This work was undertaken as a final year project in Computer Engineering, within the Department of Computer and Information Science at Linköping University. At the Department of Computer and Information Science, work oriented at testing and analyzing applications is developed to provide solution approaches to problems that arise in system product development. One of the current applications being developed is the ‘Systemics Analyst’. The purpose of the application is to facilitate for system developers with an analysis tool permitting insights on system reliability, system critical components, how to improve the system and the consequences as well as risks of a system failure. The purpose of the present thesis was to enhance the ‘Systemics Analyst application’ by incorporating an ‘automated model-based reliability prediction’ and ‘fault tree analysis’ modules. This enables reliability prediction and fault tree analysis diagrams to be generated automatically from the data files and relieves the system developer from manual creation of the diagrams. The enhanced Systemics Analyst application managed to present the results in respective models using the new incorporated functionality. To accomplish the above tasks, ‘Systemics Analyst application’ was integrated with a library that handles automated model-based reliability prediction and fault tree analysis, which is described in this thesis. The reader will be guided through the steps that are performed to accomplish the tasks with illustrating figures, methods and code examples in order to provide a closer vision of the work performed.

Testing is a critical activity in the software development process in order to obtain systems of high quality. Tieto typically develops complex systems, which are currently tested through a large number of manually designed test cases. Recent development within software testing has resulted in methods and tools that can automate the test case design, the generation of test code and the test result evaluation based on a model of the system under test. This testing approach is called model-based testing (MBT). This thesis is a feasibility study of the model-based testing concept and has been performed at the Tieto office in Karlstad. The feasibility study included the use and evaluation of the model-based testing tool Qtronic, developed by Conformiq, which automatically designs test cases given a model of the system under test as input. The experiments for the feasibility study were based on the incremental development of a test object, which was the client protocol module of a simplified model for an ATM (Automated Teller Machine) client-server system. The experiments were evaluated both individually and by comparison with the previous experiment since they were based on incremental development. For each experiment the different tasks in the process of testing using Qtronic were analyzed to document the experience gained as well as to identify strengths and weaknesses. The project has shown the promise inherent in using a model-based testing approach. The application of model-based testing and the project results indicate that the approach should be further evaluated since experience will be crucial if the approach is to be adopted within Tieto’s organization.

Software testing is an important process for ensuring the quality of the software. As the complexity of the software increases, traditional means of manual testing becomes increasingly more complex and time consuming. In most embedded systems, designing software with as few errors as possible is often critical. Resource usage is also of concern for proper behavior because of the very nature of embedded systems.  To design reliable and energy-efficient systems, methods are needed to detect hot points of consumption and correct them prior to deployment. To reduce testing effort, Model-based testing can be used which is one testing method that allows for automatic testing of model based systems. Model-based testing has not been investigated extensively for revealing resource usage anomalies in embedded systems. UPPAAL SMC is a statistical model checking tool which can be used to model the system’s resource usage. Currently UPPAAL SMC lacks the support for performing automatic test generation and test selection. In this thesis we provide this support with a framework for automatic test generation and test selection using mutation analysis, a method for minimizing the generated test suite while maximizing the fault coverage and a tool implementing the framework on top of the UPPAAL SMC tool. The thesis also evaluates the framework on a Brake by Wire industrial system. Our results show that we could for a Brake-by-wire system, simulated on a consumer processor with five mutants, in best case find a test case that achieved 100% mutation score within one minute and confidently identify at least one test case that achieved full mutation score within five minutes. The evaluation shows that this framework is applicable and relatively efficient on an industrial system for reducing continues resource usage target testing effort.

Rapid advances of computer modeling and simulation tools and computing hardware have turned Model Based Design (MBD) a more viable technology. However, using a computationally intensive, “black-box” form MBD software tool to carry out design optimization leads to a number of key challenges. The non-unimodal objective function and/or non-convex feasible search region of the implicit numerical simulations in the optimization problems are beyond the capability of conventional optimization algorithms. In addition, the computationally intensive simulations used to evaluate the objective and/or constraint functions during the MBD process also make conventional stochastic global optimization algorithms unusable due to their requirement of a huge number of objective and constraint function evaluations. Surrogate model, or metamodeling-based global optimization techniques have been introduced to address these issues. Various surrogate models, including kriging, radial basis functions (RBF), multivariate adaptive regression splines (MARS), and polynomial regression (PR), are built using limited samplings on the original objective/constraint functions to reduce needed computation in the search of global optimum. In many real-world design optimization applications, computationally expensive numerical simulation models are used as objective and/or constraint functions. To solve these problems, enormous fitness function evaluations are required during the evolution based search process when advanced Global Optimization algorithms, such as DIRECT search, Differential Evolution (DE), and Bat Algorithm (BA) are used. In this work, improvements have been made to three widely used global optimization algorithms, Divided Rectangles (DIRECT), Differential Evolution (DE), and Bat Algorithm (BA) by integrating appropriate surrogate modeling methods to increase the computation efficiency of these algorithms to support MBD. The superior performance of these new algorithms in comparison with their original counterparts are shown using commonly used optimization algorithm testing benchmark problems. Integration of the surrogate modeling methods have considerably improved the search efficiency of the DIRECT, DE, and BA algorithms with significant reduction on the Number of Function Evaluations (NFEs). The newly introduced algorithms are then applied to a complex engineering design optimization problem, the design optimization of floating wind turbine platform, to test its effectiveness in real-world applications. These newly improved algorithms were able to identify better design solutions using considerably lower NFEs on the computationally expensive performance simulation model of the design. The methods of integrating surrogate modeling to improve DIRECT, DE and BA global optimization searches and the resulting algorithms proved to be effective for solving complex and computationally intensive global optimization problems, and formed a foundation for future research in this area. / Graduate

There has been a call to reform science education to integrate scientific thinking practices, such as data interpretation and modeling, with learning content in science classrooms. This call to reform has taken place in both K-12 science education through Next Generation Science Standards and undergraduate education through AAAS initiative Vision and Change in Undergraduate Biology Education. This dissertation work examines undergraduate students' learning of multiple scientific thinking skills in a curricular format called Teaching Real data Interpretation with Models (TRIM) applied to a large-enrollment course in Cellular and Developmental Biology. In TRIM, students are provided worksheets in groups and tasked to interpret authentic biological data. Importantly, groups are tasked to relate their data interpretations to a 2D visual model representation of the relevant biological process. This dissertation work consists of two studies with the overarching question: How do students use model representations to interpret data interpretations? In the first study, we primarily describe how students learn to navigate and interpret discipline-based data representations. We found the majority of groups could construct quality written data interpretations. Qualitative coding analysis on group discourse found students relied on strategies such as decoding the data representation and noticing data patterns together to construct claims. Claims were refined through spontaneous collaborative argumentation. We also found groups used the provided model to connect their data inferences to a biological context. In the second study, we primarily target our analysis on how individual students relate their data interpretations to different modeling tasks, including student-generation of their own model drawing. I interviewed students one-on-one as they worked through TRIM-style worksheets. From iterative qualitative analysis of transcripts and collected video on hand movements, I characterize the forms of reasoning at play at the interface of data and model representations. I propose a model at the end of Study 2 describing three modes of reasoning in data abstraction into models. I found when relating between data and models, students needed to link signs in both representations to a common referent in the real-world phenomenon. Establishing this sign-referent relationship seemed to depend on bringing in outside mechanistic information about the phenomenon. Once a mechanism was established, students could fluidly move between data and model representations through mechanistic reasoning. Thus data abstraction seems to rely on mechanistic reasoning with models. The findings from this dissertation work support the feasibility of student development of multiple scientific thinking skills within a large lecture course, and provide targets for curriculum and assignment designs centered on teaching higher order reasoning skills.

Concretization is one of the most labor-intensive phases of the model-based testing process. This study concentrates on concretization of the abstract tests generated from the test models. The purpose of the study is to design and implement a structure to automate this phase which can reduce the required effort specially in every system update. The structure is completed and discussed as an extension on a model-based testing tool named ModelJUnit using adaptation approach. In this structure, the focus is mainly on bridging the gap in data-level between the SUT and the model.

Autogenous (AG) milling is utilised around the world for rst stage particle size reduction. The system exhibits highly non-linear behaviour in addition to being subject to unmeasured variability associated with most ore bodies. Anglo American Platinum aimed at improving online optimisation of the circuit by implementing industrial model predictive control to reduce system variability and continuously drive towards the optimal operating point within system constraints. A dimensional analysis of the circuit was conducted to explain the relationships between the various milling parameters discussed in the literature survey. The measured variables used in the analysis satis ed Buckingham's theorem, indicating that a complete subset of dimensionless groups were present and suitably able to describe process movement. These relationships were used as a reference point in determining the dynamic step response models between these variables necessary for model based control. The industrial dynamic matrix controller commissioned on the AG mill resulted in a 66 % reduction in power and a 40 % reduction in load. These are the main controlled variables of the mill. The controller also managed to reduce its objective function, e ective power utilisation, by 11 %. This stability improvement enabled a test campaign where the mill was controlled at various operating regions in order to establish the conditions conducive to the nest product size at a given mill feed rate. Moving the mill's operating region from the benchmarked plant to this optimal grind environment (at benchmarked variability) provided an estimated potential recovery increase of 0.27 % (absolute) due to better precious metal liberation. Stabilising the mill at this point with the model predictive controller resulted in a further 0.04 % potential recovery increase (absolute). The 0.31 % potential recovery increase is estimated at a monetary value of $93.1 million per annum. Copyright / Dissertation (MEng)--University of Pretoria, 2013. / Chemical Engineering / unrestricted

Traits are primitive units of code reuse that serve as building blocks of classes. In this research, we enhance reuse by extending the capabilities of traits; in particular, we add modeling abstractions to them. Traits have a variety of benefits, including facilitating reuse and separation of concerns. They have appeared in several programming languages, particularly derivatives of Smalltalk. However, there is still no support for traits that contain modeling abstractions, and no straightforward support for them in general-purpose programming languages. The latter is due to structural concerns that exist for them at runtime, especially traits that contain modeling abstractions. Model-driven technologies are making inroads into the development community, albeit slowly. Modeling abstractions such as state machines and associations provide new opportunities for reuse, and can be combined with inheritance for even greater reusability. However, issues with inheritance apply also when these new abstractions are inheritable units. This suggests that traits and models ought to be able to be synergistically combined. We perform a comprehensive analysis of using modeling elements in traits. We implement such traits in Umple, which is a model-oriented programming language that permits embedding of programming concepts into models. The contributions of the thesis are: a) Adding new elements including state machines and associations into traits, hence bringing more reusability, modularity, and applications to traits; b) Developing an algorithm that allows reusing, extending, and composing state machines through traits; c) Extending traits with required interfaces so dependencies at the semantic level become part of their usage, rather than simple syntactic capture; d) Adding template parameters with associations in traits, offering new applications for traits in which it is possible to define design patterns and to have a library of most-used functionality; e) The implementation of all the above concepts, including generating code in multiple general-purpose programming languages through automatic model transformation.

Rich Internet Applications (RIAs) are a new generation of web applications that break away from the concepts on which traditional web applications are based. RIAs are more interactive and responsive than traditional web applications since RIAs allow client-side scripting (such as JavaScript) and asynchronous communication with the server (using AJAX). Although these are improvements in terms of user-friendliness, there is a big impact on our ability to automatically explore (crawl) these applications. Traditional crawling algorithms are not sufficient for crawling RIAs. We should be able to crawl RIAs in order to be able to search their content and build their models for various purposes such as reverse-engineering, detecting security vulnerabilities, assessing usability, and applying model-based testing techniques. One important problem is designing efficient crawling strategies for RIAs. It seems possible to design crawling strategies more efficient than the standard crawling strategies, the Breadth-First and the Depth-First. In this thesis, we explore the possibilities of designing efficient crawling strategies. We use a general approach that we called Model-based Crawling and present two crawling strategies that are designed using this approach. We show by experimental results that model-based crawling strategies are more efficient than the standard strategies.

The purpose of this study was to explore how students view the use of representation in science classroom. Representation, as a disciplinary language of science, has long been promoted as a way to develop students’ scientific literacy and is closely linked to engaging students in scientific practices through the use of models in science. However, previous research studies have mostly focused on the use of representation and models as outcome measure of an implementation task and little is known about the learner’s perspectives. The study aimed to fill this missing gap by investigating how students view the use of representation in science classroom and how these perception are linked to the epistemic practice and cognitive/conceptual practice of science learning. In this respect, the study involved (1) developing an instrument, namely, a Representation Survey, to assess students’ views on the use of representation and (2) examining the relationship between students’ views on representation and their understanding of models in science, science content knowledge, and critical thinking skills. The Representation Survey was developed in three phases as a pencil-and-paper questionnaire with 1-5 Likert scales, and grounded in the empirical data and a literature review. An exploratory factor analysis of the Representation Survey with 619 middle school students identified two distinct ways students view the use of representation: multiple modes of representation and uni-mode of representation. Correlation analysis with a modified version of the Student’ Understanding of Models (SUMS) Survey revealed a strong relationship between students’ perception on using multiple-mode of representation and their understanding of models in science, while how students perceive uni-modal representation was shown to be related to students’ performances in assessments of science content knowledge. Lastly, students’ critical thinking skills, as measured by the Cornell Critical Thinking Test, showed no evident relationship with students’ perceptions of the use of representation. A validity argument for the newly developed Representation Survey and modified SUMS instrument is presented, followed by a discussion of broader implications and limitations of the study.