针对支持向量机发酵建模中,选择重要建模参数值的问题,提出利用全局搜索能力较强的粒子群优化算法,优化调整支持向量机建模过程中的重要参数,每一个粒子的位置向量对应一组支持向量机建模的参数。参数不断优化后,得到拟合预测效果较优的模型,预测青霉素发酵过程。仿真结果表明,该方法能使模型的预测效果较好。 In order to solve the problem of selecting the important modeling parameters in SVM fermentation modeling, a particle swarm optimization algorithm with strong global search ability is proposed to optimize and adjust the important parameters in SVM modeling. The position vector of each particle Corresponds to a set of support vector machine modeling parameters. After optimization of the parameters, a model with better fitting prediction effect is obtained and the penicillin fermentation process is predicted. The simulation results show that this method can make the prediction of the model better.