abstract: The rapid advancement of Deep Neural Networks (DNNs), computing, and sensing technology has enabled many new applications, such as the self-driving vehicle, the surveillance drone, and the robotic system. Compared to conventional edge devices (e.g. cell phone or smart home devices), these emerging devices are required to deal with much more complicated and dynamic situations in real-time with bounded computation resources. However, there are several challenges, including but not limited to efficiency, real-time adaptation, model stability, and automation of architecture design. To tackle the challenges mentioned above, model plasticity and stability are leveraged to achieve efficient and online deep learning, especially in the scenario of learning streaming data at the edge: First, a dynamic training scheme named Continuous Growth and Pruning (CGaP) is proposed to compress the DNNs through growing important parameters and pruning unimportant ones, achieving up to 98.1% reduction in the number of parameters. Second, this dissertation presents Progressive Segmented Training (PST), which targets catastrophic forgetting problems in continual learning through importance sampling, model segmentation, and memory-assisted balancing. PST achieves state-of-the-art accuracy with 1.5X FLOPs reduction in the complete inference path. Third, to facilitate online learning in real applications, acquisitive learning (AL) is further proposed to emphasize both knowledge inheritance and acquisition: the majority of the knowledge is first pre-trained in the inherited model and then adapted to acquire new knowledge. The inherited model's stability is monitored by noise injection and the landscape of the loss function, while the acquisition is realized by importance sampling and model segmentation. Compared to a conventional scheme, AL reduces accuracy drop by >10X on CIFAR-100 dataset, with 5X reduction in latency per training image and 150X reduction in training FLOPs. Finally, this dissertation presents evolutionary neural architecture search in light of model stability (ENAS-S). ENAS-S uses a novel fitness score, which addresses not only the accuracy but also the model stability, to search for an optimal inherited model for the application of continual learning. ENAS-S outperforms hand-designed DNNs when learning from a data stream at the edge. In summary, in this dissertation, several algorithms exploiting model plasticity and model stability are presented to improve the efficiency and accuracy of deep neural networks, especially for the scenario of continual learning. / Dissertation/Thesis / Doctoral Dissertation Electrical Engineering 2020

The master‘s thesis is aimed on numerical simulation of autofrettage of high pressure fuel vessel – rail in Common Rail system. First there is described Chaboche model, which is later used for simulation of autofrettage. There are described different approaches which can be used to obtain sufficient material model. Then there is observed influence of these different approaches on stress state of rail within the process of autofrettage. Suitability of Chaboche model for autofrettage and re-autofrettage simulations is assessed by comparing with more complex Jiang model. In the end there is a study of influence of autofrettage pressure on different properties, especially on residual stresses.

The main aim of this thesis is to define material parameters of Chaboche model of plasticity. Adjustment of the parameters has to correspond to the experimental datas. These datas are represented by an uniaxial strain controlled test curve for fewer than two cycles and also by cyclic stress-strain curve. After that, an cyclic tension-compresion test for various parameters of Chaboche´s model of plasticity is simulated in an ANSYS software. Finally, the most suitable configuration of Chaboche´s model of plasticity is used for cylindrical thick-walled body. Cylindrical body represents a simplified model of high-pressure tank of fuel (diesel) – rail in Common Rail system.