Search

In this paper, the integrated design problem of Fault Estimation (FE) and Fault Tolerant Control (FTC) for saturated systems is considered. The system suffers from input saturation and L2-bounded disturbances. To design an adaptive observer-based FE, the actuator output measurements are not used, when the actuator is saturated, there exist new interactions between FE and FTC. In such a case the separation principle for FE and FTC design does not hold anymore. In this paper, a method is introduced for integrated design of FE and FTC, which guarantees boundedness stability of such a system. To extract sufficient conditions for stability and minimum L2-gain performance for disturbance rejection, the Modified Sector Condition (MSC) is employed for saturation modeling. Also, Quadratic Polytopic Differential Inclusion (QPDI) approach is developed for the comparative study. All conditions are cast as Linear Matrix Inequality(LMIs) that can be evaluated in a single step. In a numerical example, the feasibility of the introduced method for stability guarantee is demonstrated and performance characteristics of FE and FTC in disturbance rejection are compared for both modeling approaches.

This study consists of using a new red algal species biomass together with Cd, as a toxic metal from synthetic contaminated wastewaters in order to test biomass potential to become a by-product to be used for the wastewater treatment. The effects of solution pH, initial Cd(II) concentration, biosorbents dosage, and contact time on the removal efficiency of Cd(II) from the synthetic solutions were investigated. Cd(II) sorption was found to be highly pH dependent, and the maximum sorption was obtained at pH 6.0. The removal efficiency of Gracilaria corticata and the agar extraction algal waste increased from 53 to 87% and 62 to 96% for an increase in biomass from 0.1 to 0.5 g L−1, respectively. Equilibrium data follow Langmuir model. The parameters of Langmuir equilibrium model are q max = 226.9 mg g−1, b = 0.009 mg L−1 and q max = 197.4 mg g−1, b = 0.012 mg L−1, respectively, for G. corticata and agar extraction algal waste. Different kinetic models such as pseudo-first-order, pseudo-second-order, and intra-particle diffusion model were tested, and the experimental data were in agreement with the pseudo-second-order model. FT-IR of biomasses revealed that amino, hydroxyl, and carbonyl were main functional groups. The results of SEM–EDX indicate that ion exchange occurred during the adsorption process. The present study suggests that G. corticata and agar extraction algal waste can be used as effective, environmentally friendly, and efficient biosorbents for Cd(II) removal from aqueous solution.

Behavior of beam-to-column connections in steel structures is characterized by the strength, stiffness and stability of them. Accurate modeling of stiffness, strength and ductility of these connections is important, especially when one considers dynamic or seismic loading. In this study, after determining the experimental behavior of the end-plate beam-to-column connection under cyclic loading, behavior of mechanical and neural network models of this connection has been evaluated. To validate the models, hysteresis response of mechanical model and computational intelligence model is compared with the experimental one. In the mechanical model, deformations of the connection are expressed only based on the geometric and material properties. On the other hand, in computational intelligence techniques with neural networks, requisite information is extracted directly from the experimental data. Hence, both models have pros and cons that are discussed in present paper. It is seen that as the mechanical model does not consider all aspects of hysteresis response, this model is not sufficiently accurate compared with neural network model. The neural network model is capable of characterizing pinching and stiffness degradation in hysteresis curve of the connection.

An optimal single DOF dynamic absorber is presented. A tremor has a random nature and then the system is subjected to a random excitation instead of a sinusoidal one; so the H2 optimization criterion is probably more desirable than the popular H∞ optimization method and was implemented in this research. The objective of H2 optimization criterion is to reduce the total vibration energy of the system for overall frequencies. An objective function, considering the elbow joint angle, θ2, tremor suppression as the main goal, was selected. The optimization was done by minimization of this objective function. The optimal system, including the absorber, performance was analyzed in both time and frequency domains. Implementing the optimal absorber, the frequency response amplitude of θ2 was reduced by more than 98% and 80% at the first and second natural frequencies of the primary system, respectively. A reduction of more than 94% and 78%, was observed for the shoulder joint angle, θ1. The objective function also decreased by more than 46%. Then, two types of random inputs were considered. For the first type, θ1 and θ2 revealed 60% and 39% reduction in their rms values, whereas for the second type, 33% and 50% decrease was observed.