探讨不同菜系在广州餐饮空间中的结构特征、抽象模型和影响因素，有助于为餐饮空间商业运营、文化交融提供地理学依据。采用地理学方法分析POI餐饮点数据的空间结构特征，并将其提炼为点轴结构模型、圈层洼地模型、尺度耦合模型和菜系竞争模型，同时揭示其经济和文化发展规律。研究发现：（1）广州餐饮空间结构呈现出多核心组团朝向东、南、北三个方向延伸；餐馆整体分布显著集聚，不同菜系集聚区域存在位置差异；餐饮圈层与城市人口和规划圈层相一致，核心-外围圈层呈四重阶梯式递减趋势；餐馆数量与多样性存在空间自相关性，高数值区域集聚在核心，而低数值区域散布外围等特征。（2）广州市餐饮空间抽象为多核组团、交通干线延伸外围的点轴结构模型；以20km为距离周期的最大辐射范围且服务中心间存在低数值洼地的的圈层衰减结构模型；不同尺度餐馆数量与类型数量存在规律差异的尺度耦合模型，即在区级尺度下，餐馆门类数几乎饱和，但门类数越多的区域，餐馆数量越多，呈现指数型趋势，而社区和街道尺度下则正态分布形式，即餐馆类型适中的区域，餐馆数量越多；空间尺度越大，本地原生菜系的空间优势越大，而在经济发达和本土文化发达的区域，各类型餐馆各具优势，呈现为整体均值和单位异质的菜系空间竞争模型。（3）受自然、经济、文化因素影响，山水阻隔是影响餐馆由核心向外围扩散的主要阻碍；不同菜系在其发展过程中表现不一，广东菜受冲击较大，本土菜发展迅猛，已在整体数量上超越广东菜，随着流动人口的增加和人口基数自身产生的多样性需求将进一步稳固其地位；外邦菜集中在消费较高、餐饮发展完善且口味包容的区域，进入经济和文化门槛相对较高；不同餐饮企业在进行空间选址的过程中，不能仅借助经济工具单纯量化人口数量及其消费能力，更应该辨析不同空间区域口味偏好、饮食包容度从而充分考虑其经营的可行性。

Exploring the structural characteristics, abstract models and influencing factors of different cuisines in Guangzhou's dining space helps provide a geographic basis for the commercial operation and cultural intermingling of dining spaces. A geographic approach is used to analyse the spatial structural characteristics of POI restaurant point data and to distil them into a point-axis structural model, a circle depression model, a scale coupling model and a cuisine competition model, while revealing their economic and cultural development patterns. The study found that: (1) The spatial structure of restaurants in Guangzhou shows multiple core groups extending towards the east, south and north; significant clustering in the overall distribution of restaurants, with location differences in the clustering areas of different cuisines; the restaurant circles are consistent with the city's population and planning circles, and the core-periphery circles show a fourfold decreasing trend; the number and diversity of restaurants are spatially autocorrelated, with high value areas clustered in the core and low value areas scattered in the periphery. (2) Guangzhou's restaurant space is abstracted into a point-axis structure model with multi-core clusters and traffic arteries extending to the periphery; a circle decay structure model with a maximum radiation range of 20km as the distance period and the existence of low value depressions between service centres; a scale coupling model with regular differences in the number of restaurants and types at different scales, i.e. at the district scale, the number of restaurant categories is almost saturated, but the more the number of categories, the more the area the more restaurants there are, showing an exponential trend, while at the community and street scales there is a normal distribution form, i.e. the more restaurants there are in areas with a moderate number of restaurant types; the larger the spatial scale, the greater the spatial dominance of local native cuisines, while in areas with a developed economy and local culture, each type of restaurant has its own dominance, presenting a spatial competition model for the overall mean and unit heterogeneity of cuisines. (3) The main obstacle to the spread of restaurants from the core to the periphery is the landscape, which is influenced by natural, economic and cultural factors; different cuisines have shown different performance in their development process; Cantonese cuisine has suffered more from the impact, while local cuisine is developing rapidly and has surpassed it in terms of overall quantity, and with the increase in the mobile population and the diversity of demand generated by the population base itself, its position will be further consolidated; foreign cuisine is concentrated in areas with higher consumption, well-developed catering and tolerant tastes. The economic and cultural barriers to entry are relatively high; in the process of selecting a location, different catering companies should not simply quantify the number of people and their consumption capacity with the help of economic tools, but should also analyse the taste preferences and food tolerance of different regions in order to fully consider the feasibility of their operations.