Model Of Parallel Structure Research Articles

Nonlinear approximation of characteristics of a fork–join queueing system with Pareto service as a model of parallel structure of data processing

A fork–join system with Pareto service time distribution is considered as a model of a parallel structure of data processing. For an approximation of the mean response time of the system and its standard deviation, we apply the approach based on a combination of simulation with linear regression and the method of nonlinear Nelder–Mead optimization. Previously, no analysis of fork–join queueing systems with M|G|1 subsystems was carried out due to the complexity of its implementation. Nevertheless, the approach proposed here is capable of delivering approximations of various types of the correlation coefficients of the sojourn times of subtasks in a given system. The analytic expressions derived below are shown to deliver good approximations to these characteristics, as evidenced by numerical experiments. Application of the proposed approach can be extended to systems with more involved architecture, and, in particular, to systems with non-Poisson input flow and various options of distributing of service times of tasks.

Ensemble empirical mode decomposition-based preprocessing method with Multi-LSTM for time series forecasting: a case study for hog prices

Drastic hog price fluctuations have a great impact on the welfare of hog farmers, people's living standards, and the macroeconomy. To stabilise the hog price, hog price forecasting has become an increasingly hot issue in the research literature. Existing papers have neglected the benefits of decomposition and instead directly utilise models to predict hog prices by capturing raw data. Motivated by this issue, the authors introduce a new robust forecasting approach for hog prices that combines ensemble empirical mode decomposition (EEMD) and multilong short-term memory neural networks (Multi-LSTMs). First, EEMD decomposes the volatile raw sequence into several smoother subsequences. Second, the decomposed subsequences are predicted separately using a parallel structure model consisting of several LSTMs. Finally, the fuse function combines all the subresults to yield the final result. The empirical results suggest that the proposed method only has minor errors and proves the effectiveness and reliability in experiments on real datasets (2.55207, 4.816, and 0.332 on MAE, MAPE and RMSLE, respectively). Reliable forecasting of hog prices is beneficial to farmers and people to allow optimisation of their production and booking rates and to moderate the adverse effects of potential shocks.

Parallel-Structure Deep Learning for Prediction of Remaining Time of Process Instances

Event logs generated by Process-Aware Information Systems (PAIS) provide many opportunities for analysis that are expected to help organizations optimize their business processes. The ability to monitor business processes proactively can allow an organization to achieve, maintain or enhance competitiveness in the market. Predictive Business Process Monitoring (PBPM) can provide measures such as the prediction of the remaining time of an ongoing process instance (case) by taking past activities in running process instances into account, as based on the event logs of previously completed process instances. With the prediction provided, we expect that organizations can respond quickly to deviations from the desired process. In the context of the growing popularity of deep learning and the need to utilize heterogeneous representation of data; in this study, we derived a new deep-learning approach that utilizes two types of data representation based on a parallel-structure model, which consists of a convolutional neural network (CNN) and a multi-layer perceptron (MLP) with an embedding layer, to predict the remaining time. Conducting experiments with real-world datasets, we compared our proposed method against the existing deep-learning approach to confirm its utility for the provision of more precise prediction (as indicated by error metrics) relative to the baseline method.

Mixing Rules of Young’s Modulus, Thermal Expansion Coefficient and Thermal Conductivity of Solid Material with Particulate Inclusion

This review paper analyzed a Young’s modulus (E), a thermal expansion coefficient (TEC, β) and a thermal conductivity (κ) of the material with simple cubic particulate inclusion using two model structures: a parallel structure and a series structure of laminated layers. The derived β equations were applied to calculate the β value of the W-MgO system. The accuracy was higher for the series model structure than for the parallel model structure. Young’s moduli (Ec) of sintered porous alumina compacts were theoretically related to the development of neck growth of grain boundary between sintered two particles and expressed as a function of porosity. The series structure model with cubic pores explained well the increased tendency of Ec with neck growth rather than the parallel structure model. The thermal conductivity of the three phase system of alumina-mullite-pore was calculated by a theoretical equation developed in this research group, and compared with the experimental results. The pores in the sintered composite were treated as one phase. The measured thermal conductivity of the composite with 0.5-25% porosity (open and closed pores) was in accordance with the theoretical prediction based on the parallel structure model. Key words: Sintering, Porous ceramics, Thermal conductivity, Thermal expansion

Discrete Planning Unit Look-Ahead Velocity Control Strategy and Parallelization Research Based on GPU

During the machining of complicated curves and surfaces, the massive microsegments are generated. The microsegments are inputted into numerical control system (CNC) to process the velocity planning and high-velocity interpolation. This whole procedure is the core algorithm of CNC. During the procedure of velocity planning, the look-ahead iterative calculation is used for several specified future microsegments and the calculation quantity is very large. Additionally, the CNC is a typical hard real-time system. Every interpolation step must be finished in specified interpolation period, so the high calculating ability of CPU is needed. The concept of GPU is quite different from the traditional CPU. It has a very strong parallel computing ability. In this project, the GPU technology is introduced into CNC algorithm. The parallel structure model is constructed by CUDA, and a discrete planning unit look-ahead velocity control strategy is proposed. The GPU parallel computing mechanism of massive microsegments is discussed and, based on the research, a high-performance velocity planning algorithm is presented. This research result can highly raise the velocity planning and interpolation efficiency. This brand new solution can help improve the performance of CNC system dramatically.

A Classification and Regression Trees (CART) Model of Parallel Structure and Long-term Prediction Prognosis of Machine Condition

This article presents a combined prediction model involving the parallel of classification and regression trees (CART) model, namely p-CART, and a long-term direct prediction methodology of time series techniques to predict the future stages of the machine’s operating conditions. p-CART model consists of multiple CART models which are connected in parallel. Each sub-model in the p-CART is trained independently. Based on the observations, these sub-models are subsequently used to predict the future values of the machine’s operating conditions separately with the same embedding dimension but different observations’ indices. Finally, the predicted results of sub-models are combined to produce the final results of the predicting process. Real trending data acquired from condition monitoring routine of compressor are employed to evaluate the proposed method. A comparative study of the predicted results obtained from traditional CART and p-CART models is also carried out to appraise the prediction capability of the proposed model.

Evaluation of Surface-Related Phenomena Using Sedimentation Tests

Abstract Sedimentation tests are suggested to assess the sensitivity of clay-water systems to environmental changes. A simple framework based on a parallel structure assumption was developed to analyze data from sedimentation tests. The slope of the relationship between the final sediment volume and the ionic concentration, plotted on a log-log scale, is related to the specific surface Sa, the double layer thickness ϑ, and to the specific gravity Gs. A new parameter, λ, the reactivity coefficient, where λ = Sa · ϑ · Gs · ρw, is introduced to help identification of reactive clay-water systems. High values of λ identify systems that are sensitive to changes in pore fluid characteristics. An experimental procedure is suggested to acquire relevant sedimentation parameters. Data collected during sedimentation tests of kaolinite, bentonite, and disaggregated Pierre shale, Mancos shale, and Queenston shale are used to verify the model. Agreement between experimental data and the model suggests that this simple technique can be used to identify reactive shales in the field. Other applications arise in tailing pond modeling, drilling mud management, and clay liner reliability assessment.