Neural Network Optimization Algorithm Research Articles

Prediction of cementation factor for low-permeability Iranian carbonate reservoirs using particle swarm optimization-artificial neural network model and genetic programming algorithm

cementation factor is a crucial parameter that has a significant influence on the estimation of reservoir parameters. Laboratory measurements for cementation factor are available for occasional cases because experimental special core analyses for determination of cementation factor values are expensive and time-consuming. While this factor plays a significant role in determining water saturation, there is no comprehensive and precise relationship for the case of Iranian carbonate reservoirs. In this article, a unique model was used based on a powerful combination of artificial neural network (ANN) and particle swarm optimization (PSO) algorithm to model the cementation factor. In the second phase of simulation, a correlation for the cementation factor was discovered by genetic programming (GP) algorithm. Both the PSO-ANN model and GP algorithm are trained by input variables such as porosity, permeability, and grain density derived from 175 routine core analysis (RCAL) samples of 21 carbonated oil fields. To determine the relative impact of the independent variables on cementation factor the sensitivity analysis was carried out for both models. The comparison between the PSO-ANN model output and the experimental cementation factor data clearly demonstrated that the built model can predict the cementation factor with great precision; the mean square error between the model predictions and the experimental data was less than 0.07. The root mean square error of training and testing data sets for the new developed correlation using GP algorithm were 0.0902 and 0.0727 respectively. Finally, to evaluate the validity and reliability of the developed models, a comparison was implemented between these two models and other empirical models over an external employment data set (21 data point). This comparison revealed that the GP algorithm and PSO-ANN model deliver a higher performance capacity compared to other proposed correlations for predicting cementation exponent.

Parallel Multi-task Cascade Convolution Neural Network Optimization Algorithm for Real-time Dynamic Face Recognition

Parallel Multi-task Cascade Convolution Neural Network Optimization Algorithm for Real-time Dynamic Face Recognition

A new prediction method of seafloor hydrothermal active field based on wavelet neural network

The study of modern seafloor hydrothermal activity and its mineralization has become one of the focuses of global geoscience. Accurate prediction of possible seafloor hydrothermal active fields is the basis of all research work. The detecting method for new seafloor hydrothermal activity still is mainly dependent on marine site investigation. This includes a series of temperature, turbidity, and geochemical anomaly investigations using submarine cameras and manned diving investigation. Both of these require expensive financial, human, and material resources. In order to realize the accurate prediction of potential hydrothermal activity areas in a low cost manner, with strong pertinence, and a wide range, we propose a prediction method of seafloor hydrothermal active region based on Wavelet Neural Network. First, we integrated the hydrothermal position information from the InterRidge Vents Database with the hydrothermal temperature information form the Argo Database to construct a data set. Then, we combined wavelet analysis with an artificial neural network to create a wavelet neural network optimization algorithm. Finally, the temperature and salinity data were input to the wavelet neural network to predict the seafloor hydrothermal active region. Sevenfold cross validation was used to evaluate the performance of the model and 90.43% prediction accuracy was achieved. The results of experiments show that salinity is not related to the existence of hydrothermal activity fields but rather that the surrounding water temperature has a strong correlation with hydrothermal existence. Therefore, it is effectively feasible to use the wavelet neural network model with an input of seawater temperature to predict seafloor hydrothermal activity fields. Although artificial neural network cannot completely replace traditional hydrothermal exploration technology, it can provide a valuable reference with a strong target.

Collective Neurodynamic Optimization for Image Segmentation by Binary Model with Constraints

Threshold method is an important image segmentation method, which has been widely used in image segmentation. For this method, it is very important to choose a good threshold. The traditional threshold segmentation algorithm is implemented by exhaustive method, which makes the solution efficiency very low. This paper presents a collective neurodynamic optimization algorithm to solve the problem of binary optimization in the image segmentation. The problem of image segmentation based on threshold is transformed into binary optimization with constraints. Then, a collective neurodynamic optimization algorithm is introduced which combined with feedback neural network and particle swarm optimization (PSO) algorithm. And the linear programming relaxation constraint method is used to relax binary constraints. It is proved by numerical simulation that the feedback neural network algorithm can converge to the exact local optimal solution of the model and the PSO algorithm can get a better local optimal solution. Finally, several sets of comparative experiments are presented. The feasibility of our proposed method is verified; the experimental results demonstrate the effectiveness of our approach in image segmentation. In this study, a collective neurodynamic optimization was proposed for the image segmentation problem. In the future, we expect that multiple centralized neurodynamic models and intelligent algorithms can be used to solve the problem and improve the convergence speed of the solved model.

Thermal error analysis, modeling and compensation of five-axis machine tools

The role of five-axis CNC machine tools (FAMT) in the manufacturing industry is becoming more and more important, but due to the large number of heat sources of FAMT, the thermal error caused by them will be more complicated. To simplify the complicated thermal error model, this paper presents a new modelling method for compensation of the thermal errors on a cradle-type FAMT. This method uses artificial neural network (ANN) and shark smell optimization (SSO) algorithm to evaluate the performance of FAMT, and developing the thermal error compensation system, the compensation model is verified by machining experiments. Generally, the thermal sensitive point screening is performed by a method in which a large number of temperature sensors are arranged randomly, it increases the workload and may cause omission of the heat sensitive point. In this paper, the thermal imager is used to screen out the temperature sensitive points of the machine tool (MT), then the temperature sensor is placed at the position of the heat sensitive point of the FAMT, and the collected thermal characteristic data is used for thermal error modeling. The C-axis heating test, spindle heating test, and the combined movement test are applied in this work, and the results show that the shark smell optimization artificial neural network (SSO-ANN) model was compared to the other two models and verified better performance than back propagation artificial neural network (BP-ANN) model and particle swarm optimization neural network (PSO) model with the same training samples. Finally, a compensation experiment is carried out. The compensation values, which was calculated by the SSO-ANN model are sent to the real-time error compensation controller. The compensation effect of the model is then tested by machining the ‘S’-shaped test piece. Test results show that the 32 % reduction in machining error is achieved after compensation, which means this method improves the accuracy and robustness of the thermal error compensation system.

Process Parameter Optimization When Preparing Ti(C, N) Ceramic Coatings Using Laser Cladding Based on a Neural Network and Quantum-Behaved Particle Swarm Optimization Algorithm

The formation process of surface coatings fabricated with laser cladding is very complicated and coating quality is closely related to laser cladding process parameters. Generally, the optimization and control of process parameters play key roles when preparing high-quality ceramic coating. In this paper, three reasonable parameters were selected for each process parameter based on the preliminary experiment. The experiment of Ti(C, N) ceramic coating prepared with laser cladding was designed via the Taguchi method. The laser power, spot diameter, overlapping ratio, and scanning velocity were selected as the main process parameters, and their effects on coating micro-hardness were analyzed using the signal-to-noise (S/N) ratio and analysis of variance (ANOVA). Then, based on the back-propagation neural network (BPNN) and quantum-behaved particle swarm optimization (QPSO) algorithm, we created the prediction model of BPNN-QPSO neural network for laser cladding Ti(C, N) ceramic coating. The mapping of process parameters to the micro-hardness of the coating was obtained according to the model and we analyzed the influence of process parameters that interacted with the coating’s micro-hardness. The results showed that the interaction of laser cladding process parameters had a significant effect on the micro-hardness of the coating. The established BPNN-QPSO neural network model was able to map the relationship between laser cladding process parameters and coating micro-hardness. The process parameters optimized by this model had similar results with ANOVA. This research provides guidance for the selection and control of ceramic coating process parameters Ti(C, N) prepared via laser cladding.

Advanced Prediction of Roadway Broken Rock Zone Based on a Novel Hybrid Soft Computing Model Using Gaussian Process and Particle Swarm Optimization

A simple and accurate evaluation method of broken rock zone thickness (BRZT), which is usually used to describe the broken rock zone (BRZ), is meaningful, due to its ability to provide a reference for the roadway stability evaluation and support design. To create a relationship between various geological variables and the broken rock zone thickness (BRZT), the multiple linear regression (MLR), artificial neural network (ANN), Gaussian process (GP) and particle swarm optimization algorithm (PSO)-GP method were utilized, and the corresponding intelligence models were developed based on the database collected from various mines in China. Four variables including embedding depth (ED), drift span (DS), surrounding rock mass strength (RMS) and joint index (JI) were selected to train the intelligence model, while broken rock zone thickness (BRZT) is chosen as the output variable, and the k-fold cross-validation method was applied in the training process. After training, three validation metrics including variance account for (VAF), determination coefficient (R2) and root mean squared error (RMSE) were applied to describe the predictive performance of these developed models. After comparing performance based on a ranking method, the obtained results show that the PSO-GP model provides the best predictive performance in estimating broken rock zone thickness (BRZT). In addition, the sensitive effect of collected variables on broken rock zone thickness (BRZT) can be listed as JI, ED, DS and RMS, and JI was found to be the most sensitive factor.

Short-term electricity price and load forecasting in isolated power grids based on composite neural network and gravitational search optimization algorithm

Electricity price forecasting is a key aspect for market participants to maximize their economic efficiency in deregulated markets. Nevertheless, due to its non-linearity and non-stationarity, the trend of the price is usually complicated to predict. On the other hand, the accuracy of short-term electricity price and load forecasting is fundamental for an efficient management of electric systems. An accurate prediction can benefit future plans and economic operations of the power systems’ operators. In this paper, a new and accurate combined model has been proposed for short-term load forecasting and short-term price forecasting in deregulated power markets. It includes variational mode decomposition, mix data modeling, feature selection, generalized regression neural network and gravitational search algorithm. A mixed data model for the price and load forecast has been considered and integrated with the original signal series of price and load and their decomposition. Throughout this model, the candidate input variables are chosen by a distinct hybrid feature selection. Two reliable electricity markets (Pennsylvania-New Jersey-Maryland and Spanish electricity markets) have been used to test the proposed forecasting model and the obtained results have been compared with different valid benchmark prediction models. Lastly, the real load data of Favignana Island's power grid have been considered to test the proposed model. The obtained results pinpointed that the proposed model’s precision and stability is higher than in other benchmark forecasting models.

Suspended sediment load prediction using artificial neural network and ant lion optimization algorithm.

Suspended sediment load (SSL) estimation is a required exercise in water resource management. This article proposes the use of hybrid artificial neural network (ANN) models, for the prediction of SSL, based on previous SSL values. Different input scenarios of daily SSL were used to evaluate the capacity of the ANN-ant lion optimization (ALO), ANN-bat algorithm (BA) and ANN-particle swarm optimization (PSO). The Goorganrood basin in Iran was selected for this study. First, the lagged SSL data were used as the inputs to the models. Next, the rainfall and temperature data were used. Optimization algorithms were used to fine-tune the parameters of the ANN model. Three statistical indexes were used to evaluate the accuracy of the models: the root-mean-square error (RMSE), mean absolute error (MAE) and Nash-Sutcliffe efficiency (NSE). An uncertainty analysis of the predicting models was performed to evaluate the capability of the hybrid ANN models. A comparison of models indicated that the ANN-ALO improved the RMSE accuracy of the ANN-BA and ANN-PSO models by 18% and 26%, respectively. Based on the uncertainty analysis, it can be surmised that the ANN-ALO has an acceptable degree of uncertainty in predicting daily SSL. Generally, the results indicate that the ANN-ALO is applicable for a variety of water resource management operations.

Type Prediction of Student’s Achievement based on Grey Neural Network Optimized by GA

Students’ learning performance is predicted based on their online learning data, which is of great significance to evaluate the learning effect of school education and the reform and development of online education. In this paper, the characteristics of learning behaviour that constitute the influencing factors of students’ achievement are analysed and screened, and then a prediction model based on neural network and intelligent optimization algorithm is proposed. In this paper, the working principle of grey neural network (GNN) is described, the genetic algorithm (GA) is used to optimize the parameters, and the performance of the model is tested by simulation experiment. The experimental results show that, compared with the reference model, the prediction model optimized by GA parameters shows good prediction performance, and the prediction accuracy remains at a high level.

Short-term Power Forecast of Wind Power Generation Based on Genetic Algorithm Optimized Neural Network

With the rapid development of economy, human consumption of fossil energy has caused a global energy crisis. Nowadays, countries all over the world are making great efforts to develop clean energy. Wind energy as one of the pollution-free, renewable and high-quality clean energy, has attracted widespread attention. However, wind power generation is easily affected by natural factors and is instable. In order to improve the quality of wind power, this paper analyzes the influencing factors of wind power, studies the prediction method of wind power forecasting, and uses genetic algorithm optimization neural network to forecast the wind power of a wind farm in Northwest China, which may provide some reference for the power generation and grid connection of wind power plants.

A combination of probabilistic neural network (PNN) and particle swarm optimization (PSO) algorithms to map hydrothermal alteration zones using ASTER data

PNN is a feed-forward neural network in which there is an important parameter called smoothing parameter. This work implemented a combination of PNN with PSO optimization in order to estimate unique smoothing parameters for each SWIR bands of ASTER image and classified the ASTER image to different hydrothermal alteration zones (argillic, phyllic, propylitic and vegetation covering). The stydy area is a part of Kerman Cenozoic Magmatic Arc (KCMA) which contains several known porphyry copper deposits. Confusion matrix was used to validate the results of PNN-PSO algorithm and it presented the overall accuracy of 76.9% for developed algorithm. Also, comparing the obtained results with traditional methods of remote sensing (SPCA) showed that PNN-PSO is superior to SPCA technique. In fact, SPCA could not dicriminate different hydrothermal alterations while the present work proved that PNN-PSO is a good tool for classfication of argillic, phyllic, propylitic and vegetation covering.

Design, development and assessment of an essential oil based slow release vaporizer against mosquitoes

Mosquitoes (Diptera; Culicidae) are a biting nuisance and are of economic and health importance, especially for people living in tropical countries like India. Given the environmental concerns and health hazards of synthetic insecticides, development of natural products for the control of mosquito and mosquito-borne diseases are needed. In view of this, an essential oil based novel liquid vaporizer formulation with citronella and eucalyptus oils has been developed using a computer aided Artificial Neural Network and Particle Swarm Optimization (ANN-PSO) algorithm approach, aiming to predict the best optimized formulation (OF). Following the development, OF was characterized by Fourier Transform-Infra Red (FT-IR) spectroscopy and gas chromatography-mass spectroscopy (GC–MS). The efficacy of the OF was assessed against two major mosquito vectors viz. Anopheles stephensi and Aedes albopictus using a Peet-Grady chamber. Finally, toxicological impacts of the OF following its inhalation were investigated as per the Organization for Economic Co-operation and Development (OECD) guidelines. The results revealed all the ideal characteristics of the OF which were found to provide a slow release of up to 450 h at room temperature. Most importantly, the OF, exhibited 50% mosquito knock down (KT50) within 11.49±1.34 and 14.15±2.15 min against An. stephensi and Ae. albopictus respectively. Toxicity assessment showed a non toxic nature of the OF following inhalation. Thus the present development would be beneficial for controlling both An. stephensi and Ae. albopictus without any associated health hazards.

Optimum utilization of grid connected hybrid renewable energy sources using hybrid algorithm

An optimal utilization of smart grid connected hybrid renewable energy sources is proposed in this paper. The hybrid technique is the combination of recurrent neural network and adaptive whale optimization algorithm plus tabu search, called AWOTS. The main objective is the RES optimum operation for decreasing the electricity production cost by hourly day-ahead and real time scheduling. Here, the load demands are predicted using AWOTS to develop the correct control signals based on power difference between source and load side. Adaptive whale optimization algorithm searching behaviour is adjusted by tabu search. The proposed technique is executed in the MATLAB/Simulink working platform. To test the performance of the proposed method, the load demand for the 24-hour time period is demonstrated. By then the power generated in the sources, such as photovoltaic, wind turbine, micro turbine and battery by the proposed technique, is analyzed and compared with existing techniques, such as genetic algorithm, particle swarm optimization and whale optimization algorithm. Furthermore, the state of charge of the battery for the 24-hour period is compared with existing techniques. Likewise, the cost of the system is compared and error in the sources also compared. The comparison results affirm that the proposed technique has less computational time (35.001703) than the existing techniques. Moreover, the proposed method is cost-effective power production of smart grid and effective utilization of renewable energy sources without wasting the available energy.

Evaluation model of enterprise operation based on BP neural network optimization algorithm

The parameter selection of the traditional BP neural network (BPNN) has randomness, which makes the network prone to local extreme values during the calculation process. In order to solve this problem, this paper introduces the bat algorithm(BA) to optimize the parameter selection process of the BPNN and apply the algorithm to evaluate the enterprises’ operating condition, a corresponding evaluation model of the enterprises’ operating condition is established, and the evaluation model is applied to the prediction of the enterprises’ future operating condition and compared with the prediction effect of the traditional BPNN model. The prediction accuracy of the BPNN optimization algorithm is higher than the prediction accuracy of the traditional BPNN. The established enterprise operation evaluation model can effectively predict the future operation of the enterprise.

Optimization Algorithms for Tibetan-Chinese Neural Machine Translation

Tibetan-Chinese neural machine translation (NMT) is facing serious resource scarcity problem. This paper compares the application effect of multiple neural network optimization algorithms under the condition of resource scarcity, and proposes an optimization method which is suitable for resource scarcity languages. Then, by which it improving the performance of the Tibetan-Chinese NMT. Experimental results show that when choosing a suitable optimization algorithm, the Tibetan-Chinese NMT can still exceed the traditional statistical machine translation (SMT) and achieve better translation performance.

Developing a novel artificial intelligence model to estimate the capital cost of mining projects using deep neural network-based ant colony optimization algorithm

This study aims to propose a novel artificial intelligence model for forecasting the capital cost (CC) of open-pit mining projects with high accuracy. It is a unique combination of a deep neural network (DNN) and ant colony optimization (ACO) algorithm, abbreviated as ACO-DNN. In this model, MineAP (annual mine production), SR (stripping ratio), MillAP (annual production of the mill), RMG (reserve mean grade), and LOM (life of mine) were used to consider the CC of open-pit mining projects. A series of simple and complex artificial neural networks (ANN) was developed for forecasting CC of 74 copper mining projects herein. Subsequently, the ACO algorithm has been applied to optimize the developed ANN and DNN models to improve the accuracy of them. Finally, an optimal hybrid model was defined (i.e., ACO-DNN 5-25-20-18-15-1) with superior performance than other models (i.e., RMSE of 130.988, R2 of 0.991, MAE of 115.274, MAPE of 0.072, and VAF of 99.052). The findings of this study showed that the DNN models could predict the CC for open-pit mining projects with more accuracy than those of the simple ANN models. In particular, the ACO algorithm played an essential role in improving the accuracy of forecasting models. Also, MineAP, MillAP, SR, and LOM have been confirmed as critical parameters that affect the accuracy of the selected model in forecasting the CC of open-pit mining projects, especially MineAP. In conclusion, this study offers a useful tool to improve resource policies of mining projects, especially copper mining projects.

Research on Anomaly Detection in Massive Multimedia Data Transmission Network Based on Improved PSO Algorithm

With the development of computer network technology and the expansion of network system, sensitive data is facing the threat of hacker attack. Intrusion detection is an active network security defense measure, which is an attempt to invade, an ongoing intrusion or an intrusion that has occurred to identify the process. At present, the detection rate of intrusion detection method is low, the false alarm rate and false alarm rate is high, and the real-time performance is poor. It needs a large number of or complete data to achieve better detection performance. In this paper, the concept, characteristics, classification, research contents and difficulties of traditional intrusion detection for mass multimedia data transmission network are described. Then, the basic principle of neural network and particle swarm optimization (PSO) algorithm and the basic idea of particle swarm optimization algorithm with quantum (QPSO) behaviour are introduced. It is emphasized that QPSO has better convergence performance than PSO algorithm in global optimization problems. In this paper, the concept, characteristics and structure of neural network are described, and the algorithm and classification of wavelet neural network are introduced. Then taking wavelet neural network (WNN) as the object, using the QPSO algorithm as the training algorithm, the concrete operation process is given. The research work in this paper shows that the performance of neural network trained by QPSO algorithm and improved QPSO algorithm is better than that of other intelligent algorithms such as PSO algorithm and genetic algorithm, and the convergence speed is faster than that of PSO algorithm or GA algorithm. QPSO is a high performance neural network training algorithm, which can play a good role in neural network anomaly detection.

Deep Neural Network and Social Ski-Driver Optimization Algorithm for Power System Restoration with VSC - HVDC Technology

Deep Neural Network and Social Ski-Driver Optimization Algorithm for Power System Restoration with VSC - HVDC Technology

Similarity Calculation of 3D Model By Integrating Improved ACO Into HNN

It is important for 3D model retrieval to compute similarity between two models accurately. In order to calculate two models’ similarity, their face matching scheme need be found. Hopfield neural network (HNN) and ant colony optimization (ACO) algorithm can be used to match source faces and target ones. But, they all drop into local optimum easily and can not get optimal face matching scheme. This paper proposes a new method of computing model similarity by integrating improved ACO algorithm into HNN. Shape similarity between source face and target one is calculated based on the difference of face’s edge number. Structure similarity is computed according to adjacency correspondence relationship between source face and target one. Two faces’ similarity is calculated based on shape similarity and structure similarity, which is introduced into transfer probability. Indirect expectation heuristic is defined to improve ACO algorithm. Then, improved ACO algorithm is integrated into HNN to search for optimal face matching scheme. Model similarity is computed based on optimal sequence of face pairs. Experimental results show that compared with improved ACO algorithm, the proposed method improves the ranking effect of 14.29% of models, which can measure two models’ difference effectively.