Photovoltaic Power Forecasting via Time History Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing power systems for rooftop photovoltaic installations face challenges in accurately forecasting energy output due to various influencing parameters, requiring complex configurations and data processing, which can be cumbersome and inefficient.
Innovation Solution
A power system that uses statistical analysis of time histories and incorporates weather and temperature forecasts to predict energy output, allowing for adaptive connection and disconnection of loads and utilizing a compact design with reduced components prone to failure, and leveraging digital communication for efficient operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If complex mathematical models and multiple parameters are used to forecast power output, then forecast accuracy is improved, but system complexity and configuration effort increase
Solution Approach 1:
The patent extracts and removes complex mathematical models and configuration parameters from the system. Instead of using intricate forecasting models, the invention relies on simple time history data collection and basic statistical analysis, eliminating the need for complex configuration while maintaining forecast capability
Solution Approach 2:
The system automatically collects time history data from the power meter and performs statistical analysis without requiring user configuration. The forecasting mechanism serves itself by utilizing readily available operational data, eliminating the need for manual parameter setup and complex model configuration
2Measurement precision
If comprehensive parameter configuration is performed during installation, then forecast accuracy is improved, but installation and commissioning time increase
Solution Approach 1:
The patent removes the time-consuming parameter configuration step entirely. By extracting the need for manual parameter input and complex model setup, the system achieves accurate forecasting using only automatically collected time history data, dramatically reducing installation and commissioning time
Solution Approach 2:
The system performs self-configuration by automatically collecting and analyzing time history data from the power meter. No manual parameter input is required during installation, as the system autonomously establishes its forecasting capability through data-driven statistical analysis
3Measurement precision
If multiple sensors and complex models are deployed, then power forecast accuracy is improved, but system reliability decreases due to more failure-prone components
Solution Approach 1:
The patent extracts and eliminates multiple sensors and complex forecasting models from the system architecture. By relying solely on time history data from the existing power meter and simple statistical analysis, the system maintains forecast accuracy while removing failure-prone components, thereby improving reliability
Solution Approach 2:
The system uses the power meter's existing time history data for self-analysis through statistical methods. This self-service approach eliminates the need for additional sensors and complex models, reducing the number of potential failure points while maintaining forecasting capability
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Control of photovoltaic systems. A power system (1) comprising: an energy conversion module (2a, 2b, 2c), a power meter (4), a switch (5) configured to disconnect the power system (1) from the load (6), a system controller (7), a power transmission bus (9a, 9b) extending from the energy conversion module (2a, 2b, 2c) to the switch (5); wherein the power meter (4) is configured to record an amount of electric power flowing through an electric power bus (9a, 9b); the system controller (7) being in operative communication with the power meter (4) and being configured to: read a first time series having a first plurality of power signals and a second time series having a second plurality of power signals from the power meter (4), each power signal being indicative of an amount of power recorded by the power meter (4).