No abstract

This paper investigates the output feedback model predictive control (OFMPC) for Takagi-Sugeno fuzzy networked control systems with bounded disturbance, where data quantization and data loss occur simultaneously. The quantization error is treated as sector bound uncertainties by using the sector bound approach and the data loss process is modeled as a time-homogeneous Markov chain. Invoking S-procedure and the notion of quadratic boundedness which can specify closed-loop stability for system with disturbance, the state observer is offline designed and the networked output feedback model predictive controller is provided which explicitly considers the satisfaction of input constraints. Two online synthesis algorithms of OFMPC are presented, one parameterizing the infinite horizon control moves into a single feedback law, the other into one free control move followed by the single feedback law based on the state observer. A new formula is introduced to refresh the ellipsoidal bound of estimation error which can guarantee the recursive feasibility of optimization problem. An example is given to demonstrate the effectiveness of the proposed new design techniques.

No abstract

In this paper, we investigate a robust constrained model predictive control synthesis approach for discrete-time Takagi-Sugeno's (T-S) fuzzy system with structured uncertainty. The key idea is to determine, at each sampling time, a state feedback fuzzy predictive controller that minimizes the performance objective function in the infinite time horizon by solving a class of linear matrix inequalities (LMIs) optimization problem. To do this, the fuzzy predictive controller is designed on the basis of non-parallel distributed compensation (non-PDC) control law, relaxed stability conditions of the closed-loop fuzzy system are developed by employing an extended nonquadratic Lyapunov function and introducing additional slack and collection matrices. In addition, the presented approach is capable of ensuring the robust asymptotic stability as well as the recursive feasibility of the closed-loop fuzzy system. Simulations on a highly nonlinear continuous stirred tank reactor (CSTR) are eventually presented to demonstrate the effectiveness of the developed theoretical approach.

Irregularity and coarse spatial sampling of seismic data strongly affect the performances of processing and imaging algorithms. Therefore, interpolation is a usual pre-processing step in most of the processing workflows. In this work, we propose a seismic data interpolation method based on the deep prior paradigm: an ad-hoc Convolutional Neural Network is used as a prior to solve the interpolation inverse problem, avoiding any costly and prone-to-overfitting training stage. In particular, the proposed method leverages a multi resolution U-Net with 3D convolution kernels exploiting correlations in cubes of seismic data, at different scales in all directions. Numerical examples on different corrupted synthetic and field datasets show the effectiveness and promising features of the proposed approach.