PV Plant Power Forecasting With Shading-Aware Matrix Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing monitoring systems for photovoltaic plants are unable to accurately forecast electrical power due to factors like shading, environmental conditions, and malfunctions, and lack automation in calibration and fault detection.
Innovation Solution
A method using sensors to collect data on solar radiation, temperature, and electrical power, processed by a control unit to populate matrices for real-time forecasting, filtering out unsuitable data, and applying bilinear interpolation for accurate power prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional monitoring systems are used to collect and display electrical power data, then basic statistical processing is achieved, but the systems cannot accurately forecast power output or detect efficiency drops due to environmental factors like shading
Solution Approach 1:
The method segments the forecasting problem by creating separate lookup tables for different environmental conditions (shaded vs. unshaded conditions, different temperature ranges). This allows the system to handle complexity through organized data structures rather than complex algorithms, improving accuracy while maintaining manageable system complexity.
Solution Approach 2:
The patent introduces an intermediary processing layer that uses astronomical algorithms to calculate solar position and determine shading conditions. This intermediary layer translates complex environmental factors into simplified lookup table selections, enabling accurate forecasting without requiring the entire system to be complex.
2Measurement precision
If manual calibration and technical analysis are performed to improve forecasting accuracy, then model accuracy improves, but automation is reduced and time consumption increases
Solution Approach 1:
The system performs self-calibration by automatically comparing forecasted power output with actual measured power output. When discrepancies are detected, the system automatically adjusts the lookup table values without requiring manual operator intervention. This maintains high accuracy while achieving full automation, eliminating the trade-off between manual calibration and automation.
Solution Approach 2:
The patent implements a feedback mechanism where actual electrical power measurements are continuously compared with forecasted values. This feedback loop enables automatic detection of accuracy deviations and triggers automated recalibration of the forecasting models, maintaining high precision while achieving complete automation.
3Difficulty of detecting and measuring
If comprehensive data collection for environmental factors is implemented to detect shading effects, then detection capability improves, but system complexity and implementation difficulty increase
Solution Approach 1:
The patent makes existing sensors multi-functional by using standard temperature and irradiance sensors for both their primary purposes and for detecting shading conditions. The system uses astronomical algorithms to universally apply shading detection logic across different times and locations. This approach improves detection capability without requiring additional specialized sensors or increasing hardware complexity.
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
The method provides fast, reliable, and automated forecasting of electrical power, considering environmental factors and plant performance, without requiring AI systems or specific software, and is suitable for cloud implementation.
Implementation Method 1
A photovoltaic plant is an electricity plant equipped with a solar energy collection system which uses photovoltaic modules and produces electricity
Data Source
AI summary
A is a method for forecasting electrical power in real time of a photovoltaic plant includes recording a set of operating data; entering for each set of operating data a respective electrical power value recorded in a cell for insertion of a power matrix and dividing the recorded electrical power value to obtain an efficiency value. The efficiency value in a cell of an efficiency matrix is entered and a power value contained in a cell of the power matrix and an efficiency value contained in a cell of the efficiency matrix identified. The power value and the efficiency value are multiplied to derive an expected electrical power value and the expected electrical power value is compared with the corresponding electrical power value acquired in real time identifying an operating condition of the photovoltaic plant as a function of the comparison.