文章以松旺镇客家话声调基频为基础,构建松旺镇客家话声调模型;利用该模型合成出客家话声调样本,并进行了听辨实验。实验结果表明,声调模型基本合理科学。该模型不仅能够在计算机上动态地、交互式地再现博白松旺镇客家话声调音高、音长和声调的凹凸变化特征的真实形态,而且还能构建和预测声调的发展变化轨迹,对方言保护和方言合成具有一定的应用价值。 Based on the tone fundamental frequency of Songwang dialect in Hakka dialect, the article builds the tone model of Songjia dialect in Hakka dialect, and uses the model to synthesize the tone samples of Hakka dialect. The experimental results show that the tone model is basically reasonable and scientific. The model can not only dynamically and interactively reproduce the true shape of the concave-convex variation features of tone pitch, pitch length and tone of Hakka Songwang Hakka dialect on a computer, but also construct and predict the development and change trajectory of tone, Protection and dialect synthesis has a certain application value.