Study of navigation sensors for agricultural vehicle

I . Introduction In the 21st century, the world population exceeded over 6 billion. It is supposed that the world food production will be short supply in the near future. The increase of food production has been required in the world. However, the self-sufficiency rate of Japan keeps on decreasing because of decrease in number of farmers, decrease in farm field and utilization rate of cultivated land. Therefore, to increase the self-sufficiency rate of food, the drastic improvement of productivity is required. Automatic navigation of agricultural vehicle and the precision farming are expected as the solutions of this problem. The required sensor elements of the agricultural vehicle automation and the precision farming are: 1) Position 2) Heading angle 3) Inclination angle Currently, there is the problem that the cost of navigation sensors which are able to apply to the agricultural vehicle is too expensave to be commercialized. The objective of this study is to develop the low cost navigation sensor module which can provide the position, the heading angle and the inclinations for the agricultural vehicle. First, the navigation system based on the geomagnetic direction sensor (GDS) was developed. This system was composed of the heading sensors such as a GDS, a fiber optical gyroscope and a vibratory gyroscope. Second, the measurement methods of the position, heading angle and inclinations by the navigation system based on the global 265 positioning system (GPS) were proposed. Finally, the navigation sensor unit composed of vibratory gyroscopes and inclinometers and the GPS was developed. II. Enhancement of Heading Accuracy of Geomagnetic Direction Sensor The precise estimation method of the vehicle heading angle by a geomagnetic direction sensor (GDS) was proposed. The GDS is frequently used as a navigation sensor. In particular, it is suitable for an off road vehicle, because it can provide the absolute direction from north-pole. But, the low signal to noise ratio of the GDS is a problem as a navigation sensor. A fiber optical gyroscope (FOG) was utilizing to improve this low Signal to noise ratio of GDS measurement. As an integration technique of the GDS and the FOG, a Kalman filter was applied in this study. To evaluate the performance of these sensors for utilizing on an automatic guidedance, and to confirm the effect of sensor fusion technique, the automatic guidance test was conducted in the field. The developed fusion technique was able to control the vehicle automatically with higher accuracy compared to that using single navigation sensor. Additionally, I proposed the dynamic compensation method against both a geomagnetic warp and a gyro-drift by constructing the adaptive guidance system. From the computer simulation, the accuracy of sensor fusion guidance system improved about 70% compared with the GDS alone. To evaluate robustness of the developed system, a field test using the developed guidance system was carried out. From the field test, the accuracy and robustness of the guidance system were improved by utilizing the proposed sensor fusion method using the GDS and the gyroscope. III. Automatic Navigation by Vibratory Gyroscope and Geomagnetic Direction Sensor In this chapter, the precise estimation method of the vehicle heading by a gyroscope and a GDS was proposed. First, the noise of the GDS and the inclinometer were eliminateded by the adaptive line enhancer (ALE). As the result of inclination correction of the GDS by ALE, the corrected GDS output was consistent with the actual orientation. Second, the drift error of the gyroscope was estimated by a least square method. Third, these sensors output were integrated to obtain the accurate vehicle orientation. Finally, to evaluate the accuracy of the proposed estimation method of orientation, the automatic navigation test was carried out in the outdoor field. As the result of the automatic navigation test, it was confirmed that proposed estimation method had the ability to navigate the vehicle to the desired straight path with high accuracy. IV. Enhancement of the Heading Angle and the Position Accuracy by Navigation System based on Global Positioning System In this chapter, the navigation system based on the global positioning system (GPS) was proposed. The measurement methods of the position and the heading angle by the GPS, geomagnetic direction sensor and vibattery gyroscope were developed. First, the real time compensation method of geomagnetic distortion effected to the GPS output was invented by using a GPS. At the same time, the coincidence method of the GPS and the GDS coordinate system was proposed. To evaluate the accuracy of the developed fusion method, the computer simulocation using the RTK-GPS when vehicle traveled to four directions (to north, south, east, and west) was conducted. From the result of the simulation, the maximum error of conventional method was up to 6.5 deg . On the other hand, the proposed method could reduce to the maximum error of 1.3 deg . The result of heading estimation in 90 deg turning navigation using the GPS indicated that proposed method could compensate the geomagnetic distortion and difference of the GDS/GPS coordinate systems in real time. Second, the precise heading estimation method using the GPS and the vibratory gyroscope was proposed. The proposed method utilizes the low-grade GPS whose sampling rate is 1Hz and accuracy is 1m. From the result of simulation test, heading angle obtained by GPS only was very noisy and the R.M.S. error was up to 2.36 deg . The heading angle obtained by the vibratory gyroscope also indicated large drift error. On the other hand, the proposed method could estimate the heading angle with 0.51 deg R. M.S. error and -0.74 deg integrated error. Third, the correction method of the GPS position error caused by vehicle inclinations was proposed. To evaluate a performance of the developed correction method, the travel test by giving roll and pitch was conducted. The average of lateral error indicated about 17.1cm, when the vehicle traveled without inclination correction, while the average error was reduced up to about 2.1cm by introducing the inclination correction. In addition, to investigate the performance under higher velocity, the traveling test in which running speed exceeding 4.0 m/s was also carried out. Without the inclination correction, the R.M.S. error in the lateral direction of the vehicle reached 10.4 cm. On the other hand, when the developed correcttion method was utilized, the R.M.S. error decreased to about 3.8cm. From this result, the importance of inclination measurement and correction of GPS position were certified. V. Development of Navigation Sensor Unit for the Agricultural Vehicle In this chapter, low cost navigation sensor unit composed of three vibratory gyroscopes and two inclinometers and the DGPS was developed. The sensor can provide the position corrected by inclination, heading angle and roll/pitch inclinations. And the measured position was interpolated by the dead reckoning, because the sampling rate of the DGPS was lHz. To measure the accuracy of developed sensor unit, the field test was conducted on the flat field, gentle slope and bumpy road. On the flat field, the developed sensor unit could estimate the roll angle with 0.24 deg R.M. S. error, the pitch angle with 0.37 deg , the heading angle with 0.63 deg and the position with 6.5cm. On the gentle slope field, it was confirmed that developed sensor unit could detect the low frequency angulations which inclined about 10 deg . The measurement error of the roll angle was 0.23 deg , pitch angle was 0.28 deg , heading angle was 0.64 deg and the position was 3.0cm. The bumpy road was consist of random inclinations about +-5 deg and the vehicle wobbled very roughly. The measurement error of the roll angle was 0.43 deg , pitch angle was 0.61 deg , heading angle was 0.59 deg and the position was 3.7cm. In spite of such rugged condition, the developed sensor unit could detect the inclination, heading angle and position at high frequency response. Judging from a viewpoint of a function, accuracy and cost, the developed sensor unit achieved the target specification as a navigation system and it is possible to utilize to the agricultural vehicle automation and the precision farming effectively.