In the NISQ era, quantum computers have insufficient qubits to support quantum error correction, which can only perform shallow quantum algorithms under noisy conditions. Aiming to improve the fidelity of quantum circuits, it is necessary to reduce the circuit depth as much as possible to mitigate the coherent noise. To address the issue, we propose PaF, a Pattern matching-based quantum circuit rewriting algorithm Framework to optimize quantum circuits. The algorithm framework finds all sub-circuits satisfied in the input quantum circuit according to the given external pattern description, then replaces them with better circuit implementations. To extend the capabilities of PaF, a general pattern description format is proposed to make rewriting patterns in existing work become machine-readable. In order to evaluate the effectiveness of PaF, we employ the BIGD benchmarks in QUEKO benchmark suite to test the performance and the result shows that PaF provides a maximal speedup of 8× by using few patterns.

中文翻译：

---

基于模式匹配的量子电路优化框架

在 NISQ 时代，量子计算机的量子比特不足以支持量子纠错，只能在噪声条件下执行浅量子算法。为了提高量子电路的保真度，需要尽可能减小电路深度以减轻相干噪声。为了解决这个问题，我们建议氟化钠， 一种帕基于ttern匹配的量子电路改写算法F优化量子电路的框架。算法框架根据给定的外部模式描述找到输入量子电路中满足的所有子电路，然后用更好的电路实现替换它们。为了扩展能力氟化钠，提出了一种通用的模式描述格式，使现有工作中的重写模式变得机器可读。为了评估效果氟化钠，我们采用大的QUEKO 基准套件中的基准来测试性能，结果表明氟化钠提供了最大的加速8×通过使用一些模式。