Performance analysis of vapor compression refrigeration system using an adaptive neuro-fuzzy inference system

In this work, an experimental investigation is carried out with R134a and LPG refrigerant mixture (composed of R134a and LPG in the ratio of 28:72 by weight) as an alternative to R134a in a vapor compression refrigeration system. Performance tests were performed with different evaporator temperatures under controlled ambient conditions. The results showed that the R134a/LPG mixture has a higher coefficient of performance (COP) than R134a by about 15.28% in the studied range. The applicability of adaptive neuro-fuzzy inference system (ANFIS) to predict the COP of R134a/LPG system was also investigated. An ANFIS model for the system was developed. The comparison of statistical analysis of mathematical and ANFIS model predictions respectively in terms of the absolute fraction of variance (0.982 and 0.994), the root mean square error (0.0056 and 0.0050) and the mean absolute percentage error (0.286% and 0.217%) showed that ANFIS model gave the better statistical prediction efficiency.     

基于阻抗模型的玻璃棉声学参数反演方法研究

为了能准确获取玻璃棉材料的声学参数,文章对玻璃棉声学参数在不同阻抗模型下的声学参数进行了反演。采用了厚度分别为22 mm和44 mm的玻璃棉样本实测吸声曲线及各声学参数,选取四种常用阻抗模型,通过遗传算法(Genetic Algorithm,GA)对玻璃棉材料进行声学参数的反演,并选择反演效果最优的模型进行敏感性分析。比较各参数反演结果的误差比,并对比不同模型描述的吸声曲线与测试曲线的一致性,最后量化并比较Johnson-Champoux-Allard (JCA)模型中各参数对吸声系数的影响程度。研究表明,使用GA结合JCA模型或Johnson-Champoux-Allard-Lafarge (JCAL)模型反演的参数值与测试值误差较小;JCA模型适用于玻璃棉材料的声学参数反演,模型中流阻率和曲折度的敏感性较高,反演过程需保证其精确度。     

Prediction of mechanical properties of a warm compacted molybdenum prealloy using artificial neural network and adaptive neuro-fuzzy models

Predictive models using artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS) were successfully developed to predict yield strength and ultimate tensile strength of warm compacted 0.85 wt.% molybdenum prealloy samples. To construct these models, 48 different experimental data were gathered from the literature. A portion of the data set was randomly chosen to train both ANN with back propagation (BP) learning algorithm and ANFIS model with Gaussian membership function and the rest was implemented to verify the performance of the trained network against the unseen data. The generalization capability of the networks was also evaluated by applying new input data within the domain covered by the training pattern. To compare the obtained results, coefficient of determination (R²), root mean squared error (RMSE) and average absolute error (AAE) indexes were chosen and calculated for both of the models. The results showed that artificial neural network and adaptive neuro-fuzzy system were both potentially strong for prediction of the mechanical properties of warm compacted 0.85 wt.% molybdenum prealloy; however, the proposed ANFIS showed better performance than the ANN model. Also, the ANFIS model was subjected to a sensitivity analysis to find the significant inputs affecting mechanical properties of the samples.     

Optimizing of ANFIS for estimating INS error during GPS outages

Global positioning system (GPS) has been extensively used for land vehicle navigation systems. However, GPS is incapable of providing permanent and reliable navigation solutions in the presence of signal evaporation or blockage. On the other hand, navigation systems, in particular, inertial navigation systems (INSs), have become important components in different military and civil applications due to the recent advent of micro-electro-mechanical systems (MEMS). Both INS and GPS systems are often paired together to provide a reliable navigation solution by integrating the long-term GPS accuracy with the short-term INS accuracy. This article presents an alternative method to integrate GPS and INS systems and provide a robust navigation solution. This alternative approach to Kalman filtering (KF) utilizes artificial intelligence based on adaptive neuro-fuzzy inference system (ANFIS) to fuse data from both systems and estimate position and velocity errors. The KF is usually criticized for working only under predefined models and for its observability problem of hidden state variables, sensor error models, immunity to noise, sensor dependency, and linearization dependency. The training and updating of ANFIS parameters is one of the main problems. Therefore, the challenges encountered implementing an ANFIS module in real time have been overcome using particle swarm optimization (PSO) to optimize the ANFIS learning parameters since PSO involves less complexity and has fast convergence. The proposed alternative method uses GPS with INS data and PSO to update the intelligent PANFIS navigator using GPS/INS error as a fitness function to be minimized. Three methods of optimization have been tested and compared to estimate the INS error. Finally, the performance of the proposed alternative method has been examined using real field test data of MEMS grade INS integrated with GPS for different GPS outage periods. The results obtained outperform KF, particularly during long GPS signal blockage.     

Electronic expansion valve mass flow rate prediction based on dimensionless correlation and ANN model

Refrigerant mass flow rate through electronic expansion valve (EEV) makes significant sense for refrigeration system intelligent control and energy conservation. Objectives of this study were to present experimental data of R134a mass flow rate through EEV and to develop models for EEV mass flow rate prediction via two approaches: dimensionless correlation based on Buckingham π-theorem and artificial neural network (ANN) model based on dimensionless parameters. The database utilized for model training and test was comprised of our experimental data and data available in open literatures including R22, R407C, R410A and R134a. Compared with three existing dimensionless correlations, the proposed dimensionless correlation and ANN model demonstrated higher accuracy. The proposed dimensionless correlation gave mean relative error (MRE) of 6.60%, relative mean square error of (RMSE) 12.05 kg h⁻¹ and correlation coefficient (R²) of 0.9810. The ANN model with the configuration of 8-6-1 showed MRE, RMSE and R² of 3.97%, 7.59 kg h⁻¹ and 0.9924, respectively.     

Performance prediction of a specific wear rate in epoxy nanocomposites with various composition content of polytetrafluoroethylen (PTFE), graphite,short carbon fibers (CF) and nano-TiO2 using adaptive neuro-fuzzy inference system (ANFIS)

Specific wear rate of composite materials plays a significant role in industry. The processes to measure it are both time and cost consuming. It is essential to suggest a modeling method to predict and analyze the effectiveness of parameters of specific wear rate. Nowadays, computational methods such as Artificial Neural Network (ANN), Fuzzy Inference System (FIS) and adaptive neuro-fuzzy inference system (ANFIS) are mainly considered as applicable tools from modeling point of view. ANFIS present integrate performance of neural network (NN) and fuzzy system (FS). Present paper investigates performance prediction of a specific wear rate of epoxy composites with various composition using ANFIS. The obtained results showed that ANFIS is a powerful tool in modeling specific wear rate. The obtained mean of squared error (MSE) for testing sets in present paper obtained 0.0071.     

An ANFIS model for the prediction of flow stress of Ti600 alloy during hot deformation process

In this paper, an adaptive network-based fuzzy inference system (ANFIS) model has been established to predict the flow stress of Ti600 alloy during hot deformation process. This network integrates the fuzzy inference system with a back-propagation learning algorithm of neural network. The experimental results were obtained from Gleeble-1500 thermal-simulator at deformation temperatures of 800–1100 °C, strain rates of 0.001–10 s^?1, and height reduction of 70%. In establishing this ANFIS model, strain rate, deformation temperature and the strain are entered as input parameters while the flow stress are used as output parameter. After the training process, the fuzzy membership functions and the weight coefficient of the network can be optimized. A comparative evaluation of the predicted and the experimental results has shown that the ANFIS model used to predict the flow stress of Ti600 titanium alloy has a high accuracy and with absolute relative error is less than 17.39%. Moreover, the predicted accuracy of flow stress during hot deformation process of Ti600 titanium alloy using ANFIS model is higher than using traditional regression method, indicating that the ANFIS model was an easy and practical method to predict flow stress for Ti600 titanium alloy.     

The discrepancy in the prediction of surface temperatures by inverse heat conduction models for different quenching processes from very high initial surface temperature

In the current work, an attempt has been made to study the effect of different parameters on the accuracy of the prediction at a very high initial surface temperature by developing two different heat conduction models. The result depicts that MSSE (minimum sum squared error) in the prediction decreases with increasing number of sensors used in the prediction. The accuracy of the prediction enhances with decreasing plate thickness and distance between the thermocouple and quenched surface. Up to a cooling rate of 60?K/s, the selection of model dimension (1-D or 2-D) does not affect, but beyond the previously mentioned cooling rate, 2-D model induces less error than 1-D. Moreover, the inclusion of thermo-physical properties in the model reduces the error in the MSSE. By using Box–Behnken methodology, the optimum conditions (d/D?=?0.81, n/Y?=?0.5 and Y*/Y?=?0.65) for the least MSSE have also been determined.     

Neuro-fuzzy model for evaluating the performance of processes using transfer function

In this work an Adaptive Neuro-Fuzzy Inference System (ANFIS) was used to model the periodic performance of some multi-input single-output (MISO) processes, namely: brewery operations (case study 1) and soap production (case study 2) processes. Two ANFIS models were developed to model the performance of the two processes under study. The results of the study show that for brewery operations, ANFIS model 2 with a correlation coefficient of 0.9972, as against 0.9956 for ANFIS model 1, had a better correlation than an equivalent MAMDANI fuzzy model. On the order hand, for soap production process, ANFIS model 1 had better correlation with an equivalent MAMDANI model. Generally, there is a general agreement among the models on the periodic performance of the processes. Thus, all the models show that for the brewery, the best performance was in the period 2010–2011 and the period 2008–2009 was the worst. Similarly, for the soap production process, the best performance was in 2011 and the worst in 2012. The results show that a combination of transfer function and ANFIS could be used effectively to model process performance.     

Prediction of hopper discharge rate using combined discrete element method and artificial neural network

An Artificial Neural Network (ANN) was developed to predict the mass discharge rate from conical hoppers. By employing Discrete Element Method (DEM), numerically simulated flow rate data from different internal angles (20°–80°) hoppers were used to train the model. Multi-component particle systems (binary and ternary) were simulated and mass discharge rate was estimated by varying different parameters such as hopper internal angle, bulk density, mean diameter, coefficient of friction (particle-particle and particle-wall) and coefficient of restitution (particle-particle and particle-wall). The training of ANN was accomplished by feed forward back propagation algorithm. For validation of ANN model, the authors carried out 22 experimental tests on different mixtures (having different mean diameter) of spherical glass beads from different angle conical hoppers (60° and 80°). It was found that mass discharge rate predicted by the developed neural network model is in a good agreement with the experimental discharge rate. Percentage error predicted by ANN model was less than ±13%. Furthermore, the developed ANN model was also compared with existing correlations and showed a good agreement.     

Prediction of the concrete compressive strength by means of core testing using GMDH-type neural network and ANFIS models

Assessment of insitu concrete strength by means of cores cut from hardened concrete is accepted as the most common method, but may be affected by many factors. Group method of data handling (GMDH) type neural networks and adaptive neuro-fuzzy inference systems (ANFIS) were developed based on results obtained experimentally in this work along with published data by other researchers. Genetic algorithm (GA) and singular value decomposition (SVD) techniques are deployed for optimal design of GMDH-type neural networks. Samples incorporated six parameters with core strength, length-to-diameter ratio, core diameter, aggregate size and concrete age considered as inputs and standard cube strength regarded as the output. The results show that a generalized GMDH-type neural network and ANFIS have great ability as a feasible tool for prediction of the concrete compressive strength on the basis of core testing. Moreover, sensitivity analysis has been carried out on the model obtained by GMDH-type neural network to study the influence of input parameters on model output.     

Determination of material parameters of a viscoplastic model by genetic algorithm

Macroscopic (continuum) and microscopic models, used for simulation of material behaviors under different loading conditions, contain a large number of material parameters and determination of these parameters is an important and difficult issue in the modeling. The aim of this work essentially deals with parameter determination procedure of any viscoplasticity model. In this study, genetic algorithm (GA) parameter optimization procedure has been proposed to determine material parameters of viscoplastic models. Parameter determination capability of the GA optimization method was tested by using VBO model which one of the viscoplasticity theory with no yield surface and loading–unloading conditions. Fourteen material parameters of VBO model are determined using uniaxial loading–unloading stress strain curves of high density polyethylene (HDPE). Using these material parameters, creep and relaxation behaviors of HDPE are simulated. A good match with experimental results is obtained. Apart from many existing studies in the literature, GA optimization procedure is applied to determine material parameters instead of trial and error procedure. This method can also be used to determine materials parameters of other viscoplasticity theories for all kinds of materials.     

Acoustic characterization and prediction of the cut-off dimensionless frequency of an elastic tube by neural networks

A neural network is developed to predict cut-off dimensionless frequencies of the antisymmetric circumferential waves (Ai) propagating around an elastic circular cylindrical shell of different radius ratio b/a (a, outer radius; b, inner radius). The useful data to train and test the performances of the model are determinated from calculated trajectories of natural modes of resonances or extracted from time-frequency representations of Wigner-Ville of the acoustic backscattered time signal obtained from a computation. In this work, the studied tubes are made of aluminum or stainless steel. The material density, the radius ratio b/a, the index i of the antisymmetric waves, and the propagation velocities in the tube, are selected like relevant entries of the model of neural network. During the development of the network, several configurations are evaluated. The optimal model selected is a network with two hidden layers. This model is able to predict the cut-off dimensionless frequencies with a mean relative error (MRE) of about 1%, a mean absolute error (MAE) of 3.10(-3) k1a, and a standard error (SE) of 10(-3) k1a (k1a is the dimensionless frequency, k is the wave number in water).     

Application of the output dependent feature scaling in modeling and prediction of performance of counter flow vortex tube having various nozzles numbers at different inlet pressures of air, oxygen, nitrogen and argon

In this study, the performance of the counter flow type vortex tube with the input parameters including the nozzle number (N), the densities of inlet gases (air, oxygen, nitrogen, and argon) and the inlet pressure (P_inlet) has been modeled with the proposed hybrid method combining a novel data preprocessing called output dependent feature scaling (ODFS) and adaptive network based fuzzy inference system (ANFIS) by using the experimentally obtained data. In the developed system, output parameter temperature gradient between the cold and hot outlets has been determined using input parameters comprising (P_inlet), (N), and the density of gases. In order to evaluate the performance of hybrid method, the mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), determination coefficient (R²), and Index of Agreement (IA) values have been used. The obtained results are 9.0670e-004 (MAE), 5.8563e-006 (MSE), 0.0024 (RMSE), 1.00 (R²), and 1.00 (IA) using the hybrid method.     

One step-ahead ANFIS time series model for forecasting electricity loads

In electric industry, electricity loads forecasting has become more and more important, because demand quantity is a major determinant in electricity supply strategy. Furthermore, accurate regional loads forecasting is one of principal factors for electric industry to improve the management performance. Recently, time series analysis and statistical methods have been developed for electricity loads forecasting. However, there are two drawbacks in the past forecasting models: (1) conventional statistical methods, such as regression models are unable to deal with the nonlinear relationships well, because of electricity loads are known to be nonlinear; and (2) the rules generated from conventional statistical methods (i.e., ARIMA), and artificial intelligence technologies (i.e., support vector machines (SVM) and artificial neural networks (ANN)) are not easily comprehensive for policy-maker. Based on these reasons above, this paper proposes a new model, which incorporates one step-ahead concept into adaptive-network-based fuzzy inference system (ANFIS) to build a fusion ANFIS model and enhances forecasting for electricity loads by adaptive forecasting equation. The fuzzy if-then rules produced from fusion ANFIS model, which can be understood for human recognition, and the adaptive network in fusion ANFIS model can deal with the nonlinear relationships. This study optimizes the proposed model by adaptive network and adaptive forecasting equation to improve electricity loads forecasting accuracy. To evaluate forecasting performances, six different models are used as comparison models. The experimental results indicate that the proposed model is superior to the listing models in terms of mean absolute percentage errors (MAPE).     

Adaptive neuro-fuzzy inference system approach to predict the mass flow rate of R-134a/LPG refrigerant for straight and helical coiled adiabatic capillary tubes in the vapor compression refrigeration system

This study deals with predicting the mass flow rate of R-134a/LPG as refrigerant inside a straight and helical coiled adiabatic capillary tube of vapor compression refrigeration system by combining dimensionless analysis and Adaptive Neuro-Fuzzy Inference System techniques. For this purpose the experimental system was designed and tested under steady state conditions, by changing the length of the capillary tube, the inner diameter of the capillary tube, the coil diameter and the degree of subcooling of the refrigerant at the capillary tube inlet. Dimensional analysis was utilized to provide generalized dimensionless parameters and to reduce the number of input parameters, while Adaptive Neuro-Fuzzy Inference System was applied as a generalized approximator of the nonlinear multi-input and single-output function. The comparison of the absolute fraction of variance (R²) (0.998 and 0.961), the root mean square error (RMSE) (0.105 kg/h and 0.489 kg/h) and the mean absolute percentage error (MAPE) (0.954% and 4.75%) demonstrated the result for combination of dimensional analysis and Adaptive Neuro-Fuzzy Inference System and dimensionless correlation model predictions respectively. The results indicated that the combination of dimensional analysis and Adaptive Neuro-Fuzzy Inference System gave the best statistical prediction efficiency.     

Prediction of the heat transfer rate of a single layer wire-on-tube type heat exchanger using ANFIS

In this paper, we applied an Adaptive Neuro-Fuzzy Inference System (ANFIS) model for prediction of the heat transfer rate of the wire-on-tube type heat exchanger. Limited experimental data was used for training and testing ANFIS configuration with the help of hybrid learning algorithm consisting of backpropagation and least-squares estimation. The predicted values are found to be in good agreement with the actual values from the experiments with mean relative error less than 2.55%. Also, we compared the proposed ANFIS model to an ANN approach. Results show that the ANFIS model has more accuracy in comparison to ANN approach. Therefore, we can use ANFIS model to predict the performances of thermal systems in engineering applications, such as modeling heat exchangers for heat transfer analysis.     

Modeling of rotary vane compressor applying artificial neural network

The thermal modeling of rotary vane compressor (RVC) was performed in this paper by applying Artificial Neural Network (ANN) method. In the first step, appropriate tests were designed and experimental data were collected during steady state operating condition of RVC in the experimental setup. Then parameters including refrigerant suction temperature and pressure, compressor rotating speed as well as refrigerant discharge pressure were adjusted.With these input values, the operating output parameters such as refrigerant mass flow rate and refrigerant discharge temperature were measured. In the second step, the experimental results were used to train ANN model for predicting RVC operating parameters such as refrigerant mass flow rate and compressor power consumption. These predicted operating parameters by ANN model agreed well with the experimental values with correlation coefficient in the range of 0.962-0.998, mean relative errors in the range of 2.79-7.36% as well as root mean square error (RMSE) 10.59 kg h⁻¹ and 12 K for refrigerant mass flow rate and refrigerant discharge temperature, respectively. Results showed closer predictions with experimental results for ANN model in comparison with nolinear regression model.     

基于RF和ANFIS算法的直驱风电机组故障预警

针对风电机组运行工况复杂和单一状态参数不能较好实现故障早期预警的特点,提出随机森林算法(RF)和自适应模糊神经网络算法(ANFIS)相结合的故障预警方法。该方法充分考虑机组运行数据高维非线性特点,应用随机森林算法,建立有功功率与运行参数的数据驱动模型,计算各运行参数影响有功功率的相关度;构建自适应网络模糊推理系统模型,以训练误差最大值作为故障预警阈值,实时监测发电机运行状态。将该方法应用于某1.5 MW直驱机组发电机故障预警分析,结果表明,该方法能够提前预警发电机健康状态,避免严重事故发生,对风电场开展预防性维护、维修具有重要的指导意义。     

Prediction of debonding strength for masonry elements retrofitted with FRP composites using neuro fuzzy and neural network approaches

This paper proposes application of neuro fuzzy and neural network for predicting debonding strength of retrofitted masonry elements. In order to achieve high-fidelity model, this study uses extensive experimental databases for bond test results between Fiber Reinforced Polymer (FRP) and masonry elements by collecting existing bond test subassemblage tests from the literature. Various influential parameters that affect debonding resistance including thickness of the FRP strip, width of the FRP strip, elastics modulus of the FRP, bonded length, tensile strength of the masonry block and width of the masonry block are considered as input parameters to the artificial neural network (ANN) and adaptive neuro fuzzy inference system (ANFIS). Test results of the ANN and ANFIS models were compared with multiple nonlinear regression, multiple linear regression and existing bond strength models. The accuracy of the optimal MNLR model was increased by 39% and 23% with respect to RMSE and MAE criteria using ANFIS. The comparison results indicated that the ANN and ANFIS models performed better than the other models and could be successfully used for prediction of debonding strength of retrofitted masonry elements.