面向电子木琴自动演奏的MIDI信息智能转换算法研究

Research on Intelligent Conversion Algorithm of MIDI Information for Automatic Performance of Electronic Xylophone

摘要: 针对电子木琴自动演奏系统中MIDI信息转换的关键问题，文章提出了两种核心算法：智能音高偏移优化算法和智能转换策略选择算法。智能音高偏移优化算法通过构建六维度加权评分体系，自动计算最优音高偏移量，将任意MIDI信息所包含的音域适配至木琴有效音域；智能转换策略选择算法基于轨道数、总音符数和主轨占比三个特征，通过规则评分自动选择全曲转换、主旋律轨转换或截取部分转换策略，以适应不同音乐类型和系统资源限制。在Python环境下开发了MIDI转换程序，实现了从MIDI信息到C语言头文件的完整转换流程。实验结果表明，两种算法能有效解决音域不匹配和策略选择问题，转换后的音高分布与原始旋律轮廓高度一致，验证了算法的有效性。研究为传统打击乐器的智能化改造提供了技术支撑。

Abstract: Regarding the key issue of MIDI information conversion in the automatic electronic xylophone playing system, this paper proposes two core algorithms: the intelligent pitch offset optimization algorithm and the intelligent conversion strategy selection algorithm. The intelligent pitch offset optimization algorithm constructs a six-dimensional weighted scoring system to automatically calculate the optimal pitch offset amount, adapting the pitch range contained in any MIDI information to the effective pitch range of the xylophone; the intelligent conversion strategy selection algorithm, based on three features - the number of tracks, the total number of notes, and the proportion of the main track - automatically selects the full piece conversion, the main melody track conversion, or the partial extraction conversion strategy to adapt to different music types and system resource limitations. The MIDI conversion program was developed in the Python environment, achieving a complete conversion process from MIDI information to C language header files. Experimental results show that the two algorithms can effectively solve the problems of pitch range mismatch and strategy selection, and the pitch distribution after conversion is highly consistent with the original melody contour, verifying the effectiveness of the algorithms. This research provides technical support for the intelligent transformation of traditional percussion instruments.