Constraint handling techniques are of great significance in efficiently solving constrained optimization problems. This paper proposes a novel ensemble framework for constraint handling techniques based on voting-mechanism, in which four popular constraint handling techniques are included. Each of the constituent constraint handling techniques votes for the solutions at each generation based on its own rules. Solutions getting more votes are regarded as promising individuals and survive to the next generation. This ensemble framework based on voting-mechanism reflects the collective wisdom in decision-making of human beings. In addition, a differential evolution (DE) variant is designed as the search engine, in which four search strategies are combined to generate new individuals and maintain the balance between diversity and convergence of the population. The proposed algorithm has been tested on 57 real world single-objective constraint optimization problems and 7 problems are selected using variable reduction strategy (VRS). The experiment shows that the proposed algorithm achieves competitive performance, indicating that the voting-mechanism ensemble constraint handling technique combine DE algorithm together can effectively deal with constrained optimization problems.

With the development of drone technology, drones have been deployed in civilian and military fields for target surveillance. As the endurance of drones is limited, large-scale target surveillance missions encounter some challenges. Based on this motivation, we proposed a new target surveillance mode via the cooperation of a truck and multiple drones, which enlarges the range of surveillance. This new mode aims to rationally plan the routes of trucks and drones and minimize the total cost. In this mode, the truck, which carries multiple drones, departs from its base, launches small drones along the way, surveils multiple targets, recycles all drones and returns to the base. When a drone is launched from the truck, it surveils multiple targets and flies back to the truck for recycling, and the energy consumption model of the drone is taken into account. To assist the new problem-solving, we developed a new heuristic method, namely, adaptive simulated annealing with large-scale neighborhoods, to optimize truck and drone routes, where a scoring strategy is designed to dynamically adjust the selection weight of destroy operators and repair operators. Additionally, extensive experiments are conducted on several synthetic cases and one real case. The experimental results show that the proposed algorithm can effectively solve the large-scale target surveillance problem. Furthermore, the proposed cooperation of truck and drone mode brings new ideas and solutions to targets surveillance problems.

Recently, the monitoring efficiency and accuracy of visible and infrared video have been relatively low. In this paper, we propose an automatic target identification method using surveillance video, which provides an effective solution for the surveillance video data. Specifically, a target identification method via multi-view and multi-task sparse learning is proposed, where multi-view includes various types of visual features such as textures, edges, and invariant features. Each view of a candidate is regarded as a template, and the potential relationship between different tasks and different views is considered. These multiple views are integrated into the multi-task spare learning framework. The proposed MVMT method can be applied to solve the ship’s identification. Extensive experiments are conducted on public datasets, and custom sequence frames (i.e., six sequence frames from ship videos). The experimental results show that the proposed method is superior to other classical methods, qualitatively and quantitatively.