Gene Expression Programming Method Research Articles

Optimal shape design of an autonomous underwater vehicle based on multi-objective particle swarm optimization

Optimization of autonomous underwater vehicle’s shape is usually a multi-objective optimization problem, which is essential for autonomous underwater navigation and manipulation. To overcome the inefficiency of computational fluid dynamics software during the optimization process and the limitations of traditional single-objective optimization, a novel strategy combining genetic expression programming and crowding distance based multi-objective particle swarm algorithm is presented. Its central idea is as follows, several underwater vehicle shapes are analysed to obtain their water resistances and determine the best underwater robot shape. Shape factor of the bow and shape factor of the stern are employed as design variables, and sample points are selected by the optimal latin hypercube design. Then gene expression programming method is used to establish the surrogate model of resistance and surrounded volume. After that, the surrogate model based on the gene expression programming method is compared with that based on the surface respond method. The results show the superiority of the GEP method. Then the resistance and surrounded volume are set as two optimized variables and Pareto optimal solutions are obtained by using multi-objective particle swarm algorithm. Finally, the optimization results are compared with the hydrodynamic calculations, which shows the method proposed in the paper can greatly reduce the cost of computation and improve the efficiency of optimal shape design for underwater vehicle.

Daily global solar radiation modeling using data-driven techniques and empirical equations in a semi-arid climate

ABSTRACTSolar radiation, moisture and temperature are the most vital meteorological variables which affect plant growth. Due to the fact that the global solar radiation (GSR) is scarcely gauged at meteorological stations in developing countries, it is commonly estimated by data-driven techniques or by empirical equations. In this study, support vector regression (SVR), model trees (MT), gene expression programming (GEP) and adaptive neuro–fuzzy inference system (ANFIS) and several empirical equations were applied to assess the relations between GSR and several meteorological variables including minimum temperature (Tmin), maximum temperature (Tmax), relative humidity (RH), sunshine hours (n), maximum sunshine hours (N), corrected clear-sky solar irradiation (ICSKY), day of year (DOY) and extra-terrestrial radiation (Ra). For this purpose, the daily GSR measured from the beginning of 2011 to the end of 2013 at Tabriz synoptic station, which is located in semi-arid regions of Iran, were used. A direct strong relationship was observed to exist between the GSR and n. For evaluating the performances of studied techniques, three different statistical indicators were used namely root mean square error (RMSE), mean absolute error (MAE), and correlation coefficient (CC). Additionally, a Taylor diagram was utilized to test the similarity between the observed and predicted GSR values. Results indicated that the SVR-6 with input parameters of Ra, RH, Tmin, Tmax, n/N had better accuracy in predicting GSR with RMSE of 1.656, MAE of 0.990, CC of 0.980 and WI of 0.990 than the other models. Moreover, MT-6 ranked as the second best model in the prediction of GSR values. As an interesting point, studied empirical equations had lower accuracies comparing with the SVR, GEP, MT and ANFIS methods. For instance, GSR values were computed by Angstrom and Prescott equation, as the best empirical equation, with RMSE of 1.786, MAE of 1.156, CC of 0.977 and WI of 0.988. Conclusively, results from the current study proved that the SVR provided reasonable trends for GSR modeling at Tabriz synoptic station. Furthermore, MT models with linear equations can be implemented with a high degree of simplicity and acceptable precision in GSR estimation.

Effect of cement strength class on the prediction of compressive strength of cement mortar using GEP method

Gene expression programming (GEP) has been widely used to predict the properties of cementation materials. In this study, 54 mix designs including six water/cement (W/C) ratios of 0.25, 0.30, 0.35, 0.40, 0.45 and 0.50, three sand/cement (S/C) ratios of 2.50, 2.75, and 3.00 as well as three cement strength classes (CSC) of 32.5, 42.5 and 52.5 MPa were first constructed and then the compressive strength of 270 constructed samples with five different ages of 3, 7, 14, 21 and 28 days was measured. The compressive strength of cement mortar was then predicted using GEP and the results were utilized to investigate the roles of linking function and CSC on the performance of GEP models. The effect of CSC on the prediction of compressive strength of cement mortar was also evaluated by comparing the prediction results obtained from the proposed GEP in the current study considering CSC as an extra input parameter with the prediction results taken from existing GEP model from the literature without considering the CSC, where the input parameters were collected from three data sets of previous studies from the literature. The results showed that the GEP model with linking function of addition has a better performance than that with linking function of multiplication. The results also showed the strong potential of proposed GEP in predicting the compressive strength of cement mortar. Furthermore the results showed that considering the CSC as an additional input data increases the prediction accuracy of compressive strength.

Investigation of Carbon Fiber- and Wollastonite-Filled Graphite/Asphalt/Cu Composite Materials Using the Gene Expression Programming

The effect of addition of carbon fibers and wollastonite on the mechanical properties of graphite/asphalt/Cu composites is described. The gene expression programming method is used to predict the effects of fillers on their properties. A multiconstraint optimization problem is solved, and the optimum composition is determined.

Proposing new relationships to estimate the pressuremeter modulus of cohesive and cohesionless media

Pressuremeter is one of the most reliable geotechnical in situ tests to be utilized in estimating different soil properties. The main parameters that can be obtained from this test are soil deformation modulus. In general, the pressuremeter test is time-consuming and costly that requires suitable equipment and experienced operators. With these limitations, it is necessary to introduce models for indirect determination of the pressuremeter modulus (EPM). Based on the literature, various models and equations have been proposed to predict the modulus of pressuremeter; however, most of them are for fine-grained soils. In this paper, multiple linear regression (MLR), artificial neural network (ANN), and gene expression programming (GEP) have been used to design new relationships for prediction of the pressuremeter modulus. The main difference between this research and other studies is the use of 62 pre-bored pressuremeter tests results in cohesive and granular soils for training and testing models. The study area is in Tehran where soils show a variation from low plasticity clay to clayey gravel. The moisture content, depth of test, and grain size distribution of soils are considered as independent variables. Comparison of the predicted EPM with the actual modulus obtained from the pressuremeter tests indicates that the proposed relationships are able to estimate the pressuremeter modulus well. The results showed that the relationships obtained from nonlinear analyses performed by smart methods are more accurate and have less error in comparison to MLR method. Comparison of the coefficient of determination (R2) values and errors related to them in the testing phase show that the obtained results from the GEP method are more reliable than the ANN method. Also, sensitivity analysis revealed that the soil moisture content is the most effective parameter among the variables on the pressuremeter modulus.

Use of adaptive neuro-fuzzy inference system and gene expression programming methods for estimation of the bearing capacity of rock foundations

PurposeThis study aims to examine the potential of two artificial intelligence (AI)-based algorithms, namely, adaptive neuro-fuzzy inference system (ANFIS) and gene expression programming (GEP), for indirect estimation of the ultimate bearing capacity (qult) of rock foundations, which is a considerable civil and geotechnical engineering problem.Design/methodology/approachThe input-processing-output procedures taking place in ANFIS and GEP are represented for developing predictive models. The great importance of simultaneously considering both qualitative and quantitative parameters for indirect estimation of qult is taken into account and explained. This issue can be considered as a remarkable merit of using AI-based approaches. Furthermore, the evaluation procedure of various models from both engineering and accuracy viewpoints is also demonstrated in this study.FindingsA new and explicit formula generated by GEP is proposed for the estimation of the qult of rock foundations, which can be used for further engineering aims. It is also presented that although the ANFIS approach can predict the output with a high degree of accuracy, the obtained model might be a black-box. The results of model performance analyses confirm that ANFIS and GEP can be used as alternative and useful approaches over previous methods for modeling and prediction problems.Originality/valueThe superiorities and weaknesses of GEP and ANFIS techniques for the numerical analysis of engineering problems are expressed and the performance of their obtained models is compared to those provided by other approaches in the literature. The findings of this research provide the researchers with a better insight to using AI techniques for resolving complicated problems.

Discharge and flow field simulation of open channel sewer junction using artificial intelligence methods

One of the most important parameters in designing of sewer structures is the ability to accurately simulation the discharge and velocity field of them. Among the various sewer receiving inflow methods, open channel junctions are mostly occurring. Because of the separation and contraction zone that occur at the open channel junctions, the fluid flow has a complex behavior. Modeling is carried out by Radial Basis Function (RBF) neural network, Gene Expression Programming (GEP), and Multiple Non-Linear Regression (MNLR) methods. Finding the optimum situation for GEP and RBF models are done by examining the various mathematical and linking functions for GEP and different number of hidden neurons and spread amount for RBF. In order to use the models in practical situations, three equations were conducted by using the RBF, GEP, and MNLR methods in modeling the longitudinal velocity. Then, the surface integral of the presented equations is used to simulate the flow discharge. The results showed that the GEP and RBF method perform significantly better than the MNLR in open channel junction characteristics simulations. The GEP method has higher performance in modeling the longitudinal velocity field compare with the RBF. However, the RBF presented more reliable results on the discharge simulations.

Long-term prediction of rockburst hazard in deep underground openings using three robust data mining techniques

Rockburst phenomenon is the extreme release of strain energy stored in surrounding rock mass which could lead to casualties, damage to underground structures and equipment and finally endanger the economic viability of the project. Considering the complex mechanism of rockburst and a large number of factors affecting it, the conventional criteria cannot be used generally and with high reliability. Hence, there is a need to develop new models with high accuracy and ease to use in practice. This study focuses on the applicability of three novel data mining techniques including emotional neural network (ENN), gene expression programming (GEP), and decision tree-based C4.5 algorithm along with five conventional criteria to predict the occurrence of rockburst in a binary condition. To do so, a total of 134 rockburst events were compiled from various case studies and the models were established based on training datasets and input parameters of maximum tangential stress, uniaxial tensile strength, uniaxial compressive strength, and elastic energy index. The prediction strength of the constructed models was evaluated by feeding the testing datasets to the models and measuring the indices of root mean squared error (RMSE) and percentage of the successful prediction (PSP). The results showed the high accuracy and applicability of all three new models; however, the GA-ENN and the GEP methods outperformed the C4.5 method. Besides, it was found that the criterion of elastic energy index (EEI) is more accurate among other conventional criteria and with the results similar to the C4.5 model, can be used easily in practical applications. Finally, a sensitivity analysis was carried out and the maximum tangential stress was identified as the most influential parameter, which could be a guide for rockburst prediction.

Forecasting copper price using gene expression programming

Forecasting the prices of metals is important in many aspects of economics. Metal prices are also vital variables in financial models for revenue evaluation, which forms the basis of an effective payment regime using resource policymakers. According to the severe changes of the metal prices in the recent years, the classic estimation methods cannot correctly estimate the volatility. In order to solve this problem, it is necessary to use the artificial algorithms, which have a good ability to predict the volatility of various phenomena. In the present work, the gene expression programming (GEP) method was used to predict the copper price volatility. In order to understand the ability of this method, the results obtained were compared with the other classical prediction methods. The results indicated that the GEP method was much better than the time series and multivariate regression methods in terms of the prediction accuracy.

A Comparative Study of Autoregressive, Autoregressive Moving Average, Gene Expression Programming and Bayesian Networks for Estimating Monthly Streamflow

In the recent years, artificial intelligence techniques have attracted much attention in hydrological studies, while time series models are rarely used in this field. The present study evaluates the performance of artificial intelligence techniques including gene expression programming (GEP), Bayesian networks (BN), as well as time series models, namely autoregressive (AR) and autoregressive moving average (ARMA) for estimation of monthly streamflow. In addition, simple multiple linear regression (MLR) was also used. To fulfill this objective, the monthly streamflow data of Ponel and Toolelat stations located on Shafarood and Polrood Rivers, respectively in Northern Iran were used for the period of October 1964 to September 2014. In order to investigate the models’ accuracy, root mean square error (RMSE), mean absolute error (MAE) and coefficient of determination (R2) were employed as the error statistics. The obtained results demonstrated that the single AR and ARMA time series models had better performance in comparison with the single GEP, BN and MLR methods. Furthermore, in this study, six hybrid models known as GEP-AR, GEP-ARMA, BN-AR, BN-ARMA, MLR-AR and MLR-ARMA were developed to enhance the estimation accuracy of the monthly streamflow. It was concluded that the developed hybrid models were more accurate than the corresponding single artificial intelligence and time series models. The obtained results confirmed that the integration of time series models and artificial intelligence techniques could be of use to improve the accuracy of single models in modeling purposes related to the hydrological studies.

A comparison of performance of several artificial intelligence methods for predicting the dynamic viscosity of TiO2/SAE 50 nano-lubricant

Since the conventional thermal fluids such as water, oil, and ethylene glycol have poor thermal properties, the tiny solid particles are added to these fluids to increase their heat transfer improvement. As viscosity determines the rheological behavior of a fluid, studying the parameters affecting the viscosity is crucial. Since the experimental measurement of viscosity is expensive and time consuming, predicting this parameter is the apt method. In this work, three artificial intelligence methods containing Genetic Algorithm-Radial Basis Function Neural Networks (GA-RBF), Least Square Support Vector Machine (LS-SVM) and Gene Expression Programming (GEP) were applied to predict the viscosity of TiO2/SAE 50 nano-lubricant with Non-Newtonian power-law behavior using experimental data. The correlation factor (R2), Average Absolute Relative Deviation (AARD), Root Mean Square Error (RMSE), and Margin of Deviation were employed to investigate the accuracy of the proposed models. RMSE values of 0.58, 1.28, and 6.59 and R2 values of 0.99998, 0.99991, and 0.99777 reveal the accuracy of the proposed models for respective GA-RBF, CSA-LSSVM, and GEP methods. Among the developed models, the GA-RBF shows the best accuracy.

Using of gene expression programming and climatic data for forecasting flow discharge by considering trend, normality, and stationarity analysis

In this research, the main hydrological characteristics (such as trend, stationarity, and normalization of hydrological data) of the Kasilian watershed are considered from 1970 to 2009. For forecasting of discharge, gene expression programming (GEP) method is applied. Normality and stationarity of time series are necessary for application of GEP method. For this purpose, third edition of Mann-Kendall trend test and skewness test are used for detection of trend and normalization of data, respectively. Also, five methods are applied for detection of stationarity of data. Modified Mann-Kendall trend test and Theil and Sen’s median slope method illustrate that annual and monthly precipitation data have slight decreasing trend, annual and monthly discharge data have insignificant decreasing trend, and annual and monthly temperature data have an increasing trend. Skewness test illustrates that annual, monthly, and daily discharge and precipitation data are not normal. By using logarithm function, skewness is minimized and symmetry of data is improved. After normalization of time series by logarithm function, five methods are applied for testing of stationarity of time series. These methods show that different normalized time series are stationarity and stationarity of time series is improved by elimination of periodic properties of data. For forecasting of daily discharge by GEP method, 85% of data are used for training and 15% of data are used for testing. By using data of 3 days ago, the GEP has the best efficiency. Coefficient of correlation (CC), root mean square error (RMSE), mean absolute error (MAE), and mean absolute relative error (MARE) are 0.9, 0.495 lit/s, 0.288 lit/s, and 0.053, respectively.

A quality control method for surface hourly temperature observations via gene‐expression programming

ABSTRACTThis article aims to find the best‐fit regression function for the quality control (QC) for the surface hourly temperature. A new QC method based on gene‐expression programming (GEP) was employed to identify potential outliers in the surface hourly temperature observations. Compared to the spatial regression test (SRT) method and inverse distance weighting (IDW) method, the results indicate that both GEP and SRT outperform IDW in all the cases. The GEP method and the SRT method display equivalent effects in most of the cases. But GEP was found to be superior to the other methods when the weather station density was low. For all cases, the GEP method could yield a smaller estimated error than the other methods, respectively. The results of the comparison led to the recommendation that the GEP method is an effective QC method in identifying the seeded errors for the surface hourly temperature.

Quantitative structure-activity relationship and molecular docking studies on designing inhibitors of the perforin.

Quantitative structure-activity relationship (QSAR) studies were performed on a series of 5-arylidene-2thioxoimidazolidin-4-ones derivatives as the inhibitors of perforin and to gain insights about the structural determinants for designing new drug molecules. The heuristic method could explore the descriptors responsible for bioactivity and gain a best linear model with R2 .82. Gene expression programming method generated a novel nonlinear function model with R2 .92 for training set and R2 .85 for test set. The predicted IC50 by QSAR, molecular docking analysis, and property explorer applet show that 42a acts as a well-pleasing potent inhibitor for perforin. This study may lay a reliable theoretical foundation for the development of designing perforin inhibitor structures.

A New Predictive Model for Rock Strength Parameters Utilizing GEP Method

The present study aims to employ modern intelligent method to predict intact rock strength parameters. This method can be used for intact rock strength parameters prediction of different extraction projects. Mechanical rock excavation projects need uniaxial compressive strength (UCS) and static modulus of elasticity (E) of the intact rock material. Many parameters affect to strength parameters, but some of them are applicable to empirical or analytical equations and the other difficulty, high-quality core samples of appropriate geometry are needed to find out these parameters. In this regard, models predicting UCS and E based on rock index tests and intact rock properties could be useful methods. This paper aims to employ Gene Expression Programming (GEP) to predict E and UCS. Out of the 44 sets of the data, 22 sets (50% of the data) were considered for training and the remaining 22 sets of the data (50%) were considered for testing. The intelligent method has been studied on the basis of data obtained from 44 different excavation projects all over the world. These parameters were collected from previous research data. The values of UCS and E are predicted by using quartz content (Q), dry density (γd) and porosity (n) of the rocks. 22 datasets (50% of the data) were utilized for modeling and the remaining 22 sets of the data (50%) were considered for evaluating theirs performance. For this purpose, writing a code was necessary, as some of the proposed relations were complex. The obtained results of this study are presented within a computer-based format in order to be easily accessible too every experts. With respect to the accuracy of the GEP method, it may be recommended for predicting intact rock strength parameters for future excavation design purpose.

Comparison of genetic programming and radial basis function neural network for open-channel junction velocity field prediction

The ability to accurately predict the velocity field in open-channel junctions has a significant impact on the designing process of many hydraulic structures, such as irrigation and drainage channels and sewer networks. The gene expression programming (GEP) and radial basis function neural network (RBF-NN) methods are developed in order to find a continuous spatial velocity description using discrete experimental measurements. By using the coordinates of each point and the tributary-to-main-discharge ratio of an open-channel junction, seven different input combinations are investigated. To find the optimum GEP model, various mathematical and linking functions are studied. The RBF-NN models are developed with various numbers of hidden layer nodes and amounts of spread. A comparison of the results indicates that both GEP and RBF-NN methods can accurately simulate flow in junctions. However, the GEP with root-mean-squared error (RMSE) of 0.2361 performs better than RBF-NN with RMSE of 0.2590. Owing to the higher accuracy and explicit output equation of the GEP, this method can be used in practical situations of predicting open-channel characteristics.

Estimating the aeration coefficient and air demand in bottom outlet conduits of dams using GEP and decision tree methods

Accurate estimates of air demand are critical when addressing the cavitation phenomenon in bottom outlet conduits of dams. In the present study, this accuracy was evaluated for air demand estimation models using gene expression programming (GEP), classification tree methods [including boosted regression trees (BRT) and random forest (RF) algorithms], and empirical relationships. Using k-fold cross validation and drawing on data from 6 dams, two different air-demand-related parameters were estimated (aerator air discharge demand and aeration coefficient) and the estimation accuracy was assessed using the Nash-Sutcliffe (NS) efficiency and other statistics. Outperforming the other models, the GEP model performed well in estimating aerator air demand (NS =0.674) and the aeration coefficient (NS =0.489). Moreover, on average, the GEP model showed 45% and 12% improvements in aeration coefficient estimation accuracy over empirical relationships and decision tree methods, respectively. Finally, in the function finding phase, the derived mathematical models were presented as non-linear equations to estimate the air demand of aerator and aeration coefficients using a non-linear relation derived from the GEP model.

Predicting the Oil Well Production Based on Multi Expression Programming

Predicting the oil well production is very important and also quite a complex mission for the petroleum engineering. Due to its complexity, the previous empirical methods could not perform well for different kind of wells, and intelligent methods are applied to solve this problem. In this paper the multi expression programming (MEP) method has been employed to build the prediction model for oil well production, combined with the phase space reconstruction technique. The MEP has shown a better performance than the back propagation networks, gene expression programming method and the Arps decline model in the experiments, and it has also been shown that the optimal state of the MEP could be easily obtained, which could overcome the over-fitting.

Development of predictive model for flood routing using genetic expression programming

AbstractFlood disasters continue to occur in many countries around the world and cause tremendous casualties and property damage. Flood peak values are required in the design of bridges, culvert waterways, spillways for dams, and estimation of scour at a hydraulic structure. When a flood wave passes through a reservoir its peak is attenuated, and the time base is enlarged due to effect of storage. Modification in the hydrograph is studied through flood routing. Flood routing is the technique of determining the flood hydrograph at a section of a river by utilising the data of flood flow at one or more upstream sections. Advances in the field of artificial intelligence (AI) offer opportunities for utilising new algorithms and models. Experiments have shown that the gene‐expression programming (GEP) model and its algorithms are 100–60 000 times faster than older genetic algorithms. This study presents the GEP method, which is an extension to genetic programming, as an alternative approach to modelling of flood routing in natural channels. Thus, novel models for prediction of flood routing using the GEP technique are developed. The proposed GEP‐based approach is compared with three hydrograph examples and the results of other solution techniques. The GEP method makes use of few hydrologic parameters (inflow, outflow, and time). The performance of the models is evaluated by two goodness‐of‐fit measures, namely, the root‐mean‐square error (RMSE) and the coefficient of determination (R2). In the results, the GEP models show superior performance to the other solution techniques based on the Muskingum model.

Intelligent analysis of global warming effects on sea surface temperature in Hormuzgan Coast, Persian Gulf

As scientific and economic interests in climate prediction and predictability have increased considerably in recent years, a need for global sea surface temperature (SST) prediction for use in global forecasts for climate variability studies have emerged. This paper examines the potential of gene expression programming (GEP) in estimation of SST from global mean temperature (GMT) in comparison with linear regression (LR), polynomial regression (PR) and exponential regressions (ER). In the present study, GMT and SST data for the point with latitude of 26.5°N and longitude of 56.5°E in the coast of Hormuzgan in the Persian Gulf were gathered for the time period of 1985 to 2010. 60% of datasets (1985-2000) were used for training of mentioned models and the residual 40% of data (2000-2010) were used for testing. Finally, the performance of the GEP was compared with regression models using some statistic parameters for error estimation. The results of GEP were satisfactory, with values of the correlation coefficient of (R) of 0.945 and root mean square error (RMSE) of 0.135 for independent test set. The performance of examined GEP method was superior in comparison with regression models that were developed in parallel (R for regression models ranging between 0.330 and 0.623 and RMSE ranging between 0.361 and 0.462). The comparison test results reveal that by using GEP method, SST can be predicted, precisely.