【摘要】 利用朴素贝叶斯分类器对芯片无氧化烘箱温度PID控制历史数据进行多代偏差和偏差变化量分析处理，来评价当前控制的有效性及参数设置的合理性；并与专家控制系统结合构成一种参数自整定机器学习方法，实现设备智能控制。基于该方法设计出自动收集数据集、进行贝叶斯学习训练及验证的程序算法。更多还原

【Abstract】 The naive Bayes classifier to analyze and process the multi-generation error and error variation of the PID temperature control without oxidation oven historical data are used to evaluate the effectiveness of the current control and the rationality of the parameter settings. Combined with expert control system, a parameter self-tuning machine learning method is proposed to realize intelligent control of equipment. Based on this method, a program algorithm for automatic data collection and Bayesian learning training and verification is designed.更多还原