Seokhyoung Lee

Statistical properties analysis of real world tournament selection in genetic algorithms

Abstract Genetic algorithms (GAs) are probabilistic optimization methods based on the biological principle of natural evolution. One of the important operators in GAs is the selection strategy for obtaining better solutions. Specifically, finding a balance between the selection pressure and diversity is a critical issue in designing an efficient selection strategy. To this extent, the recently proposed real world tournament selection (RWTS) method has showed good performance in various benchmark problems. In this paper, we focus on analyzing characteristics of RWTS from the viewpoint of both the selection probabilities and stochastic sampling properties in order to provide a rational explanation for why RWTS provides improved performance. Statistical experimental results show that RWTS has a higher selection pressure with a relatively small loss of diversity and higher sampling accuracy than conventional tournament selection. The performance tests in a traveling salesman problem further confirm that the comparatively higher pressure and sampling accuracy, which are inherent in RWTS, can enhance the performance in the selection strategy.

Distributed Estimation Fusion With Application to a Multisensory Vehicle Suspension System With Time Delays

A new distributed fusion filtering algorithm for linear multiple time-delayed systems is proposed. The multisensory distributed fusion filter is formed by the summation of local Kalman filters having time delays (LKFTDs) in both the system and measurement models. The proposed distributed filter has a parallel structure that enables processing of multisensory measurements; thereby, it is more reliable than the centralized version if some sensors turn faulty. The key contribution of this paper is the derivation of recursive error cross-covariance equations between the LKFTDs to compute the optimal matrix fusion weights. In the particular case of multisensory dynamic systems having time delays in only the measurement model, the obtained results coincide with the previous work of Sun. The high accuracy and efficiency of the proposed distributed filter are then demonstrated through its implementation on a vehicle suspension system.

Computationally Efficient Multisensor Fusion Estimation Algorithms

This paper provides two computationally effective fusion estimation algorithms. The first algorithm is based on Cholesky factorization of a cross-covariance block matrix. This algorithm has low computational complexity and is equivalent to the standard composite fusion estimation algorithm as well. The second algorithm is based on a special approximation scheme for local crosscovariances. Such approximation is useful to compute matrix weights for fusion estimation in a multidimensional-multisensor environment. Subsequent computational analysis of the proposed fusion algorithms is presented with corresponding examples showing the low computational complexities of the new fusion estimation algorithms.

Multisensor Estimation Fusion of Nonlinear Cost Functions in Mixed Continuous-Discrete Stochastic Systems

We propose centralized and distributed fusion algorithms for estimation of nonlinear cost function (NCF) in multisensory mixed continuous-discrete stochastic systems. The NCF represents a nonlinear multivariate functional of state variables. For polynomial NCFs, we propose a closed-form estimation procedure based on recursive formulas for high-order moments for a multivariate normal distribution. In general case, the unscented transformation is used for calculation of nonlinear estimates of a cost functions. To fuse local state estimates, the mixed differential difference equations for error cross-covariance between local estimates are derived. The subsequent application of the proposed fusion estimators for a multisensory environment demonstrates their effectiveness.

Distributed Fusion Filter on Images with Time Delays

This paper focuses on a distributed image fusion filtering algorithm and fusion formulas for time delayed multiple pixels received from multiple sensors (cameras). Since local cross-covariances between images are important values to implement fusion formulas, we present exact formulas for cross-covariances which are a vital factor for calculating matrix weights in image processing. Subsequent analysis of the proposed fusion algorithm is presented through a typical example demonstrating the effectiveness of the proposed fusion algorithm.

Low complexity fusion estimation algorithms in multisensor environment

This paper is focused on two fusion estimation algorithms weighted by matrices and scalars. Relationship between them is theoretically established. We present two fast algorithms addressing computation of matrix weights that arise in multidimensional estimation problems. The first algorithm is based on the Cholesky factorization. And since determination of the optimal matrix weights in real-time applications is not practical, we propose the second algorithm based on approximate calculations using special approximation for cross-covariances. Analysis of computational complexity of the both fast fusion algorithms is proposed. Examples demonstrating low-computational complexity of the fast fusion algorithms are given.

Localization performance enhancement using IMM fusion in cricket sensor network

In this paper, we propose an approach for estimating a location of moving-target over a sensor field via distributed IMM estimator with multiple pseudo measurements using trilateration positioning algorithm. Pseudo measurement which indirectly measures the position of target is defined by results of trilateration positioning coordinates. As a single node only can measure the distance to the target in Cricket sensor network, the nodes need to form groups of three to be able to perform trilateration. Also, since trilateration method ignores distance sensing error, adequate error covariance information is estimated in pseudo measurement model using self tuning method. Each distinct group can make local pseudo measurements and these data are fused with local IMM estimates. With these algorithms, localization performance enhancement of moving target is achieved.

Performance comparison of fast distributed fusion filtering algorithms

In this paper, distributed fusion filtering problem in multisensory dynamic system is considered. The approximation scheme for calculation of cross-covariances is presented, which are based on a correlation coefficient in steady-state. In addition, a new limiting cross-covariance method is also proposed, which gives the opportunity to save calculation times, and is computationally useful to implement the fusion filters in a real-time application. Numerical examples demonstrating the effectiveness of the proposed algorithm are presented.