Electricity Consumption Prediction Model Based on Bayesian Regularized BP Neural Network

The primary aim of the article is to analyze the response of the human immune system when it encounters the hepatitis B virus. This is done using a mathematical system of differential equations. The differential equation system has six components, likely representing various aspects of the immune response or virus dynamics. A Bayesian regularization neural network has been presented in the process of training. These networks are employed to find solutions for different categories or scenarios related to hepatitis B infection. The Adams method is used to generate reference data sets. The back-propagated artificial neural network, based on Bayesian regularization, is trained and validated using the generated data. The data is divided into three sets: 90% for training and 5% each for testing and validation. The correctness and effectiveness of the proposed neural network model have been assessed using various evaluation metrics. The metrics have been used in this study are Mean Square Error (MSE), histogram errors, and regression plots. These measures provide support to the neural network to approximate the immune response to the hepatitis B virus.

The Sixth International Symposium on Neural Networks (ISNN 2009)

At present, China’s power load development is facing a new situation in which policies such as the new economic norm, industrial structure adjustment, energy conservation and emission reduction, etc. are being deeply promoted, load growth in some areas begin to ease, and volatility of load gradually become prominent, which increases the difficulty of medium and long-term load forecasting. In this context, in view of multi-correlation, uncertainty of influence of policy factors on power load, in order to improve the accuracy of load forecasting under the influence of policy factors, and solve the problem that policy factors are ambiguous, difficult to be quantified, and difficult to be integrated into load forecasting model, a medium and long-term load forecasting model considering policy factors is proposed. First, by analyzing the influence of various policies on power load, a hierarchical policy influencing factor index system that combines macro and micro levels is constructed to systematically reflect the influence of economy and policies on load under the new situation. Then, in view of the traditional grey relational analysis model’s insufficient consideration of the difference of historical data and future power development situation, by respectively weighting historical periods and factor indexes, a quantification analysis model of power load influencing factors based on two-way weighted grey relational analysis is proposed to quantify the influence of various policy factors on power load, achieve the combination of subjective weighting and objective weighting,and obtain quantification weights. Finally, the weighted fuzzy cluster analysis method combined with weights is used to predict load under the influence of policy factors. The proposed model can better solve the difficulty of medium and long-term load forecasting caused by the volatility of load under the influence of policy factors, and is suitable for medium and long-term load forecasting under the background of policy changes. The analysis of calculation examples shows that compared with traditional grey relational analysis model, the quantification results of proposed methods are more realistic, compared with traditional prediction methods such as time series extrapolation and elasticity coefficient, proposed method has better prediction accuracy and engineering application value.

In order to solve the problems of insufficient accuracy of long-term power load forecasting and poor applicability of the model, this paper considers the coupling of a number of macro indicators, such as regional economic development and social development indicators, with the time series data of regional power load. BP neural network and Autoregressive integrated moving average model (ARIMA) are used to integrate and improve the forecasting model, so as to improve the trend forecasting ability of annual load forecasting model. The non parametric function method is used to forecast the periodic load data in the monthly load data, the annual load forecast is combined with the monthly load forecast to improve the overall forecasting accuracy of the model. Finally, through the comparison of grey prediction and other models and the verification of MAPE error analysis method, the prediction accuracy of the model method considering the combination of data periodicity and trend is significantly improved, which is suitable for the long-term prediction of regional power load.

By considering the characteristics of agricultural products prices such as great fluctuations, nonstationarity and nonlinearity, a BP neural network model was designed for forecasting and verified by real data from some agricultural products wholesale markets. In order to improve the prediction accuracy, a combination model of BP neural network and time series prediction was constructed to forecast prices of agricultural products in wholesale markets. After learning from sample data, it can better forecast the trend and fluctuations of the agricultural products wholesale prices, and with those real data better prediction results were attained. The combined model provides an important method for predicting prices in agricultural products wholesale markets.

Energy financial risk early warning model based on Bayesian network

An essential element of electric utility resource planning is forecasting of the future load demand in the service area. Based on the outcome of such forecasts, utilities coordinate their resources to meet the forecasted demand using a least-cost plan. In general, resource planning is performed subject to numerous uncertainties. Expert opinion indicates that a major source of uncertainty in planning for future capacity resource needs and operation of existing generation resources is the forecasted load demand. In this paper, the development and testing of a hybrid intelligent long-term load forecasting system is presented consisting of several neural networks forecasting blocks, genetic algorithms for network architecture selection and optimization, and fuzzy rules for forecast combination. This is an application of increasingly significant importance to a deregulating electric utility industry. An overview of the current practice in long term load forecasting is presented, followed by an overview of the forecasting system design process utilized in generating the long-term load forecasts and a brief description of the key building blocks of the forecasting system. This is followed by some sample long-term forecasts performed for demonstrating the feasibility of the proposed approach.

Sound quality evaluation may be conducted with multiple different objective evaluation models, but reliability of evaluation results is different.Noise samples of two ears of the driver were gathered when the car operated at the speed of 60km/h and different music was played in the car.After equal loudness, frequency band and filtering treatment of sound samples, subjective evaluation of sound quality was carried out.Based on calculation of objective parameters, MLR (Multiple Linear Regression) objective evaluation model for sound quality, objective evaluation model for sound quality based on BP neural network and objective evaluation model for sound quality based on RBF (radial basis function) neural network were established.The test samples were substitute into the three models, and their sound quality prediction effects at each frequency band were compared.The results show that MLR model is simple and has fast calculation speed, so it is suitable for linear problem analysis; BP neural network model is suitable for solving large sample problem with complex internal mechanism; RBF neural network model has high precision and small relative error, so it is suitable for analysis of nonlinear problem of small sample.Under experimental conditions, prediction effect of RBF model is superior to MLR model and BP neural network model.

Cooling fan is widely used in various fields. At present, owing to lack of reasonable and effective monitoring methods of fan's condition, the normal operation of the main system is seriously affected. The air cooling experiment is specially designed to solve this problem, and the monitoring and data acquisition lasted for a year. Through the analysis of the fan experimental data of whole life, a new method based on a combined model of time-series and BP neural networks is proposed to predict the remaining life of fan. This method utilizes the easily acquired speed signal as a parameter to assess the state of the running fan and the ARIMA model of time series is built to forecast the overall trend of fan's remaining life. It is combined with BP neural networks model which improves the prediction accuracy to make up for the disadvantage of single time series model which prediction error is large. By comparing the forecast values of the proposed model with experimental values, the results demonstrate that the method can accurately predict the remaining life of the fan while running, which provides the guidance for the condition monitoring of fan.

NMR and Bayesian regularized neural network regression for impurity determination of 4-aminophenol

Determination of pore size distribution in tight gas sandstones based on Bayesian regularization neural network with MICP, NMR and petrophysical logs

Permeability prediction of multi-stage tight gas sandstones based on Bayesian regularization neural network

The landslide caused a huge disaster to the living environment and seriously threatened the lives and property safety of nearby residents. Assess or predict landslide-susceptible the landslide displacement through monitoring are great beneficial to guide landslide control and mitigate these hazards by taking appropriate preparatory measures. In this paper, a new water cycle algorithm optimization BP neural network (BPNN) dynamic prediction model (WCA-BPNN) was established to make up for the shortcoming of BPNN convergence speed. A typical step-wise landslide——Langshuwan Landslide happened in the Three Gorges Reservoir area of China is taken as a case, and the displacement monitoring data of 4 years was used for time series analysis and modeling. The long-term creep effect of the landslide and the short-term acceleration effect of the climate are considered in the model, and the accumulative displacement is divided into two kinds of trend displacement and periodic displacement. The key influencing factors of landslide periodic displacement were screened by gray relational grade analysis method, and then used as learning data. In addition to comparing the predicted value of the model with the measured value, it also compares the accuracy of the three models of BPNN, support vector machine, extreme learning machine under the training conditions of the same learning data set. The results show that the WAC-BPNN model has faster convergence speed and higher prediction accuracy than the traditional BPNN model, and it is also the most accurate of the four models.

Pakistan, being an agricultural country, highly depends on its natural water resources which originate from the upper regions of Hindu Kush-Karakoram-Himalaya Mountains and nourish one of the world's largest Indus Basin irrigation systems. This paper presents streamflow modelling and forecasting using signal difference average (SDA) based variational mode decomposition (VMD) combined with machine learning (ML) methods at Chitral and Tarbela stations on the Indus River network. For this purpose, VMD based random forest (VMD-RF), gradient boosting machine (VMD-GBM) and Bayesian regularized neural network (VMD-BRNN) have been chosen. Moreover, traditional time series flow models, that is seasonal autoregressive integrated moving average (SARIMA) and classical decomposition approach with particle swarm optimization-based support vector regression (PSO-SVR), are considered as benchmark models for comparison. The results show that overall, VMD-BRNN performed best, followed by VMD-GBM and VMD-RF, whereas SARIMA and PSO-SVR ranked last. Overall, SARIMA and PSO-SVR are failing to capture most of the peaks even during the training period whereas hybridization of VMD and ML methods has shown increased robustness of the models. The results show that the influential role of the high dimensional components and robustness on the river flow may be explored by most optimum SDA based VMD signals hybrid with BRNN method.

The polynomial model and wood model have been extensively applied to predict the milk yield of cows, which aims to measure and partially explain the uncertainty under single factor condition. However, different sample data would affect the goodness of fit of the models. To investigate the milk yield regularities of the Chinese Holstein cows in Yunnan Standardized Pasture (YSP), data of cows with 1–3 parities using different observation periods were derived from the 401,497 records of 1,826 cows collected from YSP, and fitted in the BP neural network (BPNN) model, the Quadrinomial model and the Wood model. Prediction results of three models were compared. The results show that, for group data (mean), the Quadrinomial model was likely to over-fit under the single factor condition with small sample data, while the Wood model had a better goodness of fit; the BPNN model was more suitable for multi-factor and large sample data analysis.