基于修正型粒子群算法的智能音乐作曲研究

Study on the Intelligent Music Composition Based on Modified Particle Swarm Optimization Algorithm

针对传统作曲算法存在速度慢、工作量大的缺点,本文将修正型粒子群算法引入音乐作曲,通过音级、拍数时值和音符的编码实现音乐的数值编码,选择对口组、非对口组和专业组的加权评价结果为适应度函数,以区域内部均匀性测度(UM)、区域间对比度(GC)以及综合测度等三个指标作为评价智能音乐作曲效果指标,提出一种基于修正型粒子群算法的智能音乐作曲。结果表明,该算法具有作曲速度快、质量高的优点,有助于曲作者音乐创作,极大地降低工作量,具有一定的推广价值。

In view of the shortcomings of slow speed and large workload in a traditional music algorithm, this paperintroduced the particle swarm optimization algorithm into a music creation to realize a music numerical codes by way of thesound level, rhythm duration and note codes to select the weighted evaluation results from the counterpart, non-counterpartand professional groups as effect index, and take uniformity measure （UM） in the region, regional contrast （RC） and thecomprehensive measure as evaluation index for an intelligent music composition effect to propose the intelligent musiccomposition based on the modified particle swarm optimization algorithm. The experimental results showed that thealgorithm possessed the advantages of fast speed, high quality and can help the composer create the music and reduce theirworkload. This method has certain promotion value.