“…In fact, the search effect of evolutionary algorithm is affected by the number of search targets. For example, in literature, 24 if the number of targets is small, good effect can be obtained. In USC units, multiple influence factors from feed-water, feed-fuel to generator determine it as a multi-variable system where the total number of A, B, C up to 27, resulting in the rise of “curse of dimensionality.” In this case, the idea of learning in batches is introduced in the study, where the purpose of global identification is achieved through batch identification, and therefore, the difficulty that evolutionary algorithm can not directly apply to multi-variable system identification is flexibly solved.…”