Search

In this paper, a new mobile robot mapping algorithm inspired from the functionality of hippocampus cells is presented. Place cells in hippocampus can store a map of the environment. This model fuse...

Intrusion detection system (IDS) is one of the key security mechanisms to preserve integrity against attacks in networks. Immunity-base intrusion detection is a well-known approach to develop IDS in WSN. The Negative Selection Algorithm (NSA) is one of the immunity-base methods for classification of network attacks. The selection of the proper radius for self-samples is a challenging problem in NSA. These radiuses are used for training and constructing detectors. The incorrect selection of a radius can cause the problem of boundaries invasion and the overlap between the samples, which increase the false detection rate. To solve this problem, self-samples must be optimized before being used in the detection phase. In this paper, we propose a variable radius for self-sampling based on affinity density to reduce the false detection. The NSL-KDD data set is used to evaluate the proposed method. The results indicate that the proposed method can reduce the false detection rate and along with increasing true positive rate.

One of the challenges in recognizing handwritten texts is the individual style of writing. There is the structural similarity of the different digits to each other in writing. Along with the mentioned challenge, these similarities in the text may increase and make it difficult to correctly recognize digits and numbers. In this paper, a new model for recognizing handwritten digits is presented. The proposed model is a stacking ensemble classifier. This classifier is based on the convolutional neural network (CNN) and the bidirectional long-short term memory (BLSTM). Another innovation of the model is the use of the probability vector of images class as the input of the meta-classifier layer. One of the strengths of BLSTM is the ability to learn arrays and vectors; therefore, from a technical point of view, considering the output probability vector of the first model as the input of the meta-classifier (BLSTM) improves the accuracy of the deep learning model. The reason for using stacking ensemble classification is the sameness of the main body of some Persian/Arabic digits (e.g., “2, 3, and 4”). Also, the style the author writes makes classes that are not similar in a structure similar to each other, which causes errors incorrect recognition. This model helps to recognize the correct set of input digits by examining the structure of similarities. To achieve a reliable result in the face of this challenge, this model has been tested on a large Persian/Arabic dataset to cover a wide range of writing styles from different individuals. The dataset has a total of 102,352 data which 60,000 of them are for training data and 20,000 of them are for test data in ten classes of digits that are used in this paper. The result of using this database is to improve the recognition performance of these challenging digits. In examining the datasetpresented by the model, the accuracy rate of the training set was 99.98%. And the sample accuracy rate in the test set was 99.39%. That is, compared to experimenting with the convolutional neural network and other researches, the rate has increased. (All codes available at http://web.nit.ac.ir/ h.omranpour/ .)

Image segmentation is one of the most important and challenging problems in image processing. The main purpose of image segmentation is to partition an image into a set of disjoint regions with uniform attributes. In this study, we propose an improved method for edge detection and image segmentation using fuzzy cellular automata. In the first stage, we introduce a new edge detection method based on fuzzy cellular automata, called the texture histogram, and empirically demonstrate the efficiency of the proposed method and its robustness in denoising images. In the second stage, we propose an edge detection algorithm by considering the mean values of the edges matrix. In this algorithm, we use four fuzzy rules instead of 32 fuzzy rules reported earlier in the literature. In the third and final stage, we use the local edge in the edge detection stage to more accurately accomplish image segmentation. We demonstrate that the proposed method produces better output images in comparison with the separate segmentation and edge detection methods studied in the literature. In addition, we show that the method proposed in this study is more flexible and efficient when noise is added to an image.

In this paper, we want to solve multi-objective robot path planning problem. A new elitist multi-objective approach is proposed to determine Pareto front based on coefficient of variation. Intelligent water drops (IWD) algorithm is generalized by this approach and as a new multi-objective IWD algorithm. We call our new algorithm CV-based MO-IWD. It tried to optimize two objectives: length and safety of the path. In the CV-based MO-IWD, we want to discover solutions as close to optimal Pareto solutions as possible and find solutions as diverse as possible in the obtained Pareto front. In this way, coefficient of variation of Pareto front is determined in each objective. Then, appropriate number of heuristic operations (local search in this paper) is calculated and applied for each solution. Implementation results and comparisons with NSGA_II algorithm show the ability of the proposed approach to achieve a near optimal Pareto front with a good diversity, while the number of fitness function calls does not increase. This method is superior because of suitable distribution of heuristic operations.

The electroencephalography (EEG), which is a method for monitoring the brain signals, is a common method used to diagnose the epileptic seizures. In this study, some features are presented for the classification of the brain signals. These features are based on the texture and structure of the brain signals in the phase space representation (PSR). Due to the resonance property, the data are elliptical in the phase space. Therefore, the mentioned features are based on the calculations of the data density in the ellipses. In the first method of feature extraction, the radius values of the ellipses are assumed based on the normal distribution feature. In the two other methods of the presented features, the radius values of the assumed ellipses are calculated by the optimizer. These methods of feature extraction are based on the incremental ellipses and intersecting ellipses, respectively. The density of the data in the assumed ellipses is given to the k-nearest neighbor as a feature to classify the epileptic seizure and seizure-free EEG signals. The intended method was implemented and investigated on two databases of the Bonn university of Germany and the neurology and sleep center of New Delhi. The results indicate that the proposed features are strong tools for separation and diagnosis of this type of signal and have higher accuracy compared to the other classic and updated methods. The extraction speed of the presented features was higher in the test phase compared to the other methods.

PSO is a way of optimizing a function without knowing the function by analyzing social behavior of particles in a swarm. In PSO every particle knows it's neighbor's best and will be move toward it; if the neighbor's best is better than its own best and this effects the velocity vector of a particle. Evaluation and comparison results shows that our algorithm achieves considerably higher average. In this paper, we combine c1, c2, w parameters with the encoding space of a particle. By doing this personal best of each particle and finally the global best of whole swarm will improve remarkably.

Mapping is one of important issue in robotics. In this paper a new vision based odometery free algorithm for representing environment map is proposed. Our contribution is providing a model based on two important types of brain cells. Place cells in hippocampus play main role in mapping and entorhinal cortex layer cells reacting to image features provide main portion of positional cells input. Our contribution is modelling these two important types of cells by Kohonene self-organizing map. Considering Gabor filter compatibility in determining image texture, V1 visual layer is modelled by a Gabor filter histogram and the output of filter on input images is feeding to a Kohonene self-organizing map that models entorhinal cortex layer cells. The model is implemented on real robot equipped with an omni-directional camera in a real 10∗30 meter environment. The results showed that the robot can make an appropriate map and determine its location in drawn map.

In this paper, a novel method for mobile robot map representation is presented. We introduce a manifold based map which is constructed by mapping the color histogram of omni-directional camera images into a low dimensional space while preserving local and global geometry. The geometry is preserved using orthogonal and independent feature vectors consisting of the Euclidean data difference and the nearest and farthest neighbor data vectors. The introduced dimension reduction method has shown superior performance compared to LLE and PCA in benchmark data sets providing a hard dimension reduction of 768 into 3. This mapping matrix has been used for real time robot localization and mapping in real world experiments. In large environments, adjacent places share the same set of objects and features therefore dimension reduction may end up in a non-smooth manifold. To apply the method in large environments, first a K-Means method is used to cluster the environment and then each obtained cluster is processed separately. Successful experiments on 7 different data sets show effectiveness of this method for robotic applications such as mapping, visual place classification and kidnaped-robot localization.

Scheduling algorithms have an essential role in computational grids for managing jobs, and assigning them to appropriate resources. An efficient task scheduling algorithm can achieve minimum execution time and maximum resource utilization by providing the load balance between resources in the grid. The superiority of genetic algorithm in the scheduling of tasks has been proven in the literature. In this paper, we improve the famous multi-objective genetic algorithm known as NSGA-II using fuzzy operators to improve quality and performance of task scheduling in the market-based grid environment. Load balancing, Makespan and Price are three important objectives for multi-objective optimization in the task scheduling problem in the grid. Grid users do not attend load balancing in making decision, so it is desirable that all solutions have good load balancing. Thus to decrease computation and ease decision making through the users, we should consider and improve the load balancing problem in the task scheduling indirectly using the fuzzy system without implementing the third objective function. We have used fuzzy operators for this purpose and more quality and variety in Pareto-optimal solutions. Three functions are defined to generate inputs for fuzzy systems. Variance of costs, variance of frequency of involved resources in scheduling and variance of genes values are used to determine probabilities of crossover and mutation intelligently. Variance of frequency of involved resources with cooperation of Makespan objective satisfies load balancing objective indirectly. Variance of genes values and variance of costs are used in the mutation fuzzy system to improve diversity and quality of Pareto optimal front. Our method conducts the algorithm towards best and most appropriate solutions with load balancing in less iteration. The obtained results have proved that our innovative algorithm converges to Pareto-optimal solutions faster and with more quality.

In this paper, we propose a new algorithm for simultaneous localization and mapping in mobile robots which uses evolutionary algorithm and particle swarm optimization. The proposed method is based on both local and global heuristic search methods. In each step of robot movements, the local search is applied in the small search space of odometry errors to improve the map accuracy. A global search method is applied for loop closing. The proposed algorithm detects loops and closes them, detects and solves correspondence and avoids local extremums. With a proper representation of problem parameters in chromosome, the dimensionality of search space is reduced. The proposed algorithm utilizes occupancy grid and does not require land marks which are not available in most natural environments. A new fitness function is proposed that is computationally efficient and eliminates the need for complex statistical calculations as used in current approaches. Results of experiments on real datasets exhibit the superior performance of the proposed method compared to the current methods. DOI: http://dx.doi.org/10.11591/ijra.v1i1.274 Full Text: PDF

In this paper, generalized intelligent water drops (IWD) algorithm is proposed to solve robot path planning problem. The proposed algorithm has two levels; first level, finds best global path. Second level, performs local search at relatively near distances of global path and reduces its length and response time. IWD algorithm, like other nature-inspired algorithms, has not a mechanism to deal with constrained optimization problem. So, a mechanism has been proposed based on repair of infeasible solutions. In this mechanism, proposed local search operator repairs some of infeasible solutions. Simulation result shows ability of IWD algorithm in finding the optimal path.

The resources management in a grid computing is a complicated problem. Scheduling algorithms play important role in the parallel distributed computing systems for scheduling jobs, and dispatching them to appropriate resources. An efficient task scheduling algorithm is needed to reduce the total Time and Cost for job execution and improve the Load balancing between resources in the grid. In grid computing, load balancing is a technique to distribute workload fairly across computational resources, in order to obtain optimal resource utilization with minimum response time, and avoid overload. Load balancing is a crucial problem to grid computing. In this paper, we address scheduling problem of independent tasks in the market-based grid. In market grids, resource providers can request payment from users based on the amount of computational resource that used by them. Beside we consider Makespan and Load balancing. In this paper, NSGA II with Fuzzy Adaptive Mutation Operator is used to address independent task assignments problems in parallel distributed computing systems. Results obtained proved that our innovative algorithm converges to Pareto-optimal solutions faster and with more quality.

In this paper, a technical approach to particle swarm optimization method is presented. The main idea of the paper is based on local extremum escape. A new definition has been called the worst position. With this definition, convergence and trapping in extremumlocal be prevented and more space will be searched. In many cases of optimization problems, we do not know the range that answer is that.In the results of examine on the benchmark functions have been observed that when initialization is not in the range of the answer, the other known methods are trapped in local extremum. The method presented is capable of running through it and the results have been achieved with higher accuracy. DOI: http://dx.doi.org/10.11591/ij-ai.v1i1.367

This paper presents a new approach to clustering wireless sensor networks and determining cluster heads. Here we apply evolutionary algorithms in centralized clustering. LEACH-C is one of the effective (basic) parameters in network life time based on energy consuming value protocol. In suggested evolutionary algorithm, through proper application of evolutionary operators, we search for efficient clustering based on each node's energy. Innovation of our method is in proper representation of chromosomes and also in determining proper fitness function according to problem features based on energy measure. In this paper we propose two new fitness functions. The results of applying these functions on four different types of sensor networks are presented and compared with the results of the Simulated Annealing method.

the most important factor for a mobile robot is navigation. Success in localization is one of the four requirements of navigation which includes perception, localization, cognition and motion control. Localization is usually done in two ways: global localization problem and position tracking problem. In the global localization problem the robot does not know its initial position and it has to localize itself. In the position tracking problem for a robot given its initial position, the goal is to keep track of the position while the robot is navigating through the environment. In this paper we propose a method using an evolutionary algorithm for mobile robots' global localization. We use Imperialist Competitive Algorithm (ICA) to achieve our goal. The method that is used for global localization works in two steps: First we estimate X and Y and then we find the best direction for X and Y which we estimated in previous step. The results which have been demonstrated in the experimental results section show that our algorithm works very accurate in all of the positions that we located the robot for our tests with acceptable number of iterations.

In this paper, cable life time and its failure rate per 100 km is estimated using artificial neural network (ANN). The only parameter effecting on cable life time and its failure is considered to be operating temperature but actually cable life time is affected by several parameters. The data required for training the network was extracted from previous works being done on cable life time estimation based on mathematical models. From experimental results obtained by the ANN, it was seen that ANN is an appropriate tool for cable life time and its failure rate estimation.

In this article an algorithm is presented for edges coloring of the graph. In this algorithm the node that we discuss represented by an Agent and then all the nodes of a graph like multi-agent system. Each node independently colours its edges with respect to cellular automata using distributed and paralleling of the process. Innovation of this method is on distributing of the process for each node so that each node updates its edges color only with the use of its neighbors in some steps until all graphs are colored completely. At the end the results are being tested on some Standard graphs and the results are presented. In this method, each graph is colored using one rule repetition for each node with high speed on its edges.