可编程约瑟夫森结阵偏置组合算法研究

Research on the Bias Combination Algorithm for Programmable Josephson Junction Array

在可编程约瑟夫森量子电压标准系统中,快速并准确地计算各约瑟夫森子结阵的偏置状态是实现量子电压精准输出的前提。传统的枚举-索引算法通过遍历所有可能的偏置组合以寻找最优解,因此计算效率较低,难以适应量子电压非周期性连续变换输出等特殊应用场景。基于此,提出了一种面向三进制可编程约瑟夫森结阵的偏置组合算法。首先根据各子结阵中包含的约瑟夫森单结数量特征对子结阵进行自适应分组,并据此快速计算预偏置组合;进一步地,结合近似计算方法和量化误差补偿方法,克服预偏置组合中可能存在的子结阵失效、亚分辨率设计等特殊问题。实验结果表明,该算法在多种条件下均能找到最优偏置组合,并且计算效率明显优于传统的枚举-索引算法。

In the programmable Josephson quantum voltage standard system,fast and accurate calculation of the bias combinations of Josephson junction sub-arrays is the premise to achieve the accurate output of quantum voltage.The traditional enumeration-index algorithm searches for the optimal solution by traversing all possible bias combinations,resulting in low computational efficiency and difficulty in adapting to special application scenarios such as quantum voltage aperiodic continuous transformation outputs.Based on the above,a bias combination algorithm is proposed for the ternary programmable Josephson junction array.Firstly,the sub-arrays are grouped adaptively according to the number of Josephson junctions in each one,and then the preliminary bias combinations are quickly calculated.Furthermore,the approximate calculation method and quantization error compensation method are combined to overcome some specific problems in the preliminary bias combinations,such as sub-arrays failure,sub-resolution design,and so on.Experimental results show that the mentioned algorithm can find the optimal bias combination under various conditions,and the computational efficiency is much better than the traditional enumeration-index algorithm.