迭代学习法在开关式液压阀位置控制中的应用

Position Control with Iterative Learning Method in Hydraulic System with Switch Mode Hydraulic Valve

阐述了一种新型振动下料机的基本组成,分析了其液压送进装置的传统行程开关控制方法所存在的问题,明确指出了采用迭代学习控制的必要性和可行性.研制出了由微机、开关式液压阀及位移传感器等组成的液压送进装置的计算机智能控制系统.研究表明,在迭代学习智能控制方法中所用的控制提前量呈现正态分布,从而建立了确定最优学习因子和控制提前量等关键控制参数的计算公式.试验结果表明,整个系统的鲁棒性好,采用该智能控制方法可使送进缸停止位置的误差由原来的±5 5mm降低为±0 8mm.因此,不仅满足了实际工作的要求,而且具有重要的推广价值.

The basic constituents of a new type of vibration shear-cutting machine are described. The shortages of the traditional control method for the hydraulic feeding mechanism of the machine with stroke switches are analyzed; therefore the feasibility and the necessity of iterative learning control of feeding position are explicitly pointed out. A computer intelligent control system, which consists of microcomputer, switch mode hydraulic valve and displacement sensor, is developed. The research results indicate that the advance controlled value shows the normal distribution. And the calculating equations of variable factor and advance controlled value are established simultaneously. The results of experiment show that the system has better robustness and enables the error of feeding displacement to decrease from ±5.5 mm to ±0.8 mm. The system meets the operating requirements and is worth spreading in practical production.