Optimal Threshold Policies for Robust Data Center Control

With the simultaneous rise of energy costs and demand for cloud computing, efficient control of data centers becomes crucial. In the data center control problem, one needs to plan at every time step how many servers to switch on or off in order to meet stochastic job arrivals while trying to minimize electricity consumption. This problem becomes particularly challenging when servers can be of various types and jobs from different classes can only be served by certain types of server, as it is often the case in real data centers. We model this problem as a robust Markov decision process (i.e., the transition function is not assumed to be known precisely). We give sufficient conditions (which seem to be reasonable and satisfied in practice) guaranteeing that an optimal threshold policy exists. This property can then be exploited in the design of an efficient solving method, which we provide. Finally, we present some experimental results demonstrating the practicability of our approach and compare with a previous related approach based on model predictive control.     

Estimation of Load-Induced Damage and Repair Cost in Post-Tensioned Concrete Rocking Walls

Post-tensioned concrete rocking walls might be used to avoid severe seismic damage at the base of structural walls, decrease residual drift, and lessen post-earthquake repair costs. The prediction of load-induced damage to the rocking wall resulting from seismic loading can provide an extremely valuable tool to evaluate the status and safety of structural concrete walls following earthquakes. In this study, the behavior and the damage state of monolithic, self-centering, rocking walls, as a new type of concrete rocking wall, are investigated. The nonlinear mechanical behavior of the wall is first modeled numerically, and subsequently the mechanical parameters from the numerical simulation are used to generate the local damage index. The results from the damage index model are compared with the full-scale test results, confirming the viability of the numerically based damage index method for estimating the seismically induced damage in concrete walls. Moreover, the estimated damage can be utilized as a qualitative and quantitative scale to assess the status of the wall following seismic loading events. Finally, an equation is proposed to estimate the repair cost based on the predicted damage state for the studied structural system.     

Decomposing and Cluster Refinement Design Method for Application-Specific Network-on-Chips

Along with higher and higher integration of intellectual properties (IPs) on a single chip, traditional bus-based system-on-chips (SoC) meets several design difficulties (such as low scalability, high power consumption, packet latency and clock tree problem). As a promising solution, network-on-chips (NoC) has been proposed and widely studied. In this work, a novel algorithm for NoC topology synthesis, which is decomposing and cluster refinement (DCR) algorithm, has been proposed to minimize the total power consumption of application-specific NoC. This algorithm is composed of two stages: decomposing with cluster generation, and cluster refinement. For partitioning and cluster generation, an initial low-power solution for NoC topology is generated. For cluster refinement, the clustering is optimized by performing floorplan to further reduce power consumption. Meanwhile, a good tradeoff between power consumption and CPU time can be achieved. Experimental results show that the proposed method outperforms the existing work.     

Research on Web Page Classification Method Based on Query Log

Web page classification is an important application in many fields of Internet information retrieval, such as providing directory classification and vertical search. Methods based on query log which is a light weight version of Web page classification can avoid Web content crawling, making it relatively high in efficiency, but the sparsity of user click data makes it difficult to be used directly for constructing a classifier. To solve this problem, we explore the semantic relations among different queries through word embedding, and propose three improved graph structure classification algorithms. To reflect the semantic relevance between queries, we map the user query into the low-dimensional space according to its query vector in the first step. Then, we calculate the uniform resource locator (URL) vector according to the relationship between the query and URL. Finally, we use the improved label propagation algorithm (LPA) and the bipartite graph expansion algorithm to classify the unlabeled Web pages. Experiments show that our methods make about 20% more increase in F1-value than other Web page classification methods based on query log.     

Dynamic Behavior of Two Part Towing Cable System During Turning

An improved numerical method is used to simulate the dynamic behavior of a two part towing cable systems during turnings. In U turns and full turns, periodical heave motions are found both for the towed vehicle and for the depressor. Periodic motions of the subsea units and of the cable surface tension are closely related to the turning parameters, such as turning velocity and turning radius. System parameters, such as length of the second cable and the vehicle hydrodynamics, also damp turning instability.     

Product Quality Evaluation Method Based on Product Gene Theory

Traditional quality inspection based product quality evaluation method with complex process has high operating cost and requires more professional knowledge. To remove the above limitation, this paper leads product gene theory into product quality evaluation. Methods of quality influencing factors based modeling and encoding are established. Combined with similarity theory and product gene theory, a product gene similarity analysis based quality evaluation method is proposed. The proposed method is low cost, operates easily and requires less specialized knowledge. A case study is conducted to prove the correctness and feasibility of this method.     

TDOA Passive Location Based on Cuckoo Search Algorithm

This paper formulates a new framework to estimate the target position by adopting cuckoo search(CS)positioning algorithm. Addressing the nonlinear optimization problem is a crucial spot in the location system of time difference of arrival(TDOA). With the application of the Levy flight mechanism, the preferential selection mechanism and the elimination mechanism, the proposed approach prevents positioning results from falling into local optimum. These intelligent mechanisms are useful to ensure the population diversity and improve the convergence speed. Simulation results demonstrate that the cuckoo localization algorithm has higher locating precision and better performance than the conventional methods. Compared with particle swarm optimization(PSO) algorithm and Newton iteration algorithm, the proposed method can obtain the Cram′er-Rao lower bound(CRLB) and quickly achieve the global optimal solutions.     

Adaptive Sliding-Mode Tracking Control for an Uncertain Nonlinear SISO Servo System with a Disturbance Observer

An adaptive sliding mode controller with a disturbance observer (ASMC-DO) is proposed for the control of a single-input and single-output (SISO) servo system which has uncertain parameters, nonlinear friction, disturbance and input saturation. It is difficult to choose the suitable value of the parameters. The newly designed adaptive method is used to reduce the effects of system time-varying parameters, such as the moment of inertia and the damp coefficient. The robustness of object is improved. A DO is selected to approximate the compound disturbance and to render the estimate error convergent in finite time. The stability and the convergence of the closed-loop system are proved by using the Lyapunov theory. Experimental results show that the proposed ASMC-DO can better satisfy the influence of variable parameters and external disturbance to the control precision of the SISO servo system than other two controllers. The effectiveness of the proposed controller is showed. The control input stability and robust performances of the input saturation system are enhanced and the chattering is reduced.     

Adaptive Multi-Objective Optimization of Bionic Shoulder Joint Based on Particle Swarm Optimization

To get the movement mode and driving mechanism similar to human shoulder joint, a six degrees of freedom (DOF) serial-parallel bionic shoulder joint mechanism driven by pneumatic muscle actuators (PMAs) was designed. However, the structural parameters of the shoulder joint will affect various performances of the mechanism. To obtain the optimal structure parameters, the particle swarm optimization (PSO) was used. Besides, the mathematical expressions of indexes of rotation ranges, maximum bearing torque, discrete dexterity and muscle shrinkage of the bionic shoulder joint were established respectively to represent its many-sided characteristics. And the multi-objective optimization problem was transformed into a single-objective optimization problem by using the weighted-sum method. The normalization method and adaptive-weight method were used to determine each optimization index’s weight coefficient; then the particle swarm optimization was used to optimize the integrated objective function of the bionic shoulder joint and the optimal solution was obtained. Compared with the average optimization generations and the optimal target values of many experiments, using adaptive-weight method to adjust weights of integrated objective function is better than using normalization method, which validates superiority of the adaptive-weight method.     

Structural Optimization of Hatch Cover Based on Bi-directional Evolutionary Structure Optimization and Surrogate Model Method

Weight reduction has attracted much attention among ship designers and ship owners. In the present work, based on an improved bi-directional evolutionary structural optimization (BESO) method and surrogate model method, we propose a hybrid optimization method for the structural design optimization of beam-plate structures, which covers three optimization levels: dimension optimization, topology optimization and section optimization. The objective of the proposed optimization method is to minimize the weight of design object under a group of constraints. The kernel optimization procedure (KOP) uses BESO to obtain the optimal topology from a ground structure. To deal with beam-plate structures, the traditional BESO method is improved by using cubic box as the unit cell instead of solid unit to construct periodic lattice structure. In the first optimization level, a series of ground structures are generated based on different dimensional parameter combinations, the KOP is performed to all the ground structures, the response surface model of optimal objective values and dimension parameters is created, and then the optimal dimension parameters can be obtained. In the second optimization level, the optimal topology is obtained by using the KOP according to the optimal dimension parameters. In the third optimization level, response surface method (RSM) is used to determine the section parameters. The proposed method is applied to a hatch cover structure design. The locations and shapes of all the structural members are determined from an oversized ground structure. The results show that the proposed method leads to a greater weight saving, compared with the original design and genetic algorithm (GA) based optimization results.