Human/Machine Readable Language Representation Method in HEV Drivetrain Testing

摘要

The drivetrain for hybrid and electric vehicles (HEV) plays a key role in meeting high requirements on vehicle development which are placed by high standards for convenience and drive dynamics together with strict statutory regulations. HEV drivetrains result in high overhead for development and testing. Testing plays an important role as the feedback mechanism ensuring the quality and reliability of the hardware and software developed. Test cases are derived from norms, requirements and design parameters. In an organization with multiple teams in multiple locations working on various projects, standardization of test specifications has been of great importance to ensure comprehensibility, adaptability, and reusability. To accomplish that, a 126 VDI-Berichte Nr. 2354, 2019 human/machine readable language using a generic test specification file format have been developed to specify the test cases independent of test location, test environment, or project. The human/machine readable language is developed by taking into consideration keyworddriven testing. The idea behind the keyword driven testing approach is the ability for test engineers to design test cases combining keywords according to the rules set by the language that represent test steps. KDT architecture allows test engineers to create different variations of automated test cases using fixed set of keywords, without requiring any coding skills. The created test specifications can also be understood easily by other disciplines, which in turn makes the interdisciplinary test case review process easier. Moreover, the test specifications can easily be exchanged and adapted between locations and projects. KDT is based on three major factors: 1. A human/machine readable language consisting of keywords and phrases based on the testing terminology and needs for HEV drivetrains. 2. A parser (compiler), which is responsible for checking syntax (language rules), validating the defined test step aggregates (parameters, signals etc.), replacing the generic test step aggregates with their project specific ones or values, and mapping test steps to their corresponding automation function. 3. An abstract layer on top of an automation core framework - a set of generic functions, which is responsible for the test execution and evaluation. Human/Machine readable language combined with an abstract layer implemented on an automation core framework is independent of automation system implemented for the testing environment at a test location and therefore increases the cooperation and synergy between different test locations having various testing environments - system dyno, hardware in the loop simulation test benches, e-machine emulators, and test vehicles. This presentation mainly deals with the human/machine readable language representation of test cases for HEV drivetrains. The key points are: Definition of the language - the vocabulary and syntax rules fulfilling HEV drivetrain testing terminology. Generic test specification - Additional project files and aggregation rules introduced to ensure generic test step definitions, which are easily reused between different projects. Parser - converting the generic test steps into project/test environment specific ones and finally mapping those to their corresponding abstract automation functions. Abstraction of automation functions for test execution and result evaluation - an automation core framework dealing with various automation systems implemented on different test environment setups in multiple locations.