Yield Prediction System with Calibration Algorithms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current yield monitoring systems face challenges in accurately calibrating yield data across multiple harvesting machines and spatially within a field, leading to inaccuracies and variability due to non-linearity and crop-dependent sensor behavior, which can result in visible striping and human error in calibration adjustments.
Innovation Solution
A system that utilizes multiple yield weight measurements and aerial imagery to dynamically construct predictive yield models, aligning yield monitor data with actual yield measurements, and applying corrections to enhance accuracy, allowing for real-time predictions of machine fill levels and optimizing harvesting operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple harvesting machines are used to increase productivity, then harvest efficiency is improved, but calibration mismatch and visible striping occur due to non-linear sensor behavior
Solution Approach 1:
The patent applies parameter changes by transforming the non-linear yield monitor readings into linearized values through calibration equations. Each harvesting machine has its own calibration parameters that convert raw sensor outputs into accurate yield measurements, resolving the measurement precision issues caused by non-linear sensor behavior across multiple machines
Solution Approach 2:
The system implements feedback by using actual yield measurements from the harvesting machines to continuously refine and adjust calibration parameters. The calibrated yield data is fed back into the system to improve future measurements, eliminating calibration mismatch and visible striping artifacts in the yield maps
2Measurement precision
If manual calibration adjustment is performed to correct yield map inaccuracies, then measurement precision is improved, but human error and variability increase
Solution Approach 1:
The system implements self-service by automatically performing calibration adjustments without requiring manual intervention. The calibration equations and algorithms autonomously correct yield map inaccuracies, eliminating human error and variability while maintaining consistent measurement precision across different operations
Solution Approach 2:
The patent replaces the manual mechanical adjustment process with an automated computational system. Instead of manually adjusting calibration parameters, the system uses computer algorithms to automatically calculate and apply corrections, substituting human operators with automated software that eliminates variability and error
3Ease of operation
If simple average shifting is used to align yield data from multiple machines, then ease of operation is improved, but measurement precision deteriorates due to unaccounted variability
Solution Approach 1:
The system transforms the simple average shifting approach into a more sophisticated parameter transformation method. Instead of merely shifting averages, the patent applies comprehensive calibration equations that transform raw yield data into aligned, variability-corrected measurements, maintaining ease of automated operation while dramatically improving measurement precision
4Measurement precision
If yield monitor calibration is performed to account for non-linearity, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent replaces complex manual calibration procedures with automated computational algorithms. The system uses computer-based calibration equations that automatically account for non-linearity, substituting complex manual processes with simpler automated software that maintains high measurement precision without increasing operational complexity
Data Source
AI summary
Systems and methods for yield prediction for a field using spatial data representing yield throughout the field and multiple measurements of actual yield obtained from field regions during harvesting of the field. A yield model is generated based on the spatial data and the multiple measurements of actual yield. The yield model is used to determine information related to yield in the field.