Accurate Performance Prediction Model Research Articles

Artificial Neural Network Prediction Model of Dust Effect on Photovoltaic Performance for Residential applications: Malaysia Case Study

Dust accumulation on the photovoltaic system adversely degrades its power conversion efficiency (PCE). Focusing on residential installations, dust accumulation on PV modules installed in tropical regions may be vulnerable due to lower inclination angles and rainfall that encourage dust settlement on PV surfaces. However, most related studies in the tropics are concerned with studies in the laboratory, where dust collection is not from the actual field, and an accurate performance prediction model is impossible to obtain. This paper investigates the dust-related degradation in the PV output performance based on the developed Artificial Neural Network (ANN) predictive model. For this purpose, two identical monocrystalline modules of 120 Wp were tested and assessed under real operating conditions in Melaka, Malaysia (2.1896° N, 102.2501° E), of which one module was dust-free (clean). At the same time, the other was left uncleaned (dusty) for one month. The experimental datasets were divided into three sets: the first set was used for training and testing purposes, while the second and third, namely Data 2 and Data 3, were used for validating the proposed ANN model. The accuracy study shows that the predicted data using the ANN model and the experimentally acquired data are in good agreement, with MAE and RMSE for the cleaned PV module are as low as 1.28 °C, and 1.96 °C respectively for Data 2 and 3.93 °C and 4.92 °C respectively for Data 3. Meanwhile, the RMSE and MAE for the dusty PV module are 1.53°C and 2.82 °C respectively for Data 2 and 4.13 °C and 5.26 °C for Data 3. The ANN predictive model was then used for yield forecasting in a residential installation and found that the clean PV system provides a 7.29 % higher yield than a dusty system. The proposed ANN model is beneficial for PV system installers to assess and anticipate the impacts of dust on the PV installation in cities with similar climatic conditions.

Accurate performance prediction model for impact hammer developed using customized evolutionary algorithm

In past decades, impact hammers have played a key role in underground construction. Simply in Istanbul, impact hammers have been used to excavate more than 20 km of metro tunnels. Thus, determining the instantaneous breaking rate (IBR) of an impact hammer is attracting increasing attention. A number of IBR prediction models have been developed for impact hammers. However, there is still a demand for models that require a smaller number of easy-to-obtain rock properties as inputs and provide a reasonable level of accuracy, a wide range of applications, and high reliability. This study had the goal of developing such a prediction model based on an investigation of two subway tunnels built in Istanbul. In order to enhance the results generated by multiple linear regression analysis, a customized tool was developed for the non-linear analysis of a relatively large set of data collected for the present research. Gene expression programming (GEP) and particle swarm optimization (PSO) were merged to create a non-linear analysis tool. The GEP–PSO algorithm was trained using 80% of the available data, with the remaining 20% reserved to validate the results. The output of the algorithm was presented in the form of a mathematical equation that predicted the IBR using the uniaxial compressive strength, rock quality designation, Schmidt hammer rebound value, and machine power as input parameters. The predicted IBR values were in remarkable agreement with the recorded values. In order to verify the efficiency of the proposed prediction model, it was successfully tested against previously developed models for which input parameters were available. In addition, the proposed model was investigated under hypothetical circumstances to ensure that it legitimately described the performance changes due to changes in the input parameters. The model developed in this research is proposed as an accurate and reliable tool for predicting the performance of impact hammers over a wide application range.

Performance Prediction Model of Solid Oxide Fuel Cell System Based on Neural Network Autoregressive with External Input Method

An accurate performance prediction model for the solid oxide fuel cell (SOFC) system not only contributes to the realization of the operating condition but also plays a role in long-term prediction performance. Accordingly, a research study has been developed to suitably deal with the time-series model and accurately build the performance prediction model of SOFC system based on neural network autoregressive with external input (NNARX) method. The architecture regressor parameters of the NNARX model were efficiently determined using the Taguchi orthogonal array (OA) method for optimal sets. The identified and evaluated optimal parameter levels were used to conduct an analysis of variance (ANOVA) to prove correctness. Moreover, a series of statistics criteria and multi-step prediction were also employed for investigating the uncertainty of the predicted model and solve the overfitting and under fitting problems; further. These criteria were also used to determine the performance of the proposed model architecture. The predicted results of the current study indicated that the developed optimal model level parameters consistently had the least statistics errors and reduced workload of the trial-and-error processes.

A Minimum-Cost Modeling Method for Nonlinear Industrial Process Based on Multimodel Migration and Bayesian Model Averaging Method

With increasing drastic market competition, establishing an accurate and reliable performance prediction model for control and optimization at a minimum cost is a growing trend in industrial production. This article proposes a minimum-cost modeling method to develop the performance prediction model of a new nonlinear industrial process. The core idea of this approach is to migrate the useful information on multiple old and similar processes to develop a new process model. A multimodel migration strategy is proposed to migrate the useful information by combining the existing nonlinear process models and take full advantage of minimum data from the new nonlinear process. In order to obtain a set of optimal weights for combining the multiple old and similar process models, the Bayesian model averaging method is employed to estimate the contributions of each available old nonlinear process model to the new nonlinear process model. Moreover, a further experiment used nested Latin hypercube design (NLHD) to gather the necessary minimum data on the new nonlinear process for model migration. Finally, we apply the proposed minimum-cost modeling method to the new multistage centrifugal compressor in the combined cycle power plant, and the results show that the proposed method can develop an accurate compressor model at a minimal cost in terms of the amount of new process data. Note to Practitioners —Process optimal control and condition monitoring are vital for the stability and economic operation of industrial processes, and the basis of them is to quickly establish an accurate and reliable process performance prediction model. Traditional methods for developing process performance prediction models often require a large amount of complex calculations and rich process data, which is time- and cost-consuming. In particular, these methods focus only on the current process to be modeled, while ignoring the existing and similar process information, wasting process information. This article presents a minimum-cost modeling method for nonlinear industrial processes, which can make full use of information on multiple similar existing processes to assist the modeling of a new process to reduce the modeling cost of the new process. Specifically, a multimodel migration strategy including Bayesian model averaging is designed to migrate useful information from similar processes to the new process. The nested Latin hypercube design (NLHD) is employed to collect the necessary minimum data on the new nonlinear process. By applying the proposed approach to the industrial nonlinear process, it is possible to achieve the accurate performance prediction model with minimal new process data.

Transfer learning for pavement performance prediction

Accurate pavement performance prediction models are essential to ensure optimal allocation of resources in maintenance management. These models are developed using inventory and monitoring data regarding pavement structure, climate, traffic, and condition. However, numerous road agencies have limited pavement data. Due to the inexistence of historical data, data collection frequency, and/or quality issues, the amount of data available for the development of performance models is reduced. As a result, the resource allocati on process is significantly undermined. This paper proposes a transfer learning approach to develop pavement performance prediction models in limited data contexts. The proposed transfer learning approach is based on a boosting algorithm. In particular, a modified version of the popular TrAdaBoost learning algorithm was used. To test the proposed transfer learning approach, a case study was developed using data from the Long-Term Pavement Performance (LTPP) database and from the Portuguese road administration database. The results of this work show that it is possible to develop accurate performance prediction models in limited data contexts when a transfer learning approach is applied. All the models resulting from this approach outperformed baseline models, especially in what regards long-term forecasts. The results also showed that the transfer learning models perform consistently over different time frames, with minor performance losses from one-step to multi-step forecasts. The findings of this study should be of interest to road agencies facing limited data contexts and aiming to develop accurate prediction models that can improve their pavement management practice.

PREDICTION MODELLING APPROACH FOR CRACK PROGRESSION OF HEAVY DUTY FLEXIBLE PAVEMENTS

Pavement management at a network level requires reliable accurate performance prediction models to help road agencies make useful complex decisions about highways maintenance and rehabilitating activities. The purpose of this paper is to report the approach adopted for model development and validation for heavy duty flexible pavements representing by seven rural freeways segments. Hierarchical generalized linear modelling approach has been applied to predict multilevel model to capture the effect of variations among time series data, among road sections and among highways with same duty pavements. The estimation of pavement cracking progression has been based on longitudinal dataset contain cracking data (reported as a percent of the affected area) as dependent variable and cumulative traffic loading, pavement strength and environmental conditions as independent variables. The study illustrates how panel data can be nested to predict the probability of crack progression to capture the effect of significant unobserved heterogeneity. The significance of relevant contributing factors in predicting crack progression were presented and elucidated.The validation results indicate that the model replicates the pavement behavior well, and that the inclusion of additional factors in addition to time is improving the model prediction.

Provisioning data analytic workloads in a cloud

Data analytics applications are well-suited for a cloud environment. In this paper we examine the problem of provisioning resources in a public cloud to execute data analytic workloads. The goal of our provisioning method is to determine the most cost-effective configuration for a given data analytic workload. Provisioning a workload in a public cloud environment faces several challenges: it is difficult to develop accurate performance prediction models using standard methods; the space of possible configurations is very large so exact solutions cannot be efficiently determined, and the mix and intensity of query classes in a workload vary dynamically over time.We provide a formulation of the provisioning problem and then define a framework to solve the problem. Our framework contains a cost model to predict the cost of executing a workload on a configuration and a method of selecting configurations. The cost model balances resource costs and penalties from SLAs. The specific resource demands and frequencies are accounted for by queueing network models of the Virtual Machines (VMs), which are used to predict performance. We evaluate our approach experimentally using sample data analytic workloads on Amazon EC2.