Papers by Keyword: Convergence

Abstract: Taking an immune operator added to the framework of the genetic algorithm, individuals vaccinated to further enhance the survival ability of the individual, the immune genetic algorithm as a set of immune mechanisms and evolutionary mechanisms in one of the new evolutionary algorithm is applied to reverse genome rearrangement problem is a new attempt, the genetic algorithm can be avoided premature shortcomings, while the immune genetic algorithm parallelization can improve convergence speed.

Abstract: Relaxed technique is one of techniques for improving convergent rate of splitting iterative methods. In this paper,we study the convergence of both local relaxed parallel multisplitting method associated with USAOR multisplitting for solving a large sparse linear system whose coefficient matrix is an H-matrix.

Abstract: The problem constructed here is the convergence of bivariate trigonometric interpolation sequences, the approximation and computation would be optimal for a body of functions. Therefore some conclusions became a special cased of the results of present paper.

Abstract: Based on the requirement of small errors in measuring roundness (cylindricity) error, a leveling methodology which using a dual-point vertical layout has been put forward and analyzed. According to the direction cosine of the axis of workpiece, the amount of leveling has been defined and calculated, which overcomes the problem brought by manual adjustment technology and forms theoretical bases of fast, accurate leveling and high precise measurement. The assessment of roundness (cylindricity) error is to search for a center (cylinder axis) which satisfies the minimum condition. Due to the reliance of initial solutions and relative slowness in terms of convergence precision and convergence rate when using Nelder-Mead simplex method, a combinative method of Quasi-Newton and N-M simplex method has been proposed which achieves a fast, accurate search for global optimums. With the proof of the simulation of classical testing functions using Matlab and the measured data, the convergence rate and precision will be enhanced effectively has been certified with the combination of both methods mentioned above which ensuring and improving the measuring precision of workpiece.

Abstract: Fingerprint image preprocessing technology has been widely used with the development of intelligent identity recognition system, which studies fingerprint image restoration, enhancement and thinning based on digital image processing technology. This paper presents system design and realization of preprocessing technology for fingerprint image. Fingerprint image has restoration processing by adopting image filed calculation, image background segmentation, image equalization, image convergence and image smoothing algorithm in order at first. Then Gabor wavelet algorithm is used to enhance the contrast of fingerprint image. Image thinning algorithm is carried out to extract the skeleton of fingerprint image at last. Test result shows that the method presented has good accuracy, quick speed and strong robustness for realtime application.

Abstract: Convergence of unsteady classification model with data fluctuations has strong application value. This paper models the mathematical problem to verify the feasibility and finite convergence of unsteady classification model with data fluctuations, and verifies the feasibility and limited convergence of the model by convex optimization KKT equivalent conditions from different perspectives. Experiments validate the three real data sets collected, and results show that the proposed model is feasible and finite convergence.

Abstract: Accurate road data is critical to vehicle ride comfort research. Trigonometric series (TS) method is widely used in road roughness construction, but the accuracy of the road data in time domain is not high and the power spectrum density (PSD) exits poor convergence problem. To solve the above problems, this paper has made two improvements: generate random frequency in each band using the local PSD function, optimize phase angle using the principle of minimum energy. The simulation results show that the improved method has higher precise and better convergence of PSD.

Abstract: A tightly-coupled NFE model combined with Neural network, Fuzzy reasoning and Expert system is constructed in this paper based on the neural network-fuzzy reasoning model by introducing expert system technology. The structure and modeling process of the generalized model are presented, and it is derived in theory that this model possesses convergence property and can be regarded as a category of complete information fusion model.

Abstract: Particle Swarm Optimization (PSO) algorithm converges fast but it is easy to fall into local optimum, and bacterial chemotaxis (BC) algorithm prevents premature convergence and prevents falling into local optimum, so a new mixed bacterial chemotaxis (MBC) algorithm is proposed by combining the PSO with BC. The novel algorithm is applied to reactive power optimization on power system. First the PSO is used to find best solution, then BC is used to find the optimal solution among the selected area of previous step, the reserving elite strategy is introduced to enhance the efficiency of the algorithm, and then the optimal solution is obtained. Through the comparison with PSO and BCC in the reactive power optimization of IEEE30-bus system, the results indicate that MBC not only prevents premature convergence to a large extent, but also keeps a more rapid convergence rate than PSO and BCC.

Abstract: This paper introduces an homotopy algorithm which has convergence stability to solve the alternating current optimal power flow problem. The complicated Alternating Current Power Flow (ACPF) can simplify as simple Direct Current Power Flow (DCPF). The homotopy participation factor is introduced into the linear DCPF to make DCPF convert back into ACPF gradually to realize Alternating Current Power Flow Homotopy method (ACPFH). The homotopy curves are generated to solve a series of nonlinear problems.The traditional method can not solve the unstable points,because the calculate process always turn up Jacobian matrix.But the Homotopy method can calculate all results. It is a superiority for Homotopy,and then can explore power system problem more entirety.This novel algorithm is different from Newton - Raphson method, because it isnt sensitive to the initial point selection and has the global convergence.The homotopy algorithm is applied to IEEE - 3, 9, 14, 30, 36, 57, 118 node testing systems for power flow optional calculation, the simulation results show that the novel algorithm can solve power flow problem better and its calculating speed is much faster than the traditional algorithm, it can calculate the optimal value more direct.