Research On Automatic Accompaniment Technology For Folk Melody Of Agriculture And Animal Husbandry

In recent years, more and more farmer singers appeared along with our country development in economic and cultural. In rural areas, music creation and music playing become more and more popular. Automatic accompaniment technology was more applications in the music creation and music playing along with the development of computer science. To this end, combined with the current situation of the development of rural music, this paper proposed an automatic accompaniment algorithm for folk melody of agriculture and animal husbandry, in order to solve the problem that automatic music was accompanied.Firstly, this paper introduces the history and research status of algorithmic composition. The problems existing in the practical application of artificial intelligence common methods about automatic accompaniment aspects are analyzed. These problems include melody connotation deviation caused by frequent switching in Markov switching table, the bottleneck problem of fitness function in genetic algorithm, etc.Then, the basic music theory was combined with intelligence artificial algorithm to get a music score coding method, and to digitize the music information. Based on the music score digital coding, and combined with the advantage of hidden Markov model, an automatic accompaniment algorithm oriented to farmers and herdsmen was proposed. First, the database of harmonic progression rules was built. The sequence of accompaniment chords was generated by the combinations of the extract of the melody style and musical mode with the database of harmonic progression rules. Second, Meta-Structure was abstracted according to the melody pitch and duration, and combined with the application of hidden Markov model in algorithmic composition, the hidden Markov model of accompaniment sound pattern was built with Melody figure structure as observed value and accompaniment sound pattern as implicit partition state. Then, the Meta-Structure of accompaniment was generated by Viterbi algorithm. Finally, the sequence of accompaniment chord and the sequence of Meta-Structure was sequenced to get the complete accompany melody.Finally, an automatic accompaniment experiment system platform was built in VC++ environment, which verifies the actual effect of the above algorithm. Experimental results show that the accompaniments generated by the system can meet the needs of specific population.