【作者】 刘成洋；

【导师】 阎昌琪；

【作者基本信息】 哈尔滨工程大学 ， 核能科学与工程， 2013， 博士

【摘要】 核动力装置因其具有良好的续航力，较大的推进功率，以及高能量密度和核裂变不需要氧气等优点，被广泛应用在船舶及航天工业上。为了包容放射性，核动力装置一回路还有许多辅助系统；由于二回路使用饱和蒸汽，汽轮机组重量体积大，系统热效率低。综上原因，导致整个动力装置笨重且经济性差。因此利用优化技术降低核动力装置的重量和体积并提高其有效效率，无疑具有重要的理论和现实意义。本文以核动力装置系统为研究对象，建立系统主要设备的数学模型，开发高效的智能优化算法，开展了以减小核动力装置重量、体积和提高核动力装置有效效率为目标的优化设计工作，主要内容如下：1、建立了蒸汽发生器、稳压器、主汽轮机和主冷凝器这4个核动力设备的物理和数学模型以及核动力装置热平衡计算数学模型，采用C#语言编制了各模型的评价程序。在给定一些参数的条件下，通过评价计算得到的计算结果与母型参数相比误差较小，验证了各评价程序的可靠性。2、针对本课题研究对象的多变量、多约束和非线性的特点，综合分析局部搜索能力强的复合形算法、全局搜索能力强的遗传算法和粒子群算法，开发了一种新型混合粒子群算法。新型混合粒子群算法采用可行性规则进行约束处理并对群体排序，改进了遗传算法的交叉机制，提高了复合形算法跳出局部最优的能力。最后利用国际上广泛使用的8个无约束测试函数、13个约束测试函数和3个工程优化问题对算法进行了测试。3、以蒸汽发生器、稳压器、主汽轮机和主冷凝器为例，探讨核动力装置中单个设备的优化设计过程。在蒸汽发生器的优化设计实例中，选取了一回路工作压力、二回路饱和蒸汽压力和蒸汽发生器传热管外径等7个参数为优化变量，利用新型混合粒子群算法进行寻优，分别获得了蒸汽发生器重量、体积和双目标优化方案。在蒸汽发生器单一设备优化的基础上，考虑部分堆芯约束，对蒸汽发生器和堆芯进行了耦合优化。对其余三个设备，同样进行敏感性分析，并各自选取优化变量，给出了重量、体积和双目标优化方案。4、针对某核动力装置，选取冷凝器压力、高压缸排汽干度和低压缸排汽干度作为优化变量，以反应堆功率和蒸汽发生器总蒸汽产量在某一范围内变化作为约束条件，利用新型混合粒子群算法对核动力装置有效效率进行优化，优化后有效效率提高了3.1571%。5、通过简化核动力装置系统，并结合小组其他成员编制的反应堆堆芯和反应堆压力容器评价程序，编制了核动力装置全系统评价程序。探讨了全系统各设备之间耦合关键因素，基于系统全局性考虑，选取一回路工作压力、反应堆出口冷却剂温度、反应堆进口冷却剂温度、二回路饱和蒸汽压力和冷凝器压力这5个参数作为优化变量，对核动力装置系统重量和体积进行寻优，并对优化结果进行了分析。6、开发了核动力装置优化设计系统软件。该软件可进行核动力装置评价，优化算法测试，单一核动力设备优化和系统级核动力装置优化工作。在编程处理上，采用了最新的并行计算技术，加快计算速度。软件还为后续设备模块预留接口，具有可移植性强、界面友好和操作简单的特点。软件适用于现有系统的改进论证计算，以及新型核动力装置的初步设计计算。本文完成了核动力装置优化过程中的数学模型建立、评价程序开发、优化算法改进、优化模型完善、优化算例计算、优化结果分析和优化软件开发工作。论文研究初步形成了比较完整的核动力装置总体参数最优化设计方法，是优化技术在核动力装置设计中的有益尝试，也为进一步的优化设计理论及应用研究奠定了良好的基础。更多还原

【Abstract】 Because of its good endurance, high propulsive power, large energy density and free ofoxygen, etc., nuclear power is widely used in ship and aerospace industry. In order to protectthe environment from radioactivity, there are many auxiliary systems in primary loop ofnuclear power plant; steam turbine unit has heavy weight, big volume and low thermalefficiency due to the saturated steam used in secondary loop. In general, the performance ofnuclear power plant is bad in terms of compactness and economics. Therefore, it is significantboth theoretically and practically to reduce the weight, volume and to increase effectiveefficiency of nuclear power plant through optimization technology. In this thesis, themathematic model of main equipments were established for typical nuclear power plantsystem, intelligent optimization algorithm was developed, and optimization case study wascarried out for the purpose of reducing nuclear power plant weight, volume and increasingeffective efficiency of nuclear power plant. The main contents are as follows:1、The physical and mathematical models were set up for steam generator, pressurizer,main steam turbine and main condenser, as well as the mathematical model for thermalequilibrium of nuclear power plant. Corresponding codes were developed with C#language.Given some known parameters, the verification of the codes was performed throughcomparision between the evaluating design and the prototype.2、In the light of multivariables, many constraints and nonlinearity for the optimizationof nuclear power plant, a novel hybrid particle swarm optimization (NHPSO) algorithm wasdeveloped by incorporating the complex shape method with strong ability in local search,genetic algorithm and particle swarm optimization algorithm with good ability in globalsearch. Furthermore, the feasibility-based rule was introduced into NHPSO for the constraintshandling and sorting of the group. The crossover operator was modified and the escape abilitywas improved from local optimal point for complex shape algorithm. The NHPSO algorithmwas tested with8unconstrained test functions,13constrained test functions and3engineeringoptimization problems which are widely benchmarked in the world.3、Case studies were carried out for such single equipment optimization as steamgenerator, pressurizer, main steam turbine and main condenser. In steam generator optimizing design,7optimization variables were selected as follows: operating pressure in primary loop,saturated steam pressure in secondary loop and tube outer diameter in steam generator, etc..Weight optimization, volume optimization and multi-object optimization were performed withNHPSO algorithm. Besides, the coupled optimization of steam generator and reactor core wascarried out by adding some reactor core constraints. For the other three equipments, weightoptimization, volume optimization and multi-object optimization design schemes were alsoachieved after sensitivity analysis and the selection of their respective optimization variables.4、For a certain nuclear power plant prototype, condenser pressure, steam dryness of bothhigh pressure turbine exhaust and low pressure turbine exhaust were selected as theoptimization variables. The reactor power and total steam generated in steam generator wereconstrained within certain ranges. The efficiency was optimized with NHPSO algorithm andthe efficiency was improved by about3.16%.5、In this thesis, proper simplification was made for the nuclear power plant. Theevaluation codes of nuclear power plant were developed by the combination of the codesprovided by other members of the group for the reactor core and reactor pressure vessel. Afterthe key coupling factors were investigated for the whole system, operating pressure inprimary loop, coolant temperature at both core inlet and outlet, saturated steam pressure insecondary loop, and condenser pressure were selected as optimization variables. The weightand volume optimization of nuclear power plant was carried out, and the optimization resultswere analyzed.6、The nuclear power plant optimization design system software has been developed. Thesoftware can serve for such work as follows: nuclear power equipment evaluation,optimization algorithm testing, single nuclear power equipment optimization and system-leveloptimization. In the software, the latest.NET4.0parallel computing technology was adopted,which can accelerate the calculation. The software also preserved interface for the follow-updevice modules. Furthermore, the software has good inheritance, friendly interface and simpleoperation. The software can be applied to the demonstration computation of the existingsystem, and the preliminary design of the new nuclear power plant.The mathematic model setups, evaluating codes development, optimization algorithmimprovement, optimization model consummation, optimization case study, optimizationresults analysis and optimization software development have been accomplished in this theis. To sum up, a relatively complete framework of design optimization of parameters for nuclearpower plant is formulated. All of these are beneficial attempt to the application ofoptimization technology in nulear power plant design and good foundations to furtherresearch on the theories and application of optimization technology.更多还原