Research On The Method Of Software Product Line Evolution Based On Feature Model

Product line is a kind of by reusing reusable components in the product line to improve development efficiency of the method.The software product line development process is essentially a process of evolving new products from the software product line.There are mainly the following problems in the evolution method of software product line:(1)As the scale of software product line increases,the number of features increases,and the constraint relationship between features becomes more complex.The evolution of software product line needs to judge whether the evolved products satisfy the constraint relationship between features,and the manual management method can not meet the requirements;(2)When the domain feature model evolves to the application feature model,the application feature model may need to meet multiple evolution goals at the same time.The larger the scale of the domain feature model can evolve a larger number of application feature models.If the domain feature model is manually selected from the application feature model to meet multiple evolution goals at the same time,it is not only necessary to find all the effective application feature models,but also the application feature model with the highest comprehensive satisfaction degree for multiple evolutionary goals is found out.The efficiency of the model is low and it is easy to make mistakes.In order to solve the above problems,the main research work of this thesis is as follows:1.in order to solve the problem that the artificial management of software product line is inefficient,this thesis analyzes the importance of feature model based software product line and its evolution process,and the evolution of domain feature model to application feature model in the whole evolution process of software product line.A framework of evolution from domain feature model to application feature model is designed and proposed,In the framework of evolution,the constraints of feature models are needed and used to judge the correctness of evolution.In this thesis,the semantic representation of feature meta model and feature model is formalized by describing logical knowledge base,and the constraint rules are obtained from the semantics of feature meta model.According to the instantiation relationship between feature model elements and feature element,the constraint conditions of feature model are derived from constraint rules,and constraint conditions are expressed by combining normal form,It can be used as a logical expression to judge constraint satisfaction automatically.2.When selecting application feature models from domain feature models that meet multiple evolution goals,the manual method is inefficient and error prone.In this paper,the evolution process from domain feature model to application feature model is transformed into a multi-objective optimization problem.The constraints of the feature model are simplified by conjunctive normal form,and the multi-objective optimization algorithm is applied to automatically solve the application feature model with multiple evolution objectives.In the experiment,on the real feature model data,this paper verifies that the constraint conditions and feature elements of the feature model derived from the description logic knowledge base of the feature element model can be used as the input of the multi-objective optimization algorithm,and quickly and accurately solve the application feature model that meets multiple evolution goals.