K-Nearest Neighbors (KNN) is a classification ALGORITHM based on supervised machine learning, which works according to a voting system. The performance of the KNN ALGORITHM depends on different factors, such as unbalanced distribution of classes, the scalability problem, and considering equal values for all training samples. Regarding the importance of the KNN ALGORITHM, different improved versions of this ALGORITHM are introduced, such as fuzzy KNN, WEIGHTED KNN, and KNN with variable neighbors. In this paper, a WEIGHTED KNN based on Whale Optimization ALGORITHM is proposed for the objective of increasing the level of detection accuracy. The proposed ALGORITHM devotes a weight to each training sample of every feature by employing the WOA to explore the optimized weight matrix. The ALGORITHM is implemented and experimented on five standard datasets. The evaluation results prove that the proposed ALGORITHM performs better than both WEIGHTED KNN based on the Genetic ALGORITHM (GA) and the classic KNN ALGORITHM.

Outliers are data values that lie away from the general cluster of other data values. Detecting the outliers of a dataset is an important research topic for data cleaning and finding new useful knowledge in many research areas, i.e. data mining, pattern recognition, etc. In the past decades, many useful ALGORITHMs were proposed in the literature. In this paper, a new fuzzy kernel-clustering ALGORITHM with outliers (FKCO) is presented to locate critical areas that are often represented by only a few outliers. Theoretic analysis also shows that FKCO can converge to a local minimum of the objective function. Finally, based on the information theory, a new criterion for finding outliers is also proposed. Simulations of different types of datasets demonstrate the feasibility of this new method.

In this paper, an efficient GV1-CG is developed to optimizing the MODIFIED conjugate gradient ALGORITHM by using a new conjugate property. This is to to increase the speed of the convergence and retain the characteristic mass convergence using the conjugate property. This used property is proposed to public functions as it is not necessary to be a quadratic and convex function. The descent sharp property and comprehensive convergence for the proposed improved ALGORITHM have been proved. Numerical results on some test function indicate that the new CG-method outperforms many of the similar methods in this field.

The development of communications and telecommunications infrastructure, followed by the extension of a new generation of smart distribution grids, has brought realtime control of distribution systems to electrical industry professionals’ attention. Also, the increasing use of distributed generation (DG) resources and the need for participation in the system voltage control, which is possible only with central control of the distribution system, has increased the importance of the real-time operation of distribution systems. In real-time operation of a power system, what is important is that since the grid information is limited, the overall grid status such as the voltage phasor in the buses, current in branches, the values of loads, etc. are specified to the grid operators. This can occur with an active distribution system state estimation (ADSSE) method. The conventional method in the state estimation of an active distribution system is the WEIGHTED least squares (WLS) method. This paper presents a new method to modify the error modeling in the WLS method and improve the accuracy SVs estimations by including load variations (LVs) during measurement intervals, transmission time of data to the information collection center, and calculation time of the state variables (SVs), as well as by adjusting the variance in the smart meters (SM). The proposed method is tested on an IEEE 34-bus standard distribution system, and the results are compared with the conventional method. The simulation results reveal that the proposed approach is robust and reduces the estimation error, thereby improving ADSSE accuracy compared with the conventional methods.

This paper presents a new ALGORITHM for the identification of a specific class of hybrid systems. Hybrid System identification is a challenging problem since it involves the estimation of discrete and continuous states simultaneously. Using the method known as product of errors, this problem could be formulated such that, the identification of continuous state to be independent of discrete state estimation. We propose a new iterative WEIGHTED least squares ALGORITHM (IWLS) for the identification of switched auto regressive exogenous systems (SARX). In this method, the parameters of only one subsystem are updated at each iteration while the parameters of the other subsystems are assumed known. In the method, all four main parameters of hybrid systems, namely Subsystem degrees, Number of subsystems, Unknown parameters vector and switching signal are estimated. Simulation results shows that our proposed method has a good performance in identifying the subsystems parameters and switching signal. Also, the superiority of our ALGORITHM is shown by modeling of two SARX systems.

MODIFIED Normalized Least Mean Square (MNLMS) ALGORITHM, which is a sign form of NLMS based on set-membership (SM) theory in the class of optimal bounding ellipsoid (OBE) ALGORITHMs, requires a priori knowledge of error bounds that is unknown in most applications. In a special but popular case of measurement noise, a simple ALGORITHM has been proposed. With some simulation examples the performance of ALGORITHM is compared with MNLMS.