高计数率下的实时梯形成形算法
Real time trapezoidal shaping algorithm at high count rates

针对常见的指数型核脉冲信号，研究了高计数率下的梯形成形算法，特别针对计数率增大所带来的堆积问题进行了研究。提出了一种利用反卷积定位信号到达时刻的算法，有效提升了对堆积脉冲的识别能力，降低了堆积效应对能量分辨率的影响。结合上位机中用LabVIEW编写的采集软件，搭建了一套完整可靠的脉冲幅度分析系统，将梯形成形、堆积判别、能谱构建等一系列功能在现场可编程门阵列（FPGA）中实时实现。将该系统实际应用于使用硅漂移探测器的X射线衍射仪上，获得了良好的结果，在1.8×105 cps（每s计数）的条件下，所得能量分辨率达2.46%。
Abstract：A trapezoidal shaping algorithm was developed for common exponential nuclear signals and the pile-up effect at high count rate was studied. Deconvolution was used for time-of-arrival estimates to effectively improve the ability to recognize pile-up pulses and reduce the pile-up effect on the energy resolution. A complete pulse height analyzer was built in LabVIEW to realize real time trapezoidal shaping, pile-up recognition and spectrum construction in the field-programmable gate array (FPGA). Measurements with an X-ray diffractometer with a silicon drift detector (SDD) give a 2.46% energy resolution at a count rate of 1.8×105 cps (counts per second).