Though many numerical methods have been put for nonlinear equations, their convergence and performance are highly sensitive to the initial guesses of the solution pre-supplied. However, the selection of good initial g...
详细信息
Though many numerical methods have been put for nonlinear equations, their convergence and performance are highly sensitive to the initial guesses of the solution pre-supplied. However, the selection of good initial guess is often of hard work. Aiming at this, a novel approach is proposed to resolve nonlinear equations. It takes genetic algorithms' new achievement differential evolution algorithms as the main technique. With a function deflection technique and a novel space contraction method to re-initialize, it resolve nonlinear equations by transform them into correspondent optimization problems. Convergence reliability, computational cost and applicability of different algorithms were compared by testing several classical nonlinear equations and a benchmark mechanics problem. The numerical experiments done show that the put approach has reliable convergence probability, high convergence rate and solution precision. And DE is a successful approach in solving equations both in theory and application.
As a novel and promising algorithm, differential evolution (DE) has shown good performance in lots of optimization problems. It has been said that DE is one of the most competitive EAs for continuous optimization. As ...
详细信息
As a novel and promising algorithm, differential evolution (DE) has shown good performance in lots of optimization problems. It has been said that DE is one of the most competitive EAs for continuous optimization. As a kind of EAs, GT algorithm is a novel algorithm which based on multi-parent crossover. Compared with GT algorithm, DE performances better to find the global minima obviously. This paper presents a concept of pattern analyses to analyze the reason of the DE's highperformance. Then a new algorithm based on GT's multi-parent crossover and traditional DE's discrete recombination is presented for enhancing the performance of the obsolete and inefficient GT algorithm. According to the pattern analyses, the new algorithm obtains several patterns similar to DE. The experiments show the efficiency of the proposed new algorithm.
Particle Swarm optimization (PSO) is a new random computational method for tackling optimization functions. However, it is easily trapped into the local optimum when solving the complexity and high-dimensional problem...
详细信息
Particle Swarm optimization (PSO) is a new random computational method for tackling optimization functions. However, it is easily trapped into the local optimum when solving the complexity and high-dimensional problems, which makes the performance of PSO greatly reduced. To overcome this shortcoming, the paper proposes an Improved Particle Swarm optimization (IPSO), by adding the third particle of having a more room for progress to guide the current particles' velocity updating rule, Which can keep the diversity of the particles and reduce the probability of trapping into the local optimization .Besides, the program enhances and improves the stability and the convergence speed of the algorithm according to adjusting the particles which go beyond the default position space in each interiors. Five benchmark functions are tested, and the results indicate the effectiveness of the new program.
This paper describes the null steering method by position perturbation of selected elements with minimum sidelobe level using a proposed hybrid Enhanced Particle Swarm optimization (EPSO) / Differential Evolution (DE)...
This paper describes the null steering method by position perturbation of selected elements with minimum sidelobe level using a proposed hybrid Enhanced Particle Swarm optimization (EPSO) / Differential Evolution (DE) algorithm. EPSO and DE algorithms are proved to be high-performance evolutionary algorithms capable of solving nonlinearoptimization problems. Therefore, in this paper the two algorithms are combined and a robust EPSO/DE hybrid method is developed. This newly developed optimization algorithm is then applied to find the optimum number and the position of selected perturbed elements in the array to control the null towards the interference direction. The position perturbations of antenna elements are performed over the axial direction, elevation direction or combination of both. The hybridized algorithm has shown to give best result compared to each individual algorithm when applied to optimize different configurations.
Ensemble has been proved a successful approach for enhancing the performance of single classifiers. But there are two key factors influencing the performance of an ensemble directly: accuracy of each single member and...
详细信息
Ensemble has been proved a successful approach for enhancing the performance of single classifiers. But there are two key factors influencing the performance of an ensemble directly: accuracy of each single member and diversity between the members. There have been many approaches used in the literature to create the mentioned diversity. In this paper we add a novel approach, in which classifier type variance is utilized along with feature subset diversification to create a high diversity ensemble of different classifiers and an optimization is conducted on the initial population using a multi-objective evolutionary algorithm. The results of experiment over some standard data sets exhibit the outperformance of the suggested approach in comparison to existing ones in specific situations.
The novel compact folded Yagi antenna, for its high gain and superiority in small size, is being interested in various electromagnetic application areas. But for its complex structure, as well as the strong coupling a...
详细信息
The novel compact folded Yagi antenna, for its high gain and superiority in small size, is being interested in various electromagnetic application areas. But for its complex structure, as well as the strong coupling among the elements, making the optimization and design rather difficult. In this paper, the impact of multi-scale semi-solutions to the final performance of antenna is taken into account, a genetic algorithm applied to the multi-objective optimization of folded Yagi antenna is realized. To illustrate the feasibility and effectiveness of the algorithm, a three-element double-folded Yagi antenna is designed and fabricated, the measurements show good results in gain (7.40dBi) and perfect match in impedance (49.87-j0.49Ω), which fully verify the reliability of the algorithm.
Fuzzy control is a nonlinear control method; its performance depends on the fuzzy control rules, quantification factor and the scale factor. Because these multiple objectives are influence with each others, therefore ...
详细信息
Fuzzy control is a nonlinear control method; its performance depends on the fuzzy control rules, quantification factor and the scale factor. Because these multiple objectives are influence with each others, therefore it is very difficult to ensure the control effect only rely on expertise's experience. In this paper, dynamic evolutionary algorithm is adopted to optimize the multi-objectives in fuzzy controller. In dynamic evolutionary algorithm, based on the competitive relationship between particles' free energy and entropy in the phase space system, a new option strategy is put forward which is used to maintain species diversity; furthermore, multi-parent crossover operator is selected. This operator selects some individuals to form a space and then do searching in this space. This algorithm has strong ability to find the solutions of the problem, and it also run quickly compared with other traditional algorithms. Simulation results show that fuzzy controller optimized by the algorithm has good steady-state response time, steady-state error and overshoot.
HMM has high power to describe complex phenomena. The Baum-Welch (BW) algorithm is very popular estimation method that use for estimating HMM model parameters but it start with an initial guess and finally converge to...
详细信息
HMM has high power to describe complex phenomena. The Baum-Welch (BW) algorithm is very popular estimation method that use for estimating HMM model parameters but it start with an initial guess and finally converge to a local optimum in practice. Chaos often exists in nonlinear systems. It has many good properties such as ergodicity, stochastic properties, regularity and high sensitivity to initial states. In this paper by use of these properties of chaos, an effective hybrid CHAOS-BW optimization method is proposed that uses the Chaos optimization algorithm to optimize the initial values of Baum Welch algorithm. This algorithm not only overcomes the shortcoming of becoming trapped in local optimum of the BW algorithm, but is also fast and requires less storage than other hybrid optimizationalgorithms such as GABW, PSOBW and GAPSOBW. Experimental results on Persian digit dataset show that the propose method has both qualities of global search as well as rapid convergence. Comparison with several other more conventional approaches also reveals superior performance of the proposed model.
In this paper, we manage to use the clustering method realize sonar image segmentation. A particle swarm optimization (PSO) based FCM algorithm (PSO-FCM) is proposed which PSO incorporate with Fuzzy Clustering Method ...
详细信息
In this paper, we manage to use the clustering method realize sonar image segmentation. A particle swarm optimization (PSO) based FCM algorithm (PSO-FCM) is proposed which PSO incorporate with Fuzzy Clustering Method (FCM). The algorithm takes the clustering result of PSO as the initialization of the FCM, and uses fuzzy measures and fuzzy integrals to express the adapt function. At last, the algorithm is applied to the high resolution sonar image segmentation. Segmentation results of FCM and PSO-FCM for several sonar images are compared, which show that the PSO-FCM algorithm has better performance and fit for the sonar image segmentation better than the FCM does.
This paper proposed a whole new design for RV40(realvideo40) video decoder from FFmpeg(fast forward Moving Picture Experts Group),which is in lack of the ability to decode smoothly when decoding a period of high bit r...
详细信息
This paper proposed a whole new design for RV40(realvideo40) video decoder from FFmpeg(fast forward Moving Picture Experts Group),which is in lack of the ability to decode smoothly when decoding a period of high bit rate frames within some video streams. The design specially optimized the sub pixel interpolation algorithm, which is the core of motion compensation. The design uses detailed, different algorithms based on different frame types. Meanwhile a temporal related, flexible algorithm usage method is introduced to deal with the jamming in decoding sequence. The time consumed by calculating is reduced with an acceptable reduced frame appearance. This design is used on a test platform for testify. The result returned shows that this design can greatly promote decoding performance up to 50% while reducing a little picture performance.
暂无评论