PV Anomaly Diagnosis Using Distribution-Based Factor Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In large solar photovoltaic generation systems, it is challenging for operators to identify anomaly factors such as shadows or contamination that reduce generation capacity, leading to increased maintenance costs due to the difficulty in detecting these issues before maintenance visits.
Innovation Solution
An anomaly factor diagnosis apparatus that includes an identifier, a distribution generator, and an evaluator, which identifies measurement data corresponding to anomaly occurrences, calculates a distribution function representing feature values, and evaluates the possibility of anomaly occurrence based on this data, allowing for pre-visit preparation and effective maintenance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If operators manually inspect solar photovoltaic generation systems to identify anomaly factors, then maintenance costs increase due to detachment requirements, but anomaly detection capability is limited by human observation
Solution Approach 1:
The patent replaces manual mechanical inspection with an automated diagnostic system that uses measurement data analysis. The anomaly factor diagnosis apparatus automatically identifies shadows, contamination, and other anomaly factors by processing power, voltage, current, and irradiance data, eliminating the need for physical detachment and manual observation while improving detection precision.
2Ease of operation
If maintenance visits are performed without prior anomaly identification, then operators cannot prepare appropriate tools and parts, but systematic monitoring infrastructure is required
Solution Approach 1:
The patent implements preliminary anomaly factor identification by continuously analyzing measurement data before maintenance visits. The system calculates anomaly factor scores and identifies specific issues (shadows, contamination, etc.) in advance, allowing operators to prepare appropriate tools and parts beforehand. This preliminary diagnostic action simplifies maintenance operations without requiring complex real-time monitoring during the actual maintenance activity.
3Loss of information
If generation capacity reduction is detected using traditional measurement methods, then presence or absence of performance loss can be identified, but specific anomaly factors cannot be determined
Solution Approach 1:
The patent segments the anomaly factor identification process into distinct categories including shadows, contamination, and other factors. By dividing the diagnostic task into specific anomaly types and calculating separate scores for each, the system provides detailed information about specific anomaly factors while maintaining overall performance monitoring accuracy. This segmentation enables targeted maintenance actions for each identified issue type.
Data Source
AI summary
According to one embodiment, an anomaly factor diagnosis apparatus includes an identifier configured to identify a first measurement data item corresponding to a first time point when an anomaly factor occurs in a power generation apparatus, among measurement data items in the power generation apparatus, based on anomaly factor occurrence data identifying the first time point; a distribution generator configured to calculate a first distribution function representing a distribution of a feature value of the first measurement data item; and an evaluator configured to calculate an evaluation value of possibility of occurrence of the anomaly factor in the power generation apparatus, based on the first distribution function, and measurement data items to be tested in the power generation apparatus.