Aiming at the shortage that Particle Swarm Optimization (PSO) algorithm is easy to fall into local optima and has low precision at later evolution process, a modified Multi-Strategies synergy PSO (MSPSO) algorithm was proposed. Firstly, a probability threshold value of 0.3 was set. In every iteration, if the randomly generated probability value was less than the threshold, the algorithm with opposition-based learning for the best individual was adopted to generate their opposite solutions, which improved the convergence speed and precision of PSO; otherwise, Gaussian mutation strategy was adopted for the particle position to enhance the diversity of population. Secondly, a Cauchy mutation strategy for linearly decreasing cauchy distribution scale parameter decreased was proposed, to generate better solution to guide the particle to approximate the optimum space. Finally, the simulation experiments were conducted on eight benchmark functions. MSPSO algorithm has the convergence mean value of 1.68E+01, 2.36E-283, 8.88E-16, 2.78E-05, 8.88E-16, respectively in Rosenbrock, Schwefel's P2.22, Rotated Ackley, Quadric Noise and Ackley, and can converge to the optimal solution of 0 in Sphere, Griewank and Rastrigin, which is better than GDPSO (PSO based on Gaussian Disturbance) and GOPSO (PSO based on global best Cauchy mutation and Opposition-based learning). The results show that proposed algorithm has higher convergence accuracy and can effectively avoid being trapped in local optimal solution.

Aiming at the problem that TelemeTry and Command (TT&C) Message-Oriented Middleware (MOM) with traditional model like message queue and shared memory has the disadvantage of tightly coupling and limited extensible ability. Combining with the current characteristics of TT&C computer system, a function distributed TT&C MOM system with Publish/Subscribe (Pub/Sub) model was put forward. Centralized publish/subscribe server was canceled with its function embedded into the distributed processing unit. The working process of themes global registration, subscriptions global broadcasting and event local matching were designed. And its transmission reliability was achieved through the reliable multicast protocol, its nodes reliability was achieved by soft duplex with virtual IP mechanism and accelerated push-pull heartbeat detection. Experiments show that the average response time of Pub/Sub message is controlled within 100 ms, the packet loss rate of multicast protocol is around 0.86×10^-7, duplex switch delay is up to 56 ms. This TT&C MOM system with Pub/Sub model satisfies the requirement of highly real-time performance and reliability of TT&C applications.

To solve the problem of unbalanced energy consumption in Wireless Sensor Network (WSN), the Multi-path Routing Protocol based on Three-dimensional Space and Regional Co-evolution (MRPTSRC) was proposed. The zoning model was designed to divide the one-hop neighborhood nodes into a set of subspaces. MRPTSRC selected the local optimum node from every subspace and decided the next hop node by the Regional Co-evolution Algorithm (RCA). Weighted strategy of the forward local optimum node was proposed to escape from local optimum and accelerate the convergence speed toward the Sink node. Simulations were conducted on NS-2 platform, the time of the first node death of MRPTSRC was respectively increased by 6% and 3% of the total time, compared with DEgree COnstrained Routing (DECOR) and Forward-Aware Factor for Energy Balance Routing Protocol (FAF-EBRP). The ratio of dead nodes and the relay time of MRPTSRC respectively declined up to 38% and 30%, the standard deviation of the residual energy decreased by 16.7%, when compared with FAF-EBRP. The network lifetime of MRPTSRC increased by 30% compared with DECOR. The simulation results show that MRPTSRC can effectively improve the network performance.

Aiming at the problems of too long operation time and complex modeling of the three-dimensional roadway intersection modeling in geotechnical engineering, a method about cylinder-Axis Aligned Bounding Box (AABB) two-level bounding box detection was proposed according to the characteristics of tunnel's shape. The proposed method could quickly find out the possible intersected triangular elements and established a new approach to solve the Irregular Triangular Network (TIN) modeling problem of tunnel intersection by combining the three-Dimensional (3D) Boolean operation. The basic principles of the cylinder-AABB double bounding box collision detection and the key technologies about Boolean operation to implement intersection modeling in 3D were described, and an optimization scheme for generated entity mesh was proposed. Through the engineering examples, it is proved that, compared with the Oriented Bounding Box (OBB) hierarchical bounding box method, the modeling method by cylinder-AABB detection increases nearly 50% on the bounding box production efficiency in the roadway surface intersection modeling. The proposed method has the advantages of simple modeling, short detection time, high top detection accuracy, and so on.

Concerning the problems that exist in traditional PageRank algorithm, such as topic drifting, neglecting user browsing interests and stressing on old Web pages, an improved PageRank algorithm was proposed. To satisfy user requirements better, factors of users' clicks to links, link structure, browser time on pages, topic relevance decided by contents and existing time of pages were taken into consideration. The experimental results show that compared with the traditional PageRank algorithm, the average value of users' degree of satisfaction has been promoted approximately by 2.1% with the proposed algorithm, and ranking results has been optimized in a certain extent.

A Protein-Protein Interaction (PPI) network clustering model and an algorithm based on the mechanism of the Artificial Immune System (AIS) were proposed to improve the identification accuracy. In this algorithm, the set of cluster centers was regarded as antigens and the neighbor nodes were regarded as antibodies. The antibodies were regarded as the memory cells of clusters by calculating the affinity between the antibodies and antigens. Then excellent antibodies were selected as vaccines, and they were injected into clustering modules to get update. Finally the memory cells were updated after comparing the fitness of the modules before injection. The simulation results on PPI datasets show that, compared with FLOW algorithm, the f-measure of precision and recall value of the new algorithm have got improved.

Traditional outlier detection algorithms fall short in efficiency for their holistic nearest neighboring search mechanism and need to be improved. This paper proposed a new outlier detection method using attribute reduction techniques which enabled the algorithm to focus its detecting scope only on the most meaningful attributes of the data space. Under the reduced set of attributes, a concept of neighborhood-based outlier factor was defined for the algorithm to judge data's abnormity. The combined strategy can reduce the searching complexity significantly and find more reasonable outliers in dataset. The results of experiments also demonstrate promising adaptability and effectiveness of the proposed approach.