中文摘要：

为解决微细电火花加工过程中由于频繁出现的放电信号严重畸变、放电状态不稳定甚至突变等造成的放电状态难于准确检测的技术难点，在分析和研究传统的微细电火花加工放电状态检测方法的基础上，结合系统辨识和模糊逻辑理论，提出了微细电火花加工放电状态逐级映射检测原理和方法。对实时采集到的极间电压和电流信号，通过模糊运算判别采样点的放电状态，再将采样点放电状态值映射为放电状态矢量，并对该矢量进行统计得到“短路率”和“火花／电弧率”，经过模糊推理辨识出各分析周期的放电状态。实验表明，该检测方法准确性高、运算量低且运算速度快，与平均电压法相比，效率提高22．2％。检测结果可为微细电火花放电加工过程的实时控制提供系统放电状态的反馈输入，保证了加工控制系统的稳定性和准确性。

英文摘要：

To measure precisely the discharge state and overcome the shortcomings of the distortion of discharge signals and the instability of discharge state, the principle and method of successive mapping detection were proposed combined with the system identification and fuzzy control after analysis on traditional discharge state detection methods in micro Electrical Discharge Machining（EDM）. For real-time collected gap voltage and current signals in the process, the fuzzy operation was used to identify the discharge state of a sampling point and then to map the sampling point discharge state value into sampling point discharge state vector. Furthermore,the vector was counted to obtain the＂Short rate＂ and ＂spark/arc rate＂ and the fuzzy reasoning was used to identify the discharge state of each cycle. Test results show that the presented detection method has highly accuracy and is able to identify data and operate fast. Compared with that of the average voltage detection method, the efficiency has been increased by 22.2%. Detection results can provide feed back inputs of the system discharge states for the real-time control of the discharge process, which ensures the stability and accuracy of the process-ing control system.