自安德鲁斯和梅尔卡于2012年对欧拉五边形数定理的截断版本的研究重新焕发活力以来，截断θ级数的研究得到了广泛关注。2021年，梅尔卡提出了关于截断雅可比三重积级数的更强猜想。本文针对任意给定的$R$、$S$和$k$，提出了一种系统的方法来确定$n$的下界$N$，使得当$n\geq N$时，梅尔卡的更强猜想成立。具体来说，我们首先将上述θ级数分母中的无穷$q$-移位因子分解为以下两部分： \[ \frac{1}{\left(q^s, q^{r-s}; q^r\right)_{2}}\cdot \frac{1}{\left(q^{2r+s}, q^{3r-s}; q^r\right)_{\infty}}:=\sum_{n=0}^{\infty} p_{4}(n)q^n\cdot \sum_{n=0}^{\infty} g_{s,r}(n)q^n, \] 其中$s=S/(S,R)$，$r=R/(S,R)$互质。第一部分可以解释为仅由$s$、$r+s$、$r-s$、$2r-s$构成的整数$n$的划分函数$p_4(n)$，第二部分则可视为非模形式的无穷乘积。对于$p_4(n)$，我们利用留数定理和划分函数的性质得到了其一般的上下界。进一步结合θ级数的分子，我们推导出一个$n$的下界，使得后续各项系数均为非负值。对于$g_{s,r}(n)$，我们通过在非模形式无穷乘积上应用圆法得到了其上下界，并由此推导出$g_{s,r}(n)$的渐近公式。最后，通过对这两部分的卷积分析，我们确认了当$n\geq N$时梅尔卡的更强猜想成立。进而表明，当$k$足够大时，只需处理第一部分即可给出下界$N$，并且梅尔卡的更强猜想对任意$S$和$R$均直接成立。

The study of truncated theta series has regained vitality since Andrews and Merca's work on the truncated version of Euler's pentagonal number theorem in 2012. In 2021, Merca proposed a stronger version of the conjecture for the truncated Jacobi triple product series. In this paper, for any given $R, S$ and $k$, we provide a systematic method to determine the lower bound $N$ of $n$ such that when $n\geq N$, Merca's stronger conjecture holds. More precisely, we first treat the infinite $q$-shift factorial in the denominator of the above theta series as the product of the following two parts \[ \frac{1}{\left(q^s, q^{r-s}; q^r\right)_{2}}\cdot \frac{1}{\left(q^{2r+s}, q^{3r-s}; q^r\right)_{\infty}}:=\sum_{n=0}^{\infty} p_{4}(n)q^n\cdot \sum_{n=0}^{\infty} g_{s,r}(n)q^n,\] where $s=S/(S,R),r=R/(S,R)$ are coprime. The first part can be interpreted in terms of partition functions $p_{4}(n) $ which are the number of partitions of $n$ only with parts $s, r+s, r-s, 2r-s$. The second part can be seen as a nonmodular infinite product. For $p_{4}(n)$, we obtained the general upper and lower bounds by using the residue theorem and the properties for partition functions. Further multiplied by the numerator of the theta series, we derive a lower bound of $n$ which satisfies that the coefficients of the subsequent terms are all nonnegative. For $g_{s,r}(n)$, we obtain the lower and upper bounds by applying the circle method on nonmodular infinite products, from which we also deduce an asymptotic formula for $g_{s,r}(n)$. Then, by considering the convolution of these two parts, we confirm Merca's stronger conjecture when $n\geq N $. Consequently, we show that when $k$ is sufficiently large, it only needs to deal with the first part to provide the bound $N $ and Merca's stronger conjecture holds directly for any $S$ and $R$.