Genetic BP Algorithm Research Articles

Prediction of COD Degradation in Fenton Oxidation Treatment of Kitchen Anaerobic Wastewater Based on IPSO‐BP Neural Network

Kitchen anaerobic wastewater contains a high concentration of insoluble organic matter, and the degradation of organic matter in the wastewater is the key to treating kitchen anaerobic wastewater. The Fenton oxidation process is used to treat kitchen anaerobic wastewater, and the effects of H2O2 dosage, Fe2+ dosage, reaction time and pH value on chemical oxygen demand (COD) degradation efficiency are explored. The improved particle swarm optimization (IPSO) algorithm is used to optimize the back propagation (BP) neural network, and a prediction model of COD degradation is established based on IPSO‐BP neural network. H2O2 dosage, Fe2+ dosage, reaction time and pH value are selected as the main influencing factors of the COD degradation, and 30 groups of experimental data are selected to train the IPSO‐BP neural network. The results predicted by the trained IPSO‐BP neural network on 10 groups of test data are compared with the actual values, and the results predicted by BP model and genetic algorithm‐BP (GA‐BP) model are compared. The IPSO‐BP model has the highest fitting accuracy. The root mean square error (RMSE), RMSE coefficient of variation (CV‐RMSE), and coefficient of determination (R2) were used to evaluate the prediction performance of the model. The indicators of the IPSO‐BP model were significantly better than those of the GA‐BP model and the BP model, indicating that IPSO‐BP model had better generalization ability and could predict COD concentration more effectively. The optimal conditions for Fenton oxidation obtained from IPSO‐BP model are as follows: the dosage of H2O2 is 2000 mg/L, the dosage of Fe2+ is 1200 mg/L, the pH value is 2.8, the reaction time is 75 min, and the COD removal rate is 80.125%, which is consistent with the experimental results. Through gas chromatography‐mass spectrometry (GC‐MS) analysis, most organic compounds in kitchen anaerobic wastewater are oxidized and decomposed, indicating that the IPSO‐BP model has good predictive quality.

Multi-uncertainty river health assessment in the Loess Plateau based on a novel comprehensive similarity cloud model

With the development of related research on river health assessment, the traditional assessment methods cannot solve the uncertainty problems caused by diversified indicators and ambiguous weights. In order to comprehensively understand the river health status in the Loess Plateau and solve the uncertainty of assessment results, this study takes Wuding River as the research object, and constructs an assessment indicator system incorporating the ecological environment and social functions of the river. The genetic algorithm-BP neural network was used to establish a weight learning mechanism for assessment indicators, and the system weight was calculated by combining various weighting preferences. Based on the cloud model theory, a sample data cloud and a health level cloud were constructed, and a comprehensive similarity model considering the shape and distance of the cloud diagram was developed to identify the health level of Wuding River. The spatial and temporal variability of the river health status and the sources of pressure causing unhealthy problems were analyzed. It is shown that Wuding River is in basic health status, which gradually decreases from the upstream reaches to the downstream reaches. The health status of social service functions is better than that of the ecological environment attributes such as water resources and water quality. The sources of health pressure in the upstream and downstream reaches are different. In the upstream reaches, the pressure mainly comes from natural attributes such as successive declines in water production, slow governance of water and soil erosion, and the poor stability of the river. The health pressure in the midstream and dowmstream reaches comes from the interference of human economic activities, which include deterioration of water quality caused by the discharge of fossil energy industry, destruction of aquatic habitats, and flood safety hazards caused by insufficient standards for flood control project design. In addition, SDSCM provides river managers with accurate and comprehensive health risk information to help them manage rivers more effectively.

Application of fund analysis model to identify key factors in construction projects

This study constructs a funnel analysis model by collecting relevant data to achieve fault monitoring. Back-propagation (BP) neural networks are also used to identify structural damage in construction projects, and a genetic algorithm (GA) is used to optimise BP to improve issues such as slow convergence and long time consumption. The results indicate that the difference between the third-order frequency and the first-order curvature mode is the most suitable indicator for damage warning and identification. The difference in the first-order curvature mode of adjacent measurement points of the damaged component increases with the increase in the degree of damage. Comparing the GA–BP neural network and BP neural network, the former has a smaller error in identification and better performance. The maximum and minimum relative errors of GA–BP in identifying the damage degree of the structure are 8.06 and 1.61%, respectively, meeting the accuracy requirements of the project. The identification of the key factors in construction projects based on the funnel analysis model is beneficial for identifying structural damage and ensuring the safety of engineering projects.

Welding process optimization for blast furnace shell by numerical simulation and experimental study

At cold temperatures, the stress after welding a blast furnace shell is not completely released because of the fast cooling speed, which results in cracks. To solve this problem, an integrated method combining gray relational analysis, back propagation (BP) neural network, and genetic algorithm-BP (GA-BP) was proposed in this study. Its aim was to analyze the dominant mechanism of residual stress in multi-layer and multi-pass welding process under the heating of ceramic sheet and achieve super-objective collaborative optimization of residual stress. GRA indicated that through the correlation analysis of welding parameters, residual stress, maximum temperature of characteristic point, and deformation, welding speed was determined to be an all-important factor affecting welding quality. A BP prediction model of residual stress was established. The predicted value was in good agreement with the experimental value, and the maximum error was 8.75%. The influence of each factor on the quality of welding seam is not a single, but a complex mechanism of mutual influence. GA-BP determined the combination of low level welding speed 28 mm/min, medium level welding current 210A and medium and low level welding voltage 21V as the optimal solution of low residual stress. The maximum prediction error of residual stress was 13.76%. The experimental results show that the maximum prediction error of residual stress is 4.4%, and the optimized weld has better macro and micro structure, which proves the feasibility of the multi-objective optimization method. Most importantly, the proposed method can effectively identify the optimal welding parameters, which can provide a reliable theoretical basis for engineering practice.

A Novel Fusion Method for State-of-Charge Estimation of Lithium-Ion Batteries Based on Improved Genetic Algorithm BP and Adaptive Extended Kalman Filter

The lithium-ion battery is the power source of an electric vehicle, so it is of great significance to estimate the state of charge (SOC) of lithium-ion batteries accurately to ensure vehicle safety. To improve the accuracy of the parameters of the equivalent circuit model for batteries, a second-order RC model for ternary Li-ion batteries is established, and the model parameters are identified online based on the forgetting factor recursive least squares (FFRLS) estimator. To improve the accuracy of SOC estimation, a novel fusion method, IGA-BP-AEKF, is proposed. Firstly, an adaptive extended Kalman filter (AEKF) is used to predict the SOC. Then, an optimization method for BP neural networks (BPNNs) based on an improved genetic algorithm (IGA) is proposed, in which pertinent parameters affecting AEKF estimation are utilized for BPNN training. Furthermore, a method with evaluation error compensation for AEKF based on such a trained BPNN is proposed to enhance SOC evaluation precision. The excellent accuracy and stability of the suggested method are confirmed by the experimental data under FUDS working conditions, which indicates that the proposed IGA-BP-EKF algorithm is superior, with the highest error of 0.0119, MAE of 0.0083, and RMSE of 0.0088.

Application and Comparison of Different Models for Quantifying the Aquatic Community in a Dam-Controlled River.

In order to develop a better model for quantifying aquatic community using environmental factors that are easy to get, we construct quantitative aquatic community models that utilize the different relationships between water environmental impact factors and aquatic biodiversity as follows: a multi-factor linear-based (MLE) model and a black box-based 'Genetic algorithm-BP artificial neural networks' (GA-BP) model. A comparison of the model efficiency and their outputs is conducted by applying the models to real-life cases, referring to the 49 groups of seasonal data observed over seven field sampling campaigns in Shaying River, China, and then performing model to reproduce the seasonal and inter-annual variation of the water ecological characteristics in the Huaidian (HD) site over 10 years. The results show that (1) the MLE and GA-BP models constructed in this paper are effective in quantifying aquatic communities in dam-controlled rivers; and (2) the performance of GA-BP models based on black-box relationships in predicting the aquatic community is better, more stable, and reliable; (3) reproducing the seasonal and inter-annual aquatic biodiversity in the HD site of Shaying River shows that the seasonal variation of species diversity for phytoplankton, zooplankton, and zoobenthos are inconsistent, and the inter-annual levels of diversity are low due to the negative impact of dam control. Our models can be used as a tool for aquatic community prediction and can become a contribution to showing how quantitative models in other dam-controlled rivers to assisting in dam management strategies.

Research on Securities Portfolio Model Based on Genetic Optimization Neural Network

Portfolio is an investment management concept different from individual asset management. This consideration leads to an interesting result, that is, investors should buy a variety of securities at the same time instead of one kind of securities for diversified investment. Aiming at the limitations of BPNN (BP neural network) in traditional artificial neural network and its shortcomings such as many iterations, low convergence accuracy, and poor generalization, a portfolio method based on GA_BPNN (Genetic Optimization Neural Network) was proposed. The setting of GA (genetic algorithm) parameters and BPNN parameters are discussed in detail, and the implementation steps of genetic BP algorithm are described. The results show that the evaluation indexes of GA_BPNN prediction model are obviously better than those of the comparison prediction model, with the coincidence rate of 77.96% and the average absolute error of 12.451. The combination of GA and BPNN can effectively solve this problem. The simulation results of optimizing securities portfolio show that its optimization scheme is better than quadratic programming method, and this method is more correct, efficient, and practical.

Using Genetic Algorithm and Particle Swarm Optimization BP Neural Network Algorithm to Improve Marine Oil Spill Prediction

Using Genetic Algorithm and Particle Swarm Optimization BP Neural Network Algorithm to Improve Marine Oil Spill Prediction

Prediction of Urban Scale Expansion Based on Genetic Algorithm Optimized Neural Network Model

With the continuous development of urbanization, the urban population is becoming more and more dense, and the demand for land is becoming more and more tense. Urban expansion has become an indispensable part of urban development. This paper studies the optimization of neural network structure by genetic algorithm, puts forward the prediction model of urban scale expansion based on a genetic algorithm optimization neural network, and compares the performance of the model with the basic model. A genetic algorithm BP neural network (GA-BP) optimized by the genetic algorithm is used to shorten the running time of the algorithm and improve the prediction accuracy, but it is easy to fall into local solution. The genetic algorithm is improved by immune cloning algorithm, and the CGA-BP neural network model is established to obtain the global optimal solution. Compared with the BP neural network model and GA-BP neural network model, the CGA-BP neural network model converges faster, and the training times reach the error condition after 79 times, while the BP neural network model and GA-BP neural network model need 117 times and 100 times, respectively, and the fitness value corresponding to the number of iterations of the model is larger. Therefore, the CGA-BP neural network algorithm can make prediction more accurately and quickly and predict the expansion of urban scale through urban conditions.

Evaluation Model of Urban Smart Energy System Based on Improved Genetic Algorithm-Bp Neural Network

There are many problems such as strong subjectivity, complex calculations, and lack of intelligence in most of the current energy system evaluation models, so a design of an evaluation system for urban smart energy systems based on an improved genetic algorithm-back propagation (BP) neural network is proposed. First of all, the hierarchical structure of the indicator evaluation system for energy system was established, analytic hierarchy process (AHP) was used to assign weights to each indicator, and data samples were classified. Then, SeqGAN was used to expand the data set, which solved the difficult problem of data acquisition. Finally, the genetic algorithm was used to determine the best initial values of the weights and thresholds of the BP neural network structure parameters designed for this research. The simulation experiment results showed that the method in this paper has higher classification accuracy than the traditional method, and comprehensive evaluation model proposed can effectively evaluate the urban smart energy system.

A Wind Power Prediction Method Based on DE-BP Neural Network

With the continuous increase of installed capacity of wind power, the influence of large-scale wind power integration on the power grid is becoming increasingly apparent. Ultra-short-term wind power prediction is conducive to the dispatching management of the power grid and improves the operating efficiency and economy of the power system. In order to overcome the intermittency and uncertainty of wind power generation, this article proposes the differential evolution–back propagation (DE-BP) algorithm to predict wind power and addresses such shortcomings of the BP neural network as its falling into local optimality and slow training speed when predicting. In this article, the DE algorithm is used to find the optimal value of the initial weight and threshold of the BP neural network, and the DE-BP neural network prediction model is obtained. According to the data of a wind farm in Northwest China, the short-term wind power is predicted. Compared with the application of the BP model in wind power prediction, the results show that the accuracy of the DE-BP algorithm is improved by about 5%; compared with the genetic algorithm–BP model, the prediction time is shortened by 23.1%.

Effect of tempering drying on the physicochemical properties of paddy rice

To clarify the quantitative relationship between tempering drying parameters and physicochemical properties of paddy rice, as well as optimize the paddy rice drying process, a quadratic regression orthogonal rotational combination test was applied in this study. The initial tempering moisture content of paddy rice, tempering temperature, and tempering duration were test factors, and the amylose starch content, alkali spreading value, and gel consistency were test indicators. It is shown that under the conditions of the initial tempering moisture content of paddy rice was 20.5%, the tempering temperature was 43 °C, and the tempering duration was 3.39 h, the amylose starch content, alkali spreading value, and gel consistency were 17.83%, 7.08 grade and 86.4 mm respectively. The average error between the actual value and the optimized theoretical value was 4.06%, indicating that the regression model was accurate and the physicochemical properties of rice were improved under the optimal process conditions. Then, the visual prediction model for the physicochemical properties of rice was established based on BP (Back Propagation) and GA-BP (Genetic Algorithm-BP) neural networks respectively, and compared with the regression model established by Design-Expert software. The results showed that the root mean square error, mean absolute error, and mean absolute percentage error of GA-BP were lower.

Structural Optimization of Steel—Epoxy Asphalt Pavement Based on Orthogonal Design and GA—BP Algorithm

Fatigue cracks often occur in the deck asphalt pavement of steel bridges at the top of the longitudinal stiffening rib. To prevent this issue, the traditional design strategy of the steel bridge deck asphalt pavement structure was optimized, and a new approach is presented. This optimization technique exploits the strength simulation of the steel—epoxy asphalt pavement structure, and the stress concentration location is subsequently determined. A solid model of stress concentration including sensitive areas is then established. We examined the stress maximum point of the asphalt pavement layer at the top of the longitudinal stiffeners and the stress variation of the asphalt pavement layer at the top of the longitudinal stiffeners. To reduce the stress of the top pavement layer of the longitudinal stiffeners, an optimization method that combines orthogonal experimental design, neural network (BP), and genetic algorithm (GA) is presented. A design strategy for the steel—epoxy asphalt pavement structure and GA—BP optimization method was utilized to optimize the structure of the steel—epoxy asphalt pavement for Sutong Yangzi River Bridge. We confirmed that the presented approach improved fatigue reliability and established the efficacy of the design strategy and optimization method.

A New Neural Network with Genetic Algorithm in Searching Nonlinear Function Extremum

In order to more accurate for nonlinear function extreme, this paper used improved particle swarm optimization neural network combining with genetic algorithm method to solve the problem. In view of the particle swarm optimization algorithm is easy to appear “premature” faults, introducing the adaptive threshold, initializing particles if they were under the constraint conditions, making particles jump out to the optimal value of the position in previous search. Through the experiment, contrasts to the genetic neural network algorithm and traditional BP neural network, this method is faster in convergence and has the smallest prediction error. Finally, combining with genetic algorithm, calculating the extreme value of nonlinear function by using the above three kinds of neural network trained forecast as an individual output fitness value. The adaptive particle swarm optimization neural network proves the most close to the theoretical calculation. It shows that the method is effective.

Network Intrusion Detection Based on Sparse Autoencoder and IGA‐BP Network

Network intrusion detection system provides a better network security solution than other traditional network defense technologies. Aiming at the increasingly serious problem of Internet security in the big data environment, a network intrusion detection model based on autoencoder network model and improved genetic algorithm BP (IGA‐BP) network is constructed. In order to reduce the data dimension and eliminate redundant information, the autoencoder network model is firstly used to denoise and dedimension. A new population was formed by selecting some of the best parent individuals for cross mutation and replacing the worst parent individuals. The improved genetic algorithm and new population generation model will provide more reasonable initial parameters for BP network, namely, IGA‐BP network model. Based on IGA‐BP network model, the problems of slow detection rate and easy to get into local optimality in BP network are solved. The experiments were performed on KDD CUP99 dataset, which simulated different types of user organizations and different types of network intrusion. Compared with the existing intrusion detection methods, the experimental results show that the proposed method has a great effect on classification accuracy, false positives, and detection rate.

Remaining Useful Life Prediction and Fault Diagnosis of Rolling Bearings Based on Short-Time Fourier Transform and Convolutional Neural Network

Rolling bearings play a pivotal role in rotating machinery. The remaining useful life prediction and fault diagnosis of bearings are crucial to condition-based maintenance. However, traditional data-driven methods usually require manual extraction of features, which needs rich signal processing theory as support and is difficult to control the efficiency. In this study, a bearing remaining life prediction and fault diagnosis method based on short-time Fourier transform (STFT) and convolutional neural network (CNN) has been proposed. First, the STFT was adopted to construct time-frequency maps of the unprocessed original vibration signals that can ensure the true and effective recovery of the fault characteristics in vibration signals. Then, the training time-frequency maps were used as an input of the CNN to train the network model. Finally, the time-frequency maps of testing signals were inputted into the network model to complete the life prediction or fault identification of rolling bearings. The rolling bearing life-cycle datasets from the Intelligent Management System were applied to verify the proposed life prediction method, showing that its accuracy reaches 99.45%, and the prediction effect is good. Multiple sets of validation experiments were conducted to verify the proposed fault diagnosis method with the open datasets from Case Western Reserve University. Results show that the proposed method can effectively identify the fault classification and the accuracy can reach 95.83%. The comparison with the fault diagnosis classification effects of backpropagation (BP) neural network, particle swarm optimization-BP, and genetic algorithm-BP further proves its superiority. The proposed method in this paper is proved to have strong ability of adaptive feature extraction, life prediction, and fault identification.

An improved evolution algorithm using population competition genetic algorithm and self-correction BP neural network based on fitness landscape

The genetic algorithm-backpropagation neural network algorithm (GA-BP) makes full use of the advantages of genetic algorithm (GA) and BP neural network (BPNN). It has been widely used in practical problems, but it also has shortcomings such as slow convergence. Inspired by the generative adversarial network, a population competition mechanism (PCM) is proposed to improve the search ability of the GA, two populations are generated to compete, and the winning population can obtain the optimal individual from the failed population to ensure that the winning population has a sustainable advantage. The failed population will randomly generate new individuals and add new genes so as to get better individuals, through such a mechanism to ensure the rapid optimization of the entire population, avoid the risk of premature convergence, speed up the iterative speed and improve the stability of the GA. According to the characteristics of the fitness landscape, the learning rate of the BPNN is optimized, and it can change adaptively, which can effectively improve the network convergence speed and greatly reduce the time cost. We define the GA-BP that improves the learning rate based on fitness landscape as FL-GA-BP algorithm. On the basis of the FL-GA-BP algorithm, adding GA improved with PCM, we define this new algorithm as I-GA-BP algorithm, namely the I-GA-BP algorithm that combines PCM-improved GA and fitness landscape-improved BPNN. In this paper, we use two types of test functions with different characteristics and complexity to conduct experiments to verify the effectiveness of the I-GA-BP algorithm. By comparing the experimental data of the three algorithms GA-BP, FL-GA-BP and I-GA-BP, it is obtained that the I-GA-BP algorithm can better escape from the local optimal solution, which is more conducive to finding the global optimal solution. It also greatly improves the convergence speed of the neural network. Finally, we briefly discussed the effect of adjusting the number of neurons on the stability of the I-GA-BP algorithm.

A hybrid genetic–BP algorithm approach for thermal conductivity modeling of nanofluid containing silver nanoparticles coated with PVP

Since nanoparticles play a significant role in increasing the thermal conductivity of fluids, the present study aims to predict the thermal conductivity of silver nanofluid coated with polyvinylpyrrolidone (PVP) by the combinational model of multilayer perceptron artificial neural network and genetic algorithm. For modeling, the results of experimental measurements have been used for thermal conductivity of nanofluid containing PVP-coated silver nanoparticle-based deionized water at 25–55 °C in volume fraction of 250 ppm, 500 ppm and 1000 ppm. Henceforth, genetic algorithm is applied to improve learning process in the artificial neural network. It is in this way that the masses were chosen for each neuron’s communications as well as their bias happens according to optimization performed by the genetic algorithm. To evaluate the accuracy of the model in predicting thermal conductivity of nanofluid, mean absolute percentage error, root mean square error, coefficient of determination (R2) and mean bias error have exerted indices which are 1.202, 0.345, 0.989 and − 0.016, respectively. The results of the indices and predictions, compared to the experimental results, show high accuracy and reliable combinational model of artificial neural network and genetic algorithm.

Clustering of the body shape of the adult male by using principal component analysis and genetic algorithm–BP neural network

In order to improve the efficiency and accuracy of human body shape prediction, principal component analysis method (PCA) is proposed to reduce the dimension of related variables and eliminate the multicollinearity among variables. Then, the transformed variables are input into genetic algorithm and BP neural network, and a new method of human body shape prediction is designed. To avoid the problems that slow convergence speed and easy falling into local minima of BP neural network, the genetic algorithm is used to optimize the weights and thresholds of BP neural network. Moreover, to prove the superiority of PCA–GA–BP model, the prediction results are compared with those of other algorithms. Body sizes of 18–25-year-old, 26–44-year-old and 45–59-year-old males were selected as experimental data to analyze these models. The prediction results of GA–BP, PCA–BP, BP, SVM and K-means were compared with PCA–GA–BP neural network. The results show that the prediction effect of PCA–GA–BP neural network is significantly better than that of GA–BP, PCA–BP, BP, SVM and K-means prediction models, which can accurately predict and cluster the human body shape. The model has better prediction and classification and simpler structure.

Inversion Method of In-situ Stress and Rock Damage Characteristics in Dam Site Using Neural Network and Numerical Simulation—A Case Study

In mountain and ravine region of southwest China, several hydropower stations have been built in this area. It is the main national base for hydropower energy development, but under the influence of geological structure and surface erosion, the in-situ stress environment in this area is more complex. The main purpose of this paper is to establish the relationship between measured in-situ stress data and numerical calculation by means of neural network. Therefore, it can be used to analyze the distribution characteristics of large-scale in-situ stress and the failure of regional geological bodies under the action of in-situ stress in Xiluodu area. This paper analyzes the geological conditions and the measured data of in-situ stress in Xiluodu area. The results show that there is obvious surface weathering phenomenon in this area, the stress environment is complex. The depth has a positive impact on the level of in-situ stress, but the impact degree is different. The genetic algorithm-BP artificial neural networks (G-P) method is trained by the measured in-situ stress data. Based on the field measurement data, neural network algorithm and numerical simulation technology, the three-dimensional in-situ stress field distribution characteristics of the project area are carried out. The research shows that the scheme of combining the actual measurement, numerical analysis and neural network inversion is reliable; Depth is an important factor affecting the maximum horizontal stress value in Xiluodu area; Under the action of in-situ stress, the fissures of geological body are relatively developed, and three types of original water storage and flow spaces are formed before dam storage period; The flow and storage space of groundwater includes limestone fracture, basalt fracture and slope fracture aquifer; After the impoundment of the reservoir area, the reservoir water is transmitted downward through the vertical fissures, and finally the reservoir water forms a hydraulic connection with the groundwater. After impoundment, the hydrogeological conditions of the reservoir area would be changed.