Code-Generation For Autonomous Vehicles Based On Ptolemy â…¡

According to the unceasing increment of functional requirement and system complexity of the autonomous vehicles control systems, defect and deficiency of traditional software development methodology is becoming manifest. Given guaranteeing the correctness and safety of autonomous vehicles control systems software and increasing the efficiency of software development, this thesis proposes a kind method of autonomous vehicles control systems software development based on Ptolemy II after summarized and compared M MBD (Model-based Development) with MDD (Model Driven Development), and analyzed the shortcomings of traditional software development methodology. This method realized code generation from system model and communication model, which do increase the e efficiency of autonomous vehicles control systems software development. Further more, this method verified the correctness of design by deploy the generated code onto distributed computation platform performing HILS (Hardware-in-The-Loop Simulation). The main contains and contributions of this thesis are as follows.First, we put forward the autonomous vehicle control system software development method based on Ptolemy II. Model-based design methodology is introduced to solve the problem of correctness and dependability in autonomous vehicle control system software design. Synchronous dataflow is a model of computation suitable for modeling control systems. In a synchronous dataflow model, the underlying model of computation governs the interaction and communication of the components in the model. MIC (Model-Integrated Computing)is utilized to build CME meta-model by UML in GME. Based on the CME meta-model, the code interpreter is built for generating communication code, and the model transfer interpreter is built for transferring CME models to models which can be parsed in Ptolemy II simulation environment. Code interpreter and model transfer interpreter have the universal properties of code generation and model transformation form different models of different systems.Second, Ptolemy II is used to model autonomous vehicles control system,which is a system level heterogeneous model design and simulation platform. Model code can be g enerated from the Ptolemy II after we improved on some defects of the code generator. The code of communication for fulfills distribution data transformation is automatically generated by CME code interpreter. The autonomous vehicle control system software development method based on Ptolemy II increase of efficiency of software development by automation code generation.Third, we take controlling fly the quid-rotor helicopter performing distributed HILS by autonomous vehicle control system as an instance to explore the feasibility and validity of the autonomous vehicle control system software development method. The experimental results validate the correctness of the software designed by the method.