基于人工智能作曲技术的民族音乐作品创作应用研究

Application and Research of Folk Music Composition Based on Artificial Intelligence Technology

刘炼(Liu Lian)

《社会科学理论与实践》 2卷2期 (2020/06) Pp. 35-42

https://doi.org/10.6914/TPSS.202006_2(2).0004

摘要

音乐是人类情感表达的重要形式之一。以人工的方法和技术让机器快速识别光学乐谱和实时乐音，通过音乐情感模型的推理和优化，获取人类音乐情感的表达模式，主动与用户完成人机协同的智能作曲等相关服务，对促进基于多源感知的情感型人工智能发展具有重要的研究价值和实践意义。本文主要融合多种模型现有人工智能模型（分型音乐模型、马尔可夫链模型、遗传算法模型、神经网络模型等），对规则知识进行改进，而形成实用、有效的混合模型。其创新点在于结合公式化和矩阵数据组合等方法来对自动作曲中非音符单元（重现结构）来智能化民族曲目创作，从而达到乐曲的和谐性、可听性和智能性。

關鍵字

人工智能 ； 作曲 ； 民族音乐 ； 创作

並列摘要

Music is one of the important forms of human emotion expression. Using artificial methods and technologies to allow the machine to quickly identify optical scores and real-time music sounds, through the inference and optimization of music emotion models, obtain human music emotion expression patterns, and actively complete human-machine collaboration with intelligent composition, and other related services. The development of multi-source perception of affective artificial intelligence has important research value and practical significance. This article mainly integrates multiple models of existing artificial intelligence models (typed music models, Markov chain models, genetic algorithm models, neural network models, etc.) to improve the rule knowledge to form a practical and effective hybrid model. Its innovation lies in combining formulas and matrix data combinations to intelligently compose non-note units (reproducing structures) in automatic composition, so as to achieve harmony, audibility and intelligence of music.

並列關鍵字

Artificial intelligence ； composition ； folk music ； composition