概念数据模型

概念数据模型(CDM)帮助你分析信息系统的概念结构，然后识别出需要描述的主要实体，这些实体中的属性和这些实体间的关系。概念数据模型比逻辑或者物理数据模型更为的抽象。

逻辑数据模型

逻辑数据模型(LDM)帮助你分析信息系统的结构，它独立于任何具体的物理数据库实现。ldm已经涉及到实体的标识符，它没有cdm那么抽象，但并不允许你设计视图、索引和其他更为具体的物理数据模型的元素。

物理数据模型

物理数据模型(PDM)帮助你分析表、视图、和其他数据库对象，包括数据仓库所需的多维度对象。pdm比cdm和ldm都更为具体。你可以进行建模、逆向工程、和生成所有最为流行的数据库。

原文

原文见官方文档《powerdesigner.chm》中Getting Started with Data Modeling章节。
Conceptual Data Models
A conceptual data model (CDM) helps you analyze the conceptual structure of an information system, to identify the principal entities to be represented, their attributes, and the relationships between them. A CDM is more abstract than a logical (LDM) or physical (PDM) data model.

Logical Data Models
Logical Data Models A logical data model (LDM) helps you analyze the structure of an information system, independent of any specific physical database implementation. An LDM has migrated entity identifiers and is less abstract than a conceptual data model (CDM), but does not allow you to model views, indexes and other elements that are available in the more concrete physical data model (PDM).

Physical Data Models
A physical data model (PDM) helps you to analyze the tables, views, and other objects in a database, including multidimensional objects necessary for data warehousing. A PDM is more concrete than a conceptual (CDM) or logical (LDM) data model. You can model, reverse-engineer, and generate for all the most popular DBMSs.

ouyida3的csdn博客
2015.5.12