Search

The mathematical models of many engineering problems can often come down to two-dimensional optimization problems,it has been one of the hot issues of seeking the high effective and precision optimizing method.The characteristics of harmonic function and Green theory were used to study the relations between the values of harmonic function and its boundary values,and demonstrate the non-locality of its extreme value.When the objective function of the optimization is harmonic function,the two-dimensional optimization can be simplified as a one-dimensional optimization on the boundary of its feasible region.At last,three examples were given to show the feasibility of the method here.The research ideas and the conclusions can be generalized to the solution of the three-dimensional optimization problems.

The optimization problems of binary functions widely exist in many practical engineering problems,their global optimization methods have always been one of the hot research issues. A high effective method of global optimization to nonnegative binary functions was put forward basing on the maximum modulus theorem of the analytic function of complex function,this method can simplify the optimization problem of the object function on a two-dimension bounded area as an one-dimension global optimization problem,the theory basis for the optimization method was given,and three numerical examples were given to show the effectiveness of the method. On one hand,the conclusions obtained in this paper can be directly used to solve the optimization problem in which the analytic function theory could be used,on the other hand,it can be used to obtain the global optimization solutions of the two-dimension mathematical optimization problems in which the method could be used in a high precision and high effective manner.

The application of image recognition technology in apple fruit automatic picking and selecting process is an important way for the development of apple industry. At present, the apple image identification and detection take long time and its recognition rate is low. The paper analyzes the image denoising, feature analysis, feature extraction and other key issues, and focuses on occlusion of apple fruit recognition method. Based on the Otsu adaptive threshold segmentation method, an apple detection technology based on local maxima is proposed. Experimental results show7 that the method can accurately extract the characteristics of apple fruit compared with the traditional identification method and local maxima method, especially, for adjacent and occluded the apple image recognition, it maintains a high rate of accuracy. The device can greatly improve automatic level of picking robot for apple. [ABSTRACT FROM AUTHOR]