超市导购路径规划方法研究

发布时间：2018-05-25 03:01

本文选题：空间环境建模 + 最短路径规划　；参考：《合肥工业大学》2015年硕士论文

【摘要】：随着国民经济的不断发展,去大型超市购物已成为现代都市居民日常生活中不可缺少的一部分。大型超市里商品琳琅满目,消费者有了更多的选择,也容易因为商品种类的繁多、空间布局的复杂而迷失在室内。由此,在为消费者提供称心如意的商品的同时,给予他们便捷的购物路线指导,已是提升消费者购物体验的一种可行方法。本文旨在研究在大型超市内进行导购路线的个性化推荐,包括室内空间环境的建模、最短路线的规划和最优路线的设计三方面。在对超市的室内布局做出分析后,结合商品的摆放特点,将可行走的过道部分离散化为若干个区域,建立用节点和无向边分别表示可行走区域和两个相邻区域之间可行走路线的无向图模型。在已知消费者的采购清单的条件下,根据商品的货架位置将采购清单上的商品与具体的区域进行映射,最短路线规划的目的是用遗传算法优化一条联结超市入口、采购清单上的商品所在的货架区域和结账柜台的最短路线。最优路线的设计是在最短导购路线规划的基础上引入基于内容的商品个性化推荐技术。首先根据商品的自然属性、消费者的历史购物记录等数据,采用关联规则等挖掘方法在商品的小类维度上分别构造商品模型和消费者模型,用余弦相似度度量方法计算消费者-商品兴趣度矩阵,将消费者可能感兴趣的商品加入到导购路线的规划中,给消费者带来意外的惊喜。通过在Matlab软件上分别对最短和最优的导购路线规划进行实例分析,仿真结果显示本文提出的方法简单、稳定、高效,能够快速推荐出一条最佳的导购路线,供消费者行进参考。
[Abstract]:With the development of national economy, shopping in large supermarkets has become an indispensable part of the daily life of modern urban residents. Large supermarkets have a wide range of goods, consumers have more choices, but also easily because of the variety of goods, the complexity of the space layout and lost in the room. Therefore, it is a feasible method to improve the consumer's shopping experience by providing consumers with satisfactory goods and giving them convenient shopping route guidance at the same time. The purpose of this paper is to study the personalized recommendation of shopping guide routes in large supermarkets, including the modeling of indoor space environment, the planning of the shortest route and the design of the optimal route. After making an analysis of the interior layout of the supermarket, and combining the characteristics of the display of goods, the walkable aisle is discretized into several areas. An undirected graph model with nodes and undirected edges to represent walkable regions and walkable routes between two adjacent regions is established. Under the condition that the consumer's purchase list is known, the goods on the purchase list are mapped to the specific area according to the shelf location of the goods. The purpose of the shortest route planning is to optimize a connecting supermarket entrance with genetic algorithm. The shortest route to the shelf area and checkout counter where the goods on the purchase list are located. The design of the optimal route is to introduce the content-based personalized recommendation technology on the basis of the shortest guide route planning. Firstly, according to the natural attributes of commodities and the historical shopping records of consumers, we use association rules mining methods to construct commodity models and consumer models on the subclass dimension of commodities, respectively. This paper calculates the consumer-commodity interest matrix by using cosine similarity measure method, and adds the goods that consumers may be interested in to the planning of the purchase guide, which brings the consumers a pleasant surprise. Through the analysis of the shortest and optimal route planning on Matlab software, the simulation results show that the method proposed in this paper is simple, stable and efficient, and can quickly recommend the best purchase guide route for consumers' reference.
【学位授予单位】：合肥工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：F717.6;F252;TP18

【相似文献】