Mohammed Ouassaid

Moroccan long term electricity demand forecasting using Wavelet neural Networks

This article aims at developing long term electricity forecasts for Morocco, using Non Linear Autoregressive model (NLARX) based on a special class of Artificial Neural Networks (ANN), namely Wavelet Networks. This model provides more accurate forecasting results thanks to the nature of the used neural networks structure that has been developed in other fields. The obtained results will help design adequate strategies in order to meet the electricity demand with the most optimal energy mix and to evaluate the financial support, if needed, to state electricity companies that generally overestimate their needs to get more financial support from the government. For investors, reliable electricity forecasting constitutes one of the most important parameters that allow evaluating the viability of their investments.

A Non-linear Decentralized Control of Multimachine Power Systems Based on a Backstepping Approach

Power system requires high-performance control techniques due to their elevated complexity, high nonlinearity and almost continuously time-varying nature. Also, power systems are often subjected to small and large disturbances. To enhance the multimachine power system stability, a new approach to designing decentralized nonlinear control scheme is proposed. The approach seeks first build a novel mathematical model of multimachine power systems. The main characteristic of this model is that interactions between generators and changes in operating conditions are represented by time-varying parameters. More important, those parameters are update online, using only local measurements. Second, it develops a decentralized controller for the transient stabilization and voltage regulation. The controller consists of two controllers, known as the terminal voltage regulator and rotor speed stabilizer. The methodology adopted is based on backstepping design strategy. The proposed stabilizing feedback laws for the power system are shown to be globally asymptotically stable in the context of Lyapunov theory. Case studies are achieved in a two-area four machine power system to verify the effectiveness of the approach. Numerical results are presented to illustrate the usefulness and the performance of the proposed control scheme, under different contingencies.

Optimization of the subscribed maximum demand for the high voltage tariff consumers in Morocco using Nelder Mead method

This paper presents an optimization method for the power contract under the High Voltage tariff structure in Morocco. The optimization is based on minimization of the billing cost, formulated as an objective function that takes into account the power charges, the excess of the power contract charges and the bi-seasonality of the tariff structure. This method is applied to the real case study of the Bouregreg pumping station in Morocco, which feeds potable water to Rabat and Casablanca and their respective surroundings. The financial gains are then evaluated for the reference year (2013) and validated for the year 2014. This method can be used as a first step of an energy efficiency project, and will allow substantial savings without important investment.