Solving Traveling Salesman Problem with Time Windows Using Hybrid Pointer Networks with Time Features

This paper introduces a time efficient deep learning-based solution to the traveling salesman problem with window (TSPTW). Our goal is reduce total tour length traveled by -*the agent without violating any limitations. will aid in decreasing required supply type of service, as well lowering emissions produced automobiles, allowing our planet recover from air pollution emissions. The proposed model variation pointer networks that has better ability encode TSPTW problems. this inspired previous work hybrid context encoder and multi attention decoder. primarily comprises transformer graph encoder; these encoders feature vector before passing it decoder layer. consists well. output attentions two decoders are aggregated used select following step trip. To best knowledge, network first neural be able solve medium-size Moreover, we conducted sensitivity analysis explore how performance changes width training testing data changes. experimental shows outperforms state-of-the-art for sizes 20, 50 100 nodes/cities. We expect become methodology solving