Junaid Ali Khan

Solution of Fractional Order System of Bagley-Torvik Equation Using Evolutionary Computational Intelligence

A stochastic technique has been developed for the solution of fractional order system represented by Bagley-Torvik equation. The mathematical model of the equation was developed with the help of feed-forward artificial neural networks. The training of the networks was made with evolutionary computational intelligence based on genetic algorithm hybrid with pattern search technique. Designed scheme was successfully applied to different forms of the equation. Results are compared with standard approximate analytic, stochastic numerical solvers and exact solutions.

AMPLITUDE AND DIRECTIONAL OF ARRIVAL ESTIMATION: COMPARISON BETWEEN DIFFERENT TECHNIQUES

In this work, we propose a method based on Genetic algorithm hybridized with Pattern Search for joint estimation of Amplitude and Direction of Arrival, azimuth as well as elevation angles using L-type array. Four other schemes, i.e., the Genetic algorithm, Pattern Search, Simulated Annealing and Simulated Annealing hybridized with Pattern Search are also discussed and compared with Genetic algorithm hybridized with Pattern Search. Multiple sources are taken in the far fleld of sensors array and Mean Square Error is taken as a fltness function. This fltness function is optimal in nature and requires only a single snapshot. It avoids any ambiguity or required permutation as in some other methods to link it with angles found in the previous snapshot. The reliability and efiectiveness of the proposed scheme is tested on the basis of Monte- Carlo simulations and its statistical analysis.

Identification of Input Nonlinear Control Autoregressive Systems Using Fractional Signal Processing Approach

A novel algorithm is developed based on fractional signal processing approach for parameter estimation of input nonlinear control autoregressive (INCAR) models. The design scheme consists of parameterization of INCAR systems to obtain linear-in-parameter models and to use fractional least mean square algorithm (FLMS) for adaptation of unknown parameter vectors. The performance analyses of the proposed scheme are carried out with third-order Volterra least mean square (VLMS) and kernel least mean square (KLMS) algorithms based on convergence to the true values of INCAR systems. It is found that the proposed FLMS algorithm provides most accurate and convergent results than those of VLMS and KLMS under different scenarios and by taking the low-to-high signal-to-noise ratio.

Hybrid evolutionary computational approach: Application to van der pol oscillator

We presented a method for van der Pol oscillators using artificial neural network optimized by evolutionary computational approach. A trail solution of the oscillator is written as a feed-forward neural network containing adjustable adaptive parameters. The optimization of the networks is performed by genetic algorithms in an unsupervised way. The proposed scheme is tested successfully by applying on both the stiff and non-stiff problems. A Monte Carlo simulation is performed for the reliability and effectiveness of the scheme. It is shown that the obtained results are in good agreement with Runge Kutta numerical method.

An Application of Artificial Intelligence for the Joint Estimation of Amplitude and Two-Dimensional Direction of Arrival of Far Field Sources Using 2-L-Shape Array

An easy and efficient approach, based on artificial intelligence technique, is proposed to jointly estimate the amplitude, elevation, and azimuth angles of far field sources impinging on 2-L-shape array. In these proposed artificial intelligence techniques, the metaheuristics based on genetic algorithm and simulated annealing are used as global optimizers assisted with rapid local version of pattern search for optimization of the adaptive parameters. The performance metric is employed on a fitness evaluation function depending on mean square error which is optimum and requires single snapshot to converge. The proposed approaches are easy to understand, and simple to implement; the genetic algorithm specifically hybridized with pattern search generates fairly good results. The comparison of the given schemes is carried out with 1-L-shape array, as well as, with parallel-shape array and is found to be in good agreement in terms of accuracy, convergence rate, computational complexity, and mean square error. The effectiveness and efficiency of the given schemes are examined through Monte Carlo simulations and their inclusive statistical analysis.

An application of hybrid computing to estimate jointly the amplitude and Direction of Arrival with single snapshot

In this paper, utilization of hybrid computational approach is evaluated for the joint estimation of amplitude and Direction of Arrival of far field sources impinging on a uniform linear array. In this hybrid approach, swarm intelligence based on Particle swarm optimization is exploited as a global optimizer assisted with pattern search technique as a rapid local search technique. The optimization of adaptive parameters depending upon the amplitudes and direction of arrival is performed using the fitness function based on Mean Square Error that defines an error between desired response and estimated response. The interest in this function is due to its ease in implementation, efficiency and simplicity of concept. It is derived from Maximum Likelihood and requires only single snapshot to converge. The proposed algorithm is robust enough to produce fairly good results even in the presence of low signal-to-Noise Ratio and requires relatively less number of antenna elements in the array. The results of hybrid technique are much better as compared to Particle Swarm Optimization and pattern search alone. A number of test cases are discussed on the basis of different number of sources impinging on the array with different number of sensors in the array. The accuracy and reliability of the proposed scheme is tested on the basis of Monte-Carlo simulations and its superior statistical analysis.

Car Number Plate Recognition (CNPR) system using multiple template matching

Image processing algorithms are used for converting textual image to editable text. One application could be a car number plate recognition (CNPR) system where these type of algorithms automatically detect the car registration number by capturing an image. In this paper, we improve one of the existing CNPR algorithms. The contributions include: multiple template matching; considering light intensity and recognizing the car number plate even in low intense light and independence of distance of car number plate to camera. The involvement of different conditions and external factors actually improve the CNPR system efficiency. We run different experiments on different car number plates. Our proposed improvements yield better results in terms of false positive and false negative values for CNPR.

Enhanced car number plate recognition (ECNPR) system by improving efficiency in preprocessing steps

In this paper, we develop an enhanced car number plate recognition (ECNPR) system that is capable of recognizing different design and font styles of car number plates. Our contribution is threefold. Firstly, we use multiple templates matching for recognition of characters, i.e. of various types of font and size. Secondly, the removal of noise by dynamically adjusting the pixel value. Lastly, we create and make use of a more challenging dataset which is comprised of different font style and design of number plates. The results show that the enhancement of CNPR proposed system yields better results in terms of false positive and false negative values.

LLR-based Erasure Decoding of SFH-MFSK in the presence of tone jamming

Error and Erasure decoding (E2D) is known to yield better performance than error-only decoding. Therefore, Erasure insertion (EI) has been applied to noncoherent M-ary frequency shift keying (MFSK) with slow frequency hopping (SFH) and Reed Solomon (RS) coding, when the system operates in the presence of jamming and fading. In this context, various criteria have been proposed to declare erasures in MFSK. One of these criteria is the Ratio Threshold Test (RTT). In this contribution, we employ symbol probabilities based RTT as an EI criteria i.e. an erasure is declared when ratio of the second largest to the largest symbol probabilities exceed a given threshold. Through simulation results we demonstrate that this scheme outperforms the conventional RTT in terms of combating tone jamming (MTJ) and Nakagami-m fading.