Hybrid Power Plant AI Reporting for Lost Production Event Categorization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for categorizing and communicating lost production events (LPEs) in wind turbines are inefficient and prone to inaccuracies due to manual data parsing, language barriers, and inadequate vocabulary, leading to repeated issues.
Innovation Solution
A method utilizing a combination of machine learning models and a categorization AI system to analyze various data sources, including SCADA data, maintenance data, and weather data, to determine a final categorization of LPEs, with a large language model generating a textual explanation for discrepancies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual parsing of data is used for LPE categorization, then human judgment can be applied, but the process is tedious and inaccurate
Solution Approach 1:
The patent replaces manual mechanical data parsing with automated machine learning models and AI systems. Multiple ML models process different data types (SCADA, maintenance, weather) automatically, eliminating the need for human manual analysis while improving both speed and accuracy of LPE categorization.
Solution Approach 2:
The patent uses AI systems to create accurate representations (copies) of the categorization process. The AI generates textual descriptions that replicate human explanation capabilities, preserving the nuance of human judgment while automating the process entirely.
2Reliability
If manual communication of LPE results is used, then human understanding can be conveyed, but language barriers and inadequate vocabulary cause pieces of explanation to be lost in translation
Solution Approach 1:
The patent replaces manual human communication with automated large language models. These AI systems generate precise textual descriptions of LPE categorizations, eliminating language barriers and vocabulary limitations while maintaining clear, consistent communication across all stakeholders.
3Measurement precision
If multiple data sources are processed for LPE categorization, then categorization accuracy improves, but processing complexity increases
Solution Approach 1:
The patent segments the complex processing task by assigning different machine learning models to different data types. Each model specializes in processing specific data sources (SCADA, maintenance, weather), and their results are integrated by an AI system, making the overall complex process manageable and scalable.
Solution Approach 2:
The patent creates a universal AI system that can handle multiple data types and generate both categorization results and textual explanations. This multi-functional approach consolidates what would otherwise require separate systems for processing, analysis, and communication.
Data Source
AI summary
Embodiments herein describe improved techniques to evaluate and effectively communicate an LPE categorization. An initial LPE categorization may be generated from, for example, SCADA data collected at the wind turbine by a SCADA system. The initial LEP categorization and different categorization predictions from other auxiliary data sources also collected at the wind turbine may be evaluated by an LPE categorization AI system. This LPE categorization AI system is configured with categorization ML models. The LPE categorization system outputs a final LPE categorization which may differ from the initial LPE categorization.