Simulation based tool for error propagation analysis of simulink models

Safety is a growing demand for all types of today's systems. Whether it is a nuclear or a transportation system, reliability and safety are the most urgent needs for the design of the entire system. Engineers have developed different standards and models to ensure the desired function of the system without any human life or material loss. All these safety and economy related issues make the topic "model-based error detection analysis" an important and interesting topic for different areas of engineering. This growing demand makes it compulsory for the companies in the industry to form dedicated departments just for the safety assessment and certification. Different types of methods and standards have been developed to calculate the reliability of such safety-critical systems. Various safety standards for various fields are also given by examples in the thesis. These methods can be qualitative or quantitative depending on the type of the system. Analyzes of these safety-critical systems require time and careful work by the designers and testers. As it is necessary to fulfill some criteria based on reliability, engineers developed methods such as fault tree analysis, failure mode and effect analysis, and hazard analysis. To validate the analytical results, a simulation-based approach is introduced in this thesis. After giving background information and mathematical models for the analytical approach, the benefits of a simulation-based approach and its difference from the analytical methods are explained. Simulation-based approach is developed for the environment of MATLAB Simulink, by implementing different types of faults and fault injection methods. These fault types are classified as the sub-elements of the general fault definitions such as sensor faults, hardware faults, and network faults. Every fault type can be injected by using a different method that consists the parameters of occurrence rates and the duration of the fault effect. The correct values that are used for the comparison by the tool are obtained through a fault-free run. The comparison of the faulty run values and the fault-free run values provides important information about the reliability and the performance of the system. In the next phase, the methodology to obtain these reliability and performance metrics are explained along with the features offered to the user by ErrorSim. How to use these features and how to interpret the results obtained by these features are introduced by explanatory examples. Apart from safety-related issues, the architecture of the developed tool is described. It gives the user an insight into the main algorithm which is behind fault injection and error detection. It is important for the user to know the listeners call-back functions and a few technical details for the correct usage of the tool. Results of the simulation are studied in different perspectives in order to show that the tool that is developed can also illustrate performance-related issues in control systems. Two different case studies are taken into consideration for the interpretation of the tool application results. In the conclusion section, the achievements and benefits of ErrorSim are discussed. The general usage of ErrorSim and its importance is explained. Possible improvements on ErrorSim and future works are also discussed.