Anomaly Alarms – Get the most out of your solar

Do you know if your solar system is producing the maximum energy expected? If you don’t know the answer to this question, it’s highly likely that you may be missing out on your maximum earning potential from your existing system due to undetected low performance. You have already invested a lot, so you deserve to get the most out of it.

inavitas anomaly alarms keep you informed and tell you which inverter is being slack or any other issues in the system. inavitas energy intelligence will learn your production patterns, send you an anomaly alarm, and help you decide what to do.

In a nutshell, the aim of anomaly alarms is detecting failures at an early stage on solar power plants which is one of the core concepts of inverters’ monitoring. The goal is to detect likely to occur failures as early as possible so that shortening the maintenance time and limiting the times when generation capacity drops. Real-time monitoring is carried out by inavitas system via sensors on a unit such as a power plant, solar modules, inverters. Sensitive alarms can be generated if the values coming from these sensors exceed the limits based on expert knowledge.

The solar panels that are similarly located in solar farms and similarly affected by solar radiation are expected to show similar power generations. It is assumed that the time-dependent trend of these generations will preserve this structure on consecutive days. Dynamic time warping, which is a method used in calculating the distances of time-dependent distributions according to their properties, is utilized. According to this method, the distances of the inverters are calculated according to the patterns of the time series (even if they are of different length) including the energy generations. The k-means clustering algorithm, which increases the distance between clusters while decreasing the variance within the cluster, clusters the inverters using the calculated DTW distances.

Sample outputs of this algorithm are given in Figure 1.

Figure 1: k-means with Dynamic Time Warping Distances

As being inavitas users, you don’t worry about this technical knowledge as the platform makes it understandable for all users. For example, normally, solar production of inverters on a good sunny day should be like a big curve, as seen in the Figure 2 presented in the inavitas energy intelligence platform. It is very easy to understand that some inverters, like a light blue one, are under performance, and there is an unusual production pattern.  It is quite easy to pick up which inverter is not working as expected.

Figure 2

Our customers, specifically solar installers or operation and maintenance companies, benefit from having anomaly alarms in place to manage their resources efficiently.

If you would like to learn more about the benefits of anomaly detection and alarms, please contact us.

‍