<p>The precise measurement of the antineutrino spectra produced by isotope fission in reactors is of great significance for studying neutrino oscillations, refining nuclear databases, and addressing the reactor antineutrino anomaly. In this paper, we report a method that utilizes a feedforward neural network (FNN) model to decompose the prompt energy spectrum observed in a short-baseline reactor neutrino experiment and extract the antineutrino spectra produced by the fission of major isotopes such as <InlineEquation ID="IEq1"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41365_2025_1746_Article_IEq1.gif" Format="GIF" Height="10" Rendition="HTML" Resolution="72" Type="Linedraw" Width="19" /> </InlineMediaObject> <EquationSource Format="TEX">$^{235}$</EquationSource> <EquationSource Format="MATHML"><math> <mmultiscripts> <mrow /> <mrow /> <mn>235</mn> </mmultiscripts> </math></EquationSource> </InlineEquation>U, <InlineEquation ID="IEq2"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41365_2025_1746_Article_IEq2.gif" Format="GIF" Height="10" Rendition="HTML" Resolution="72" Type="Linedraw" Width="19" /> </InlineMediaObject> <EquationSource Format="TEX">$^{238}$</EquationSource> <EquationSource Format="MATHML"><math> <mmultiscripts> <mrow /> <mrow /> <mn>238</mn> </mmultiscripts> </math></EquationSource> </InlineEquation>U, <InlineEquation ID="IEq3"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41365_2025_1746_Article_IEq3.gif" Format="GIF" Height="10" Rendition="HTML" Resolution="72" Type="Linedraw" Width="19" /> </InlineMediaObject> <EquationSource Format="TEX">$^{239}$</EquationSource> <EquationSource Format="MATHML"><math> <mmultiscripts> <mrow /> <mrow /> <mn>239</mn> </mmultiscripts> </math></EquationSource> </InlineEquation>Pu, and <InlineEquation ID="IEq4"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41365_2025_1746_Article_IEq4.gif" Format="GIF" Height="10" Rendition="HTML" Resolution="72" Type="Linedraw" Width="19" /> </InlineMediaObject> <EquationSource Format="TEX">$^{241}$</EquationSource> <EquationSource Format="MATHML"><math> <mmultiscripts> <mrow /> <mrow /> <mn>241</mn> </mmultiscripts> </math></EquationSource> </InlineEquation>Pu in the nuclear reactor. We present two training strategies for the model and compare them with the traditional <InlineEquation ID="IEq5"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41365_2025_1746_Article_IEq5.gif" Format="GIF" Height="19" Rendition="HTML" Resolution="72" Type="Linedraw" Width="19" /> </InlineMediaObject> <EquationSource Format="TEX">$\chi ^2$</EquationSource> <EquationSource Format="MATHML"><math> <msup> <mi>χ</mi> <mn>2</mn> </msup> </math></EquationSource> </InlineEquation> minimization method by applying them to the same set of pseudo-data corresponding to a total exposure of <InlineEquation ID="IEq6"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41365_2025_1746_Article_IEq6.gif" Format="GIF" Height="19" Rendition="HTML" Resolution="72" Type="Linedraw" Width="246" /> </InlineMediaObject> <EquationSource Format="TEX">$(2.9\times 5\times 1800)~\mathrm {GW_{th}\cdot tons\cdot days}$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mrow> <mo stretchy="false">(</mo> <mn>2.9</mn> <mo>×</mo> <mn>5</mn> <mo>×</mo> <mn>1800</mn> <mo stretchy="false">)</mo> </mrow> <mspace width="3.33333pt" /> <mrow> <msub> <mi mathvariant="normal">GW</mi> <mi mathvariant="normal">th</mi> </msub> <mo>·</mo> <mi mathvariant="normal">tons</mi> <mo>·</mo> <mi mathvariant="normal">days</mi> </mrow> </mrow> </math></EquationSource> </InlineEquation>. The results show that the FNN model not only converges faster and better during the fitting process but also achieves relative errors of less than 1% in the <InlineEquation ID="IEq7"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41365_2025_1746_Article_IEq7.gif" Format="GIF" Height="14" Rendition="HTML" Resolution="72" Type="Linedraw" Width="39" /> </InlineMediaObject> <EquationSource Format="TEX">$2-8$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mn>2</mn> <mo>-</mo> <mn>8</mn> </mrow> </math></EquationSource> </InlineEquation> MeV range in the extracted spectra, outperforming the <InlineEquation ID="IEq8"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41365_2025_1746_Article_IEq5.gif" Format="GIF" Height="19" Rendition="HTML" Resolution="72" Type="Linedraw" Width="19" /> </InlineMediaObject> <EquationSource Format="TEX">$\chi ^2$</EquationSource> <EquationSource Format="MATHML"><math> <msup> <mi>χ</mi> <mn>2</mn> </msup> </math></EquationSource> </InlineEquation> minimization method. The feasibility and superiority of this method were validated in the study.</p>

Extraction of fissile isotope antineutrino spectra using feedforward neural network

期刊论文
发表日期：

2025年7月16日

查看全文

Jian Chen,
Jun Wang,
Wei Wang,
Yue-Huan Wei

摘要

The precise measurement of the antineutrino spectra produced by isotope fission in reactors is of great significance for studying neutrino oscillations, refining nuclear databases, and addressing the reactor antineutrino anomaly. In this paper, we report a method that utilizes a feedforward neural network (FNN) model to decompose the prompt energy spectrum observed in a short-baseline reactor neutrino experiment and extract the antineutrino spectra produced by the fission of major isotopes such as $^{235}$ $^{235}$ U, $^{238}$ $^{238}$ U, $^{239}$ $^{239}$ Pu, and $^{241}$ $^{241}$ Pu in the nuclear reactor. We present two training strategies for the model and compare them with the traditional $\chi ^2$ $χ^{2}$ minimization method by applying them to the same set of pseudo-data corresponding to a total exposure of $(2.9\times 5\times 1800)~\mathrm {GW_{th}\cdot tons\cdot days}$ $(2.9 \times 5 \times 1800) {GW}_{th} \cdot tons \cdot days$ . The results show that the FNN model not only converges faster and better during the fitting process but also achieves relative errors of less than 1% in the $2-8$ $2 - 8$ MeV range in the extracted spectra, outperforming the $\chi ^2$ $χ^{2}$ minimization method. The feasibility and superiority of this method were validated in the study.