<p>An artificial tree (AT) algorithm has been proposed recently, and the performance of AT has been enhanced because of the introduction of improved AT algorithm with two-population (IATTP). However, the branch update operators of IATTP cannot effectively balance exploration and exploitation, which limits the optimization accuracy and efficiency of IATTP. To further improve the performance of IATTP, this work proposes a two-population artificial tree algorithm based on adaptive updating strategy for dominant populations (TATAD). In TATAD, six operators named self-evolution operator 2, crossover operator 2, improved self-evolution operator, gradient descent update operator, Gauss and Cauchy variational operator, and random traceless Sigma variational operator are applied to form an operator library. A dominant operator dynamic following mechanism is proposed to assign these six operators to the two populations in an optimal pairing scheme. Both populations and operators compete with each other, and the advantages of all operators are fully utilized. Moreover, the combination of diverse operators and dominant operator dynamic following mechanism can effectively balance the exploration and exploitation of TATAD. The performance of TATAD is compared with four AT algorithms and six efficient algorithms through typical test functions, and their results are bested by Wilcoxon rank sum test (WRST) and Friedman ranking test. It is found that TATAD is the most competitive algorithm among these algorithms for solving these optimization problems.</p>