Contribuições à análise de robustez de sistemas de controle usando redes neurais

This work develops a robustness analysis with respect to the modeling errors, being applied to the strategies of indirect control using Artificial Neural Networks - ANN s, belong to the multilayer feedforward perceptron class with on-line training based on gradient method (backpropagation). The presented schemes are called Indirect Hybrid Control and Indirect Neural Control. They are presented two Robustness Theorems, being one for each proposed indirect control scheme, which allow the computation of the maximum steady-state control error that will occur due to the modeling error what is caused by the neural identifier, either for the closed loop configuration having a conventional controller - Indirect Hybrid Control, or for the closed loop configuration having a neural controller - Indirect Neural Control. Considering that the robustness analysis is restrict only to the steady-state plant behavior, this work also includes a stability analysis transcription that is suitable for multilayer perceptron class of ANN s trained with backpropagation algorithm, to assure the convergence and stability of the used neural systems. By other side, the boundness of the initial transient behavior is assured by the assumption that the plant is BIBO (Bounded Input, Bounded Output) stable. The Robustness Theorems were tested on the proposed indirect control strategies, while applied to regulation control of simulated examples using nonlinear plants, and its results are presented

Caracterização do conteúdo de fossas e tanques sépticos na cidade de Natal

Only 32% of the population of Natal is attended by sewage, while the remaining population use pits and septic tanks. The characterization of the contents of septic tanks and pits contributes to the performance of such system and may guide the decision on treatment of these contents. The main of this research is to characterize the contents of interior residential pits and septic tanks in the greater Natal, with the following specific goals: to develop and manufacture a sampler capable of collecting a representative sample of the entire column (the surface scum, the clarified liquid and sludge bottom); to compare the contents of the tanks with the pits; to compare the contents of the septage from vacuum trucks; to relate the composition content with socioeconomic characteristics of households; to compare the content in both chambers of the septic tanks in series; to assess the situation of the content before and six months after the cleanness; and ultimately propose a pilot scale plant for treatment of septage. Once the sampler was developed, samples were collected within 14 septic tanks and 10 pits in many districts of Natal. Medians of the 24 systems were obtained: temperature, pH, conductivity, oil and grease, total solids, total suspended solids and sediments of 28.0 °C, 6.95; 882 mS/cm, 75.2 mg/L; 10,169 mg/L, 6,509 mg/L and 175 mL/L respectively; 111.0 mgN/L for ammonia, 130.5 mgN/ L for organic nitrogen, 0.2 mgN/L for nitrite, 0.4 mg/L for nitrate; 8935 mgO2/L for COD, 29.2 mgP/L for total phosphorus, thermotolerant coliforms from 9.95 E +06 CFU/100mL helminth eggs and 9.2 eggs/L with a maximum concentration of 688 eggs/L and minimum of 0 eggs/L. Medians of organic nitrogen and TKN were significantly different between groups of tanks and pits. The systems with cleanness gap from 11 and 20 years presented the higher concentrations for most variables. The effluent from the toilets and bathrooms participate more effectively in contributing fractions of solids, alkalinity, nitrogen, COD, total phosphorus, thermotolerant coliforms and helminth eggs. The systems used by socioeconomics class with income from R$ 3,700.00 to R$ 7,600.00, presented higher concentrations for COD, nitrogen, solids and helminth eggs. The first of the two chambers had always presented higher concentrations over the second compartment. The analysis of variance for most variables, showed that the values of septic tanks, pits and septage from vacuum trucks belong to the same group. In the samples taken after cleanness, the median of pH and temperature increased, while alkalinity, COD, organic nitrogen, total phosphorus, ammonia and helminth eggs decreased. The oils and greases and thermotolerant coliforms had slightly varied due to the continuous release of sewage into the systems that maintained their steady state concentrations.

Contribuições para estratégia de controle aplicada à geração fotovoltaica interconectada à rede elétrica

Generation systems, using renewable sources, are becoming increasingly popular due to the need for increased use of electricity. Currently, renewables sources have a role to cooperate with conventional generation, due to the system limitation in delivering the required power, the need for reduction of unwanted effects from sources that use fossil fuels (pollution) and the difficulty of building new transmission and/or distribution lines. This cooperation takes place through distributed generation. Therefore, this work proposes a control strategy for the interconnection of a PV (Photovoltaic) system generation distributed with a three-phase power grid through a connection filter the type LCL. The compensation of power quality at point of common coupling (PCC) is performed ensuring that the mains supply or consume only active power and that his currents have low distorcion. Unlike traditional techniques which require schemes for harmonic detection, the technique performs the harmonic compensation without the use of this schemes, controlling the output currents of the system in an indirect way. So that there is effective control of the DC (Direct Current) bus voltage is used the robust controller mode dual DSMPI (Dual-Sliding Mode-Proportional Integral), that behaves as a sliding mode controller SM-PI (Sliding Mode-Proportional Integral) during the transition and like a conventional PI (Proportional Integral) in the steady-state. For control of current is used to repetitive control strategy, which are used double sequence controllers (DSC) tuned to the fundamental component, the fifth and seventh harmonic. The output phase current are aligned with the phase angle of the utility voltage vector obtained from the use of a SRF-PLL (Synchronous Reference Frame Phase-Locked-Loop). In order to obtain the maximum power from the PV array is used a MPPT (Maximum Power Point Tracking) algorithm without the need for adding sensors. Experimental results are presented to demonstrate the effectiveness of the proposed control system.