The development and utilization of low-altitude airspace are gradually becoming a strategic pillar for advancing the low-altitude economy. However, the massive influx of low-altitude aircraft poses unprecedented safety challenges over urban areas. How to scientifically, comprehensively, and accurately quantify the risk distribution in complex urban low-altitude environments, and subsequently generate safe and efficient air routes, has become a critical scientific problem. Existing research often overlooks the multi-dimensional coupled risks between urban low-altitude and ground space, and air route generation algorithms are typically confined to local search, making it difficult to achieve accurate risk quantification and globally optimal route generation. To this end, this paper proposes a low-altitude risk quantification method and an intelligent route generation model constrained primarily by ground risk. Firstly, by fusing three types of data—population density, building density, and land use type—we propose a low-altitude risk factor quantification and combined weighting method to construct a high-medium-low multi-level low-altitude risk distribution map. Secondly, an intelligent air route generation model based on adaptive particle swarm optimization (APSO-AR) is proposed. By incorporating an adaptive air route node construction, a risk perception sampling mechanism, and an interval penalty mechanism, the model achieves global risk perception and optimal route generation for low-altitude aircraft within complex urban scenes. Finally, air route generation experiments were conducted using Wuhan city as the research area. The results demonstrate that, compared to ant colony optimization (ACO), genetic algorithm (GA), risk A* algorithm, and the traditional particle swarm optimization (PSO), the air routes generated by APSO-AR show an average reduction of 12.0% in route length and an average reduction of 30.7% in risk level, while the average code runtime is reduced by 77.4%. The model simultaneously ensures route optimality, risk minimization, and global rationality. This paper provides a new technical path for “risk quantification—route generation”, providing methodological support for urban low-altitude risk map construction and route planning.
