Seepage and evaporation are the most serious forms of water losses in an irrigation canal network. Seepage loss depends on the channel geometry, whereas evaporation loss is proportional to the area of the free surface. In this investigation, a methodology has been devised which describes the optimal canal dimensions to convey a particular discharge. The objective nonlinear water loss function, for the canal, which comprises seepage and evaporation losses, is developed. Two constrains, minimum permissible velocity as a limit for sedimentation and maximum permissible velocity as a limit for erosion of canal, have been taken into consideration in the canal design procedure. Using Lagrange's method of undetermined multipliers, the optimal canal dimensions are obtained which give the least water loss. Using a random search method, a simple computer program was developed to carry out design calculation and provide the optimal canal dimensions. A set of design charts has been prepared by plotting the results. The proposed charts facilitate easy design of the optimal canal dimensions guarantying minimum water loss and computation of water loss from the canal section without going through the conventional and cumbersome trial and error method. A design example with sensitivity analysis has been included to demonstrate the simplicity and practicability of the proposed method.

This paper presents an effective load forecasting model for the western area of Saudi Arabia (WESA). Weather, load demand, wind speed, wind direction, heat, sunlight and so on are quite different in a desert land than other places. Thus this model is different from typical forecasting model considering inputs and outputs. Two models are implemented: firstly, a load forecasting model for prediction, however, is not sufficient for accurate forecasting, and, secondly, an optimization process to improve the results to be at least better than existing results.

This paper presents a comparative study between three Linear Matrix Inequality (LMI)-based iterative multivariable Proportional-Integral-Derivative (PID) controllers; PID design using H-norm, named Hi, PID design using H2-norm, named H2, of the system transfer function, PID design with Maximum Output Control (MOC), named Max, and the classical LMI-based robust output feedback controller using H-norm, named ROB. Multivariable PID is considered here because of its wide use in the industry, simple structure and easy implementation. It is also preferred in plants of higher order that cannot be reduced and thus require a controller of higher order such as is the case for the classical robust H output feedback controller whose order is the same as that of the plant. LMI technique is selected because it allows easy inclusion of divers system constraint requirements that should be fulfilled by the controller, and thus make its design very efficient. The duty of each of the controllers is to drive a single- generator connected to a large power system via a transformer and a transmission line. The generator is equipped with its speed/power (governor) and voltage (exciter) control-loops that are lumped in one block. The errors in the terminal voltage and in the output active power, with respect to their respective references, represent the controller inputs and the generator-exciter voltage and governor-valve position represent the controller outputs. A comparative study is carried out using the named controllers (Hi, H2, Max, ROB). Divers tests are applied, namely, step-change and tracking in the references of the controlled variables, and variation in some plant parameters, to demonstrate the controllers effectiveness. Encouraging results are obtained that motivate for further investigations.

The architecture and the performance of a network interconnecting interfaces are of great importance. Indeed, the majority of these interfaces function have time-critical constraints that must be considered when designing such a network. Performance can be evaluated in different ways using parameters measurement: end to end delay, time of request-response protocols, data losing or network throughput etc. In case of industrial environments time is a critical parameter, and the performance can be assessed in terms of estimated time of the request-response services.