Abstract:Volume rendering is an effective method to visualize complex physical features in large-scale scientific data with high expressiveness, the difficulties in processing huge amount of data and capturing complex features, however, are still a great challenge to volume rendering. To address the challenges and improve efficiency and effect of volume rendering, researchers conducted in-depth research on volume rendering algorithms from three aspects. On the one hand, it is an effective way to improve the efficiency of volume rendering by sharing computation with lots of processor cores so as to reduce the computational amount of one processor core. On the other hand, by fully exploring the intrinsic characteristics of three-dimensional data fields, data reduction methods can greatly decrease the amount of data in the rendering process and thus reduce the overhead of a volume rendering algorithm. In addition, feature analysis and enhancement techniques can also be integrated into volume rendering algorithms, thus complex physical features are highlighted from the data fields and high-quality rendering of scientific data is achieved. A survey of recent advances on volume rendering techniques was presented and various research methods were analyzed. In the end, this paper makes prospection for future research directions on volume rendering of large-scale scientific data, including application-driven feature volume rendering, feature-based data reduction in volume rendering, hardware-adapted multi-level acceleration of volume rendering and in-situ intelligent volume rendering.

Abstract:A novel asynchronous memory access path, which supports highly concurrent and out-of-order off-chip memory requests was proposed. In order to satisfy the requirements of graph applications, a software-defined interface in our proposed memory access path to handle hundreds of kinds of off-chip memory requests with arbitrary granularity via hardware-software co-design methodology was implemented. A custom memory semantic interconnect was designed for fine-grained remote memory access among various computing nodes leveraged in future distributed graph processing scenarios. Last but not least, we integrate our proposed novel memory access path into a RISC-V instruction set architecture-based SoC(system-on-chip) architecture and implement an FPGA prototype. Based on our custom random access microbenchmarks, preliminary evaluation results show that performance of array-based and random address-based off-chip memory access is improved by 3.5x and 2.7x respectively using our proposed asynchronous memory access path, and accessing 4 bytes data from remote memory only takes 1.63 μs.

Abstract:Three key factors exert big influence upon the application′s I/O performance, including the I/O programming interface, the performance characteristic of I/O sub-system( both architecture and system software), and the I/O configuration parameters at user-level. From the user′s perspective, this paper discussed the user-level parallel I/O configuration optimize space toward large scale cluster. Besides, we proposed a method of testing and analyzing the I/O characteristic of large scale cluster. Based on this method, the I/O performance portrait of a domestic super computer was built up and several user-level parallel I/O optimize suggestions were put forward. With these carefully selected I/O configuration parameters, the time of restart data write operation was cut down by 15 percent under 8192 processes in a real application environment, while the program′s initial time is shortened from 10 minutes to 5 seconds at the scale of 4096 processes.

Abstract:An Offload-mode heterogeneous parallel algorithm via inner-out subdomain dividing was proposed for CFD(computational fluid dynamics) program CNS. Combined with the characteristics of finite difference computing and fourth order Runge-Kutta method in structure mesh, the scheme of ghost region was introduced, based on which a Ghost-Region-Shrinking computing scheme was designed, significantly reducing the overhead of data movement between heterogeneous computing resources, making the computing and MPI communication on CPU absolutely overlap with the accelerator computing under load balance condition, bringing better heterogeneous synergetic parallelism. Parameter of the ghost region for the computing validity was given and load balance tuning was demonstrated. On a server with CPU (Intel Haswell Xeon E5-2670 12 cores×2)+MIC (Xeon Phi 7120A ×2), an averaged performance improvement of 5.9× was gained over the algorithm of using accelerator with task blocks integrally. Compared with MPI/OpenMP two-level parallel algorithm running on 24 Intel Haswell CPU cores, the proposed method achieved an accelerating of 1.27× with one MIC and 1.45× with two MICs. Finally the bottleneck and disadvantage were discussed.

Abstract:In some low precision applications, pipelined floating-point reciprocal operation is required actually. Based on SRT-4 algorithm, a pipelined floating-point reciprocal operation unit was designed and implemented, which is constructed as a 6-stage pipeline unit, resulting in an 8-bit valid fractions. In order to support hardware process of denormal numbers, the unit was improved to get higher performance, which is constructed as a 8-stage pipeline unit, adding source operand pre-normalization and result post-normalization function components and supporting hardware process of denormal numbers. After logic synthesis, the area of the unit was increased by 19.23%, which is reasonable. The timing of the unit was not affected obviously and met the expected frequency goal of 1.6 GHz.

Abstract:As the increasing of the scale of parallel systems, some problems such as large measurement cost and memory overhead exist in the diagnostic methods of communication waiting phenomenon. With the deep analysis on the existing diagnostic methods, and considering the actual demand of controllable measurement, a diagnosis model for communication waiting based on hotspot function was established, and a tidy and practical diagnostic method based on the above model was presented. The above diagnostic method was applied to the diagnostic process of the communication waiting phenomenon in the large-scale MPI parallel programs, such as the LARED integration, the LARED-S, the LAP3D. The application results show that this method can accurately identify the key code segment leading to communication waiting and the proposed optimization solution and performance improvement space has reference value for the subsequent program improvement. The optimized LARED-S program, according to the diagnostic result, can increase performance by 32% and reduce communication waiting time by 44%.

Abstract:The traditional method of alarming high-speed rail traffic in gale is based on an instantaneous threshold. Although it covers all alarm events, there are a lot of unnecessary alarms, which affect the efficiency of high-speed rail traffic. An early warning method based on sequence pattern was proposed. It aimed at mining frequent patterns in the preorder data and finding out the changing rules of alarm events. The unique sequence characteristics of early warning sequences were obtained by filtering out the public frequent patterns of non-early warning sequences, and a database of early warning patterns was constructed. Through the verification of monitoring data along Lanzhou-Urumchi high-speed railway, the method can improve the accuracy of prediction, and reduce the rate of missing reports concurrently. It reduces the time required for pattern matching effectively, and reserves sufficient time windows for early warning, which can accord more with the practical application requirements.

Abstract:The analysis of the existing typical hybrid store architectures in big data center found that they cannot fully take advantage of both hybrid storage management systems and hybrid storage devices. Thus a hardware and software co-design data placement strategy was proposed, which simultaneously considers the hybrid storage management systems at software level and the hybrid storage devices at hardware level, and figures out the trail of data placement on both storage systems and devices regarding application characteristics. Moreover, the static placement pattern before running and the dynamic placement pattern during running were proposed on the basis of different application scenes. An experiment was implemented by running three kinds of workloads on simulated data placement strategies based on the model according to the performance parameters of storage management systems and storage devices. The results show that the proposed design outperforms traditional ones that either consider storage management systems or storage devices separately by up to 30%.

Abstract:According to the characteristics of supersonic and subsonic flow, the Prandtl-Meyer expansion wave theory, and the Mach number after expansion waves based on the back pressure and static pressure, the bleed mass flow rate was obtained, the conditions for the occurrence of choking were derived, and a two-dimensional supersonic bleed model was established. According to the impulse theorem, based on the assumption that the state change of gas from the main flow area to the bleed chamber was due to the impulse effect of the pressure difference, a subsonic bleed model was established and the mechanism of subsonic bleed chocking was revealed. CFD(computational fluid dynamics) verification shows that the deviation of the subsonic bleed impulse model is considerable, so the subsonic bleed flow was described by a numerical fitting model. This study provides a reference for three-dimensional bleed research.

Abstract:Based on traditional corrugated core structure,a structure-thermal protection integrated structure was designed for missile in severe aerodynamic and thermodynamic environment. The coupled thermo-mechanical numerical model was established to analyze the structural loading capacity and thermal protection performance of the integrated structure, and results show that the integrated structure has good loading capacity at high temperature and the thermal protection efficiency can be improved by increasing thickness of structural thermal insulating layer and inner wall. Aimed at maximum thermal protection efficiency and minimum structural mass, non-dominated sorting genetic algorithm-Ⅱmethod was adapted to achieve multi-objective optimization. The optimization results show that the thermal protection efficiency factor and structural mass of the integrated structure interact with each other, so it needs comprehensive consideration when designed.

Abstract:When the shooting range′s optical theodolite tracks the target in real time, the camera will randomly jitter, causing the target large-scale motion in the image. In dealing with large-scale motion, the tracking method based on the search window is easy to lose the target, and the tracking method based on the full-image search is time-consuming. Considering these problems, an improved TLD(tracking-learning-detection) framework combining KCF(kernelized correlation filter) and target position prediction was proposed. An orthogonal polynomial optimal linear filter and camera angle information were utilized to predict the position of the next frame of the target, and KCF was used for fast-tracking in this area, which can improve the success rate and save the tracking time, and can detect when the tracking fails. Simulation experiments demonstrate that the optimal linear filter can accurately predict the target position and provide KCF with a more accurate search position. Besides, the algorithm consumes only 1.1ms per frame, and the positioning accuracy is better than that of TLD and KCF, which can effectively copes with camera′s jitter. The actual task verification proves that this method can improve the automatic interpretation level of the shooting range and reduce manual intervention.

Abstract:In order to solve the problem of quantitative analysis of aviation unsafe events under uncertain conditions, based on the random and fuzzy theory, a mixed uncertain description of the subjective and objective variables that affect flight safety was proposed. To describe the subjective variables which were hard to be expressed by probability, the random variables that follow a uniform distribution in the λ-cut set of membership were introduced. Then, an aviation safety index and a numerical computational method under mixed uncertain conditions were proposed, which can reflect the aviation safety level concisely, intuitively and accurately. Through the Bow-tie model which is commonly used in the field of aviation safety, the relationship of the safety index and the membership of tire burst accident was obtained by simulation analysis. The quantitative results show that the aviation safety indices can be significantly improved with proper control of the membership degree.

Abstract:In order to analyze the attitude motion of end-bodies in deployment of STS(space tether system), based on the Lagrangian equations of the second kind, the mathematical model of deployment of the system and the angular motion of end-bodies were developed, to analyze the attitude dynamics of end-bodies in deployment. Since the sizes of satellites are not negligible in actual space missions, the mother-satellite and sub-satellite connected by tether were regarded as rigid bodies with certain sizes, and the mass of the mother-satellite is much larger than the sub-satellite in modeling. Besides, the tether is regarded as a rigid rod with a certain mass. The mathematical model was used to analyze the deployment process of STS, and to analyze the main influence factors to the attitude motions of end-bodies including the initial disturbance of attitude angles, and the dynamic or static asymmetry of end-bodies. Simulation results demonstrate that either the initial disturbance or the dynamic and static asymmetry of end-bodies in deployment can cause the attitude angles instability, or even cause the tether enwinding satellites. The simulation results can provide reference for the attitude control of satellites in deployment.

Abstract:The interstage bolted flange connection is a common connection mode of the missiles (rockets), but it destroys the continuity of the whole missiles (rockets) structure and is of weak load bearing capacity that the whole missiles (rockets) structure is liable to fail so as to lose the strength under external load. Based on the bolted flange connection structure between stages of the actual missiles (rockets), multiple experimental specimens were simplified and manufactured, the quasi-static loading experimental system and the special bolt force signal acquisition sensor were designed, two quasi-static failure experiments were carried out, and the finite element simulation model was established by ABAQUS simultaneously. According to the experimental results and the measured data, the failure mechanism of the connection structure under quasi-static load was analyzed, and the numerical simulation results and precision of the finite element simulation model were proved good. The results in this research can be used as a reference for the loading bearing capacity and experimental design of connection structure between stages of missiles (rockets) under quasi-static load.

Abstract:Orbit maneuvers can cause satellite orbital anomalies and are one of the key concerns of the space situational awareness. A method to detect historical orbital maneuvers from satellite TLE(two line element) data was proposed. By analyzing the prediction errors, the abnormal cataloging values of the selected orbital parameter were detected, and then the corresponding historical maneuvering information was obtained. Firstly, a large number of sample data of prediction errors were obtained from the historical TLE data, and the probability distribution model of the prediction errors expressed by the Gaussian mixture model was fitted from these sample data by the EM(expectation maximization) algorithm. Then, based on the fitted distribution model, the threshold of detecting abnormal cataloging values of orbital parameter through prediction errors was determined. Finally, taking into account the relationship between orbital maneuvers and abnormal cataloging values of the orbital parameter, the method to detect orbital maneuver from the cataloging sequence of the orbital parameter was determined. The results of maneuver detection on typical active satellites show that the proposed method can detect historical maneuvers accurately while maintaining a low false detection rate.

Abstract:Dynamics model of the rocket and its debris were established, and the quaternion algorithm was used to compute the carrier attitude. A prediction method of safety control zone based on infill-sampling Kriging model was proposed, which combined the characteristics of Monte Carlo and Kriging surrogate models. Then a flow chart of prediction was provided furthermore, and one booster rocket′s debris safety control zone was simulated. Simulation results show that, compared with the Monte Carlo method without losing computational accuracy, the proposed method has equivalent computational accuracy and higher computational efficiency:it can meet the rapid iteration requirement for engineering design. While compared with the conventional extreme deviation overlying method, it can reduce the area of safety control zone dramatically and has relatively good engineering application value.

Abstract:In order to meet the real-time field and repeatable detection requirements for the signal coverage effectiveness of wireless communication network, a distributed wireless coverage detection algorithm based on wireless sensor network was proposed. The received signal strength of the wireless communication network was perceived and preprocessed by wireless sensor nodes randomly deployed in the target area. The variogram was used to construct a new BP(back propagation) neural network objective function, and the initial weight and threshold were optimized by the modified particle swarm algorithm. The trained network model was used to estimate the interpolation of the target area with detection blind zone, and the data collected by the sensor nodes were combined to generate the equal signal strength line of the wireless communication network. The simulation results show that the proposed algorithm has higher prediction accuracy than other classical algorithms, and can effectively detect the signal coverage situation of the wireless communication network in the target area.

Abstract:Considering the moving target detection in reverberation background, the high-order statistics obtained by the output of received data after beamforming and matched filtering were regarded as statistical observation space. Based on the high-order statistical characteristic difference of reverberation and target echoes, the high-order statistics can be used to build characteristic vectors by multiple ping output. The Mahalanobis distance between characteristic vectors of the target and reverberation was used as the quantified standard to measure difference between the target and reverberation. The threshold was based on the maximal constant conditional power test. ROC(receiver operating characteristic) curves were obtained under different signal-reverberation ratio conditions by Monte Carlo simulations. Simulations and sea trial results show that the new method achieves higher performance than traditional detection using single ping.The output signal-reverberation ratio, which ensures the false alarm lower than 0.01 and the detection probability higher than 0.5, is reduced to 3 dB, approximately 6 dB less than that of the traditional method.

Abstract:A unified diagnosis model based on probabilistic graphical theory was studied. The model constructing method and its variations in different scenarios were discussed. Complex problems such as fault diagnosis in multimode systems, diagnosis with coupling faults, dynamic faults and fault prognosis were solved by using the framework. In order to combine the advantages of the model-based method and the data-driven method, a model learning algorithm was proposed, by means of which the diagnostic result was improved. In the end, the possible model developing directions and research focus were discussed, which can provide a reference for the follow-up theory research.

Abstract:The muzzle disturbance is an important factor of affecting the naval gun firing′s accuracy. In order to reduce the muzzle disturbance and optimize the barrel structure, the ABAQUS finite element model of flexible barrel was established, and the modal calculation results of the model were compared with the modal test values, which proves that the established flexible finite element model is effective. A method of new quantum chaos particle swarm optimization algorithm and dynamic joint optimization was proposed to carry out multi-objective optimization of the artillery recoil mechanism. Optimization results indicate that the line speed, angular velocity and angular displacement of optimized muzzle center are significantly reduced when compared with that before optimization, the barrel mass is reduced, and the structure is more reasonable. This indicates that the optimization method is effective and feasible, and can provide reference for the overall design of the whole naval gun.

Abstract:The DSM (Delta-Sigma modulator) can get a better performance by increasing the quantizer′s levels. But the multi-bit DSM output cannot drive the SMPA (switch-mode power amplifiers) directly due to the multi-levels of DSM output in the class-S DPA (digital power amplifier). So the multi-level SMPA with multiple SMPA cells is needed. Several level synthesis strategies for 4-level class-S DPA based on 2 bit DSM were proposed and analyzed. The theory and simulation results show that, with an appropriate level synthesis scheme, the 4-level class-S DPA obtains a much better performance. Especially, the effect of level synthesis strategy is more remarkable in power back-off condition.

Abstract:The ASLR (address space layout randomization) is a defense mechanism to prevent the control-flow hijack. The lack of analysis of the impact of ASLR in existed automatic vulnerability analysis and exploit technologies makes the test cases difficult to be used in actual environment. Aimed at the defects of address randomization and features of its bypass technologies, an analysis method was proposed to deal with the vulnerability of ASLR based on program states transition. The FSM (finite states machine) was used to describe the transition of each key state on the program path, the constraints for some common scenes of memory leakage and control-flow hijack were built, and the vulnerability of ASLR was analyzed by solving the compatibility of memory leakage state constraints and control-flow hijack state constraints. Experimental results show that the proposed method can effectively detect ASLR bypass and control-flow hijack attacks caused by memory leakage, realize the automatic vulnerability analysis of ASLR, and improve the efficiency of software security analysis.

Abstract:BITE (built-in test equipment) is widely used in many fields such as fault diagnosis, equipment prognosis and health management. The problems encountered in the process of BITE design and update, including the classifiers update, samples imbalance and hardware limitation, were analyzed, and the initial solutions were proposed. The density-based cluster and artificial immune system were applied to process the raw data; the delegates-based hybrid learning methods were proposed. The evaluation of the solution was validated by the numerical and experiment examples with support vector machine. Results show that the proposed solution can solve the mentioned problems well and is helpful for data based fault diagnosis design and update in the process of BITE maturation.

Abstract:For the nonlinear degradation equipment which widely exists in practice, the current RUL (remaining useful lifetime) prediction methods ignore the effect of random failure threshold. Therefore, the RUL prediction method based on nonlinear degradation processes with random failure threshold was proposed by analyzing equipment′s degradation processes. A nonlinear degradation model based on Wiener process with the individual difference and measurement error was built in this work. Next, the degradation states were updated synchronously by applying the Kalman filtering algorithm and constructing the state-space model. And then, the estimation method of failure threshold distribution parameters based on maximum likelihood estimation was proposed to obtain the probability distribution of the random failure threshold. Finally, an analytical and closed-form RUL distribution based on random failure threshold was derived, and the RUL prediction can be adaptively updated with the available observed data. The case study shows that the presented method can significantly improve the accuracy of RUL prediction and thus it has a certain engineering application value.

Abstract:A sensor scheduling method based on the minimum risk was proposed for different target detecting tasks. The risk of active sensor being intercepted and the risk of target detecting were combined to build the sensor scheduling model based on the minimum risk under the form frame of target detecting. Three different kinds of target detecting task, for example, target tracking, target identification, and estimation of targets′ threat level, were considered separately to crystallize the sensor scheduling model based on the minimum risk, with three calculating methods of the target detecting risk given. For solving the model, an improved artificial bee colony algorithm was raised on the basis of the chaos theory, the backward learning and the two-way roulette. Simulations were made to illustrate that the proposed model is feasible and the algorithm is effective.

Abstract:The support operation process of ship-borne aircraft is constrained by multiple resources, and the interference of emergency may occur at the same time. In view of this characteristic, the amendment strategy of model under the interference condition was analyzed and worked out. The Tabu search operator was introduced into the genetic algorithm to improve the mutation operation of the traditional process. And the simulation results show that the improved genetic algorithm is more efficient than the traditional genetic algorithm in optimizing the multi-carrier aircraft support problem with interference events. The rescheduling scheme is reflected intuitively by Gantt chart. As a result, this method lays a theoretical foundation for effectively dealing with the interference events in the process of ship-borne aircraft support in the real situation.

Abstract:In magnetic levitation control system, the elasticity of orbit is often neglected to design control algorithm. Under this control algorithm, when the track stiffness is small, the system is prone to vibration. To solve this problem, the track elasticity can be added to the suspension model, and then the control algorithm is designed. Considering the track elasticity, the model of the suspension system will be more complex, and the control algorithm is difficult to realize in engineering. Therefore, the Hankel norm approximation method is used to reduce the order of the model considering the track elasticity, and the control algorithm is designed on the basis of the reduced order model to solve the system vibration problem caused by the track elasticity. Moreover, this method has the advantage of easily being implemented in engineering. Finally, the feasibility of the reduced order method is verified by the simulation results.