【摘要】 本文根据状态的分类特征,利用PNN神经网络作为检测系统的程序基础,结合前期研究的电压电流信号,进行编程学习和训练得到合格的算法,然后利用实验条件和参数搭建LABVIEW虚拟平台,并与程序共同实现对间隙放电状态的实时检测。实验部分主要是采用15 ml/min蒸汽水雾介质切割中厚度工件单因素实验,研究精加工5个主要加工参数(脉宽、峰值电流、脉冲间隔比、偏移量和工作台空载速度)对3个加工质量评定指标(表面粗糙度、切割速度和火花率分布)影响趋势。结果表明:脉冲宽度与切割速度成正比例对应趋势,但是随着脉宽的递增,粗糙度也将变大;脉冲间隔比对粗糙度和切割速度的影响变动不如脉宽强烈,变化趋势是先随着间隔比的增加先减少后变大;峰值电流的递增使切割速度和粗糙度数值上同样正比例递增,但表面粗糙度受峰值电流影响很大;切割速度随着偏移量的增加而减少,而粗糙度则随着偏移量的增加先减少后增加;粗糙度随着空载速度的增加先减少后增加。通过以上结论得出最佳的加工参数。更多还原

【Abstract】 According to the classification characteristics of the state, by using the PNN neural network as the program basis of the detection system, and combined with the voltage and current signals studied in the previous period, the programming learning and training were carried out and a qualified algorithm was obtained. Then, the LABVIEW virtual platform was built with experimental conditions and parameters, and the real-time detection of the gap discharge state was realized together with the program. To study the impact trend of the five main processing parameters(pulse width, peak current, pulse interval ratio, offset and workbench load speed) on the three machining quality assessment indicators(surface roughness, cutting speed and distribution of spark rate), the single factor experiments were conducted to workpiece with moderate thickness in the 15 ml/min steam mist medium. The results show that pulse width corresponds to the trend of cutting speed, but the roughness will increase with the increase of pulse width. The influence of pulse interval ratio on roughness and cutting speed is not as strong as that of pulse width. The increase of peak current increases the cutting speed and roughness in the same positive proportion, but the surface roughness is greatly affected by the peak current. The cutting speed decreases with the increase of offset and the roughness decreases with the increase of offset. The roughness decreases first and then increases with the increase of the empty load speed. Through the above conclusion, the optimum processing parameters are obtained.更多还原