一种基于混合任务集的高能效调度算法

动态电源与频率调整技术能够帮助实时系统显著减少能耗,之前的研究大多聚焦于基于周期性任务的线程调度算法,却很少考虑周期性与非周期性任务混合的模型。同时,尽管基于CPU利用率的DVS算法可以从系统级上减少能耗,但不能保证实时性。本文提出一种新的算法,它结合减慢因子的DVFS调度算法与系统级的DVS技术,融合PID控制器与自适应的权衡策略为软实时系统提供更好的能耗减少方法。该算法的能耗在服务器利用率低于25%的情况下比加州大学提出的算法下降了14.2%²5.9%,周期性任务超过时限率低于3%。     

Aging and performance of home tasks

Household behaviors must be understood in terms of a hierarchy of behavioral competence within which complexity determines the location of behaviors that can be evaluated in terms of basic health and social-normative criteria. Older people's household behaviors as well as higher-order behaviors are schematized in terms of physical and instrumental activities of daily living (ADL). Data on impairment rates for ADLs and time budget studies are cited to emphasize the importance of the home environment. Other behaviors that represent needs for autonomy, support, and proactivity are discussed, as are research needs, both for development of measures of microbehaviors and for the study of behaviors designed to achieve alternative routes to an instrumental goal whose preferred means of attainment has been disturbed by physical impairment.     

Validity of self-assessed reports of occurrence and duration of occupational tasks

To obtain quantitative estimates of the physical workload in epidemiological and intervention studies of musculoskeletal disorders, there is a need to extend task based exposure data to job exposure profiles. For this purpose a work task diary was developed and evaluated. This was validated against direct observations of a day's work for twenty-two female office workers and twenty female hospital cleaners. There was a good agreement regarding the occurrence of the main tasks. However, the less time-consuming tasks were under-reported. Moreover, about two thirds of the changes between tasks were not reported. The difficulties of defining tasks that function as occupational entities seems to be a major reason for the lack of agreement. The underestimation of the duration of breaks/pauses was most pronounced for the cleaners. Still, the diary would be useful for the calculation of job exposure, by time-weighting task exposure data, when the tasks and/or their duration vary between days.     

Age differences in the performance of information retrieval tasks

In two experiments younger (18-25 years) and older (60-70 years) participants performed an information retrieval task in which they searched for the answers to questions in a hierarchical menu structure. Participants' movement speed, spatial ability, spatial memory, working memory capacity and reasoning speed were measured. Results showed older participants to be slower than younger participants on overall latencies on the information retrieval task. This slowing increases with each consecutive step in the menu structure. Regression analysis showed that movement speed, reasoning speed and spatial ability predicted the overall latencies accurately. Modelling the consecutive steps showed that latencies on the first selection are predicted by movement speed and reasoning speed. Memory and spatial measures are predictors for latencies on steps further into the menu structure only. This finding is consistent with increased slowing of older participants for later selections and suggests that deep menu structures are less suited for older users.     

Age differences in the performance of information retrieval tasks

In two experiments younger (18-25 years) and older (60-70 years) participants performed an information retrieval task in which they searched for the answers to questions in a hierarchical menu structure. Participants' movement speed, spatial ability, spatial memory, working memory capacity and reasoning speed were measured. Results showed older participants to be slower than younger participants on overall latencies on the information retrieval task. This slowing increases with each consecutive step in the menu structure. Regression analysis showed that movement speed, reasoning speed and spatial ability predicted the overall latencies accurately. Modelling the consecutive steps showed that latencies on the first selection are predicted by movement speed and reasoning speed. Memory and spatial measures are predictors for latencies on steps further into the menu structure only. This finding is consistent with increased slowing of older participants for later selections and suggests that deep menu structures are less suited for older users.     

Determination of reachable set for a linear discrete system

An easily implemented method for obtaining the reachable setR_{N}, N > n, of annth-order discrete system with constraints on the input is described. This method uses elementary matrix calculations at each step to construct the boundary hyperplanes of the reachable set. The result can be applied to determine the minimum control time required for a given terminal state. IfN leq n, the method of echelon matrix analysis is employed to solve the same problem.     

Analysis of hand movements during the performance of positioning tasks

An electroacoustic method was developed to analyse hand movements in space. This method was used to analyse the control error of the hand-arm system during the performance of a simple cyclic positioning task. It could be shown that the traces of the hand movements become significantly more dispersed with increasing time at work.     

Configuration space as a means for augmenting human performance inteleoperation tasks

This paper considers an approach to operator-guided real time motion control of robot arm manipulators that's based on the use of configuration space (C-space). The goal is to improve operator performance in a complex environment with obstacles, In such tasks, traditional teleoperation techniques, which are all based on control in work space (W-space), suffer from human errors tied to deficiencies in human spatial reasoning. The C-space approach transforms the problem into one humans are much better equipped to handle-moving a point in a maze-and results in a significant improvement in performance: shorter path, less time to complete the task, and virtually no arm-obstacle collisions. Versions of the approach are described for two-dimensional (2-D) and three-dimensional (3-D) tasks, and tools are developed to efficiently interface the human and machine intelligence. Effectiveness of the C-space approach is demonstrated by a series of experiments, showing an improvement in performance on the order of magnitude in the 2-D case and a factor of two to four in the 3-D case, compared to usual work space control.     

Improving the performance of speaker and language identification tasks using unique characteristics of a class

In classification tasks, the error rate is proportional to the commonality among classes. In conventional GMM-based modeling technique, since the model parameters of a class are estimated without considering other classes in the system, features that are common across various classes may also be captured, along with unique features. This paper proposes to use unique characteristics of a class at the feature-level and at the phoneme-level, separately, to improve the classification accuracy. At the feature-level, the performance of a classifier has been analyzed by capturing the unique features while modeling, and removing common feature vectors during classification. Experiments were conducted on speaker identification task, using speech data of 40 female speakers from NTIMIT corpus, and on a language identification task, using speech data of two languages (English and French) from OGI_MLTS corpus. At the phoneme-level, performance of a classifier has been analyzed by identifying a subset of phonemes, which are unique to a speaker with respect to his/her closely resembling speaker, in the acoustic sense, on a speaker identification task. In both the cases (feature-level and phoneme-level) considerable improvement in classification accuracy is observed over conventional GMM-based classifiers in the above mentioned tasks. Among the three experimental setup, speaker identification task using unique phonemes shows as high as 9.56 % performance improvement over conventional GMM-based classifier.     

The performance of KIMS in image recognition tasks

KIMS is an acronym for a nowledge-Based mage anagement ystem developed in the Robotics and Artificial Intelligence Laboratory (RAIL) at North Carolina A&T State University. KIMS model architecture consists of rules which are developed through statistical experimentation with thresholding and quality control chart algorithms. The control architecture of KIMS is driven by the pattern of these rules. KIMS can analyze features of an X-ray image of a manufactured product such as printed circuit board in a real-time mode and make decisions on whether there are defect symptoms in the product. In this paper we present the current performance of KIMS in product inspection decision.     

The design of a planar robotic manipulator for optimum performance of prescribed tasks

The application of a general optimization methodology, previously proposed by the authors, is extended here to the design of a three link revolute-joint planar manipulator performing more practical and complicated prescribed tasks. In particular a tool moving task and a spray painting task are considered. Both the minimization of average torque and energy usage required for execution of the tasks are addressed and the optimization is carried out with the link lengths and base coordinates taken as the five design variables. In addition to simple physical bounds placed on the variables, the maximum deliverable torques of the driving motors represent further constraints on the system. Joint angle constraints, not previously considered but of great practical importance, are also imposed in this study. This results in significantly more challenging optimization problems than those previously tackled. The complications arising from lock-up and nonassembly are handled by specially devised procedures. The optimization is carried out via penalty function formulations of the constrained problems to which the Snyman unconstrained trajectory optimization algorithm is applied in a special way. For both tasks and for both objective functions, with the full complement of constraints imposed, feasible designs with low objective function values are obtained by using, in each case, four different infeasible designs as starting points for the algorithm.     

The efficacy of using a taxonomic approach in predicting performance of Navy tasks

The objective of this research was to validate the use of Fleishman's human abilities taxonomy approach in predicting human performance in loading the U.S. Navy's Close‐In Weapon System (CIWS). An experiment was conducted that compared actual and predicted performance of CIWS loading operations. Twenty‐one male students and staff from the Close‐In Weapon System School, Fleet Training Center, were participants in the experiment. They ranged in age from 20 to 35 years (mean = 27.2, standard deviation = 4.88) with an average of 81.1 months (standard deviation = 51.83) of military experience and 23.95 months of CIWS loading experience. Each participant was tested in the human abilities identified for CIWS loading operations, and the test results were correlated with recorded times of the loading tasks. Correlations were performed using bivariate regression analysis. The results of the experiment showed high linear relations between human abilities scores and CIWS loading times. The results show that the participants who scored highest on the human abilities tests completed the loading tasks in the shortest time. © 2010 Wiley Periodicals, Inc.     

Efficient calculation of the space-time performance of the working set

The performance of working set memory management is often used as a benchmark against which other algorithms are compared. Efficient algorithms which can generate exact performance curves for the mean working set size and fault rate have appeared in the literature. These algorithms exploit the fact that the curve of fault rate or mean resident set size versus the working set parameter value may be computed by gathering various statistics during one pass across the reference string and then by substituting those statistics into relatively simple formulae.

This paper develops an efficient technique for the determination of the exact space time product versus parameter value curve. The usual simulation approach to compute a space-time performance curve exhibits worst case time complexity of O(nn²) where p is the number of pages referenced by the program and n is the total number of memory references generated by the program. The one-pass technique presented here requires O(pn) time. Given that n may be quite large in practice (10⁷ references is not unusual), this reduction in complexity is necessary for the practical determination of space-time performance of the working set.     

Optimizing the stretch of independent tasks on a cluster: From sequential tasks to moldable tasks

This paper addresses the problem of scheduling non-preemptive moldable tasks to minimize the stretch of the tasks in an online non-clairvoyant setting. To the best of the authors’ knowledge, this problem has never been studied before. To tackle this problem, first the sequential subproblem is studied through the lens of the approximation theory. An algorithm, called DASEDF, is proposed and, through simulations, it is shown to outperform the first-come, first-served scheme. Furthermore, it is observed that machine availability is the key to getting good stretch values. Then, the moldable task scheduling problem is considered, and, by leveraging the results from the sequential case, another algorithm, DBOS, is proposed to optimize the stretch while scheduling moldable tasks. This work is motivated by a task scheduling problem in the context of parallel short sequence mapping which has important applications in biology and genetics. The proposed DBOS algorithm is evaluated both on synthetic data sets that represent short sequence mapping requests and on data sets generated using log files of real production clusters. The results show that the DBOS algorithm significantly outperforms the two state-of-the-art task scheduling algorithms on stretch optimization.     

Measures for evaluating the decision performance of a decision table in rough set theory

As two classical measures, approximation accuracy and consistency degree can be employed to evaluate the decision performance of a decision table. However, these two measures cannot give elaborate depictions of the certainty and consistency of a decision table when their values are equal to zero. To overcome this shortcoming, we first classify decision tables in rough set theory into three types according to their consistency and introduce three new measures for evaluating the decision performance of a decision-rule set extracted from a decision table. We then analyze how each of these three measures depends on the condition granulation and decision granulation of each of the three types of decision tables. Experimental analyses on three practical data sets show that the three new measures appear to be well suited for evaluating the decision performance of a decision-rule set and are much better than the two classical measures.     

The effects of grouping and cueing in the performance of location tasks

In location tasks such as assembly of a control panel, operators respond to stimulus information by locating a given target in an extensive set of response alternatives. Arrangement of the response alternatives and the method of presenting the location information (cueing), as well as the interaction between these factors, were hypothesised to influence performance in this type of task. To test these hypotheses, a factorial experiment involving 60 subjects was performed in which five levels of grouping and four levels of cueing were investigated. Grouping appeared to affect location accuracy more than location time, whereas the effects of cueing were found to be significant for both location accuracy and time. The absence of an interaction effect between grouping and cueing suggested that the effects of these factors on performance were independent.