Search

The paper presents a comparative analysis of nature inspired Artificial Bee Colony and Particle Swarm Optimization algorithms for the enhancement of images. The paper briefly reviews the swarm-based algorithms. It describes the dynamics of ABC and Particle Swarm Optimization algorithm and mapping for image enhancement. The key purpose of Image processing is to extract essential features from images. During an image processing operation, the input given is an image, and its output is an enhanced high-quality image based on some characteristics/features related to that Image. The mapping of the ABC and PSO algorithms for image enhancement was implemented using the Matlab image processing toolbox. For both algorithms ABC and PSO, the parameters of Contrast Enhancement alpha was set in the range of 0-255 and beta was set in the range of 1-109. Finally, the Comparative Analysis of ABC and PSO Algorithms was done using the metrics of Entropy, MSE and PSNR values for the Images. It is found that ABC yields a better performance as compared to PSO.

Modern energy systems often regarded as smart grid (SG) systems are cyber-physical systems (CPS) equipped with advanced metering and smart sensing devices, leading to a high-dimensional data generation in large volumes. To address this challenge, we propose a new variational autoencoder (VAE)- based dimensionality reduction technique for SG data to enable renewable energy generation forecasting with improved accuracy. The proposed method integrates bidirectional long short-term memory (BiLSTM) deep neural networks with variational inference, to generate an encoded representation of the high-dimensional time-series energy data. The encoded data is further utilized as low- dimensional representation of the original data for the application of energy forecasting, which leads to the reduced computational cost and more accurate forecasting results. The efficacy of the proposed VAE-BiLSTM method is evaluated using python programming and tensorflow library on the data traces taken from the Ausgrid solar generation dataset. Moreover, a comparative analysis of the proposed technique is presented with the benchmark autoencoder (AE) and VAE-based methods. Our result analysis illustrates that the proposed VAE-BiLSTM outperforms VAE-RNN, VAE-LSTM, as well as standard AE- based methods using evaluation metrics such as reconstruction error, pinball score, root-mean square error (RMSE), and mean absolute error (MAE).

The world is on the brink of an era of Unmanned Aerial Vehicles (UAVs), widely known to public as drones, where we will get to experience multiple UAVs flying in the national airspace carrying out diverse tasks such as monitoring, surveillance, product deliveries, law enforcement, fertilizing crop fields, aerial photography, and transport. In such scenarios, there is a possibility of collisions, path overlaps, mix-ups, and uncertainties as far as their flying routes are concerned. The routes to which aircraft are restricted to fly are called air corridors. These flying routes could be inside constructed air corridors where the UAVs would be allotted to fly, similar to the air corridors of commercial aircraft. There is a growing need to identify and construct these air corridors for UAVs to fly in their respective corridors to avoid such mishaps. This served as the primary motivation for proposing a novel estimation algorithm and a method to visualize air corridors for autonomous multi-UAV flights in an airspace. In addition, we studied various popular uncertainty visualization techniques to incorporate uncertainty in the visualization of the air corridors. Furthermore, a standalone web application with a user-friendly graphical user interface (GUI) was developed using HTML5, CSS3, JavaScript and CesiumJS. We present the estimation algorithm and related visualization results of our proposed algorithm, and discuss the summarized research findings to conclude this paper.

Abnormal activity recognition and its detection becomes a major task in surveillance videos. In the context of security purposes, it becomes more challenging to identify the abnormal events or activities in video surveillance. Surveillance closely observes, monitors and identify the behaviour of any object like people, suspicious thing or any abnormal activity. Technology is increases very fast, it becomes easier to identify and detect the abnormal activities. There are a lot of techniques and methods are available in field of detection of abnormal activities. In this paper, we discuss about the various methods and techniques and discuss about their advantages and disadvantages. And also compare these technologies based upon their performance. In related work, there are many techniques like Spatio-temporal Saliency Detection, Graph Formulation, Oriented GMM, Based upon Sparse reconstruction and using Statistical Hypothesis Detector etc.

In this paper, Convolution Neural Network (CNN) and a special variant of Recurrent Neural Network (RNN) named Long Short-Term Memory Model (LSTM) with peep hole connection is developed for optical character recognition (OCR). Data-set of mathematical equations known as Image to Latex 100K is retrieved from OPEN-AI and used for testing the model. First, the mathematical equations from the images are converted to Latex texts. Then this Latex text is used to render the mathematical equations. The proposed method uses the tokenized data, which is sequentially given to the deep learning network. The sequential process helps the algorithms to keep track of the processed data and yield high accuracy. A new variant of LSTM called "LSTM with peephole connections" and Stochastic "Hard" Attention model was used. The performance of the proposed deep learning neural network is compared with INFTY (which uses no RNN) and WYGIWYS (which uses RNN). The proposed algorithm gives a better accuracy of 76% as compared to 74% achieved by WYGIWYS.

Backpropagation Neural Network Algorithm is a long-standing supervised algorithm, with applications in many domains. This method is useful when there is a predefined set of input and output vectors, and aiming to produce the output for the given input. However, to reduce the computational time latency, many optimization techniques that are available in Vivado HLS tool were applied to generate a downable code for FPGA platform to enhance the performance of the algorithm. The overall latency has been reduced from 3.44ms without optimization to 2.42ms with optimization. Therefore, the latency reduced by about 1.4 times. This paper also implemented and paralyzed the same algorithm on CPU and GPU for comparison purposes. The result shows the FPGA implementation gives better results than CPU and GPU on the comparison of performance due to the applied optimization techniques in the implementation.

Representing a given dataset with a mathematical model is a very useful tool in many engineering applications. Several techniques exist to evolve a mathematical model for a given data set. Non-Linear regression being the most often used. The curves can be generated from these mathematical models, which provide a visualization of how that model fits the data. In this paper, three algorithms to train artificial neural networks are used to develop a good fit for the data. The three algorithms are Levenberg-Marquardt, Bayesian Regularization, and Scaled Conjugate Gradient. These algorithms were applied to the housing data set. The comparative performance of these algorithms was compared using Mean Square Error (MSE), which represents the best curve fitting for these data sets. The mean squared error was computed for each algorithm. Levenberg-Marquardt had MSE of 7.0902. Scaled Conjugate Gradient at 15.2932, and Bayesian Regularization at 5.3480. It was found that Bayesian Regularization gave the best accuracy at 96.78%, followed by Levenberg-Marquardt at 94.53% and Scaled Conjugate Gradient at 90.51%.

In this paper, a new swarm intelligence based algorithm called Grey Wolf Optimizer (GWO), mimicking the social hierarchy and hunting behavior of grey wolves, is used to optimally schedule the operation of Energy Storage Unit (ESU) in smart homes. The aim is to reduce the cost of power consumption and balance the load on the power grid. This is achieved by drawing the power from the grid when the price is cheap and using the excess power to store is the battery (ESU) and drawing from the ESU when the price is expansive. The algorithm determines when the power should be drawn from the grid and when the power should be drawn from the stored unit for each hour of the day. The algorithm was tested on the data collected from the Chicago region by the department of energy, United States. The results were compared with Particle Swarm Optimization (PSO) on the same dataset. GWO provides up to 25.57% of cost saving compared to PSO. In addition, GWO does not impose any restrictions on the consumer to use appliance at any time providing total freedom and still saving money.

The accurate modeling of a time series using a Sugeno Fuzzy Inference System (FIS) requires an algorithm that can train the FIS to minimize the error of seen and unseen data points. Many researchers have used genetic algorithms to optimize the parameters of the FIS membership functions with a great deal of success. It is presented here that incorporating FIS structure identification into the training process can greatly improve accuracy of predicting future time series data, by using the well-known Mackey-Glass time series as a benchmark. The main structural identification consists of optimizing the number of membership functions per input and total number of rules in the rule base.

In this paper, a model to reflect the immunological properties of white blood cells is created. White blood cells kill infectious bacteria. Two kinds of white blood cells — CHO-IIA and THP1 are studied. Bacteria of sizes 5μm and 0.05μm, and an amount of catalytic agent MβCD is added to the white blood cell. Experiments are conducted first by keeping the bacteria size fixed and varying the amount of MβCD to observe their effects on bacteria bound to a cell, and then bacteria internalized by a cell. Later the same experiment is repeated with a different bacteria size. The data collected is used to create a model with inbuilt rules, based on fuzzy logic. The model is trained with back propagation using neuro-fuzzy techniques, in order to improve the model. The results obtained by the model are compared with the actual observed data to compare the efficiency of the model. With training, the Root Mean Squared Error reduced from 0.4099 to 0.3705. Complete software with a Graphical User Interface is built to easily display the output of the model, from user-given input.

The growth in number of fuzzy logic applications led to the need of finding efficient ways to implement them. The advantage of using Field Programmable Gate Arrays (FPGA) is the ease with which controllers can be re-designed based on requirements without incurring major non recurring engineering (NRE) costs. In this paper, the implementation of a Mamdani Fuzzy Inference System has been demonstrated with the application of a Home Water Sprinkler System. The design of the Home Water Sprinkler System uses two inputs, Temperature and Soil Humidity, to give a single output, Watering Duration. VHDL programming language is used for creating the design in which simulations are carried out to check the functional verification of individual blocks. The design created on VHDL platform was synthesized, verified and implemented using Xilinx Integrated Software Environment (ISE) Version 6.1. The bit stream created is downloaded onto the Xilinx Virtex 2 FPGA which is mounted on a V2MB1000 board. For the verification of FIS, DIP switches and push buttons were used to simulate inputs. The outputs and inputs were displayed on the two seven segment LED displays. The accuracy of the model on FPGA was compared with simulation results obtained using MATLAB & Fuzzy Logic Tool Box. The results obtained from the two models differed with a Root Mean Square (R.M.S.) error of 0.8%.

In this paper we have developed fuzzy inference system models for a field-effect transistor. The hope is to see if such techniques can be used for inventing future semiconductor based devices. Three modeling techniques have been used. Neuro Fuzzy based on grid partitioning and Neuro Fuzzy based on cluster partitioning create Sugeno Fuzzy Inference Systems, which are trained with a neural network back propagation method. The third modeling technique is based on Grammatical Evolution, where a grammar template in the form of rules is evolved using genetic algorithms based evolutionary techniques. This grammar template is based on the Mamdani Fuzzy Inference System. Experimental results indicate that all models produce acceptable levels of performance, some even have an error rate that is nearly negligible.

In this paper, a novel hybrid genetic algorithm for solving Traveling Salesman Problem (TSP) is presented based on the Nearest Neighbor heuristics and pure Genetic Algorithm (GA). The hybrid genetic algorithm exponentially derives higher quality solutions in relatively shorter time for hard combinatorial real world optimization problems such as Traveling Salesman Problem (TSP) than the pure GA. The hybrid algorithm outperformed the NN algorithm and the pure Genetic Algorithm taken separately. The hybrid genetic algorithm is designed and experimented against the pure GA and the convergence rate improved by more than 200% and the tour distance improved by 17.4% for 90 cities. These results indicate that the hybrid approach is promising and it can be used for various other optimization problems. This algorithm is also independent of the start city of travel whereas the result of NN algorithm are based on start city.

Achieving a robust system for control of a non-linear process is the driving factor of adaptive controls. The design presented in this paper attempts to utilize a genetic supervisor on a fuzzy gain schedule controller to obtain performance similar to a self tuning regulator. This design is simulated through the use of the Matlab Genetic Algorithm toolbox.

The use of Grammatical Evolution for automating the structure identification of fuzzy rule based classification systems is investigated in this paper. To demonstrate its usefulness, we apply this method to the iris species identification problem. We achieve favorable results by evolving the rule base of a fuzzy system with fixed membership functions.

The goal of this paper is to investigate the use of fuzzy logic in game playing environment. First, the card game of Euchre was chosen to be simulated. After that a game strategy that a human goes through needed to be implemented through an algorithm the computer could understand. This was achieved by using several basic Surgeno's fuzzy models in neural networks by Yen and Langari, (1999). In addition to the fuzzy logic involved, classical logic was used for various playing schemes to create a hybrid logical interface. One final unique aspect of this project was the use of the palm operating system. In the end, the palm played an intelligent and competitive game of cards.

Laser lap welding quality is a nonlinear response based on a host of material and process variables. This is inclusive of both categorical and numeric variables. An adaptive neuro fuzzy inference system in combination with a classification and regression tree is used to characterize this process. This characterization is shown to be more accurate than that achieved by a nonlinear regression model.

This paper attempts to study the behavior of asphalt pavements using the fuzzy logic technique. The aim is to be able to determine the pavement condition after a certain time period, given the pavement construction materials, the age and the traffic to which the pavement is/will be exposed. The Long-Term Pavement Performance (LTPP) database is made use of in this paper. The data extracted from this database is used in the designing and testing of the Fuzzy Inference System. The membership functions of all the input parameters are built in accordance with these database records. The rules defining the behavior of the fuzzy inference system are written by taking into account the data available as well as by using logical reasoning. The end result is a system wherein if you enter the pavement data like construction materials, traffic and other pertinent data It can predict the rut depth after the required number of years. The resulting system can predict possible rut depths up to 15 years after construction.

The paper presents the design of a talking thermometer based on microcontroller MC68HC11. It announces the temperature in both Fahrenheit and Celsius using a SPO 256-AL2 speech synthesis integrated circuit from General Instruments. The E port of MC68HC11 is used as an input port, which converts the voltage received from the transducer into an 8-bit number. The thermometer was built and tested, it exhibits an accuracy of /spl plusmn/1/spl deg/F. >

This paper presents the design and development of a device, called the MIDI (Musical Instrument Digital Interface) Inline Transposer, which uses an MC68HC11 microcontroller. The device is used in conjunction with electronic musical instruments and follows the MIDI digital interface specifications and is intended to transpose musical notes from one key to another in real time.

A new approach for determining multiple output terminal reliability of communication circuits is proposed. The approach represents the multiple output terminal paths of given circuits in the form of minterms; and then a switching theoretic approach is used so as to result in non-overlapping simplification of multiple output Boolean functions. The technique will give disjoint terms in which the branch reliability value can be substituted and the multiple output reliabilities are determined. One detailed example using the proposed methodology is then presented.

The paper describes a unique way to generate different musical notes using Motorola's 8 bit microcontroller MC68HC11. These are twelve basic notes in music which have a proportional frequency distribution. The notes have been generated using the output compare port 3 (OC3) in the timer unit. A dedicated output port B of the microcontroller is used to synchronize the colored lights which indicate the note being played. A selection of a well known music is generated for illustration.

This paper describes the implementation of an intelligent chat group based on fuzzy logic. The chat group is based on client server model. The expert system based on fuzzy logic is developed in Java and monitors parameters, like number of clients, the idle time of each client etc., and then determines the priority of each client.

Presents an expert system based on fuzzy logic to predict the rut depth of asphalt pavements, given the parameters of construction materials, pavement thickness, road age and traffic count. A GUI in Visual BASIC was built, which invokes the fuzzy inference engine when submitted with the required input data. The result obtained is sent to the GUI window as rut depth. The aim is to be able to determine the pavement condition after a certain time period, given the pavement construction materials, the age and the traffic to which the pavement is/will be exposed. The long-term pavement performance (LTPP) database is made use of in this paper. The data extracted from this database is used in the designing and testing of the fuzzy inference system. The membership functions of all the input parameters are built in accordance with these database records. The rules defining the behavior of the fuzzy inference system are written by taking into account the data available as well as by using logical reasoning. The resulting system can predict possible rut depths till 15 years since construction. A correlation of 0.700829 was obtained between the measured readings available in the data and the output obtained by the fuzzy inference system. The regression analysis done on the data gave an R square factor of 0.92 and higher for the fuzzy model and 0.78 and lower for the actual data. This indicates that the fuzzy expert system gave a much better estimate of the prediction of the performance of pavement than the original data.