Deep Learning

Machine Learning Advanced Service

"Deep Learning" advanced machine learning service

In-depth physical examination to understand the health status of old solar power plants

Major Analysis Items

Shengqi cooperated with an Israeli artificial intelligence software company to launch the "Deep Learning" advanced machine learning service. Just install the Pixel View monitoring system, you can apply for an upgrade and use the existing power plant data for data analysis. Through the extremely complex algorithm, Interpret all index data of solar energy and analyze the reasons for the decline in power generation efficiency, deeply study the power generation behavior of power plants in the past year, and analyze 20 reasons for the decline in power generation indicators:

Inverter wakes up late

Wire loss and connector breakage

MPPT/Converter Efficiency

Efficiency measurement of the tracking system

frequency deviation Converter data integrity
Abnormal temperature coefficient Converter Clipping
Device disconnection Voltage deviation
Solar Meter Data Integrity Tandem power generation performance
Environmental sensor failure Tandem Generation Data Integrity
Environmental Sensor Data Consistency MPPT exception
Solar module clean/dirty Converter efficiency below specification
Vegetation needs pruning The relative efficiency of the converter is abnormal
Inverter wakes up late Converter failure
Abnormal power factor Abnormal temperature alarm
Converter/AC meter power generation ratio Transformer is inefficient
Transformer temperature is abnormal
frequency deviation Converter data integrity Inverter wakes up late
Abnormal temperature coefficient Converter Clipping Converter failure
Device disconnection Voltage deviation Abnormal power factor
Solar Meter Data Integrity Tandem power generation performance Abnormal temperature alarm
Environmental sensor failure Tandem Generation Data Integrity Converter/AC meter power generation ratio
Environmental Sensor Data Consistency MPPT exception Transformer is inefficient
Solar module clean/dirty Converter efficiency below specification Transformer temperature is abnormal
Vegetation needs pruning The relative efficiency of the converter is abnormal
Including: Converter Clipping, Data Collection Integrity, Voltage and Frequency Deviation, Module Temperature Coefficient Abnormal, MPPT/Inverter Efficiency Abnormal, MPPT Accuracy, Disconnected Serial, Inverter Conversion Efficiency Poor, Inverter Failure or Due to factors such as getting up late, environmental sensor failures, etc., a quantifiable renovation and maintenance plan is proposed, and preventive abnormal maintenance and corrective measures are carried out, which can increase the average power generation of old power plants by 3-8% and reduce O&M costs by 10-12%.

Deep Learning AI Monitoring Steps

AI monitoring steps

Case Study

Inverter Efficiency Analysis
download
Finding Disconnected Strings
download
Inverter Efficiency Analysis
download
Finding Disconnected Strings
download