This result has been achieved by partner MPEG and coordinator VUB. Below you can find an explanation of what the self-healing algorithm does accompanied by a result and a Figure. More information can be found in the reports of Work Package 4.
The self-healing machine learning EMS optimizes the usage of the hybrid battery energy storage system (HBESS) towards longer lifetime and higher reliability. For example, in UC1 (picture a) it causes less power fluctuation between the battery modules minimizing the power losses and in consequence increasing the overall lifetime.
Going directly to the results, you can see some results (picture b) of the UC2. Here our rule-based method and the SHML method. The results are really clear, with the SHML there is less power fluctuation minimizing the power losses and in consequence increasing the lifetime of the battery modules.
(a) Rule-Based EMS method SoC variation for HE and HP batteries.