氘氚聚变反应历史测量系统优化设计及亚纳秒信号恢复方法研究

针对传统光路追迹设计方法的不足,建立了遗传算法调用Geant4程序优化设计方法并应用于气体切伦科夫系统的优化设计,取得了较传统光路追迹设计方法优良的系统时间响应和探测效率,对于GCD(gas Cherenkov detector)系统,优化设计后探测效率增加20%且时间响应压缩7.2%,对于GRH(gamma reaction history)系统,优化结果效率较传统光路追迹方法相对提高91.2%。针对聚变测量过程,通过比较研究,确立了MRNSD(modified residual norm steepest descent)方法为聚变反应历史测量过程信号恢复方法,该方法可实现在响应函数半高宽10倍于待恢复信号半高宽,且存在相对信号峰值10%高斯白噪声情况下的亚纳秒信号恢复,适合于氘氚聚变反应历史测量。

Optimization Design on Fusion Reaction History Diagnostic System and Study on Recovery Method of Sub-nanosecond Signal

An optimization method combining Geant4 code with genetic algorithm was established and applied to designs of two types of gas Cherenkov detectors (GCDs) because of deficiencies of the traditional optical ray-tracing method.Improved efficiencies and time responses were obtained compared with results of traditional optical ray-tracing method.The efficiency of GCD system through optimization is enhanced by 20％ while time response is eliminated by 7.2％.The efficiency of GRH (gamma reaction history) system through optimization is enhanced by 91.2％.Aiming at fusion diagnostic process,the MRNSD (modified residual norm steepest descent) method was ascertained as signal recovery method for fusion reaction history diagnosis through comparison study.The MRNSD method can recover sub-nanosecond signal distorted by response function of which the FWHM (full-width at half maximum) is 10 times of the signal to be recovered and with 10％ Gaussian white noise added,which is suitable for fusion reaction history diagnosis.