Controlled-swap-based and Knowledge Navigated Quantum-inspired Computational Intelligence for Quantum Circuit Optimization

Quantum and reversible circuit optimization stand as pivotal areas in quantum computing, crucial for error mitigation, size reduction, and cost minimization. In this scenario, computational intelligence (CI) is an efficient approach for improving circuit optimization issues and constructing equivalent small-scale quantum algorithms. In this paper, we propose a novel quantum-inspired computational intelligence method to automatically synthesize quantum circuits, benefiting reliable quantum computing realization. The proposed algorithm uses the feature of quantum mechanics to enhance the exploration ability of computational intelligence. It then detects situations when encountering local optima and reacts appropriately, drawing inspiration from the controlled-swap gate to navigate candidate solutions effectively and escape local optima with preference. Given the uncertain landscape of future quantum technologies and the diverse preferable gate sets with different platforms, our approach is a flexible algorithm that can simply adapt to different gate sets. The experimental results demonstrate that our method achieves superior performance compared to previous synthesis methods, with fewer circuit costs and more stable efficiency. Furthermore, we provide all circuits synthesized by our novel method to facilitate future studies to conduct in-depth research, simplify circuits, and map to different logic gate sets. This study advances reliable quantum computing through a quantum-inspired CI to mutually benefit both realms, promoting insight into interdisciplinary collaboration.