Abstract: Highlights•We effectively combine EAX-GA and LKH through a special individual.•Combine reinforcement learning with key steps of EAX-GA and LKH to boost performance.•Extensive experiments show the superiority of our algorithm over advanced baselines.
Loading