Target value calculation
The calculation of a simulated value for the PV plants initially entails a physical simulation. With a self-learning algorithm it is time to reach the next level. Why?
A lot of PV plants have shortcomings. Shading or clipping are only two examples that can lead to an erroneous simulation. With the algorithm we improve the target value automatically and let it learn the characteristics of the specific plant. Thus false alarms can be avoided, even without intensive configuration work. Deviations between the target and the real value will then be a valid indicator for technical issues.
Taking snow into account
False alarms are common in winter due to solar modules covered by snow. This causes unnecessary O&M teams and additional costs. It is difficult to simulate the slipping and melting of snow on modules only based on physical simulations. However, with the use of artificial neuronal networks, we are already able to considerably optimize the solar power forecast and to enhance the alarming in the future.
There is more to come
With the VCOM we offer you an advanced alarm system. You always have the performance of your plants under control. In combination with the ticket system, it´s an ideal tool to organize your O&M processes.
Machine learning and big data is already making O&M more efficient but it’s just the beginning. Stay tuned for more data intelligence features!