Design and management optimization of trickle irrigation systems using non-linear programming

A non-linear model is presented which optimizes the lay-out, as well as the design and management of trickle irrigation systems, to achieve maximum net benefit. The model consists of an objective function that maximizes profit at the farm level, subject to appropriate geometric and hydraulic constraints. It can be applied to rectangular shaped fields, with uniform or zero slope. The software used is the Gams-Minos package. The basic inputs are the crop-water-production function, the cost function and cost of system components, and design variables. The main outputs are the annual net benefit and pipe diameters and lengths. To illustrate the capability of the model, a sensitivity analysis of the annual net benefit for a citrus field is evaluated with respect to irrigated area, ground slope, micro-sprinkler discharge and shape of the field. The sensitivity analysis suggests that the greatest benefit is obtained with the smallest microsprinkler discharge, the greatest area, a square field and zero ground slope. The costs of the investment and energy are the components of the objective function that had the greatest effect in the 120 situations evaluated. (C) 1996 Academic Press Limited

Efficient linear programming algorithm for the transmission network expansion planning problem

The transmission network planning problem is a non-linear integer mixed programming problem (NLIMP). Most of the algorithms used to solve this problem use a linear programming subroutine (LP) to solve LP problems resulting from planning algorithms. Sometimes the resolution of these LPs represents a major computational effort. The particularity of these LPs in the optimal solution is that only some inequality constraints are binding. This task transforms the LP into an equivalent problem with only one equality constraint (the power flow equation) and many inequality constraints, and uses a dual simplex algorithm and a relaxation strategy to solve the LPs. The optimisation process is started with only one equality constraint and, in each step, the most unfeasible constraint is added. The logic used is similar to a proposal for electric systems operation planning. The results show a higher performance of the algorithm when compared to primal simplex methods.

EFFECT OF EXERCISE DURING PREGNANCY AND DAM AGE ON MATERNAL BLOOD-CHEMISTRY AND FETAL GROWTH

In order to determine the effect of maternal exercise on maternal nutritional status and fetal growth, young (Y = 45-50 days old) Wistar rats were divided into 4 groups of 5 to 8 animals: control pregnant (CP), control non-pregnant (CNP), exercise-trained (swimming 1 h/day, 5 days/week, for 19 days) pregnant (TP) and exercise-trained non-pregnant (TNP). Four equivalent groups of adult rats (A - 90-100 days old) were also formed. Serum glucose, total protein, albumin, hematocrit and liver glycogen were determined in female rats and pups. There were no statistical differences in serum glucose, total protein and albumin levels, litter size ot birth weight among exercise-trained animals, controls and their respective pups. Hematocrit was significantly lower in pups of exercise-trained young rats than in all other groups (YCP = 38.6 +/- 3.0; YTP = 32.6 +/- 2.1; ACP = 39.0 +/- 2.5; ATP = 39.2 +/- 2.9%). Liver glycogen levels were lower in pregnant than in non-pregnant rats but similar in exercise-trained and control rats of the same age and physiological status (YCNP = 4.1 +/- 0.2; YCP = 2.7 +/- 0.9; YTNP = 4.9 +/- 0.8; YTP = 2.7 +/-0.4; ACNP = 6.1 +/- 0.6; ACP = 3.1 +/- 0.8; ATNP = 6.6 +/- 0.8; ATP = 2.2 +/- 0.9 mg/100 mg). We conclude that pups of adult female rats are spared from the effects of this kind of exercise training during pregnancy. on the other hand, it appears that maternal adaptations to exercise training in young rats are able to preserve only some aspects of pup metabolism.

Effect of grazing system on fetal development in Nellore cattle

Intensive grazing systems for beef females, based on abundant availability of high quality forages and supplementary concentrates, may affect fetal development. The objective of this study was to determine the effect of grazing system on length of gestation, fetal development, and characteristics of the calf at birth. Twenty-four pregnant (bred to Nellore bulls) Nellore females were allocated into two groups. The control group (G1) grazed Brachiaria decumbens (signal grass) in a traditional (extensive) grazing system and the second group (G2) were managed on Panicum maximumcv. Tanzania 1 (Tanzania grass) in an intensive grazing system. Fetal development was evaluated by ultrasonography on days 31, 45, 59, 94, 122, 220, and 255 of gestation. The diameter of the amniotic and allantoic cavities, crown-rump length, circumference, and diameter of the head and ocular orbit were determined. At birth, calves were weighed and height, length, thoracic circumference, and ocular orbit and bi-parietal diameters were measured. There were no differences (P > 0.05) in fetal development. The G1 cows had a longer gestation period (4.5 days; P < 0.05) and their calves had greater (P < 0.05) weight, height, length, and thoracic circumference at birth. In conclusion, Nellore females raised under intensive pasture management conditions, had significantly shorter gestation and smaller calves at birth than those raised under extensive pasture management conditions. Therefore, adoption of new management practices (e.g. intensive pasture management), should take into consideration animal behavior and productivity. (C) 2003 Elsevier B.V. All rights reserved.

Transmission-expansion planning using the DC model and nonlinear-programming technique

The paper presents a constructive heuristic algorithm (CHA) for solving directly the long-term transmission-network-expansion-planning (LTTNEP) problem using the DC model. The LTTNEP is a very complex mixed-integer nonlinear-programming problem and presents a combinatorial growth in the search space. The CHA is used to find a solution for the LTTNEP problem of good quality. A sensitivity index is used in each step of the CHA to add circuits to the system. This sensitivity index is obtained by solving the relaxed problem of LTTNEP, i.e. considering the number of circuits to be added as a continuous variable. The relaxed problem is a large and complex nonlinear-programming problem and was solved through the interior-point method (IPM). Tests were performed using Garver's system, the modified IEEE 24-Bus system and the Southern Brazilian reduced system. The results presented show the good performance of IPM inside the CHA.