Search

This paper proposes a novel harmonics cancellation approach for DC-DC boost converters by using two-phase sliding mode (SM). On the concept of multiphase SM and hysteresis modulation (HM), the circuit topology of boost converter is extended into two parallel channels so that the two phase SM controllers can be applied for the direct ON/OFF control of converters. The trajectory of system state is analyzed in the neighbourhood of equilibrium state, and the influence relationship between the HM width and harmonics is determined. Finally three cases are considered and the cooperation strategy of the two SM controllers are deduced for the quantitative harmonics cancellation. Simulations validate the proposed approach.

This paper investigates stability and stabilization analyses issue for T-S fuzzy-model-based systems involving distributed time-delay. The novel imperfect premise matching methodology is employed to derive fuzzy controller, unlike conventional PDC method, this novel technique allows the designed fuzzy controller and the T-S fuzzy model to have distinct membership functions and number of rules, as a result, great flexibility and better robustness can be achieved. Moreover, with the introduction of one novel inequality which is tighter than other existing ones, relaxed LMI-based stability and stabilization criteria are obtained in term of the Lyapunov stability theory. Lastly, two numerical examples are offered to prove the valadity of the derived results.

In this paper, based on the traditional Partial Least Squares(PLS) algorithm, a new way to select the most effective data sets for the PLS regression is proposed. The reason why we apply this new approach is that it could maintain or even surpass the original performance of control and diagnosis of a certain process while keep the data sets as less as possible to enhance the conciseness. The most significant advantage of the proposed data set selection method is that it identifies the data sets with the most typical characteristics of the group of data, excluding less informative data. Based on the ordinary PLS algorithm and with the improvement of conciseness, a better performance on prediction and fitting could be achieved. To achieve these goals, the ordinary and the improved PLS algorithm are introduced, after which a new method of data set selection is proposed. Specifically, the enhanced effectiveness of the proposed approach could be revealed by the simulation results of a numerical case.

Achieving precise and robust human detection and tracking over camera networks is a very challenging task in the research of intelligent video surveillance. Its difficulties mainly result from abrupt human object motion, object occlusion and object scale change, and changing object appearance due to changes in illumination and viewpoint, non-rigid deformations, intra-class variability in shape and posture, and potential camera movement, non-overlapping field of views between cameras. This paper surveys the frameworks of human detection and tracking systems over camera networks with non-overlapping field of views. There are three crucial functional modules discussed in this survey paper, namely, human detection under a single camera, human tracking under a single camera as well as human tracking across multiple cameras with non-overlapping field of views. Existing problems, challenges and future research directions are also addressed based on the analyses of the research status of each function module.

The navigation error of a platform inertial navigation system is a key factor to evaluate comprehensive performance of inertial navigation system. This thesis built an error model of the platform inertial navigation and used a proximal support vector regression (PSVR) method to quickly evaluate navigation errors of the platform inertial navigation system. The error model established in the thesis considered 36 error sources and was almost able to completely reflect the degree of influence between each one of the error sources and comprehensive performance of the platform inertial navigation system. Compared with the navigation error of the system obtained through the solution of inertial navigation system error differential equations, the assessment method in the thesis can meet timeliness requirements. The result of simulation showed that the evaluation methods of navigation errors in platform inertial navigation system based on PSVR features high evaluating accuracy with less time consumption and has more obvious advantages in massive data processing compared with the standard support vector regression.

Background subtraction and frame differencing algorithm are widely used approaches for detecting moving objects from video sequences. Many different methods have been proposed over the recent years and many related researchers can be confused about their benefits and limitations. In order to overcome this problem, this paper provides an improved algorithm based on these two techniques. It uses surendra background algorithm to get background model and uses frame differencing algorithm to update it, then extracts the moving objects from the fusion of images extracted from these two methods. Because the moving objects we used in the video are bees which always fly in high speed and they are very small, the moving target becomes difficult to extract. In order to solve this problem and get more moving information, the circle segmentation dynamic threshold is used in this paper. It is shown through the result that the method can satisfy the precision and accuracy in the video sequences without affecting the requirement of real-time.