جستجوی مقالات مرتبط با کلیدواژه « kalman filter » در نشریات گروه « مکانیزاسیون کشاورزی »

Nowadays, the growth of in-vehicle and off-road vehicle technology is significant and the need to develop its related infrastructure to maximize their capacity for land, sea and air transportation is essential and unavoidable. It is time consuming and costly to monitoring the environment in water, earth and air conditions for a safe operation, without the use of robots. With automatic monitoring, operations can be performed with the least error and 24 hours a day. So the purpose of this research is to develop and evaluate a composite algorithm for navigating an off-road vehicle (Surface Vehicle) and compare it with the results obtained from computer simulations, to check the accuracy of this algorithm. This robot boat is designed and developed for hydrographic construction could navigate and perform hydrographic operations around the clock and fully autonomous without any supervision. Comparison between experimental and simulation results showed that the simulated algorithm had acceptable accuracy and could navigate experimentally. The Standard Deviation (SD) for practical test was below 0.5 m.

The most widely used machine for precision sowing of cucumber and sorghum seeds is vacuum type. The performance and distribution accuracy of the vacuum type metering unit is of great importance. Therefore, the experiments at three pressures of 25, 35, 45 mbar and two level of forward speed 3 to 4.5 km/h and 6 to 8.5 km/h were conducted and multiple planting, feeding quality and miss planting were measured. In the meantime, a motion model is constructed using Kalman filter to track and draw seed trajectories. Based on the analysis of seeds falling trajectories, it was found that the there was a close relation between of seeds falling trajectories and seeding uniformity. The optimum levels of vacuum pressure and forward speed for precision seeding were found to be 35 mbar, 3 to 4.5 km/h, and 35 mbar, 3 to 4.5 km/h for cucumber and sorghum seeds, respectively. Finally, it was perceived that with increasing forward speed at inappropriate vacuum pressure, percentage of deviated seeds from straight movement was increased, which it decreased seed distribution uniformity.

In this research، four different positioning methods were compared in order to evaluate their accuracy، using a remotely controlled robot on a specific route. These methods included: using a single GPS module، combining the data from three GPS modules، using an Inertial Measurement Unit (IMU)، and GPS/IMU data fusion. The comparison of these four methods showed that GPS/IMU data fusion along with a Kalman filter was the most precise method، having a root mean square error of 23. 4cm. Integrating the data acquired simultaneously from three GPS modules with fixed and equally spaced position and far enough from each other، had a root mean square error of 31. 3cm was the second most precise method.. Also analysis of the IMU data showed that due to cumulative errors، it was not a suitable method using a single IMU for positioning.