This paper focuses on the numerical scheme for multiple-delay stochastic differential equations with partially H\"older continuous drifts and locally H\"older continuous diffusion coefficients. To handle with the superlinear terms in coefficients, the truncated Euler-Maruyama scheme is employed. Under the given conditions, the convergence rates at time $T$ in both $\mathcal{L}^{1}$ and $\mathcal{L}^{2}$ senses are shown by virtue of the Yamada-Watanabe approximation technique. Moreover, the convergence rates over a finite time interval $[0,T]$ are also obtained. Additionally, it should be noted that the convergence rates will not be affected by the number of delay variables. Finally, we perform the numerical experiments on the stochastic volatility model to verify the reliability of the theoretical results.

翻译：本文关注具有部分Hölder连续漂移项和局部Hölder连续扩散系数的多延迟随机微分方程的数值格式。为处理系数中的超线性项，我们采用了截断的Euler-Maruyama格式。在给定条件下，通过Yamada-Watanabe逼近技术，我们证明了在时间$T$处$\mathcal{L}^{1}$和$\mathcal{L}^{2}$意义下的收敛速率。此外，还得到了有限时间区间$[0,T]$上的收敛速率。值得注意的是，收敛速率不受延迟变量个数的影响。最后，我们在随机波动率模型上进行数值实验，以验证理论结果的可靠性。