Multi-Agent Photovoltaic Diagnosis via Segmented Reference Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing solutions for managing complex photovoltaic systems are inadequate for real-time or near-real-time diagnosis and efficient failure detection, particularly in systems with multiple modules.
Innovation Solution
A method utilizing a multi-agent system with local and global management agents, sensors, and communication devices for real-time measurement, estimation, and comparison of operating quantities against reference models, allowing for fault detection and verification across similar modules, and adaptive grouping for comprehensive diagnosis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing management solutions are used for photovoltaic systems, then system operation continues, but real-time diagnosis and efficient failure detection are insufficient
Solution Approach 1:
The system segments the photovoltaic installation into multiple groups based on geographical location, orientation, and inclination. Each group is monitored independently with its own reference model, allowing localized diagnosis without requiring complex centralized processing of all modules simultaneously. This segmentation enables reliable failure detection while managing system complexity through distributed intelligence.
Solution Approach 2:
The system dynamically adjusts reference quantities based on changing environmental parameters such as irradiance, temperature, and time of day. Reference models are updated to reflect seasonal variations, weather conditions, and operational states. This parameter adaptation allows the system to distinguish between normal performance variations and actual failures, improving detection reliability without requiring overly complex fixed-threshold systems.
2Reliability
If a multi-agent system with verification steps is implemented, then fault detection reliability improves, but processing time and computational load increase
Solution Approach 1:
The system pre-calculates and stores reference quantities for various environmental conditions and operational states. When a module deviates from expected performance, the verification process compares against pre-computed reference data rather than performing complex real-time calculations. This preliminary preparation maintains high verification accuracy while reducing the time required for actual fault diagnosis.
Solution Approach 2:
The multi-agent system performs self-verification by comparing each agent's measurements against reference models and neighboring agents' data. The system automatically identifies and flags anomalies without requiring external intervention or complex centralized analysis. This self-service approach enables thorough verification while minimizing processing time through distributed autonomous decision-making.
3Difficulty of detecting and measuring
If comprehensive monitoring of all photovoltaic modules is performed, then detection capability improves, but system complexity and computational requirements worsen
Solution Approach 1:
The system applies different monitoring strategies to different groups of modules based on their specific characteristics such as location, orientation, and inclination. Each group has tailored reference models and verification criteria appropriate to its operational conditions. This localized approach enables comprehensive detection capability for each group while avoiding the complexity of a single monolithic monitoring system for all modules.
Solution Approach 2:
The multi-agent architecture uses a universal framework where each agent performs multiple functions: self-monitoring, neighbor comparison, reference model evaluation, and fault verification. This multi-functional design enables comprehensive monitoring capability while reducing overall system complexity by eliminating the need for separate specialized systems for each monitoring task.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables reliable and efficient detection of faults in photovoltaic systems, improving energy production management by identifying issues promptly and adjusting models for accurate diagnosis and maintenance.
Implementation Method 1
a photovoltaic system comprising at least one photovoltaic field (10)
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A diagnostic method for a photovoltaic system (1) comprising at least one photovoltaic field (10), module agents (12, 13, 14) associated with subsets (2, 3, 4) of at least one photovoltaic module and the photovoltaic system (1) further comprising a local management agent (30; 30', 30"), characterized in that it comprises a diagnostic phase (P2), which includes the repetition of the following steps: (E5): measurement and/or estimation of one or more operating quantities of a subset; (E6): calculation of one or more reference quantities of said subset; (E7): comparison of a measured or estimated quantity with the same reference quantity; (E8): in case of a difference exceeding a predefined threshold, transmission of a fault message and verification (E9) of the fault at the level of the local management agent.