combine-perf

This is the repository for hosting a set of Jupyter notebooks for demonstrating the data processing software for the Journal of ASABE manuscript: A New Methodology For Combine Performance Analyses in Wheat Harvests With GNSS Data.

Software Dependency

• python==3.9.1
• pandas==1.4.4
• geopandas==0.9.0
• numpy==1.21.5
• shapely==1.7.1
• sklearn==1.1.2
• scipy==1.7.3

Organization

There are a total of 8 Jupyter notebooks. Each of them corresponds to different parts of the Methodology section in the paper.

step0-preprocessing-get-data.ipynb

This corresponds to steps described in Raw Positioning Data Preprocessing.

step1-imm.ipynb

This corresponds to the IMM algorithm utilized in Step 1: Removal of Random Noise V(k). It specifies the state-space models and how IMM is run.

step2-classification.ipynb

This corresponds to the steps taken in Combine State Classification that employs ST-DBSCAN clustering.

step3-smooth-state-based.ipynb

This corresponds to how a Kalman-based smoother is set up and run in State-Based Track Smoothing.

step4a-optimization.ipynb

This corresponds to the steps taken in Optimized Solution Search.

step4b-effective-swath-estimation.ipynb

This corresponds to computations of metrics defined in Part III: Wheat Harvest Performance Metrics.

step5a-efficiency-calculations.ipynb

This corresponds to more computations of metrics defined in Part III: Wheat Harvest Performance Metrics.