Performance Prediction Methodology Research Articles

A Performance Prediction Method for Talent Team Building Based on Integrated ISA-BP Neural Networks

INTRODCTION: Objective, accurate and fair development of research and effective performance prediction methodology for the construction of the talent team is the current needs of the new era of innovation and reform and development of university management, as well as the need to improve the quality of scientific research and teaching level of the talent team.OBJCTIVES: To address the problems of irrational principle of indicator selection, incomplete system and imprecise methodology in the current research on performance prediction of talent team building.METHODS:This paper proposes a talent team construction performance prediction method based on intelligent optimization algorithm improving neural network with integrated learning as the framework. First of all, through the analysis of the current talent team construction performance prediction influencing factors selection principles, analyze the talent team construction performance management process, select the talent team construction performance prediction influencing factors, and construct the talent team construction performance analysis system; then, with the integrated learning as a framework, improve the neural network through the internal search optimization algorithm to construct the talent team construction performance prediction model; finally, through the simulation experiments to analyze and verify the effectiveness and superiority of the proposed method. The effective type and superiority of the proposed method are verified.RESULTS: The results show that the proposed method satisfies the real-time requirements while improving the prediction accuracy.CONCLUSION: This paper addresses the lack of precision in forecasting the performance of the talent pipeline and the lack of a sound analytical system.

A methodology for performance prediction: Hydrodynamic investigation of spiral grooved thrust bearing

AbstractThis paper presents the design and numerical optimization of oil‐lubricated spiral grooved thrust bearing (SGTB) for its application at high speed and axial loading conditions. A numerical model is developed using nonlinear incompressible Reynold's equation and is solved using the finite volume method (FVM) to determine the static characteristics over the bearing surface. Further, the influence of groove parameters such as spiral angle, groove angle, film thickness ratio, number of grooves and speed on the static characteristics of the bearing has been investigated. The result shows that the designed oil‐lubricated SGTB can operate at high‐speed conditions and withstand high axial load. Further, the characteristic data sets acquired from the numerical analysis are trained using an artificial neural network (ANN), and their performance is evaluated through the computation of the regression coefficient. Then adaptive neuro‐fuzzy interface system (ANFIS) surface plot is obtained to determine the optimum bearing parameters.

Using Machine Learning for Loss Prediction in a Hybrid Meanline Modeling Method to Deliver Improved Radial Turbine Performance Prediction

Abstract Low fidelity modeling approaches remain attractive due to an unrivaled ability to predict full turbine performance maps quickly compared to high-fidelity approaches such as computational fluid dynamics (CFD), especially in the preliminary design process. As improvements in performance on a component level approach a point of diminishing returns, the ability to efficiently optimize the complete charging system for a given duty is a topic attracting significant research interest. In the case of turbocharging applications, existing engine and powertrain simulations require turbine maps to calculate the turbine performance, which are usually obtained from experimental testing. Unfortunately, the need for extrapolation is unavoidable because of the limited range of testing data available, leading to inaccuracies especially at off-design conditions. To enable intensive modeling and optimization of complete vehicle powertrains for different drive cycles, the current piece of work seeks to combine the advantages of machine learning techniques and physical meanline modeling to facilitate faster, more accurate predictions of complete turbocharger maps. This paper presents a novel methodology for turbocharger turbine rotor and nozzle performance prediction based on hybrid modeling. The turbine rotor and nozzle were parameterized to conduct CFD simulations for a wide variety of turbine geometries, which were used to form a database to train an artificial neural network (ANN). The predicted losses provided by the ANN were then utilized in the meanline code, substituting for the conventional empirical loss models. As well as removing the need for empirical loss models, modifications were undertaken to the meanline approach to further enhance modeling accuracy. First, in order to accurately characterize the stage mass flow capacity, the losses occurring in the nozzle and rotor were subdivided into those occurring before and after the throat. A second novel aspect is that the aerodynamic blockage level at the rotor throat was implemented as a variable rather than a constant value. By training the ANN to predict the variation of blockage with geometry and operating condition, a more accurate depiction of the changing secondary flow fields could be achieved. The capabilities of the hybrid meanline modeling method were evaluated on several unseen test cases. The resulting predictions of efficiency and mass flowrate demonstrated strong correlation with CFD results and experimental test results. The hybrid meanline modeling method therefore displays great potential in wide range radial turbine performance prediction with enhanced accuracy in comparison to traditional approaches.

An investigation on tool wear rate in the electrical discharge face grinding process for the machining of Monel 400

This paper attempts to model the electrical discharge face grinding (EDFG) process, while machining the Monel 400 superalloy, using the response surface methodology (RSM) and artificial neural network (ANN) for process performance prediction. Initially, the experiments were designed using the central composite design of RSM. The grinding wheel speed, peak current, pulse-on-time, and pulse-off-time were chosen as input process parameters, while tool wear rate (TWR) as the performance attributes. Analysis of variance was to identify the significant model terms and their influence on TWR. The grinding wheel speed was identified as the most influencing parameter, and a 61% reduction in TWR was recorded when increasing the speed from 200 rpm to 600 rpm. The developed ANN model (4-20-1) reduced the mean square prediction error up to four times, outperforming the RSM model.

Turbofan engine performance prediction methodology integrated high-fidelity secondary air system models

This research studies the performance of an ultra-high bypass ratio turbofan engine, and specifically its secondary air system (SAS). A co-simulation methodology is explored whereby a high-fidelity SAS model and an engine performance code featuring flexible modules can be coupled and interactively executed. The percentage of bleed flows and boundary conditions for the SAS are updated at each iteration step. For this purpose, a SAS model including different elements is developed. Furthermore, a commercial computational fluid dynamics (CFD) solver is adopted to capture the complex flow field in the pre-swirl system. The credibility of cycle calculation and SAS elements is validated by comparing with publicly available data. Subsequently, an elaborately designed SAS is modeled and co-simulated with the AGTF30 engine using a flow network simulation method. The coupling effect between the engine performance and the SAS is studied for eight different flight conditions. The correlation and prediction of engine performance due to seal clearance change is presented. The co-simulation approach clarifies the mutual interactions between the engine overall parameters and the SAS. The results reveal that the enhanced flow network model can improve the simulation accuracy of engine performance over a wide range of operating conditions.

OA-PU Algorithm-to Enhance WSN Life Time with Cluster Head Selection

Clustering is the most popular strategy for increasing the lifetime of a wireless sensor network, which immediately leads to a stronger routing process. This process requires the processing of sensor nodes into clusters and assigning relevant cluster heads to each cluster. This paper aims to implement a new hybrid algorithm called Over taker Assisted Political Update (OA-PU) for selecting an efficient cluster head. This cluster head is selected based upon four factors, namely energy, distance, cluster radius, and time. The hybrid algorithm is a combination of the Fusion Rider Optimization Algorithm (F-ROA) as well as a nature-inspired Political Optimization Algorithm (PO). Fusion-ROA depends on groups of riders that are attempting to reach the goal and PO is a state of the art meta-heuristic optimization methodology for global and structural analysis performance prediction in the WSN environment. In terms of alive node analysis, cost function analysis, and energy analysis, the proposed OA-PU outperforms conventional approaches. Consequently, the lifetime of WSN is increased and an optimal system is developed.

A methodology for performance prediction: aerodynamic analysis of axially loaded gas foil bearing

A methodology for performance prediction: aerodynamic analysis of axially loaded gas foil bearing

A New Methodology of Production Performance Prediction for Strong Edge-Water Reservoir

Abstract Material balance is a basic principle in reservoir engineering, which is still used as a quick and easy analytical tool for reservoir evaluation. In this article, a new methodology of production performance prediction for water-flooding reservoir was proposed based on the material balance principle, which considers the water saturation change caused by water injection and natural water influx, and its effect on transient gas–oil ratio. Among them, the cumulative water production was calculated based on Tong’s water-driver performance curve; the cumulative water influx was obtained by the Fetkovitch method; the transient gas–oil ratio can be acquired by Darcy’s law and Baker’s relative permeability model. Comparisons have been made between the new methodology and commercial reservoir simulator for two different reservoirs. The results show that there is good similarity between these two tools, which verifies the correctness of the new methodology.

A reduced-order model of twin-entry nozzleless radial turbine based on flow characteristics

Twin-entry radial turbine has evident advantages on energy recovery of internal combustion engine because of better utilization of exhaust pulse energy. One-dimensional performance prediction plays an important role in automobile industry because of the cost-effective feature. However, as twin-entry turbine is confronted by out-of-phase pulses, to capture the flow distortion by twin-entry turbine model is challenging, especially under partial admission conditions. This paper establishes a reduced-order parallel-rotor model for performance prediction of twin-entry nozzleless radial turbine based on internal flow characteristics. Firstly, twin-entry turbine experiment and 3D simulation are applied to analyze the different flow features under partial admission conditions. The flow distortions in the spanwise and circumferential directions are both demonstrated. Based on these flow characteristics, a reduced-order twin-entry turbine model with parallel rotor passages is established for predicting the rotor at distorted flow field. Finally, the model is carefully validated against the results of CFD. The results show that satisfied consistency can be obtained and the maximum discrepancy of turbine performance is 3.8% and the performance difference and flow distortion under partial admissions can be captured, thus prove the credibility of the model. This investigation provides a reliable methodology for performance prediction and behavior analysis of twin-entry turbine.

Shrinkage cracking model for cementitiously stabilized layers for use in the mechanistic-empirical pavement design guide

Abstract Shrinkage cracking is one of the primary distresses found in cementitiously stabilized layers (CSL) of pavements. Shrinkage can cause cracking on the surface layer due to the bond between the surface layer and the CSL. Shrnkage of CSL includes autogenous shrinkage due to hydration, drying shrinkage due to loss of moisture, and thermal shrinkage due to low temperature contraction. The current methodology for analysis and performance prediction of CSL pavements in the Mechanistic-Empirical Pavement Design Guide (MEPDG) does not account for the prediction of shrinkage cracking. Given the importance of shrinkage cracking, incorporation of a shrinkage cracking model for CSL into the MEPDG will improve its analysis and design capability. This study uses dimensional analysis to develop a shrinkage cracking model to predict crack spacing and widths for CSL for stabilized fine and coarse-grained soils. The specific functional form of the shrinkage cracking model was developed based on field data as well as pilot-scale laboratory experimental results. It is found that the developed model for shrinkage cracking is effective in predicting shrinkage crack spacing and widths. Further calibration of the shrinkage cracking model is needed using additional field data.

Towards a Realistic Estimation of the Powering Performance of a Ship with a Gate Rudder System

This paper presents an investigation on the scale effects associated with the powering performance of a Gate Rudder System (GRS) which was recently introduced as a novel energy-saving propulsion and maneuvring device. This new system was applied for the first time on a 2400 GT domestic container ship, and full-scale sea trials were conducted successfully in Japan, in 2017. The trials confirmed the superior powering and maneuvring performance of this novel system. However, a significant discrepancy was also noticed between the model test-based performance predictions and the full-scale measurements. The discrepancy was in the power-speed data and also in the maneuvring test data when these data were compared with the data of her sister container ship which was equipped with a conventional flap rudder. Twelve months after the delivery of the vessel with the gate rudder system, the voyage data revealed a surprisingly more significant difference in the powering performance based on the voyage data. The aim of this paper, therefore, is to take a further step towards an improved estimation of the powering performance of ships with a GRS with a specific emphasis on the scale effect issues associated with a GRS. More specifically, this study investigated the scale effects on the powering performance of a gate rudder system based on the analyses of the data from two tank tests and full-scale trials with the above-mentioned sister ships. The study focused on the corrections for the scale effects, which were believed to be associated with the drag and lift characteristics of the gate rudder blades due to the low Reynolds number experienced in model tests combined with the unique arrangement of this rudder and propulsion system. Based on the appropriate semi-empirical approaches that support model test and full-scale data, this study verified the scale effect phenomenon and presented the associated correction procedure. Also, this study presented an enhanced methodology for the powering performance prediction of a ship driven by a GRS implementing the proposed scale effect correction. The predicted powering performance of the subject container vessel with the GRS presented an excellent agreement with the full-scale trials data justifying the claimed scale effect and associated correction procedure, as well as the proposed enhanced methodology for the practical way of predicting the powering performance of a ship with the GRS.

Hydraulic Performance Prediction Methodology in Regenerative Pumps Through CFD Analysis

Peripheral pump efficiency is a factor of performance defined by geometrical parameters and characteristic curves of the system. Due to the need to determinate an appropriate pump operating point in order to obtain an adequate hydraulic system performance in industrial and home applications, this paper presents a methodology that simulates the flow behavior of vortex or regenerative pumps using geometrical factors that directly define its optimal performance. An approximated real behavior of the flow phenomena is determined through the application of numerical methods and turbulence models based on Computational Fluids Dynamics with the aim to link and validate the results obtained in the predictive approach. The technological advance of the computational methods used in the analysis of flow phenomena enables an approximated description of the flow studied by means of computational tools. Therefore, numerical analysis is developed by using OpenFOAM software. This computational tool is an open-source code computed in the C++ library characterized by the development of complex analysis related to the flow phenomena applied in engineering applications based on the performance of turbomachinery used in energy transformation systems. The methodology used in this paper improves the prediction of the flow in a peripheral pump using statistical models based on the multiple linear regression model determined with the Taguchi Orthogonal Array method. A deviation of less than 5 percent between the predictive approach and the numerical model compared has been reached. Geometrical optimization of the peripheral centrifugal pump, taking into account the angular velocity of the impeller, inlet, and outlet angles of the blade, and height impeller blades have been developed.

Performance prediction of pump and pumping system based on combination of numerical simulation and non-full passage model test

The design scheme of open intake will often increase the shaft length and the height of wet-pit pump house when the range of intake water level varies relatively large. And if the pump house is far away from the discharge bay, a longer discharge pipe may be inevitable, and the arrangement of full passage model pumping system may exceed the allowed size of a specific test bench. Aiming at a long-shaft vertical pumping system with open intake design, larger water level variation and longer discharge pipe, a methodology of performance prediction for the pump and pumping system was proposed by combining numerical simulation and non-full passage model test because of less experiment funds, shorter research periods and restricted test bench. Through numerical simulation, the hydraulic losses and drag coefficients of sump and discharge pipe were calculated and followed by the experiment design and model test of non-full passage pumping system. Based on the calibration accuracy of measurement instruments and test results, the experimental uncertainties were analyzed. Through combination of numerical simulation and non-full passage model test, the prediction of model pump and pumping system performance was realized. The performance of prototype pump and pumping system was obtained by applying the similarity law of pump. When operated under the designed head of 13.59 m, the flow rate and efficiency of the prototype pumping system reached 11.86 m3/s and 86.56%, respectively, exceeding the requirements stipulated in the bidding documents and showing that it possessed greater pumping capacity and higher efficiency. The methodology proposed in this paper can be referenced and applied to the engineering design of similar pump stations.

Experimental and numerical assessment of a methodology for performance prediction of Pumps-as-Turbines (PaTs) operating in off-design conditions

In this work, an extensive assessment of a predicting model used to evaluate Pumps-as-Turbines’ (PaTs) characteristic curves is presented, with specific attention to the off-design operating conditions. The novelty of the proposed model consists in the possibility to reconstruct the performance curves of a PaT only by knowing a limited number of operating data in turbine mode at the Best Efficiency Point (BEP), which, in many applications, represents a design constraint. The availability of the off-design performance curves supplies important indications for technical and economic evaluations in those applications where a constant flow rate cannot be granted. The predicting model was derived by re-elaborating a wide experimental data-set based on the most relevant scientific literature related to several PaTs operating in turbine mode. The prediction’s capability of the model was validated with experimental tests and confirmed by numerical simulations. The experimental tests were carried on in both direct and reverse modes by inspecting several flow rates. The model data were compared with the experimental ones in order to validate the Computational Fluid Dynamics (CFD) analyses. Subsequently, the numerical model was used to investigate the performances of other two PaTs operating in turbine mode. The study of the performance obtained with the CFD analyses allowed to evaluate the effectiveness of the predicting model, highlighting its pros together with its possible improvements. In general, it is possible to conclude that the proposed model is able to predict the performances of the studied PaTs with errors included in the range of ±7% with respect to the Best Efficiency Point (BEP) in turbine mode.

Transformation of User Interface to Activity Models and Assessing Performance of WA/WS

Quality of Service is an important attribute of a software system. In retrospect, performance assessment based on user interaction with the system has given a better understanding of underlying disciplines of the product. In this paper, we capture user interaction with the prototype/User Interface (UI). An approach for developing activity model from the user interface model is presented using workflows and functional elements. A methodology is proposed to transform UI into activity diagram. The approach is validated by an experimental setup using Amazon service. The performance of Amazon service is assessed using activity based performance prediction methodology, and the simulation results are obtained using SMTQA.

Performance prediction methodology and analysis of a variable pitch fan turbofan engine

The objective of this paper is development and application of a methodology for preliminary analysis of variable pitch fan (VPF), both as a separate component and as a module integrated into a short-medium range geared turbofan engine developed within European FP7 project ENOVAL. For this purpose, a high bypass ratio two spool geared turbofan engine model was constructed in software PROOSIS. A VPF performance modeling methodology was then developed using 3D steady RANS CFD produced fan maps as baseline; the CFD maps characterized five discrete fan pitch angle settings. In order to represent those maps in PROOSIS and add the pitch angle as a degree of freedom, they were transformed into the Map Fitting Tool (MFT) reference frame. Once the complete VPF turbofan model was in place, engine mission optimization experiments were carried out. The resulting performance is characterized by a good capability to control the fan surge margin, without degrading the engine fuel consumption. This paper represents a new contribution on the topic firstly by coupling a 0D engine performance code with a 3D RANS calculation, and then by introducing the concept of MFT maps with an additional degree of freedom as the interface between the two.

Development of a Program Definition Rating Index for the Performance Prediction of Construction Programs

Program management is the structured and strategic process of managing multiple projects at a high level to maximize benefits. The essentials of programs include high costs and long implementation periods, and thus, the negative impacts caused by the failure of program management are more significant and greater than that of a project. Therefore, to achieve high program performance, it is essential for program management to be well defined during the early stages. However, the existing research is mainly focused on the performance prediction methodologies for projects, while the research pertaining to programs has concentrated on identifying the qualitative critical success factors (CSFs). Thus, this study developed a methodology for predicting the program performance. Forty-five CSFs were identified herein from literature review and expert interviews, then grouped through factor analysis. In addition, the Program Definition Rating Index (PgDRI) was developed by calculating the weights of the proposed CSFs through structured equation modeling in order to evaluate the quantitative program performance. For validation, the PgDRI was applied to three in-progress cases, and the PgDRI scores were compared with the actual performance of each case. The PgDRI developed in this study can contribute to the body of knowledge pertaining to program management by quantifying the performance management of a program. In addition, the PgDRI can be utilized in the performance management of a program in terms of the cost and schedule by allowing practitioners to apply the PgDRI repeatedly to the major decision-making processes during the early stages of a program.

A review of applications of data mining techniques for prediction of students’ performance in higher education

This paper brings in an insight of the number of ways the students’ performance can be predicted. The performance of a student plays a vital role in any institution. It is the measure for assessing the academic excellence of any esteemed educational body. Not only the institute but the academic performance matters to the student as well. There is a tough competition in the field of academics. Students spend their sleepless nights turning up the pages of their books. Well, amidst of this, a student tussle with many hardships. At present, there are various methodologies to predict the student's performance and Data mining provides the most scientific and dependable method for predicting the same. Educational data mining is the process behind the evaluation of useful patterns out of huge education databases. These patterns help the educators in predicting the performance of the student. In a way it also helps in advancing effective teaching. Data mining techniques can be useful in culling out the data concerted to specific needs. This paper presents the study to find out the gap in the existing performance prediction methodologies.

Incorporation of GPR and FWD into pavement Mechanistic-Empirical design

A Mechanistic-Empirical (M-E) design based pavement quality or performance prediction methodology incorporating nondestructive test such as Ground Penetrating Radar (GPR) and Falling Weight Deflectometer (FWD) tests is demonstrated in this study. To facilitate, three pavement sites, namely, US285, US54, and I-40 in New Mexico are selected for nondestructive tests such as GPR and FWD. It is observed that the GPR predicted AC layer thickness is more consistent than base layer thickness. The FWD test was also conducted on the same locations, and like GPR, backcalculated layer moduli also shows varying level of inconsistency in different pavement sections. Based on the GPR and FWD tests, each of these three sites is divided into two major areas: one with the mean and the other with the minimum layer thicknesses and modulus. Layer thicknesses as well as moduli are incorporated to the AASHTOWare-ME software to determine load related distresses such as bottom-up crack and rut. Based on the amount of bottom-up cracks, the US54 pavement section is predicted to fail during 15th year of its service life, and the time of failure can be very early in the area with minimum layer thickness and modulus. Based on the rut prediction, pavement sections in US54 and I-40 is predicted to fail early. Finally, this methodology is recommended to implement to perform pavement quality assessment incorporating GPR and FWD tests in both project and network levels.

Systematic scalability assessment for feature oriented multi-tenant services

Recent software engineering paradigms such as software product lines, supporting development techniques like feature modeling, and cloud provisioning models such as platform and infrastructure as a service, allow for great flexibility during both software design and deployment, resulting in potentially large cost savings. However, all this flexibility comes with a catch: as the combinatorial complexity of optional design features and deployment variability increases, the difficulty of assessing system qualities such as scalability and quality of service increases too. And if the software itself is not scalable (for instance, because of a specific set of selected features), deploying additional service instances is a futile endeavor. Clearly there is a need to systematically measure the impact of feature selection on scalability, as the potential cost savings can be completely mitigated by the risk of having a system that is unable to meet service demand.In this work, we document our results on systematic load testing for automated quality of service and scalability analysis. The major contribution of our work is tool support and a methodology to analyze the scalability of these distributed, feature oriented multi-tenant software systems in a continuous integration process. We discuss our approach to select features for load testing such that a representative set of feature combinations is used to elicit valuable information on the performance impact and feature interactions. Additionally, we highlight how our methodology and framework for performance and scalability prediction differs from state-of-practice solutions. We take the viewpoint of both the tenant of the service and the service provider, and report on our experiences applying the approach to an industrial use case in the domain of electronic payments. We conclude that the integration of systematic scalability tests in a continuous integration process offers strong advantages to software developers and service providers, such as the ability to quantify the impact of new features in existing service compositions, and the early detection of hidden feature interactions that may negatively affect the overall performance of multi-tenant services.