These days, the accuracy of the measured tilt angle is extremely important and it is always a challenge especially for robotics application. MEMS (micro-electro-mechanical system) gyroscope and accelerometer has been widely used to measure the tilt angle along the axis. However, various studies highlighted that MEMS gyroscope and accelerometer has weakness to gain accuracy of orientation. In order to cater the low accuracy, data from the MEMS gyroscope and accelerometer require a filter where the data from both sensors will be combined by using the complementary filter. Thus, digital complementary filter is developed in such a way that the strength of one sensor will be used to overcome the weaknesses of the other sensor which is complementary to each other. The filtering will be demonstrated using MEMS gyroscope and accelerometer in different position throughout the measurement for analysis and angle estimation via complementary filter. From the simulation result, it is evident that the complementary filter method is achieved with responsive and accurate angle of 0 to 90 degrees with same as actual position at factor value of 0.97. In addition, it is not sensitive to horizontal accelerations or to gyroscope drift with microcontroller-friendly filtering system.

Low cost MEMS gyroscope and accelerometer implementation without Kalman Filter for angle estimation

Sensor Fusion (Complementary and Kalman filters) and Moving Average filter are implemented on an Arduino microcontroller based data acquisition of rotation degree from Inertial Measurement Unit (IMU) sensor for stabilized platform application. Stable platform prototype is designed to have two degrees of freedom, roll and pitch rotation. Output data from gyro and accelerometer were combined to take the advantage of each sensor. Digital filter algorithm was embedded into microcontroller programming. This paper analyzes overshoot percentage, rise time, and data series smoothness of Sensor Fusion (Complementary and Kalman filter) and Moving Average filter response in IMU data acquisition from step input of 20-degreerotation. Moving-average filter resulted in the smallest overshoot percentage of 0% but produce theslowest responsewith 0.42 second rise time. Overall best results are obtained using Complementary filter (alpha value 0.95) by overshoot percentage of 14.17%, 0.24 second rise time, and 0.18 data series smoothness.

Embedded sensor fusion and moving-average filter for Inertial Measurement Unit (IMU) on the microcontroller-based stabilized platform

UASs (Unmanned Aerial Systems) have become increasingly popular for both military and civil applications, thanks to improved batteries, motors, and rapidly falling prices. They are mostly implemented during various operation tasks, as search and rescue, disaster assessment, urban traffic monitoring, 3D mapping, etc., that would be risky or impossible for a human to perform. DAA (Detect and Avoid) is a new technology necessary for UASs to safely avoid obstacles or other aircrafts. In this paper low-cost sensors, such as a 10-DOF (Degree Of Freedom) MEMS (Micro Electro-Mechanical System) IMU (Inertial Measurement Unit) and a LIDAR (Light Detection and Ranging), synchronized at 10-Hz measurement rate, were installed on a small unmanned rotorcraft to estimate the position of the platform and its distance from an obstacle or a landing field. To correct the IMU data from systematics errors (bias) and measurement noise, and to derive estimated positions from accelerometer data, Kalman filtering was implemented. The optimal estimation algorithm, developed on an onboard microcontroller (Arduino Mega 2560), allows for low-cost hardware implementations of multiple sensors for use in aerospace applications.

UAS for positioning and field mapping using LIDAR and IMU sensors data: Kalman filtering and integration

This work presents a multi-sensor fusion approach for an electronic navigation aid for the blind and visually impaired persons. This approach proposes to intelligently fuse the surrounding information senses via ultrasonic sensors and vision sensors. The intelligent component of the prototype serves in several facets including object detection and recognition. Extended Kalman filter is used to fuse the data emerging from homogeneous sensors and rule-based fusion is used to fuse the data from heterogeneous sensors. Feedback is provided via tactile and audio feedback. Critical obstacles of blind navigation like staircases are recognized by the Hough line detection in image processing. Rotations, which occur due to the body movements in the camera, correct using the fusion of data obtain by the inertial measurement unit which is connected to the camera. The results of the evaluations proved that the use of fusion of multiple homogeneous sensors improve the detection of a particular obstacle and fusion of vision and ultrasonic sensors improve the object detection identification of the obstacles. The current status of the work and the future developments are presented in this paper.

http://ieeexplore.ieee.org/iel7/7939934/7946517/07946537.pdf

Sensor fusion for visually impaired navigation in constrained spaces

In this paper, the dynamic model of a pneumatic system which comprises a pneumatic actuator and a proportional directional electrical valve is derived. Accordingly, an experimental method is proposed to the end of modeling unknown parameters which are unmeasurable directly from the available pneumatic system. Friction force and viscous coefficient in dynamic equation of the actuator and mass flow rate in dynamic equations of electrical proportional valve are unknown parameters which are subject to be identified. The dynamic behavior of the unknown parameters are modeled by using exponential, polynomial and fourier series transform functions. Pressure of the pneumatic actuator and supply pressure are measured by the means of pressure sensors. Displacement encoder and IMU-GY80 sensor are employed to obtain the value of position and acceleration of the cylinder, respectively. Viscous coefficient is calculated in such a way that pneumatic force is not exerted which is based on the so-called Free-Fall test. By solving a second order differential equation of cylinder dynamics, viscous coefficient and Coulomb friction are obtained. Finally, the identified model is validated by comparing the state space variables of the pneumatic system measured by experimental tests and the identified model. From the experimental results, it has been revealed that the mean absolute percentage error for estimating the state variables are 19.94 (position), 26.78 (velocity), 0.78 (pressure 1) and 0.21 (pressure 2). The latter certifies the exactitude and performance of the identified dynamic model.

An experimental study on friction identification of a pneumatic actuator and dynamic modeling of a proportional valve

/pdf/low-cost-imu-implementation-via-sensor-fusion-algorithms-in-3jcerflz7t.pdf

Brandon McCarron

Papers

Low-Cost IMU Implementation via Sensor Fusion Algorithms in the Arduino Environment