PowerBrainTM Technology Summary
The solar inverter is the heart of the PV solar system, converting solar energy into revenue. infiniRel’s inverter health scanner serves as the check-engine light for your control center, assesses failure risks in minutes, and protects your revenue generating assets. Once the infiniPadTM scanner establishes the inverter’s health signature by non-invasive and real-time measurement of its internal vital signals, unique to each solar plant, our alert system offers the control center operator the insight to optimize power allocation for maximizing uptime and availability, and cutting unexpected maintenance costs.
Critical inverters may be put on a maintenance schedule before an outage, saving downtime costs including loss of revenue, expedite charges, extended clean-up, and potential legal consequences.
Traditional predictive maintenance systems have been limited in their success to mechanical systems, such as gas turbines and gear boxes. In contrast, power electronics, such as the IGBT engine in your inverter, can experience a degradation to failure in milli-seconds, whereas mechanical systems may take minutes to hours after initial degradation detection. Condition-Based Maintenance for modern energy transformation requires a fundamentally different approach:
Similar to an EKG measuring the heartbeat signature of a patient, proprietary signal-processing technology combined with novel machine learning techniques extract each inverter’s health signature from the power grid the inverter connects to.
By transforming the switching noise produced by the inverter into a signature that is indicative of its health degradation, infiniPadTM senses the stress level before inverter failure. We overcome time and cost constraints of a producing solar plant by testing on live equipment under actual operating conditions, integrating with existing maintenance processes.
Our proprietary and patented technology
1. samples data 50 times faster than conventional power systems.
Traditional inverter monitoring systems measure the power quality of the grid frequency of 60-Hz. We measure the quality of its internal switching frequency of kilo-Hertz. This compares to the difference between using a EKG over a blood pressure meter to predict heart failure;
2. uses power and code-efficient data correlation techniques to reduce data streams by a factor of 100,000, that is 100-times more efficient than the latest schemes for data compression. This allows our inverter health scanners to instantly learn from new field knowledge;
3. integrates a proprietary Machine Learning technique that promises to be 1000-times more resource efficient for training data, processing time, and processing power.
This enables us to detect anomalies in time for corrective action, using a fraction of data center power.
infiniPadTM offers risk score and confidence level information so that control center operators can optimize inverter power allocation manually, or automatically, across the plant, including the option for automatic, safe and controlled inverter shutdown and avoid escalating downtime and maintenance costs.