Research And Implementation Of Gait Measurement Methods Based On Inertial Sensors

Walking is one of the most common human physical activities. Evaluation of time and distance parameters during walking is helpful in assessing abnormal gait, to quantify improvement resulting from interventions, or to predict subsequent events such as falls. An important requisite for a portable system is the ability to provide kinematic information about gait. Gait analysis system based video and force plate are effective tools to provide accuracy gait parameters. Unfortunately, these measures are not often used in routine functional assessments of gait because specialist training, a dedicated gait laboratory, and expensive motion analysis equipment is required. With the fast development of technology, gait analysis based on wearable sensors is promoted and attracts a lot of researchers. Researchers paid more attention to the gait analysis based on accelerometers and gyroscopes. Gait data during walking were collected using a gait data acquisition system based on inertial measurement unit and processed through corresponding algorithms, and objective results were obtained in the end.In this paper, two inertial measurement sensors are placed on outside of each shoe of subjects to collect gait data, and the data are imported to computer. Next, the gait data are segmented into individual gait cycle using proposed algorithms and the time-domain parameters are calculated. Then quaternion-based Extend Kalman filter is used to estimate the orientation of the foot, and the acceleration is transform into local reference system using the orientation cosine matrix. In the last, the converted acceleration is integrated to obtain velocity, and ZUPT is applied to eliminate accumulative error, and then the corrected velocity is integrated to compute space parameters, such as stride length, maximum foot clearance. Results show that the proposed algorithm can detect each gait cycle and the stride length error is less than 2%, which imply the feasibility and effectiveness of this proposed algorithm. From the result we can see that this gait analysis is effective and can be applied into different field.