For the special working conditions such as harsh environment and complex geology in the downhole automatic control process of the anchor drilling arm, it is necessary to consider the parameter selection of PID, and for this reason, the stochastic fractal search algorithm (SFS) is used for the adjustment of the PID parameters. Due to the problems of insufficient local search capability and easy to fall into local optimization, the random fractal search algorithm is ineffective in the face of many parameters to be calibrated and serious coupling between parameters of the robotic arm. Therefore, chaotic initialization, positive cosine strategy, and adaptive K value selection strategy are introduced to improve the local search ability and global search speed of the algorithm; Cauchy and differential directional variation strategies are adopted to make the algorithm jump out of the local optimum. The standard test function test shows that the improved stochastic fractal search algorithm has improved the convergence speed and accuracy. Finally, through the simulation experiment of anchor drilling arm parameter adjustment, it is shown that: the overall adaptability value of the PID parameters adjusted by the improved stochastic fractal search algorithm is increased by 21%; and the quantitative indexes of the joints indicate that the improved stochastic fractal search algorithm jumps out of the local optimum, and improves the overall searching accuracy and speed.optimal and improves the overall optimization accuracy and speed.