In gas field environments, the scheduling of inspection drones becomes highly complex and difficult to implement efficiently due to the coexistence of multiple inspection tasks and the large distances between task points. To address this issue, an improved genetic algorithm was proposed, incorporating an elitist retention strategy. A matrix element encoding method was adopted, which significantly reduced computational time and ensured the integrity of chromosomes in subsequent calculations. After encoding, the fitness of each individual was calculated, and individuals were grouped based on their fitness, thereby enhancing the accuracy of the results. Furthermore, an improved roulette wheel selection method was introduced to ensure the diversity of the population, thereby reducing the risk of the algorithm falling into local optima. Simulation experiments were conducted based on a gas field in western Sichuan and a gas field in the Ordos Basin. The results show that the optimized genetic algorithm saves, on average, 17.1% and 11.3% in cost compared to traditional genetic algorithms and ant colony algorithms, respectively, when calculating the scheduling of inspection drones.