未知放射源辐射场构造与智能随机搜索及其优化研究
发布时间：2018-08-23 19:33
【摘要】：人类合理开发利用核能过程中,就开始研究核放射源安全问题。对于丢失、被盗放射源等安全事故发生情况下,如何还原未知核放射源辐射场以及如何在高辐射环境下,快速搜索并清除散落区域内的核放射性物质是核技术运用中重要的安全保障问题。对核安全应急响应策略具有重大现实意义。丢失、被盗的核放射性物质搜索大多局限于从事核技术专业的工作者,借助伽马射线检测仪器或其他射线探测仪,凭经验来搜索放射源的位置。但由于放射源搜索难度大或者特殊环境情况下的搜索效率低,放射源危害人们的身体安全等问题是制约放射源搜索的技术发展瓶颈问题。在核能产业迅猛发展和核技术广泛应用的今天,核放射源的安保研究,尤其是未知放射源的智能搜索研究,显得更加重要。本文从研究核放射源辐射场的物理特性入手,考虑到核辐射场自身的特点,利用数学构造方法,构造了稀疏数据下具有核放射源特性的辐射场函数。采用还原的未知放射源辐射场数据,设计了启发式智能搜索算法,实现了计算机辅助智能搜索,提出了人机互动模式下放射源搜索技术路线。有效降低了搜索放射源的辐照风险和搜索成本；有利于快速降低和污染区域辐射水平；有利于保护公众的人身安全和环境生态安全。本研究主要从以下四个方面展开：(1)对未知放射源存在的区域进行进行网格划分,并采集网格节点数据。在充分考虑到辐射场内任意点辐射剂量值与放射源距离平方成反比关系,运用数学构造方法还原放射源辐射剂量场,构造了网格线上斜率逐段连续变化的网格线上的辐射剂量曲线,同时结合邻点辐射剂量占优方法构造了辐射剂量场来反演还原辐射剂量场,得到了具有放射源特征的尖陡的凹函数辐射场,从而得到辐射场内任意点近似辐射剂量值。模拟实验计算得反演构造的辐射场与实际真实辐射场误差3.19%(见P47误差分析实验),反演构造辐射剂量函数形态更具放射源特征,因此构造辐射场方法切实可行且更接近真实辐射剂量场。提出了数学构造法反演还原核辐射场的思路。(2)在核放射源存在的区域范围,对放射源搜索区域进行了定义。同时针对放射源搜索问题中如何判断搜索到的辐射剂量极大值是否是放射源进行了研究,构造了适合不同情况的阈值算法理论来进行判断。以放射源辐射剂量最高处为对称的凹函数特性设计的确定放射源判断的数学算法为主。(3)在建立核辐射剂量场的理论基础上,为了能在较短的时间内快速搜索到未知放射源,本文选用了启发式算法,设计了线扫搜索算法。鉴于放射源仅出现在辐射剂量极大值点处,改进了启发式线扫算法,设计了仅局部拟合波峰段辐射剂量曲线的二次拟合线扫算法,简化了搜索过程,提高了搜索的运行效率；针对由于数据误读可能造成放射源搜索失败,设计了三次拟合线扫搜索算法,提高了搜索成功率。同时模拟实验验证了线扫算法,二次拟合和三次拟合线扫算法及其优化算法均能在合理时间内较好实现放射源搜索。(4)提出随机搜索概率优化模型。模仿典型常规搜索放射源的经验,并在计算机上实现典型的搜索过程。将启发式搜索思想与随机搜索结合,使得随机搜索适应实际搜索中各种复杂情况,设计了带有自动识别校正数据误读、大步长快速搜索以及跳出低辐射剂量区域等功能的智能随机搜索算法。特别对于多个放射源并存区域情况进行了深入讨论,设计了单链式和多链式智能启发式随机搜索算法。对于有障碍物的情况,设计了具有追踪辐射剂量极大方向随机搜索算法,使得随机搜索算法具有较强的适应性。最后在计算机上模拟了上述多种复杂情况,利用计算机智能搜索处于障碍区内的多点放射源,验证了随机搜索算法能有效解决放射源搜索问题,较好实现了计算机对放射源的准确定位功能。随机搜索算法节约数据采集量,降低辐照风险,同时在计算机上实现人机互动智能随机搜索。
[Abstract]:In the process of rational development and utilization of nuclear energy, people begin to study the safety of nuclear radiation sources. How to restore the radiation field of unknown nuclear radiation sources and how to quickly search and remove the radioactive materials in scattered areas under high radiation environment are important in the application of nuclear technology in the case of loss and stolen radioactive sources. Safety and security issues. It is of great practical significance for nuclear safety emergency response strategies. Lost, the search for stolen nuclear and radioactive materials is mostly confined to workers engaged in nuclear technology, with the help of gamma ray detection instruments or other ray detectors, to search the location of radioactive sources by experience. However, the search for radioactive sources is difficult or special. Nowadays, with the rapid development of nuclear energy industry and the wide application of nuclear technology, the security research of nuclear radioactive sources, especially the intelligent search of unknown radioactive sources, is becoming more and more important. The physical properties of radiation field of nuclear radiation source are studied. Considering the characteristics of nuclear radiation field itself, the radiation field function with nuclear radiation source characteristics under sparse data is constructed by using mathematical construction method. This paper puts forward the technical route of searching radioactive sources under man-machine interaction mode, which effectively reduces the radiation risk and searching cost of searching radioactive sources, helps to rapidly reduce and pollute the radiation level in the region, and helps to protect the public's personal safety and environmental ecological safety. The radiation dose field of the radiation source is restored by mathematical construction method, and the radiation dose curve on the grid line with continuous gradient variation is constructed. A radiation dose field is constructed by combining the adjacent radiation dose dominance method to retrieve the reduced radiation dose field, and the sharp concave function radiation field with the characteristics of the radiation source is obtained. The approximate radiation dose value at any point in the radiation field is obtained. The error between the retrieved radiation field and the actual radiation field is 3.19% (see P47). The error analysis experiment shows that the inversion of the structure radiation dose function is more characteristic of radiation source, so the method of constructing radiation field is feasible and closer to the real radiation dose field. This paper studies how to judge whether the maximum radiation dose is a radioactive source in the problem of searching for radioactive sources, and constructs a threshold algorithm theory suitable for different situations to judge whether the maximum radiation dose is a radioactive source. Based on the theory of radiation dose field, in order to search the unknown radiation source quickly in a short time, this paper chooses the heuristic algorithm and designs a line scan algorithm. The sub-fitting line sweeping algorithm simplifies the search process and improves the search efficiency; aiming at the possible failure of radioactive source search due to data misreading, a cubic fitting line sweeping search algorithm is designed to improve the search success rate. Simultaneously, the line sweeping algorithm, the quadratic fitting and the cubic fitting line sweeping algorithm and its optimization algorithm are verified by simulation experiments. (4) A random search probability optimization model is proposed, which imitates the experience of typical conventional search for radioactive sources and realizes the typical search process on a computer. Intelligent random search algorithm with the functions of automatic identification and correction of data misreading, large step-size fast search and jumping out of low radiation dose area is designed. Especially for the case of multiple radiation sources coexisting in the region, a single-chain and multi-chain intelligent heuristic random search algorithm is designed. Random search algorithm for maximum direction of radiation dose tracing makes the random search algorithm have strong adaptability. Finally, the above-mentioned complex situations are simulated on a computer, and multi-point radiation sources in obstacle areas are searched intelligently by computer. The results show that the random search algorithm can effectively solve the problem of radiation source searching and better realize the design. Random search algorithm saves the amount of data acquisition and reduces the risk of radiation. At the same time, it realizes intelligent random search of man-machine interaction on computer.
【学位授予单位】：南华大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TL73
本文编号：2199695
[Abstract]:In the process of rational development and utilization of nuclear energy, people begin to study the safety of nuclear radiation sources. How to restore the radiation field of unknown nuclear radiation sources and how to quickly search and remove the radioactive materials in scattered areas under high radiation environment are important in the application of nuclear technology in the case of loss and stolen radioactive sources. Safety and security issues. It is of great practical significance for nuclear safety emergency response strategies. Lost, the search for stolen nuclear and radioactive materials is mostly confined to workers engaged in nuclear technology, with the help of gamma ray detection instruments or other ray detectors, to search the location of radioactive sources by experience. However, the search for radioactive sources is difficult or special. Nowadays, with the rapid development of nuclear energy industry and the wide application of nuclear technology, the security research of nuclear radioactive sources, especially the intelligent search of unknown radioactive sources, is becoming more and more important. The physical properties of radiation field of nuclear radiation source are studied. Considering the characteristics of nuclear radiation field itself, the radiation field function with nuclear radiation source characteristics under sparse data is constructed by using mathematical construction method. This paper puts forward the technical route of searching radioactive sources under man-machine interaction mode, which effectively reduces the radiation risk and searching cost of searching radioactive sources, helps to rapidly reduce and pollute the radiation level in the region, and helps to protect the public's personal safety and environmental ecological safety. The radiation dose field of the radiation source is restored by mathematical construction method, and the radiation dose curve on the grid line with continuous gradient variation is constructed. A radiation dose field is constructed by combining the adjacent radiation dose dominance method to retrieve the reduced radiation dose field, and the sharp concave function radiation field with the characteristics of the radiation source is obtained. The approximate radiation dose value at any point in the radiation field is obtained. The error between the retrieved radiation field and the actual radiation field is 3.19% (see P47). The error analysis experiment shows that the inversion of the structure radiation dose function is more characteristic of radiation source, so the method of constructing radiation field is feasible and closer to the real radiation dose field. This paper studies how to judge whether the maximum radiation dose is a radioactive source in the problem of searching for radioactive sources, and constructs a threshold algorithm theory suitable for different situations to judge whether the maximum radiation dose is a radioactive source. Based on the theory of radiation dose field, in order to search the unknown radiation source quickly in a short time, this paper chooses the heuristic algorithm and designs a line scan algorithm. The sub-fitting line sweeping algorithm simplifies the search process and improves the search efficiency; aiming at the possible failure of radioactive source search due to data misreading, a cubic fitting line sweeping search algorithm is designed to improve the search success rate. Simultaneously, the line sweeping algorithm, the quadratic fitting and the cubic fitting line sweeping algorithm and its optimization algorithm are verified by simulation experiments. (4) A random search probability optimization model is proposed, which imitates the experience of typical conventional search for radioactive sources and realizes the typical search process on a computer. Intelligent random search algorithm with the functions of automatic identification and correction of data misreading, large step-size fast search and jumping out of low radiation dose area is designed. Especially for the case of multiple radiation sources coexisting in the region, a single-chain and multi-chain intelligent heuristic random search algorithm is designed. Random search algorithm for maximum direction of radiation dose tracing makes the random search algorithm have strong adaptability. Finally, the above-mentioned complex situations are simulated on a computer, and multi-point radiation sources in obstacle areas are searched intelligently by computer. The results show that the random search algorithm can effectively solve the problem of radiation source searching and better realize the design. Random search algorithm saves the amount of data acquisition and reduces the risk of radiation. At the same time, it realizes intelligent random search of man-machine interaction on computer.
【学位授予单位】：南华大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TL73
【相似文献】
中国重要会议论文全文数据库 前1条
1 冯yN;李霞;;一种K最近邻分类的改进算法及应用[A];2011年全国通信安全学术会议论文集[C];2011年
中国博士学位论文全文数据库 前1条
1 张敏;未知放射源辐射场构造与智能随机搜索及其优化研究[D];南华大学;2016年
中国硕士学位论文全文数据库 前3条
1 吕丹彦;群体约束情境下随机搜索行为建模与分析[D];东南大学;2016年
2 符国庆;一种随机搜索优化算法[D];北京交通大学;2007年
3 解瑞飞;基于启发式搜索的生物特征辨识算法研究[D];杭州电子科技大学;2012年
，本文编号：2199695
本文链接：https://www.wllwen.com/kejilunwen/anquangongcheng/2199695.html
