19^th International CODATA Conference
Category: Infoscience

Fuzzy Logic Expert Rule-based Multi-Sensor Data Fusion for Land Vehicle Attitude Estimation

Mr. Jau-Hsiung Wang (wangjh@ucalgary.ca) and Dr. Yang Gao
Department of Geomatics Engineering, University of Calgary, Canada, www.geomatics.ucalgary.ca


GPS-based land vehicle navigation system can provide a cost-effect navigation solution with acceptable accuracy. But due to the signal fading in urban area, it requires aids from other enabling sensors. A popular solution to this problem is to integrate GPS with complementary navigation sensors such as Inertial Navigation System (INS), which is based on dead-reckoning methodology to obtain the position state. Based on INS mechanization, the error of velocity and position estimations will be mainly governed by the accuracy of estimated attitude. Therefore, a fine estimation and an effective correction of attitude error are very important for using INS to successfully assist GPS-based navigation system. As the advent of MEMS (Micro-Electro-Mechanical System) technology, the low-cost and small-size accelerometers and gyroscopes are now available and adoptable for vehicular navigation. But the trade-off is the poorer performance of relatively high instrument bias, drift and noise. Unlike MEMS accelerometers with more stable performance, MEMS gyroscopes currently is still limited due to high gyro drifts and complex procedures of initial alignments for attitude determination. In contrast to gyro, a magnetometer is able to provide absolute heading information relative to the magnetic north without time-accumulated errors and complex initialization processes. Even though the magnetometer measurement is still distorted by local magnetic field and external interference, the errors not accumulating with time provide some complementary characteristics to gyroscopes. For tilt sensing, when vehicle is static, the accelerometer measurement containing gravity field only can directly derive pitch and roll angle without time-accumulated errors. This paper presents a new fuzzy logic expert rule-based multi-sensor data fusion to estimate vehicle attitude based on the complementary characteristics of the low-cost GPS, MEMS inertial sensors and compass. Since the performance and characteristics of each sensor are related to vehicle dynamics, the correlation between raw measurements and vehicle dynamics is first investigated. Then, a fuzzy logic rules-based decision-making system is developed for the classification of vehicle dynamics. The knowledge of specific physical shortcomings and strengths of each sensor modality in the corresponding status of vehicle motion will be used as a teacher or an expert to design the fusion rules for weighting and combing the attitude estimations from each sensor. Field test of vehicle runs on several routes with different ruggedness will be performed to examine the accuracy of vehicle attitude estimated by the proposed system. The performance improvement of the proposed system comparing to the use of stand-alone MEMS inertial sensors would also be discussed.