2002 Annual Report
Research
Development of Machinery Systems and Sensors for Site-Specific Farming
Jonathan Chaplin, Associate Professor
Pierre Robert, Professor, Soil, Water, and Climate
Carl Rosen, Professor, Soil, Water, and Climate
Funding Source
Minnesota Agricultural Experiment Station
Objective
- To investigate the precision needed for sensing biological parameters in the landscape including crop residues, soil nutrients, crop yield (for grain crops and root crops), and soil moisture.
- To develop machines and assist others in developing machines for use in precision farming activities.
- To investigate the economics of using variable-rate technologies.
Project Description
Interest in site-specific farming over the past few years has been driven
by newly developed agricultural technology. Equipment is now available
that can locate machinery on the field and vary the inputs being applied,
or record the yield of the crop being harvested. This equipment will help
farmers improve economic competitiveness.
We have developed techniques for computer-enhanced rapid measuring and
mapping of crop residue. We have also developed a sensor that determines
the effect of tillage on the soil. This sensor should help farmers reduce
erosion and select machinery, crops, and husbandry methods. We have developed
an automatic soil sampler for use in precision farming and soil mapping.
We have also initiated research on the use of nuclear magnetic resonance
as a possible method for measuring soil nutrients, texture, and moisture
content.
Results
Grain Yield Sensor Analysis
A laboratory test system has been used to measure the response of combine operating parameters to variations in grain throughput. A personal computer controls material flow in a stepwise manner. A combine (IH 915) equipped with a grain flow sensor was tested. The sensor was an impact plate type, which has become the norm for many companies marketing such instrumentation. Grain flow into the machine was controlled using servo-actuated gate valves. The grain flow was split so that both constant and modulated flows could be investigated. Data was collected at 200 Hz and analyzed using DADDISP and MacANOVA. The lowest controlled flow rate was 2 lb/s which represents approximately 30 bu/A assuming a forward speed of 3 mph and a 4x36" corn row head. Confidence intervals were evaluated for various flow rates and equivalent yield. Here is a list of yields followed by 95% confidence intervals; 60 bu/A (CI 18-82 bu/A), 90 bu/A (CI 60-130 bu/A), 120 bu/A (CI 90-165 bu/A), 180 bu/A (CI 125-210 bu/A), 240 bu/A (CI 160-250 bu/A). Errors of observation were also computed and these range from 80 percent at low flow rates to 10 percent at high flow rates. Information gathered and analyzed shows that the widely accepted impact plate grain yield sensor may be able to measure trends in grain yield. Maps are generated using grain yield data from this type of sensor. The maps should reflect the precision and confidence levels of the original data to be of practical use.