• Journal of Sensor Science and Technology
• /
• v.18 no.6
• /
• pp.449-455
• /
• 2009

In this paper, we developed a digital gait analyzer using the triaxial accelerometer(TA). An approach for normal gait detection employing decay slope peak detection(DSPD) algorithm was presented. The TA was attached to the center of the waist of a subject. The subject walked a bare floor at 60, 92 and 120 steps/minute. We analyzed vertical axis acceleration signal for gait detection. At 60, 92, 120 steps/minute walking, detection accuracy of gait events were over 99 % accuracy.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.48 no.2
• /
• pp.127-137
• /
• 2011

We have developed a wearable device that can convert sensor data into real-time step counts and activity levels. Sensor data on gait were acquired using a triaxial accelerometer. A test was performed according to a test protocol for different walking speeds, e.g., slow walking, walking, fast walking, slow running, running, and fast running. Each test was carried out for 36 min on a treadmill with the participant wearing a portable gas analyzer (K4B2), an Actical device, and the device developed in this study. The signal vector magnitude (SVM) was used to process the X, Y, and Z values output by the triaxial accelerometer into one representative value. In addition, for accurate step-count detection, we used three algorithms: an heuristic algorithm (HA), the adaptive threshold algorithm (ATA), and the adaptive locking period algorithm (ALPA). A regression equation estimating the energy expenditure (EE) was derived by using data from the accelerometer and information on the participants. The recognition rate of our algorithm was 97.34%, and the performance of the activity conversion algorithm was better than that of the Actical device by 1.61%.

• Journal of Internet Computing and Services
• /
• v.12 no.3
• /
• pp.17-26
• /
• 2011

We have developed a wearable device that can convert sensor data into real-time step counts. Sensor data on gait were acquired using a triaxial accelerometer. A test was performed according to a test protocol for different walking speeds, e.g., slow walking, walking, fast walking, slow running, running, and fast running. Each test was carried out for 36 min on a treadmill with the participant wearing an Actical device, and the device developed in this study. The signal vector magnitude (SVM) was used to process the X, Y, and Z values output by the triaxial accelerometer into one representative value. In addition, for accurate step-count detection, we used three algorithms: an heuristic algorithm (HA), the adaptive threshold algorithm (ATA), and the adaptive locking period algorithm (ALPA). The recognition rate of our algorithm was 97.34% better than that of the Actical device(91.74%) by 5.6%.

• Journal of Institute of Control, Robotics and Systems
• /
• v.20 no.2
• /
• pp.149-156
• /
• 2014

A magnetic compass module must be calibrated accurately before use. Moreover, the calibration process must be performed taking into account any magnetic dip if the magnetic compass module has tilt angles. For this, a calibration method for a magnetic compass module is explained. Tilt error of the magnetic compass module is compensated using a biaxial accelerometer generally. The accelerometer error causes a tilt angle calculation error that gives rise to an azimuth calculation error. For error property analysis, error equations are derived and simulations are performed. In the simulation results, the accuracy of derived error equations is verified. If a biaxial magnetic compass module is used instead of a triaxial one, the magnetic dip and z-axis magnetic compass data must be estimated for tilt compensation. Lastly, estimation equations for the magnetic dip and z-axis magnetic compass data are derived, and the performance of the equations is verified based on a simulation.

• Korean Journal of Applied Biomechanics
• /
• v.18 no.4
• /
• pp.109-114
• /
• 2008

The purpose of this study was to estimate the energy expenditure simply and practically during physical activities. The physical activity is quantified by the integration of the accelerometer signals obtained from the triaxial accelerometer attached at the waist level of the human body. To find a relationship between energy expenditure and accelerometer data, 6 male and 5 female subjects walked and ran on the treadmill with speeds of 1.5, 3.0, 4.5, 6.0, 6.5, 7.0, and 8.5 km/hr. Each subject performed walking at the speed lower than 6.0 km/hr and running at the speed higher than 6.5 km/hr. Actual energy expenditure was determined by a continuous direct gas analyzer. Two predictive equations of walking and running mode for energy expenditure which includes gender, body mass index(BMI) and data from accelerometer were developed using multiple regression analysis. The correlation coefficients and coefficients of determination between the estimated and measured energy expenditure were R=0.936, R2=0.876 and R=0.881, R2=0.776 in walking and running mode, respectively. For further study, experiments on a larger scale of test subjects are essential for acquiring more reliable results.

• Smart Structures and Systems
• /
• v.21 no.4
• /
• pp.407-420
• /
• 2018

In structural health monitoring, it is meaningful to comprehensively utilize accelerometers and strain gauges to obtain the modal information of a structure. In this paper, a modal estimation theory is proposed, in which the displacement modes of the locations without accelerometers can be estimated by the strain modes of selected strain gauge measurements. A two-stage sensor placement method, in which strain gauges are placed together with triaxial accelerometers to obtain more structural displacement mode information, is proposed. In stage one, the initial accelerometer locations are determined through the combined use of the modal assurance criterion and the redundancy information. Due to various practical factors, however, accelerometers cannot be placed at some of the initial accelerometer locations; the displacement mode information of these locations are still in need and the locations without accelerometers are defined as estimated locations. In stage two, the displacement modes of the estimated locations are estimated based on the strain modes of the strain gauge locations, and the quality of the estimation is seen as a criterion to guide the selection of the strain gauge locations. Instead of simply placing a strain gauge at the midpoint of each beam element, the influence of different candidate strain gauge positions on the estimation of displacement modes is also studied. Finally, the modal assurance criterion is utilized to evaluate the performance of the obtained multitype sensor placement. A bridge benchmark structure is used for a numerical investigation to demonstrate the effectiveness of the proposed multitype sensor placement method.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.28-28
• /
• 2000

Low-cost solid-state inertial sensors of three rate Gyroscopes and a triaxial Accelerometer are evaluated in static and dynamic environments. As a interim result, error models of each inertial sensors are generated. Model parameters with respect to temperature are acquired in static environment. These error models are included in an Extended Kalman Filter(EKF) to compensate bias error due to temperature variation. Experimental results in dynamic environment are included to show the validity of the each error model and the performance improvement of a compensated low cost inertial sensors for a navigational application

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.61 no.6
• /
• pp.851-858
• /
• 2012

In this paper, we design a seismic data acquisition system(SDAS) and implement it. This system is essential for development of a noble local earthquake disaster preventing system in population center. In the system, we choose a proper MEMS-type triaxial accelerometer as a sensor, and FPGA and ARM processor are used for implementing the system. In the SDAS, each module is realized by Verilog HDL and C Language. We carry out the ModelSim simulation to verify the performances of important modules. The simulation results show that the FPGA-based data acquisition module can guarantee an accurate time-synchronization for the measured data from each axis sensor. Moreover, the FPGA-ARM based embedded technology in system hardware design can reduce the system cost by the integration of data logger, communication sever, and facility control system. To evaluate the data acquisition performance of the SDAS, we perform experiments for real seismic signals with the exciter. Performances comparison between the acquired data of the SDAS and the reference sensor shows that the data acquisition performance of the SDAS is valid.

• ETRI Journal
• /
• v.45 no.6
• /
• pp.1079-1089
• /
• 2023

Stroke is one of the leading causes of long-term disability worldwide, placing huge burdens on individuals and society. Further, automatic human activity recognition is a challenging task that is vital to the future of healthcare and physical therapy. Using a baseline long short-term memory recurrent neural network, this study provides a novel dataset of stretching, upward stretching, flinging motions, hand-to-mouth movements, swiping gestures, and pouring motions for improved model training and testing of stroke-affected patients. A MATLAB application is used to output textual and audible prediction results. A wearable sensor with a triaxial accelerometer is used to collect preprocessed real-time data. The model is trained with features extracted from the actual patient to recognize new actions, and the recognition accuracy provided by multiple datasets is compared based on the same baseline model. When training and testing using the new dataset, the baseline model shows recognition accuracy that is 11% higher than the Activity Daily Living dataset, 22% higher than the Activity Recognition Single Chest-Mounted Accelerometer dataset, and 10% higher than another real-world dataset.

• Korean Journal of Applied Biomechanics
• /
• v.30 no.2
• /
• pp.145-154
• /
• 2020

Objective: The purpose of this study was to determine the relationship between impact and shear peak force, and tibia-accelerometer variables during running. Method: Twenty-five male heel strike runners (mean age: 23.5±3.6 yrs, mean height: 176.3±3.3 m/s, mean mass: 71.8±9.7 kg) were recruited in this study. The peak impact and anteroposterior shear forces during treadmill running (Bertec, USA) were collected, and impact shock variables were computed by using a triaxial accelerometer (Noraxon, USA). One-way ANOVA was used to test the influence of the running speed on the parameters. Pearson's partial correlation was used to investigate the relationship between the peak impact and shear force, and accelerometer variables. Results: The running speed affected the peak impact and posterior shear force, time, slope, and peak vertical and resultant tibial acceleration, slope at heel contact. Significant correlations were noticed between the peak impact force and peak vertical and resultant tibia acceleration, and between peak impact average slope and peak vertical and resultant tibia acceleration average slope, and between posterior peak (FyP) and peak vertical tibia acceleration, and between posterior peak instantaneous slop and peak vertical tibial acceleration during running at 3 m/s. However, it was observed that correlations between peak impact average slope and peak vertical tibia acceleration average slope, between posterior peak time and peak vertical and resultant tibia acceleration time, between posterior peak instantaneous slope and peak vertical tibial acceleration instantaneous slope during running at 4 m/s. Conclusion: Careful analysis is required when investigating the linear relationship between the impact and shear force, and tibia accelerometer components during relatively fast running speed.