摘要
将Smith预估器与PID控制结合应用到芯片高温固化设备温控系统中,用以改善设备的大惯性、时变时滞问题;针对Smith预估参考模型不准确带来的控制劣化效应,引入机器学习中的BP神经网络理论建立在线参考预估模型,最大限度地消除了系统特征方程中的滞后环节,优化了Smith预估器的前馈补偿能力;在成本不变的前提下,改进了设备的综合性能,具备较高的推广价值。
The Smith predictor combined with PID control is applied to the temperature control system of the chip high-temperature curing equipment to improve the large inertia and time-varying delay problems of the equipment.Aiming at the control deterioration effect caused by the inaccuracy of the Smith prediction reference model,the BP neural network theory in machine learning is introduced,and the online reference prediction model is established to minimize the lag in the system characteristic equation and optimize the Smith predictor.On the premise of constant cost,the comprehensive performance of the equipment is improved,and it has high promotion value.
作者
梁达平
王鸿斌
赵利民
LIANG Daping;WANG Hongbin;ZHAO Limin(School of Electronic Information and Electrical Engineering,Tianshui Normal University,Tianshui 741000,China;Tianshui Huatian Technology Co.,Ltd.,Tianshui 741000,China)
出处
《电子工业专用设备》
2022年第3期21-27,33,共8页
Equipment for Electronic Products Manufacturing
基金
甘肃省教育厅教育揭榜挂帅项目(2021jyjbgs-06)
甘肃省高等学校创新基金项目(2021A-102)。
关键词
PID控制
智能控制
封装测试
芯片烘箱
Proportional-integral-derivative(PID)control
Smith prediction
Neural network control
Chip oven