Optimized Back Propagation Neural Network Research Articles

Predicting the Growth Value of Technology Enterprises with an Optimized Back-propagation Neural Network

Predicting the Growth Value of Technology Enterprises with an Optimized Back-propagation Neural Network

Optimized bp Neural Network Based on Improved Dung Beetle Optimization Algorithm to Predict High-Performance Concrete Compressive Strength

In modern architecture, the structural safety of buildings largely depends on the compressive strength of high-performance concrete (HPC), which is determined by the complex nonlinear relationships between its components. In order to more accurately forecast HPC’s compressive strength, this paper proposes a prediction model based on an improved dung beetle optimization algorithm (OTDBO)-optimized backpropagation neural network (BPNN). Extreme Gradient Boosting (XGBoost) is employed to determine the inputs for the BPNN, enhancing the computational efficiency under high-dimensional data feature conditions. To address the issues of local optima entrapment and slow convergence in the dung beetle optimization algorithm (DBO), four improvements were made to enhance its performance. In the initial population generation stage, the optimal Latin hypercube method was used to increase the population diversity. In the rolling stage, the osprey optimization algorithm’s global exploration strategy was introduced to improve the global search capability. The variable spiral search strategy was employed in the reproduction stage, and an adaptive t-distribution perturbation strategy was combined in the foraging stage to enhance the algorithm’s adaptability and search efficiency. The improved dung beetle optimization algorithm (OTDBO) outperformed other algorithms in performance tests on the CEC2017 benchmark functions. In terms of predicting the compressive strength of HPC, the XG-OTDBO-BP model developed in this study outperformed models optimized by other algorithms in terms of fitting outcomes and prediction accuracy. These findings support the XG-OTDBO-BP model’s superiority in the compressive strength of HPC prediction.

Comparison of Positioning Error Prediction Results of Industrial Robots Based on Three Different Types of Neural Networks

ABSTRACTWith the increasing development of industry, the market demand for manufacturing has shifted to large‐scale customized production. This poses new challenges to the production flexibility of industrial robots. The offline programming method can perfectly meet this challenge. But its disadvantage is that it relies heavily on the absolute positioning accuracy of industrial robots. In recent years, there has been an increasing number of studies using neural networks (NN) to predict the positioning errors of industrial robots to improve their absolute positioning accuracy. However, most of these studies only focus on the application of NNs, and do not compare the prediction results and performance of different kinds of NNs. This paper selects three typical network models: backpropagation neural network (BPNN), particle swarm algorithm optimization BPNN (PSO‐BPNN), and radial basis function neural network (RBFNN). Through in‐depth experiments and analysis of these networks, the purpose is to reveal their respective prediction effects and characteristics and to summarize their advantages and disadvantages. Experimental results show that BPNN performs poorly in predicting positioning errors. As an optimization method, the particle swarm algorithm can effectively improve the prediction performance of BPNN. In contrast, the RBFNN performs well, which makes it very suitable for predicting the positioning error of industrial robots.

Optimization method of cable structure demolition driven by digital twin evolution model

The large-span cable structure has the problems of complex component connection, cumbersome demolition process and difficult process coordination, which lead to high construction safety risk and low efficiency. Moreover, the optimization of the construction process scheme is difficult, and the convergence and visualization of the finite element analysis are insufficient, which hinders the construction efficiency. This study proposes a cable structure demolition optimization method driven by the digital twin (DT) evolution model. From the perspective of virtual-real mapping, the DT evolution process of cable structure demolition is clarified. The virtualization method of physical demolition process is formed by integrating visual modeling, simulation modeling and data modeling. Taking full account of the temporal and spatial evolution of the demolition process, a DT evolution model of the demolition process is established. Starting from the information flow, component flow, control flow and their relationship, the problems existing in the process of demolition execution are analyzed. Driven by the twin evolution process, a data mapping method for the demolition process is proposed. The safety risk is characterized by the mechanical properties (MPs) of the structure, and the expression language of the data association relationship is established. The particle swarm optimization back propagation neural network (PSO-BPNN) is improved, and an intelligent analysis method of MPs in the whole process of demolition is proposed. The best construction scheme is obtained by integrating the changes in MPs between the various demolition steps. The information of the demolition evolution process is divided into functional modules, and a DT platform is developed to efficiently guide the actual construction. Taking the demolition process of a cable truss experimental model as an example, the effectiveness of the proposed method is verified. The research results show that the DT evolution model can accurately map the actual demolition process. The deep learning method analyzes the mechanical properties of the structure and obtains the best construction scheme quickly and accurately. The development and application of the twin platform improves the degree of visualization and realizes the efficient guidance of the virtual model to the actual construction.

Research on Failure Pressure Prediction of Water Supply Pipe Based on GA-BP Neural Network

The water supply pipeline is regarded as the “lifeline” of the city. In recent years, pipeline accidents caused by aging and other factors are common and have caused large economic losses. Therefore, in order to avoid large economic losses, it is necessary to analyze the failure prediction of pipelines so that the pipelines that are going to fail can be replaced in a timely manner. In this paper, we propose a method for predicting the failure pressure of pipelines, i.e., a genetic algorithm was used to optimize the weights and thresholds of a BP neural network. The first step was to determine the topology of the neural network and the number of input and output variables. The second step was to optimize the weights and thresholds initially set for the back propagation neural network using a genetic algorithm. Finally, the optimized back-propagation neural network was used to simulate and predict pipeline failures. It was proved by examples that compared with the separate back propagation neural network model and the optimized and trained genetic algorithm-back propagation neural network, the model performed better in simulation prediction, and the prediction accuracy could reach up to 91%, whereas the unoptimized back propagation neural network model could only reach 85%. It is feasible to apply this model for fault prediction of pipelines.

Fault Diagnosis of Distributed Energy Distribution Network Based on PSO-BP

With the increasing scale of distribution network at distribution time, its complexity grows geometrically, and its fault diagnosis becomes more and more difficult. Aiming at the slow convergence and low accuracy of traditional backpropagation neural network in dealing with single-phase ground faults, the study proposes a backpropagation neural network based on improved particle swarm optimization. The model optimizes the weights and acceleration constants of the particle swarm algorithm by introducing dynamic coefficients to enhance its global and local optimization seeking ability. It is also applied in optimizing the parameters of backpropagation neural network and constructing the routing model and ranging model for fault diagnosis about distributed energy distribution network. The simulation results revealed that the maximum absolute error of the improved method is 0.08. While the maximum absolute errors of the traditional backpropagation neural network and the particle swarm optimized backpropagation neural network were 0.65 and 0.10, respectively. The fluctuation of the relative errors of the research method was small under different ranges of measurements. At 8.0 km, the minimum relative error was 0.39% and the maximum relative error was 2.81%. The results show that the improved method proposed in the study significantly improves the accuracy and stability of fault diagnosis and localization in distribution networks and is applicable to complex distribution network environments. The method has high training efficiency and fault detection capability and provides an effective tool for distribution network fault management.

Productivity Prediction Model of Tight Oil Reservoir Based on Particle Swarm Optimization–Back Propagation Neural Network

Single-well productivity is a crucial metric for assessing the effectiveness of petroleum reservoir development. The accurate prediction of productivity is essential for achieving the efficient and economical development of oil–gas reservoirs. Traditional productivity prediction methods (empirical formulae and numerical simulation) are limited to specific reservoir types. There are few influencing factors, and a large number of ideal assumptions are made for the assumed conditions when predicting productivity. The application scenario is ideal. However, in tight oil reservoirs, numerous factors affect productivity, and their interactions exhibit significant complexity. Continuing to use traditional reservoir productivity prediction methods may result in significant calculation errors and lead to economic losses in oilfield development. To enhance the accuracy of tight reservoir productivity predictions and achieve economical and efficient development, this paper investigates the tight reservoir in the WZ block of the Beibuwan area, considering the impact of geological and engineering factors on productivity; the random forest tree and Spearman correlation coefficient are used to analyze the main influencing factors of productivity. The back propagation neural network optimized by particle swarm optimization was employed to develop a productivity prediction model (PSO-BP model) for offshore deep and ultra-deep tight reservoirs. The actual test well data of the oilfield are substituted into this model. The analysis results of the example application yielded an RMSE of 0.032, an MAE of 1.209, and an R2 value of 0.919. Compared with traditional productivity prediction methods, this study concludes that the model is both reasonable and practical. The calculation speed is faster and the calculation result is more accurate, which can greatly reduce productivity errors. The model constructed in this paper is well suited for predicting the productivity of tight reservoirs within the WZ block. It offers substantial guidance for predicting the productivity of similar reservoirs and supports the economical and efficient development of petroleum reservoirs.

BP neural network multi-module green roof thermal performance prediction model optimized based on sparrow search algorithm

Green roofs, as a passive building energy-saving technology, offer a range of benefits including reduced energy consumption and mitigation of the urban heat island effect. To analyze the thermal performance of green roofs more accurately and efficiently, this study introduces a novel predictive model that utilizes a Sparrow Search Algorithm (SSA) optimized Backpropagation Neural Network (BPNN) for forecasting the thermal performance of green roofs in subtropical regions. Compared to traditional BPNN, the SSA-BPNN model enhances the network's convergence and generalization capabilities by optimizing initial weights and thresholds. The model, established using meteorological data from Guangzhou, China, and simulated green roof temperature data, underwent rigorous training and validation. The results indicate that predictive accuracy has significantly improved across all seasons, with substantial reductions in all 12 evaluation metrics, and the model GR3 (SSA-BP) emerging as the optimal model. Particularly in summer, the SSA-BPNN model exhibits the highest accuracy, underscoring its effectiveness for subtropical climates. The findings suggest that the SSA-BPNN model is a powerful tool for sustainable urban planning and green roof design in subtropical regions, marking a significant advancement in the integration of artificial intelligence in environmental engineering.

Prediction Model of Diabetes Complications Based on Genetic Engineering Improved Genetic Algorithm Optimized BP Neural Network

Nowadays, with the rapid development of science and technology, people’s living standards have been greatly improved, and many chronic diseases have also been brought, including diabetes. The occurrence of diabetes not only poses a serious threat to human body, but also poses a threat to human life with its development. BP (Back Propagation) neural network model can well solve the logic regression problem of single factor and multiple factors, and also better solve the collinearity problem of multiple factors. BP neural network optimized based on improved genetic algorithm can reflect the influence mode and degree of various factors, and can be predicted from the perspective of patients’ diet, exercise, and doctors’ application of insulin. In this paper, the patients in a hospital were taken as the research object, and the BP neural network method was used to analyze the causes of the disease. The prediction model was used to screen out the high-risk groups of diabetes patients and reduce their incidence rate. Secondly, according to the collected data, the relationship between diabetes related complications was analyzed in depth by using genetic engineering technology, thus providing a theoretical basis for the prevention and treatment of diabetes and its complications. The prediction accuracy of BP neural network optimized by genetic algorithm can reach 94.1%. In the high-risk group of diabetes, taking appropriate diet and behavioral measures can reduce the probability of diabetes. The prediction scheme of diabetes complications proposed in this paper is simple and its cost is low, which can greatly reduce the cost of prevention and treatment of diabetes and the probability of diabetes.

The surrogate model for short-term extreme response prediction based on ANN and Kriging algorithm

Numerical simulation is a common method for calculating the short-term extreme response of floating offshore wind turbines (FOWTs). However, it requires significant computational resources. This study presents a dynamic response database for a 5MW semi-submersible FOWT under complex environmental conditions, including wind speed, effective wave height, and wave spectral peak period, using a numerical model. The peak over threshold (POT) method can be used to obtain the parametric database of short-term extreme responses, which includes the short-term extreme response distribution parameters for four responses: float surge, mooring tension, outward bending moment at the leaf root surface (OoPBM) and tower base pitching moment (TBPM). And the parameter database is applied to train models such as the Genetic Algorithm optimization Back Propagation neural network (GA-BP) and Kriging algorithm models. The research indicates that a correlation can be established between environmental conditions and short-term extreme response parameters using two algorithms. The accuracy of surrogate model prediction for some parameters can be improved by grouping the data based on wind speed and training separately. Additionally, selecting the appropriate surrogate model for each parameter separately can improve the accuracy of short-term extreme response prediction.

Prediction of flash flood peak discharge in hilly areas with ungauged basins based on machine learning

ABSTRACT Peak discharge is an essential element of hydrological forecasting. Due to rapid outbreaks of flash floods in hilly areas and the lack of measured data, the fast and accurate estimation of peak discharge is crucial for flash flood hazard management. Three machine learning algorithms were applied to estimate peak discharge; this estimation was compared with the results of hydrological–hydraulic models, and the results were verified with measured watershed data. In this paper, 10 hydrological and geomorphological parameters were selected to predict the flood peak discharge in 103 watersheds in Taiyi Mountain North District. The results show that the particle swarm optimization backpropagation (PSO-BP) neural network model outperforms the BP neural network and random forest regression in prediction performance. PSO-BP has a lower mean absolute error (2.51%), root mean square error (3.74%), and mean absolute percentage error (2.74%) than the other models, which indicates that PSO-BP has high prediction accuracy. Importance analysis revealed that rainfall, early impact rainfall, catchment area, and rain intensity are the key input parameters of PSO-BP. The proposed method was confirmed to be a fast and relatively accurate algorithm for estimating the peak discharge of flash floods in ungauged basins.

Research on gravity compensation control of BPNN upper limb rehabilitation robot based on particle swarm optimization

AbstractA four‐degree‐of‐freedom upper limb exoskeleton rehabilitation robot system with a gravity compensation device is constructed. The objective is to address the rehabilitation training needs of patients with upper limb motor dysfunction. A BP neural network adaptive control method based on particle swarm optimization is proposed. First, the degrees of freedom of the human body are analyzed, and a Lagrange method is employed to construct a dynamic model. Second, a particle swarm optimization back propagation neural network adaptive control algorithm based on particle swarm optimization is presented. Subsequently, the range of motion of the upper limbs is analyzed with reference to muscle anatomy and a three‐dimensional motion capture system. And the robot structure design is analyzed in detail. Finally, simulation experiments were conducted, and the results demonstrated that the proposed method exhibited high effectiveness and accuracy.

Stability prediction of roadway surrounding rock using INGO-RF

In order to more accurately classify the stability of roadway surrounding rock and identify dangerous areas in a timely manner to prevent roadway collapse and other disasters, an INGO-RF roadway surrounding rock stability prediction model is proposed. This model combines the improved Northern Gok algorithm (INGO) and Random Forest (RF) based on the analysis of influencing factors of roadway surrounding rock stability. First, three strategies were employed to enhance the Northern Gob algorithm (NGO): Logistic chaotic mapping, refraction reverse learning, and improved sine and cosine. Subsequently, INGO was utilized to optimize the number of decision trees and the minimum number of leaf nodes for RF species in order to improve the prediction accuracy of RF. Secondly, a data set consisting of 34 groups of roadway surrounding rock data is selected. The input indexes of the model include roof strength, two-wall strength, floor strength, burial depth, roadway pillar width, ratio of direct roof thickness to mining height, and surrounding rock integrity. Meanwhile, surrounding rock stability is considered as the output index. Particle swarm optimization backpropagation neural network (PSO-BPNN), genetic algorithm optimization support vector machine (GA-SVM), Sparrow search algorithm optimization RF (SSA-RF) models were introduced to compare the prediction results with the INGO-RF model, and the results showed that: INGO-RF model has the best performance in the comparison of various performance indicators. Compared with other models, the accuracy rate (Ac) in the test set has increased by 0.12∼0.4, the accuracy rate (Pr) has increased by 0.07∼0.65, and the recall rate (Re) has increased by 0.08∼0.37. The harmonic mean (F1-Score) of the recall rate increased by 0.08∼0.52, the mean absolute error (MAE) decreased by 0.1428∼0.4285, the mean absolute percentage error (MAPE) decreased by 7.15%∼28.57%, and the root mean square error (RMSE) decreased by 0.1565∼0.3779. Finally, the data on surrounding rock conditions of roadways in multiple mining areas in Shanxi Province were collected to test the INGO-RF model. The results indicate that the predicted outcomes closely align with the actual results, demonstrating a certain level of reliability and stability. It shows that it can better meet the practical needs of engineering and avoid the occurrence of mine disasters.

Inverse Kinematics of Large Hydraulic Manipulator Arm Based on ASWO Optimized BP Neural Network

In order to solve the problem of insufficient end positioning accuracy due to factors such as gravity and material strength during the inverse solution process of a large hydraulic robotic arm, this paper proposes an inverse solution algorithm based on an adaptive spider wasp optimization (ASWO) optimized back propagation (BP) neural network. Firstly, the adaptability of the SWO algorithm is enhanced by analyzing the phase change in population fitness and dynamically adjusting the trade-off rate, crossover rate, and population size in real time. Then, the ASWO algorithm is used to optimize the initial weights and biases of the BP neural network, effectively addressing the problem of the BP neural network falling into local optima. Finally, a neural network mapping relationship between the actual position of the robotic arm’s end-effector and the corresponding joint values is established to reduce the influence of forward kinematic errors on the accuracy of the inverse solution. Experimental results show that the average positioning error of the robotic arm in the XYZ direction is reduced from (91.3, 87.38, 117.31) mm to (18.16, 24.67, 27.21) mm, significantly improving positioning accuracy by 80.11%, 71.78%, and 76.81%, meeting project requirements.

Optimization of a radial inflow turbine rotor with splitter blades based on entropy production theory and artificial neural network

The radial inflow turbine is the core component of the organic Rankine cycle, and the application of splitter blades is an effective way to reduce its flow loss. It is helpful to study the energy loss characteristics of radial turbines with splitter blades and the optimal geometrical parameters of splitter blades for further improving their aerodynamic performance. In this paper, an optimization framework based on the combination of back-propagation neural network and non-dominated sorting genetic algorithm II is developed. Considering the turbine efficiency as the objective, the geometric parameters of splitter blades, including relative length of the blade, relative distance of the leading edge, circumferential offset, and outlet angle of the blade, are optimized. In addition, the entropy production method is used to locate and quantitatively evaluate the energy loss of the radial turbine with the optimal splitter blades. The results indicate that the genetic algorithm optimized back-propagation neural network model can accurately predict turbine efficiency with a maximum deviation of 2.47 %. In addition, the optimal geometrical parameters of the splitter blade are related to the turbine geometry. In this case, the optimal splitter blade has a relative blade length of 0.639, a relative leading edge distance of 0.3, a circumferential offset of 0.442, and an outlet angle of 38°. Compared with the original turbine, the efficiency of the radial turbine with optimal splitter blades increases by 4.65 %. The total entropy production of rotor and turbine is reduced by 33.3 % and 44.1 % respectively by the installation of optimal splitter blades. Compared with the case without splitter blades, the turbine with the optimal splitter blades can achieve higher efficiency at a flow rate 40 % higher than the design value, and it expands the efficient working range of the turbine by 60 %. This work can provide guidance for the optimization design of radial inflow turbines.

Position prediction of underwater gliders based on a new heterogeneous model ensemble method

An underwater glider (UG) is a promising unmanned underwater vehicle driven by buoyancy that can perform long-period and large-scale ocean observations. Due to the lack of inertial sensors and the low-speed relative to the ocean current, it is difficult to determine the exact position of the UG. Besides, the position of the UG is of great significance for acoustic applications. Therefore, this paper proposes a new heterogeneous model ensemble method (HMEM) based on the whale algorithm (WOA) optimized backpropagation neural network with complement ensemble empirical mode decomposition (CEEMD-BPNN) and the physical model. The physical model can deduce the underwater position of the UG considering the ambient environment, and the WOA-CEEMD-BPNN is adopted to predict the surface position with historical data. Then, the new heterogeneous model ensemble method integrates these two with an unscented Kalman filter according to the predicted surface position. To evaluate the performance of the HMEM, the evaluation index is introduced, and the sea trial conducted by the Petrel-II glider is used as a case study. The experiment demonstrates that the MAE and RMSE of the HMEM are 0.51 km and 0.76 km, respectively, and the mean time difference is 3.08 × 102 s, which outperforms the dynamic model and can achieve accurate position prediction for long periods.

Long-term water demand forecasting using artificial intelligence models in the Tuojiang River basin, China.

Accurate forecasts of water demand are a crucial factor in the strategic planning and judicious use of finite water resources within a region, underpinning sustainable socio-economic development. This study aims to compare the applicability of various artificial intelligence models for long-term water demand forecasting across different water use sectors. We utilized the Tuojiang River basin in Sichuan Province as our case study, comparing the performance of five artificial intelligence models: Genetic Algorithm optimized Back Propagation Neural Network (GA-BP), Extreme Learning Machine (ELM), Gaussian Process Regression (GPR), Support Vector Regression (SVR), and Random Forest (RF). These models were employed to predict water demand in the agricultural, industrial, domestic, and ecological sectors using actual water demand data and relevant influential factors from 2005 to 2020. Model performance was evaluated based on the Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE), with the most effective model used for 2025 water demand projections for each sector within the study area. Our findings reveal that the GPR model demonstrated superior results in predicting water demand for the agricultural, domestic, and ecological sectors, attaining R2 values of 0.9811, 0.9338, and 0.9142 for the respective test sets. Also, the GA-BP model performed optimally in predicting industrial water demand, with an R2 of 0.8580. The identified optimal prediction model provides a useful tool for future long-term water demand forecasting, promoting sustainable water resource management.

Optimizing Active Disturbance Rejection Control for a Stubble Breaking and Obstacle Avoiding Control System

In order to improve the obstacle avoidance control performance and anti-interference ability of a stubble breaking device of a no-tillage planter, a back-propagation neural network (BPNN)-optimized fuzzy active disturbance rejection control (ADRC) controller was designed to optimize the control performance of a servo motor. Firstly, a negative feedback mathematical model was established for the obstacle avoidance control system. Then, the nonlinear state error feedback (NLSEF) parameters in the fuzzy ADRC were intelligently optimized by the BPNN algorithm. In this way, a fuzzy ADRC controller based on BPNN optimization was formed to optimize the control process of a servo motor. Matlab/Simulink (R2022b) was used to complete the simulation model design and parameter adjustment. Consequently, the response time was 0.089 s using the BPNN fuzzy ADRC controller, which was shorter than the 0.303 s of the ADRC controller and the 0.100 s of the fuzzy ADRC controller. The overshoot was 0.1% using a BPNN fuzzy ADRC controller, which was less than the 2% of the ADRC controller and the 1% of the fuzzy ADRC controller. After noise signal interference was introduced into the control system, the regression steady state time of the BPNN fuzzy ADRC controller was 0.22 s, which was shorter than the 0.56 s of the ADRC controller and the 0.45 s of the fuzzy ADRC controller. A hardware-in-the-loop simulation experimental platform of the obstacle avoidance control system was constructed. The experiment results show that the servo motor control system has a fast dynamic response, small steady-state error and strong anti-interference ability for obstacle avoidance at the target height. Then, the control system error was within the allowable range. The servo motor control effect of the BPNN fuzzy ADRC was better than the ADRC and fuzzy ADRC. This optimized servo motor control method can provide a reference for improving the obstacle avoidance control effect problem of no-tillage seeders in stubble breaking operations on rocky desertification areas.

Unveiling land use-carbon Nexus: Spatial matrix-enhanced neural network for predicting commercial and residential carbon emissions

Carbon emissions play a pivotal role in driving global warming, with commercial and residential sectors ranking as the third-largest source after industrial and transportation sectors. Consequently, establishing a framework for spatial characterizing and predicting carbon emissions from these sectors is vital for effective carbon reduction planning. This study proposes an optimized Backpropagation Neural Network, incorporating a Spatial Weight Matrix, to model the relationship between land use and carbon emissions, taking spatial effects into account. Additionally, an optimized Random Forest model with variable search step lengths is developed for large-scale and multi-class land use prediction. Beijing, China serves as the study site, yielding the following findings: (1) The optimized Backpropagation Neural Network outperforms traditional models, achieving goodness of fit of 0.999 and 0.997 in training and testing datasets, respectively. (2) The optimized Random Forest model, while reducing prediction precision for ecological areas (Root Mean Square Error increases by 4.12 %), enhances overall model performance by 73.95 % in a single iteration. (3) By 2025, commercial and residential carbon emissions in Beijing are projected to reach 8.91 million tons. (4) The study characterizes the spatial patterns of commercial and residential carbon emissions at a 1 km resolution for the next 15 years, revealing statistical features. (5) Expansion of mixed commercial and residential land relies on service-oriented development, positively impacting regional carbon emission intensity control.

Optimizing reverse osmosis desalination from brackish waters: Predictive approach employing response surface methodology and artificial neural network models

A mathematical model frame has been developed to optimize the performance of reverse osmosis (RO) desalination in brackish waters with medium to high salt concentrations using response surface methodology (RSM) and artificial neural network (ANN), specifically catered towards efficiently operating RO desalination systems. The RSM model aims at predicting RO performance with second-order polynomials based on 15 experimental datasets. Inputs for RSM model include feed TDS concentration, feed water flow rate, and concentration-to-desalination ratio, while the outputs consider rejection, membrane flux, performance index, and specific energy consumption. The Genetic Algorithm optimized back-propagation Neural Network was employed to update network weights and biases, leading to an optimal ANN network structure of 3–10-4. A ratio of 70 % training, 15 % validation, and 15 % test data was chosen to establish an ANN model that predicts the performance of RO desalination. The ANN model demonstrated significantly higher prediction accuracy than the RSM model, with impressive correlation coefficients of 0.9936 for the performance index and 0.9932 for SEC prediction, and low RMSE of 0.85 for rejection. To maximize the performance index and minimize specific energy consumption, a two-step optimization method was further set up by applying the RSM model to initially determine optimal process conditions followed by the ANN model to enhance the accuracy of performance forecasts. Optimal process conditions were proposed for specific feed TDS scenarios, which involved setting a feed TDS concentration limit to meet drinking water standards.