000
This paper proposes a fuzzy approach to classify single-site electromyograph (EMG) signals for multifunctional prosthesis control. While the classification problem is the focus of this paper, the ultimate goal is to improve myoelectric system control performance, and classification is an essential step in the control. Time segmented features are fed to a fuzzy system for training and classification. In order to obtain acceptable training speed and realistic fuzzy system structure, these features are clustered without supervision using the Basic Isodata algorithm at the beginning of the training phase, and the clustering results are used in initializing the fuzzy system parameters. Afterwards, fuzzy rules in the system are trained with the back-propagation algorithm. The fuzzy approach was compared with an artificial neural network (ANN) method on four subjects, and very similar classification results were obtained. It is superior to the latter in at least three points: slightly higher recognition rate; insensitivity to overtraining; and consistent outputs demonstrating higher reliability. Some potential advantages of the fuzzy approach over the ANN approach are also discussed.