In this paper, we propose the application of a new fault detection approach with a sequential updating function under new operating conditions or natural evolving degradation processes. The proposed approach is based ...
详细信息
ISBN:
(纸本)9781509003839
In this paper, we propose the application of a new fault detection approach with a sequential updating function under new operating conditions or natural evolving degradation processes. The proposed approach is based on Online Sequential Extreme Learning Machines (OS-ELM). OS-ELM have the advantages of a strong learning ability, very fast training, and online learning. The concept of applying OS-ELM to fault detection is demonstrated on an artificial case study. We expect that OS-ELM can contribute to improve the fault detection and also the associated initiation of maintenance activities for engineering components working in an evolving environment such as electric components, bearings, gears, alternators, shafts and pumps, in which the monitored signals are not only significantly affected by working load and surrounding environment but may also experience some modifications due to a slow aging and degradation process.
We propose a new extension for index based scheduling rules which adds a global system perspective by estimating how much a job could benefit from an immediate processing. We implement this criterion in the ATC rule a...
详细信息
ISBN:
(纸本)9781424448692
We propose a new extension for index based scheduling rules which adds a global system perspective by estimating how much a job could benefit from an immediate processing. We implement this criterion in the ATC rule and compare the performances of the standard ATC and the extended rule ATCG. It is shown that ATCG performs better than ATC, especially when system loads are high and processing times follow a uniform distribution.
The output of 25 models used in the Coupled Model Intercomparison Project phase 3 (CMIP3) were evaluated,with a focus on summer precipitation in eastern China for the last 40 years of the 20th *** mod-els failed to re...
详细信息
The output of 25 models used in the Coupled Model Intercomparison Project phase 3 (CMIP3) were evaluated,with a focus on summer precipitation in eastern China for the last 40 years of the 20th *** mod-els failed to reproduce rainfall associated with the East Asian summer monsoon (EASM),and hence the seasonal cycle in eastern China,but provided reasonable results in Southwest (SW) and Northeast China (NE).The simula-tions produced reasonable results for the Yangtze-Huai (YH) Basin area,although the Meiyu phenomenon was underestimated in *** typical regional phe-nomenon,a seasonal northward shift in the rain belt from early to late summer,was completely missed by most *** long-term climate trends in rainfall over eastern China were largely underestimated,and the ob-served geographical pattern of rainfall changes was not reproduced by most *** extremes were evaluated via parameters of fitted GEV (Generalized Ex-treme Values) *** annual extremes were grossly underestimated in the monsoon-dominated YH and SW regions,but reasonable values were calculated for the North China (NC) and NE *** results suggest a general failure to capture the dynamics of the EASM in current coupled climate ***,models with higher resolution tend to reproduce larger decadal trends and annual extremes of precipitation in the regions studied.
The pricing of knowledge based services should be based on the three following components: the cost structure,the competition and the perceived value by the ***,it is mainly based on the cost structure which does not ...
详细信息
The pricing of knowledge based services should be based on the three following components: the cost structure,the competition and the perceived value by the ***,it is mainly based on the cost structure which does not account for the real value provided to the *** on an integrated optimization model combining an aggregate planning model with a share of choice model,we produce implicit values of *** results will be presented about a travel agency.
We investigate the temporal behaviour of completion time estimates derived from scheduling algorithms that are based on the set of already known and accepted orders (static job shop problem). In a dynamic job shop env...
详细信息
We investigate the temporal behaviour of completion time estimates derived from scheduling algorithms that are based on the set of already known and accepted orders (static job shop problem). In a dynamic job shop env...
详细信息
We investigate the temporal behaviour of completion time estimates derived from scheduling algorithms that are based on the set of already known and accepted orders (static job shop problem). In a dynamic job shop environment, these estimates change over time, typically increasing. This leads to a systematic error in the completion time estimation based on static job shop algorithms. We propose an approach for calculating this systematic error. Furthermore, a method for compensating this error is derived. For some simple examples, the theory is worked out and compared with simulation results. In all cases, the estimates based on the static algorithm can substantially be improved.
Recent research has demonstrated the utility of using supervised classification systems for automatic identification of low quality microarray data. However, this approach requires annotation of a large training set b...
详细信息
Recent research has demonstrated the utility of using supervised classification systems for automatic identification of low quality microarray data. However, this approach requires annotation of a large training set b...
详细信息
Recent research has demonstrated the utility of using supervised classification systems for automatic identification of low quality microarray data. However, this approach requires annotation of a large training set by a qualified expert. In this paper we demonstrate the utility of an unsupervised classification technique based on the Expectation-Maximization (EM) algorithm and naive Bayes classification. On our test set, this system exhibits performance comparable to that of an analogous supervised learner constructed from the same training data.
暂无评论