【作者】 殷巧玉；

【导师】 李伟力；

【作者基本信息】 哈尔滨理工大学 ， 电机与电器， 2011， 博士

【摘要】 量子计算是结合了信息科学和量子力学的新兴交叉科学，而以量子算法为代表的量子计算理论，由于其高度的并行性、指数级存储容量和对经典启发式算法的指数级加速作用，因而具有极大的优越性并且蕴涵着强大的生命力，现在已经成为世界各国相关学者研究的前沿热点领域。而将量子计算理论引入传统智能优化算法，改变了传统智能优化算法的迭代寻优方式，提高了传统智能优化算法的迭代收敛速度和全局寻优能力等性能。因此，研究量子智能优化算法及其在电机优化中的应用有着重要的理论和现实意义。本论文主要研究连续量子蚁群优化、连续量子粒子群优化、连续量子免疫克隆优化，以及量子智能优化算法在电机优化应用中的研究。具体可归纳如下。1．将量子计算原理和蚁群优化相融合，提出连续量子蚁群优化算法，数值仿真结果表明连续量子蚁群优化算法的迭代收敛速度和全局寻优能力明显优于蚁群优化算法。以一台117kW高速永磁发电机为例，通过温度场计算分析和量子智能优化算法对其冷却结构进行了优化设计研究。使电机内温度分布趋于均匀，并研究了流道高度和通道截面变化位置对电机内温度分布的综合作用影响。建立了双目标函数双维度变量的流道优化设计的连续量子蚁群优化数学模型，通过优化算法得到了定子绕组最高温度和轴向温差均为最小的流道结构方案。得出的规律性结论可为高速永磁发电机冷却系统的改进设计与研究提供参考。2．将量子计算原理和粒子群优化相融合，提出连续量子粒子群优化算法，仿真结果表明引入量子计算的相关理论，可明显提高算法的优化效率。建立了多优化目标的流道优化设计连续量子粒子群优化数学模型，以降低电机定子绕组最高温度和定子铁心轴向最高温度，减小定子绕组轴向温差和定子铁心轴向温差为目标。根据连续量子粒子群优化确定的流道优化最优方案，电机不同位置轴向最高温度和温差均显著减小，为高速永磁发电机内冷却结构的优化设计提供一种新的优化方法。3．将量子计算理论和免疫克隆算法相融合，提出连续量子免疫克隆优化，仿真结果表明连续量子免疫克隆优化的优化性能明显优于免疫克隆算法。深入研究了高速永磁发电机冷却结构双维度多目标优化。以降低电机定子绕组最高温度、转子最高温度和定子铁心轴向最高温度，减小定子绕组轴向温差、转子轴向温差和定子铁心轴向温差为目标，对流道dh和dl参数优化进行了研究。根据连续量子免疫克隆优化确定的流道优化最优方案，使得电机内轴向温度分布更趋均匀。更多还原

【Abstract】 Quantum computation is an emerging interdisciplinary science whichcombines information science and quantum mechanics. And quantumcomputation theory, represented by quantum algorithm, has great superiority andstrong vitality because it has a high degree of parallelism, the index level ofstorage capacity, and can speed up the classic heuristic algorithms, so it hasbecome the frontier research field for many relevant scholars around the word.The integration of traditional intelligent optimization algorithms and quantumcomputation theory changes iterative optimization method of traditionalintelligent optimization algorithms, improve iterative convergence rate andglobal search capability. Hence, the research on quantum intelligent optimizationalgorithm and its application to motor optimization has important theoretical andpractical significance. This dissertation mainly studied the continuous quantumant colony optimization、continuous quantum particle swarm optimization、continuous quantum immune clonal optimization、and the applications to motoroptimization, the specific content of the paper can be summarized as follows.1．Through integration of the quantum computation theory and the antcolony optimization, the continuous quantum ant colony optimization is proposed,the simulation results show that its convergence speed and approximation abilityare evidently superior to the ant colony optimization. The optimal design of a117kW level high speed permanent magnetic generator (HSPMG) cooingstructure is studied through thermal analysis and the quantum intelligentoptimization algorithm. The new cooling systems can make temperaturedistribute more evenly in HSPMG, and the influences of groove height and theaxial variation position on HSPMG temperature distributions are studied. Basedon the continuous quantum ant colony optimization, a mathematical optimizationmodel with dual objective functions and two dimensional variables for statorslots grooves optimal design is proposed, and a groove structure which could make both the windings axial largest temperature and the axial temperaturedifference to be the minimum ones is obtained by optimization algorithm. Theobtained conclusions may provide useful reference for the design and research ofcooling structure in HSPMG.2．Through integration of the quantum computation theory and the particleswarm optimization, continuous quantum particle swarm optimization isproposed, the simulation results show that the induction of quantum computationtheory can significantly improve the optimization efficiency of algorithm. Amathematical optimization model with multi-objective optimization for statorslots grooves optimal design is proposed based on the continuous quantum antcolony optimization, which could make the stator windings axial largesttemperature, the stator core axial largest temperature, the stator windings axialtemperature difference and the stator core axial temperature difference to be theminimum ones. According to the optimal solution of stator slots groovesdetermined by continuous quantum ant colony optimization, the motor axiallargest temperature and axial temperature difference in motor different positionsare decreased significantly, a new kind of optimization method is proposed forHSPMG cooling structure.3．Through integration of the quantum computation theory and the immuneclonal algorithm, the continuous quantum immune clonal optimization isproposed, and the simulation results show the optimization ability of continuousquantum immune clonal optimization is evidently superior to the immune clonalalgorithm. Multi-objective optimization with two dimensional variables forcooling structure in HSPMG is studied in depth. The stator windings axial largesttemperature, the stator core axial largest temperature, the rotor axial largesttemperature, the stator windings axial temperature difference, the rotor axialtemperature difference and the stator core axial temperature difference areselected as optimization goals, the stator slots grooves parameters dh and dloptimization are studied. The stator slots grooves optimal solution based on thecontinuous quantum immune clonal optimization make the motor axialtemperature distribution more uniform.更多还原