Search Results

Computer algorithms were developed for generating the guidance parameters necessary to steer an agricultural tractor. A variety of field operations were considered in order that the guidance program be suited for general applications including travel in curved rows and following a single edge. Testing of the guidance algorithm was performed in the laboratory using simulated and videotaped images of rowcrops and tilled soil. From the images, yaw angle change of the tractor, direction value and offset error were computed. Prediction of the direction value and offset error compared well to measured values. Accuracy of the direction value was within +/- 0.5 degrees while the offset error was within +/- 0.05 meters. Good performance was observed for straight and curved rows as well as following a single edge.

A heuristic line detection algorithm is described for computing guidance information from row crop images. The technique processes binary images representing crop rows against a soil background. Points along the centers of crop rows are enhanced using a modified run-length encoding procedure. The properties of lines in images can be improved by filtering based on characteristics of the object run-length. A clustering algorithm was used to aggregate pixels that fall on the same crop row. The technique was compared with the Hough transform, a common line detection technique in image processing. Both procedures accurately represented lines measured manually in a set of images representing a range of expected field conditions.

A vision based tractor guidance system was designed and developed using distributed control techniques. The components and their interaction with each other are discussed. A geometric steering model was used to predict offset error based on heading error. Simulated straight crop rows were developed to test the system. Results show that the overall performance of the steering correction system was good for sufficient initial tractor orientation. The system proved to be unstable for large errors and that the geometric model incorporated was not capable of handling large heading and offset errors.

The feasibility of using exhaust temperatures and injector nozzle needle lift duration to predict power output and fuel consumption of a diesel engine was investigated using an instrumented John Deere 4440 tractor. Using data obtained during a series of PTO dynamometer experiments, regression models were determined correlating the needle lift duration, exhaust temperatures, and engine speed with the parameters of interest. The models were subsequently tested under normal tractor operating conditions on a commercial farm. Fuel flow and engine power output were accurately predicted from engine speed and needle lift duration and both these variables could be measured using one needle lift sensor. Engine power estimation from exhaust temperature measurements proved unsatisfactory due to the slow speed of response to changes in engine load.

Image processing techniques were investigated for developing a guidance signal for a tractor operating on agricultural row crops. The guidance signal was computed from thresholded images segmented by a Bayes classifier. The distribution of crop canopy and soil background pixels in an image was approximated with a bimodal Gaussian distribution function. The parameters of the distribution were estimated by regression to systematically subsampled images. Run-length encoding was used to locate center points of row crop canopy blobs in the thresholded images. A heuristic line detection algorithm was used to determine the parameters defining crop row location on the image plane. Row parameters were used to compute a tractor guidance signal. Results are presented on the performance of the individual components of the algorithm.

A limiting factor in tractor performance research is the lack of an easily implemented method of accurately measuring three-point hitch forces. This paper describes two methods that were investigated to measure these forces. The first method used extended ring load cell transducers to measure three-dimensional forces in the lower links of a three-point hitch. The second method used load sensing clevis pins that replaced the existing pins that connected the upper and lower links to the tractor. The three-point hitch (TPH) dynamometers were compared with respect to ease of implementation and accuracy. It was concluded that the clevis pin load cell dynamometer was easier to implement while both dynamometers were found to be equally accurate. A Dynamic Weight Transfer Coefficient (DWC) was found for several mounted implements. The DWC was categorized according to type of mounted implement and method used to control implement depth.