Deconvolution-based real-time neutron flux reconstruction for Self-Powered Neutron Detector

• Published in: Nuclear engineering and design Vol. 326; pp. 261 - 267
• Main Authors: Zhang, Qingmin, Deng, Bangjie, Liu, Xinxin, Li, Chengyuan, Sang, Yaodong, Cao, Liangzhi, Bi, Guangwen, Tang, Chuntao, Zhang, Peng, Tong, Dayin, Li, Yang
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01.01.2018; Elsevier BV
• Subjects: Compensation; Convolution; Deconvolution; Delay compensation; Dynamic models; Fluctuations; Impulse response; Initial conditions; Iterative methods; Neutron counters; Neutron flux; Neutrons; Nuclear engineering; Nuclear reactors; Nuclides; Reactors; Real time; Real-time neutron flux reconstruction; Response functions; Ruggedness; Self-Powered Neutron Detector; Delay compensation; Deconvolution; Self-Powered Neutron Detector; Real-time neutron flux reconstruction
• ISSN: 0029-5493; 1872-759X
• DOI: 10.1016/j.nucengdes.2017.11.024

•A deconvolution-based method was developed to compensate SPND response delay.•The compensation performance for jump neutron flux is only 0.3 seconds.•Stability analysis shows this method is robust even if noise current is added.•This method has been compared and validated, showing its advantages.•This method has high intuitiveness, good simplicity and strong applicability.•This method can even predict SPND current for known neutron flux. Self-Powered Neutron Detector (SPND) is useful for in-core neutron flux measurement in nuclear reactors due to its tiny size, simple structure, ruggedness and self-powered feature. One type of SPNDs with delayed current from instable intermediate nuclides cannot directly represent the real-time in-core neutron flux Φ(t) by their current I(t), which should be avoided during reactor control and protection. In this paper, we proposed a deconvolution-based method to reconstruct real-time neutron flux for SPND. Following the establishment of dynamic model, the unit-impulse response function h(t) was easily obtained when neutron flux was unit-impulse. Then, the iterative compensation relations were established for delay compensation according to the convolution relationship I(t) = Φ(t) * h(t). In the meanwhile, determination methods for initial values were also proposed and the compensation performance for jump neutron flux was demonstrated to be only 0.3 s. Furthermore, the dependences on initial conditions and sampling time interval were studied systematically, indicating our method is effective and robust. Finally, our method has been compared with a typical compensation method and validated with measured current, showing its advantages. This method is very attractive due to its obvious simplicity, high intuitiveness and general applicability.