基于量子遗传算法优化的空间魔法波形装置动态网格坐标误差分析

Analysis of dynamic grid coordinate error of spatial magic waveform device based on quantum genetic algorithm optimization

研究了一种基于量子遗传算法的空间魔法波形装置的动态网格坐标误差优化计算。空间魔法波形是一种由等长木棍在三维空间拼接成动态网络并由绳索驱动的用于科普演示的装置,木棍网络的每个节点坐标的计算对于装置整体的设计和组装有重要意义。引入一种将量子计算概念(量子态、量子门、量子状态特性、概率幅等)融合到传统遗传算法中的优化算法——量子遗传算法,用量子位表示基因,再利用量子信息的叠加性使每个量子位表达基因的可能性相同,进而通过基因位的旋转完成基因的变异,每次完成变异后判断适应度是否满足条件,满足条件的解将持续更新,在不停地迭代计算下求出最终的最优解。利用量子遗传算法计算变化后的空间魔法波形木棍网络节点的z坐标比使用传统遗传算法和蒙特卡洛算法时的平均精度分别降低了0.027 0%和0.045 8%,该算法的使用提高了在机械设计、制造工艺、计量测试中坐标求解的精度。

In this paper,the optimal calculation of dynamic grid coordinate error for a kind of spatial magic waveform device based on quantum genetic algorithm(QGA)is studied.The space magic waveform is a dynamic network of equal-length wooden sticks stitched together at the 3D and driven by ropes for popular science demonstrations.The calculation of the coordinates of each node of the stick network is very important for the design and assembly of the whole device.The QGA optimization algorithm integrates the concepts of quantum computation(quantum states,quantum gates,properties of quantum states,probability amplitudes,etc.)into the traditional genetic algorithm.The superposition of quantum information leads to a fact that each qubit expression gene has the same possibility,and then the gene variation can be completed through the rotation of the qubit.In each completion of the mutation,the variated gene judges whether the fitness meets the conditions.The solution is updated continuously to satisfy the condition,and the final optimal solution is obtained by iterative calculation.Experimental results indicate that average error of the proposed QGA method decreases by 0.0270%and 0.0458%as compared with that of the traditional genetic algorithm and Monte Carlo method,respectively.The proposed algorithm shows great potentials in the optimization calculation in the fields of mechanical design,manufacturing process,and measurement.