Photovoltaic Array State Detection Using String Current Anomaly Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the operating state of photovoltaic arrays, such as using infrared images, are prone to inaccuracies due to environmental temperature fluctuations, leading to missed or false alarms and reduced power generation efficiency.
Innovation Solution
A method and apparatus that acquire present output current values from photovoltaic strings, use trained abnormality score predicting models to determine a present abnormality score, and compare it with a target score to assess the operating state, thereby avoiding temperature-related interference.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If infrared images are used to determine operating state of photovoltaic modules, then the method can detect temperature differences between normal and failed modules, but the accuracy is reduced due to environmental temperature fluctuations affecting the infrared images
Solution Approach 1:
The patent changes the detection parameter from temperature (infrared imaging) to electrical current characteristics. By monitoring output current values and comparing them against predicted values from a trained model, the system avoids temperature-related interference entirely while maintaining high accuracy in detecting photovoltaic module operating states and failures.
Solution Approach 2:
The patent replaces the thermal detection mechanism (infrared imaging) with an electrical detection mechanism (current measurement and analysis). This substitution eliminates the harmful effect of environmental temperature fluctuations by using electrical parameters that are not influenced by ambient temperature changes.
2Ease of operation
If infrared image acquisition devices on drones are used, then the operating state can be assessed remotely, but the discrimination between failed and normal modules is low leading to false alarms
Solution Approach 1:
The patent implements a feedback mechanism where historical output current values are used to train an abnormality score predicting model. The model continuously learns from past data to improve its ability to distinguish between normal variations and actual failures, thereby reducing false alarms while maintaining remote assessment capability.
Solution Approach 2:
The patent performs preliminary training of the abnormality score predicting model using historical data before actual operation. This preliminary action establishes a baseline of normal behavior patterns, enabling the system to accurately discriminate between failed and normal modules during remote assessment without being affected by environmental conditions.
Data Source
AI summary
A method including: acquiring present output current values of photovoltaic strings included in a photovoltaic array that, wherein includes at least two photovoltaic strings in parallel; determining a target abnormality score predicting model based on the present output current values. Determining a present abnormality score corresponding to the photovoltaic array by inputting the present output current values into the target abnormality score predicting model; and determining the operating state of the photovoltaic array by comparing the present abnormality score with a target abnormality score. The present abnormality score is output by inputting the present output current values acquired in real-time into the abnormality score predicting model, so as to determine the operating state of the photovoltaic array.