The accuracy of a knowledge extraction algorithm in a large database depends on the quality of the data preprocessing and the methods used. The massive amounts of data that we collect every day are putting storage capacity at a premium. In reality, many databases are characterized by attributes with outliers, redundant, and even more missing values. Missing data and outliers are ubiquitous in our databases, and imputation techniques will help us mitigate their influence. To solve this problem, as well as the problem of data size, this paper proposes a data preprocessing approach based on the k-nearest neighbor (KNN) completion for imputation of missing data and principal component analysis (PCA) for processing redundant data, thus reducing the data size by generating a significant quality sample after imputation of missing and outlier data. A rigorous comparison is made between our approach and two others. The dissolved gas data from Rio Tinto Alcans transformer T0001 were imputed by KNN, where k equals 5. For 6 imputed gases, the average percentage error is about 2, 17.5 after average imputation, and 23.65 after multiple imputations. For data compression, 2 axes were selected based on the elbow rule and the Kaiser threshold.

Cette étude s'appuie sur des ressources bibliographiques issues d'ouvrages, de documents et de rapports techniques d'institutions locales et internationales. L'offre d'énergie au Cameroun reste inférieure à la demande, même si elle est en croissance constante depuis 1980. L'un des objectifs de croissance du Plan de Développement du Secteur de l'Energie de 3 GW de production en 2020 n'ayant pas été atteint, il a semblé opportun de lancer une étude pour recueillir informations afin d'estimer le potentiel énergétique conventionnel du Cameroun et de mettre en évidence les sources solaires, éoliennes et de petite hydraulique. Au total, environ 90 articles et rapports techniques liés à l'énergie électrique ont été consultés. Après avoir éliminé les doublons et les documents sans rapport, les quelque 50 matériaux restants ont été sélectionnés pour des critères d'éligibilité tels que la pertinence par rapport à la situation de l'énergie conventionnelle au Cameroun, se concentrer sur l'une des sources renouvelables ciblées, le nombre de citations et la date de publication. A l'issue de ce tri, 30 références ont été définitivement retenues et ont servi de base à cette étude. Il en ressort que, les productions hydroélectrique et thermique du pays sont respectivement de 1414,2 MW et 605,622 MW pour un total de 2019,822 MW, le potentiel exploitable en énergies renouvelables étudié est estimé à 1339,038 MW ce qui représente 66,29% du potentiel énergétique national actuel.

This paper presents an artificial intelligence direct torque control (DTC) method for an electric vehicle (EV) drive system. The architecture of the proposed electric vehicle is that of four wheels each with an induction motor (IM). A comparative study of the different torque and speed controllers proposed in this paper is made. An electronic differential is used to control the speed of each wheel as well as a variable master-slave control (VMSC) for the management of the magnetic quantities because the motors on the same side are fed by the same converter. This study allows highlights the performance of the propulsion system in terms of dynamics and safety of the vehicle and better stability. The different controllers are implemented by the MATLAB/Simulink software and the simulation results obtained show better flexibility in the control of the vehicle. It is worth noting that direct torque control with fuzzy logic (DTFC) performs better than DTC associated with neural networks in terms of a time reduction increase of 1.47%, an overshoot of less than 5.33, and a static steady-state error close to zero.

The aim of this work is to optimize the alkaline hydrolysis of rice straw lignocellulose using kanwa, with the goal of producing bioethanol. The alkaline pretreatment was carried out using a saline rock from North Cameroon (kanwa) under the following experimental conditions: temperature between 170 and 200°C, and duration between 20 and 60 minutes. Characterization revealed that the local material contained 91.75 ± 0.48% dry matter, 8.25 ± 0.48% water, 31.22 ± 2.04% cellulose, 9.92 ± 2.46% lignin, and 17.72 ± 0.51% hemicellulose. The optimization design (central composite design) used to validate the reducing sugar models was based on the following indicators: Adjusted R² (99.36), AADM (12%), Bf (1.004), and Af (1.049). The optimal pretreatment conditions were 1.681, -1, and 0.808 in terms of temperature, time, and kanwa concentration, resulting in the maximum extraction of 49% cellulose, 2.4% hemicellulose, and 19.3% lignin. A concentration of 4.50 ± 0.02 mg/mL of reducing sugars was obtained under the conditions of 55°C for 72 hours, followed by centrifugation at 4000 rpm for 15 minutes. This resulted in an ethanol yield of 32% and a lower calorific value of 9184 ± 0.01 J/g.

The aim of this work is to optimize the alkaline hydrolysis of rice straw lignocellulose using kanwa, with the goal of producing bioethanol. The alkaline pretreatment was carried out using a saline rock from North Cameroon (kanwa) under the following experimental conditions: temperature between 170 and 200°C, and duration between 20 and 60 minutes. Characterization revealed that the local material contained 91.75 ± 0.48% dry matter, 8.25 ± 0.48% water, 31.22 ± 2.04% cellulose, 9.92 ± 2.46% lignin, and 17.72 ± 0.51% hemicellulose. The optimization design (central composite design) used to validate the reducing sugar models was based on the following indicators: Adjusted R² (99.36), AADM (12%), Bf (1.004), and Af (1.049). The optimal pretreatment conditions were 1.681, -1, and 0.808 in terms of temperature, time, and kanwa concentration, resulting in the maximum extraction of 49% cellulose, 2.4% hemicellulose, and 19.3% lignin. A concentration of 4.50 ± 0.02 mg/mL of reducing sugars was obtained under the conditions of 55°C for 72 hours, followed by centrifugation at 4000 rpm for 15 minutes. This resulted in an ethanol yield of 32% and a lower calorific value of 9184 ± 0.01 J/g.