Abstract In this paper, the robustness of a typical control scheme for Wiener systems is studied. These systems consist of the cascade connection of a linear time invariant system and a static nonlinearity. To control this kind of systems, several approaches were discussed in the literature. Most of these control schemes involve transformation of the measured variable as well as the setpoint, by the inverse of the nonlinear gain. The approach followed in this work uses the inverse model of the static nonlinear gain, while the uncertainty in the Wiener model is treated as a partitioned problem. The linear block is considered as a parameter-affine-dependent model and, on the other hand, the nonlinear block uncertainty is analyzed as a conic-sector. The robustness analysis is performed using µ-theory. The results are evaluated on the basis of a simulation of a pH neutralization process.

An approach for compensating non-linearities in a continuous-time first-order sigma-delta converter with one bit quantization is presented. The proposed compensators are parallel nonlinear dynamic systems using piecewise linear static functions. A reduction of an order of magnitude is obtained in the measured squared error when compared to the uncompensated sigma-delta converter. Our results confirm a significant improvement in signal to noise and distortion ratio for single tone input signals and in spurious free dynamic range for multi-tone inputs

An inadequate water quality in epicontinental waterbodies could provoke risks on aquatic species especially in the endemic species that are less tolerant to several kinds of pollutants. In this paper the results of artificial wetland (AW) with a puliment filter (PF) to obtain a good water quality for aquaculture activities are presented. An AW of 55 m² was connected to an intermitent vertical flow, with four doses daily of 1,000 L of water, that is equivalent to 72 mm³/mm² day of superficial hydraulic charge (SHC). The finality of this AW, was to remove organic material, nitrogen and phosphorous. A PF of 1.8 m² was installed in order to eliminate PO4(3-). During 12 months of operation, the total remotion reached 92% of chemical oxygen demand (COD), 85% of N-NH4 and 80% of PO4(3-). The final water quality obtained in this system can be considered good for aquaculture purposes.

La baja calidad del agua en cuerpos acuáticos epicontinentales podría ocasionar riesgos en las especies acuáticas, especialmente en las de tipo endémico, que son menos tolerantes a la presencia de determinados contaminantes. En este documento, se presentan los resultados obtenidos en un humedal artificial (HA), acoplado a un filtro de pulimento (FP), diseñado para obtener agua apta para la preservación de vida acuática silvestre. El HA tiene una superficie de 55 m² y se alimenta con un flujo vertical intermitente, repartido en 4 dosificaciones diarias de 1,000 L cada una, equivalentes a una carga hidráulica superficial (CHS) de 72 mm³/mm² d. La finalidad del HA es la de remover materia orgánica carbonosa, nitrógeno y fósforo. El FP, tiene una superficie de 1.8 m² y se instaló con el objeto de promover el proceso de desnitrificación y alcanzar una mayor eliminación de PO4(3-). Durante 12 meses de operación, se logró una reducción global de hasta 92% de la demanda química de oxígeno (DQO), 85% de N-NH4 y 80% de PO4(3-). La calidad final del agua obtenida por este sistema puede ser considerada como buena para propósitos de acuicultura.

The water column in a fish farm was sampled at three depth layers to determine large and short term changes in water quality. The overall goal of this study was to determine the variation of nutrient concentrations in the water column in open-ocean floating cages with cultures of Lutjanus peru (Pacific red snapper) and L. guttatus (spotted rose snapper). Nutrient concentration (ammonia-N, nitrite-N, nitrate-N, and phosphate) were evaluated every fifteen days in the water column at three stations (cages) and two depths; likewise, several water quality parameters were analyzed daily (water temperature, salinity, dissolved oxygen and Secchi disk). Analyses of water in general showed that no significant differences existed between the monthly concentration of nutrients from the control stations and depths. Nitrate was the nutrient with the highest concentration (2.83 to 3.56 mg/L); however, these values were relatively low and normal for these waters. The results show that when working with fish densities of 2500 to 3500 fish/cage and small quantities of food (60 ton/year) no impact is made on the water column quality in the floating cages culture system.

La columna de agua de una granja de peces fue muestreada a tres profundidades para determinar los cambios en la calidad del agua a largo y corto plazo. El objetivo de este estudio fue determinar las variaciones en la concentración de nutrientes en la columna de agua del cultivo en jaulas flotantes de Lutjanus peru (huachinango) y L. guttatus (flamenco). La concentración de nutrientes (amonio, nitritos, nitratos y fosfatos) fue evaluada cada 15 días en la columna de agua en tres estaciones (jaulas) y dos profundidades. También, varios parámetros de la calidad del agua fueron analizados diariamente (temperatura del agua, salinidad, oxígeno disuelto y profundidad de visión del disco de Secchi). Los análisis del agua muestran en general que no hubo diferencias significativas mensualmente entre la concentración de los nutrientes de la estación control y las estaciones experimentales. Los nitratos fueron el nutriente con la mayor concentración (2.83 a 3.56 mg/L), sin embargo, esas concentraciones fueron relativamente bajas y normales para ese tipo de aguas. Los resultados muestran que cuando se trabaja con densidades de peces de 2500 to 3500 peces/jaula y bajas cantidades de alimento (60 ton/año) no se impactan la calidad de la columna de agua del sistema de cultivo de jaulas flotantes.

A nonlinear adaptive time series predictor has been developed using a new type of piecewise linear (PWL) network for its underlying model structure. The PWL Network is a D-FANN (Dynamical Functional Artificial Neural Network) the activation functions of which are piecewise linear. The new realization is presented with the associated training algorithm. Properties and characteristics are discussed. This network has been successfully used to model and predict an important class of highly dynamic and non-stationary signals, namely speech signals.

Nitrification and denitrification rates of inorganic nitrogen were studied in a closed recirculating-water system, comparing three load rates of rainbow trout Oncorhynchus mykiss (89, 156 and 194 kg in each tank with two repetitions). Six self-cleaning water circular fish tanks with a volume of 4.3 m³ were used, maintaining a 3.94 m³/day of average flow rate and constant aeration. A total of 371 rainbow trout, 524 ± 8 g initial wet weight were introduced in the system and fed with a commercial feed that contained 38% of protein. A total study time of 44 days was divided into three phases of 14, 17 and 13 days according to the load fish rate. Temperature, dissolved oxygen, pH, total ammonia nitrogen (TAN), un-ionized ammonia, nitrite and nitrate were daily evaluated at four monitoring sites: fish tank (FT), settling tank (ST), biofilter (B) and reconditioning tank (RT). Water physicochemical characteristics and their fluctuations played an important role in treatment efficiency. Water temperature varied between 18 °C and 20.5 °C and dissolved oxygen from 4.6 to 7.7 mg/l. The lowest values of these two variables were registered in the ST where all wastes accumulate. No significant differences (p<0.05) were observed in pH values (8.3-8.6). These conditions allowed good nitrification and denitrification rates. TAN varied from 0.2 to 1.96 mg/l; however, this value was 80% lower in the outlet (RT) as compared to the inlet (ST). The load fish rate caused a significant difference (p<0.05) in TAN and non-ionized ammonia in the FT with the lowest value for 89 kg load density as compared to 156 and 194 kg respectively. Conversely, nitrite concentration did not show a significant difference (p>0.05) among load fish rate. Nitrate concentration had an accumulative tendency at 156 kg load rate batch up to 30 days with a further decrease. The results showed that a reduction of load rate did not change apparently the equilibrium of bacteria population. Therefore, it is possible to control variables such as TAN, non-ionized ammonia and nitrite concentration, hence maintaining an adequate water quality for rainbow trout.

Se estudiaron las tasas de nitrificación y desnitrificación del nitrógeno inorgánico, en un sistema cerrado de recirculación de agua, comparando tres cargas de biomasa de trucha arco iris Oncorhynchus mykiss (89, 156 y 194 kg por estanque con dos repeticiones). Se utilizaron seis estanques circulares de autolimpieza con volumen de 4.3 m³ de volumen, con un flujo promedio diario total de agua de 10.93 m³ y aireación constante. Un total de 371 truchas arco iris con peso inicial de 524 ± 8 g fueron introducidas en el sistema y alimentadas con alimento balanceado, que contenía 38% de proteína. El estudio duró 44 días continuos, divididos en tres fases de 14, 17 y 13 días respectivamente, de acuerdo con la carga de biomasa de peces. La temperatura, oxígeno disuelto, pH, nitrógeno amoniacal total (NAT), amoniaco, nitrito y nitrato fueron evaluados diariamente en cuatro sitios de monitoreo: estanque de peces (EP), estanque de sedimentación (ES), biofiltro I (BI) y estanque de reacondicionamiento (ER). Las características fisicoquímicas del agua y la fluctuación de los parámetros jugaron un importante papel en la eficiencia del tratamiento. La temperatura del agua varió de 18 °C a 20.5 °C y el oxígeno disuelto de 4.6 a 7.7 mg/l. Los valores más bajos de estas dos variables fueron registrados en el ST donde los desechos se acumulan. No se observaron diferencias significativas (p<0.05) en los valores de pH (8.3-8.6). Estas condiciones permitieron una buena tasa de nitrificación y desnitrificación. El NAT varió de 0.2 a 1.96 mg/l, sin embargo, este valor fue 80% más bajo en la salida (ET) comparada con la entrada al sistema (ES). La carga de biomasa de peces causó una diferencia significativa (p<0.05) en los valores de NAT y amoniaco en el EP, con los valores más bajos para 89 kg, comparado con 156 y 194 kg respectivamente. Por su parte la concentración del nitrito no mostró diferencias significativas (p>0.05) entre las diferentes cargas. Las concentraciones de nitrato tuvieron una tendencia acumulativa a 156 kg hasta los 30 días con un rápido decremento. Los resultados mostraron que la reducción de la carga de biomasa de peces, no cambia aparentemente el equilibrio de la población bacteriana del biofiltro. Además, es posible controlar las variables como el NAT, el amoniaco y la concentración de nitrito, manteniendo una adecuada calidad del agua para la trucha arco iris.

In this paper we present a nonlinear infinite impulse response (NIIR) model structure for black-box identification of nonlinear dynamic systems. The proposed model structure allows the implementation of an identification algorithm in which the degrees of freedom of the Nonlinear Output Error (NOE) model can be easily increased or decreased during the identification process. This property is very attractive to find the appropriate NIIR model, avoiding overfitting. This is done using High Level Canonical Piecewise Linear (HL CPWL) functions with an increasing (decreasing) grid division. Therefore, the algorithm may start using a linear estimation of the model. The parameters of the HL CPWL functions are updated using a simple algorithm based on a modified steepest descent method with an independently adaptive learning rate.

Se utilizaron cuatro grupos experimentales con 60 hembras de trucha arco iris (Onchorhynchus mykiss) cada uno, con un peso promedio de 467 g, las cuales fueron sometidas a un ensayo aleatorio para comparar la eficiencia de cuatro dietas conteniendo diferentes concentraciones de pigmentos carotenoides y determinar sí la concentración y el origen del mismo presentaba una influencia sobre el desarrollo gonadal y la maduración de los ovocitos. La primera dieta fue una dieta control no pigmentada (C); la segunda contenía 100 mg kg-1 de Carofil Rosa (CR); la tercera y cuarta contenía 200 (CR200) y 250 (CR250) mg kg-1 de oleorresina saponificada de chile rojo (Capsicum annuum), respectivamente. Los resultados mostraron que existieron diferencias significativas (p <= 0.05) entre las dietas C y CR250 con respecto al índice gonadosomático (IGS), peso promedio de la gónada (PG) y el diámetro de los ovocitos (DO), presentando la última dieta los valores más altos de estos factores. Los tratamientos de CP y CR200 presentaron valores similares. Al final del experimento solamente se observó el 10% de ovocitos maduros en la dieta CR250.

Four groups of rainbow trout (Onchorhynchus mykiss) each with 60 females with an average body weight of 467 g were submitted to an aleatory assay in order to compare the efficiency of four diets containing different concentrations of carotenoid pigments, and to determine if pigment concentration and source had an influence on female gonad development and oocyte maturity. The first diet was the non-pigmented control (C). The second diet contained 100 mg kg-1 of Carophyll Pink (CR). The third and fourth diets contained 200 (RC200) and 250 (RC250) mg kg-1 of saponified red chili oleoresin, respectively. The results indicated significant differences (p <= 0.05) between C and RC250, with respect to the gonado-somatic index (IGS), average gonad weight (PG) and average diameter of oocyte (DO), with CR250 showing higher IGS, PG and DO values. Treatments CR and CR200 presented similar values. At the end of the experimental period, only 10% of oocytes were mature in the RC250 diet.

The desired operating point in Model Predictive Control is determined by a local steady-state optimization, which may be based on an eco-nomic objective. In this paper we proposes the solu-tion of a linear dynamic back-off problem to obtain a hierarchical scheme that ensures feasible operation in despite of disturbances. This is performed by computing the critical disturbances and expanding the optimization problem to ensure the existence of a control action that ensures the rejection of each perturbation.