Sharif Digital Repository

Deep Neural Networks (DNNs) are widly used in various domains, such as medicine, engineering, industry, financial markets, mathematics and management. DNNs are composed of several layers, such as convolutional and fully connected layers. Increasing the number of layers in DNNs provides different application with their required accuracy. In recent years, there have been many accelerator aiming to execute DNNs. However, the high computation and memory demands in DNNs are the main challenges to execute DNNs. To reduce computation and memory requirements, various methods such as pruning and quantization, have been proposed. Pruning and quantization methods make the DNNs sparse and increase the... 

Time delay and bandwidth limitations are common characteristics of communication channels that cause problems in direct use of classic control strategies in networked control systems. As networked systems become larger in scale and the need to share communication channels by multiple processes and controllers, this issue becomes more apparent. In recent decades, the stabilization of time-delay systems has attracted much attention. One of the most common control methods in such cases is based on the use of predictor feedback controllers. Predictor feedback controllers apply a delay-dependent or delay-independent control signal to the system. On the other hand, event-based control has many... 

Deep Neural Network (DNN) compression is a highly effective technique for reducing the computational burden and energy consumption associated with neural network inference, which is particularly important for low-power, embedded, and real-time systems. Weight pruning and quantization are among the most effective methods for neural network compression. Nonetheless, DNN compression poses various challenges, such as preserving network accuracy, particularly when dealing with adversarial attacks. Network compression can also lead to irregularities in the network structure and imbalanced distribution of workloads, which in turn can result in reduced utilization from the potential compression... 

The objective of this research is designing a controller for the event-triggered control of multi-agent systems in the presence of common constraints. In this regard, an adaptive decentralized event-driven controller is designed for the agents of a multi-agent system. The dynamics of the agents are nonlinear and unknown and they are subject to external disturbances. This controller is capable of achieving the objective of consensus with the desired performance in the presence of sensor and actuator faults, input nonlinearity, input and output quantization, and unmeasured states. The stability of the closed-loop system has been established via Lyapuonov theory. The considered multi-agent...