Niu Dongxiao

Application of Principal Component Regression Analysis in power load forecasting for medium and long term

This paper deals with the power load forecasting for medium and long term using based on Principal Component Regression Analysis. The paper first reviews the research achievement of the load forecasting and its relationship with economic development, then introduces the basic theory of the principal component analysis and principal component regression analysis model. Finally, taking Beijing as an example, the paper extracts the principal components from the relevant economic factors related power consumption in Beijing, then establishes a multi-parameter regression prediction model (Principal component regression model) on the principal components. The results show that the error is small between prediction load and actual load, proving that the model is a feasible and effective method of load forecasting.

An ARMA Cooperate with Artificial Neural Network Approach in Short-Term Load Forecasting

Short-term load forecasting is important for electricity load planning and dispatches the loading of generating units in order to meet the electricity system demand. The precision of the load forecasting is related to electricity company’s economic. This paper presents a approach named an autoregressive moving average (ARMA) cooperate with BP Artificial Neural Network (BPNN) approach, which can combine the linear component and nonlinear component at the same time. the experiment result shows that the MAPE of this method is 0.92%, and MSE is 17.07, compared to single ARMA’s MAPE 2.08% and MSE 47.65 or BPNN’s MAPE 2.63% and MSE 56.91, this method is outperform the single ARMA and BPNN forecast method.

Combined models for day‐ahead electricity price forecasting based on improved gray correlation methodology

Purpose – To combine the forecasting by single method using influence information fully, other than regular combined methods only focusing on historical forecasting errors. Design/methodology/approach – To combine the single methods based on the analysis of improved gray correlation, with more related information being considered to enhance the price forecasting precision, such as the trend of the prices, the historical forecasting errors, and the temporal influence factors on prices. Findings – A case of PJM market of USA shows that the proposed method has better performance than any other combined methods, and all single models as well. Research limitations/implications – The combined performance depends on the forecasting precision of single methods, and the correlation between the single methods, as well as the number of single method that to be combined. Practical implications – It is a novel idea for combined method to forecasting the time series data, such as electricity prices, electric power loads. Originality/value – The proposed method considers all the following factors: the similarity between the trends of the single forecasting, the errors of the single models and the temporal influence.Purpose – To combine the forecasting by single method using influence information fully, other than regular combined methods only focusing on historical forecasting errors.Design/methodology/approach – To combine the single methods based on the analysis of improved gray correlation, with more related information being considered to enhance the price forecasting precision, such as the trend of the prices, the historical forecasting errors, and the temporal influence factors on prices.Findings – A case of PJM market of USA shows that the proposed method has better performance than any other combined methods, and all single models as well.Research limitations/implications – The combined performance depends on the forecasting precision of single methods, and the correlation between the single methods, as well as the number of single method that to be combined.Practical implications – It is a novel idea for combined method to forecasting the time series data, such as electricity prices, electric power loads.Or...

Combining simulate anneal algorithm with support vector regression to forecast wind speed

Accurate wind speed forecasting is essential for predicting the wind power output. The wind speed is randomness, so the forecasting is very difficult. Least squares support vector machines (LSSVM) for load forecasting requires the identification of relevant parameters by expert experiment, this paper proposed a combination of adaptive particle swarm optimization the relevant parameters of least square support vector machine to forecast the wind speed. Compare using the default parameters of LSSVM method, the experimental results show that the proposed method can effectively select the parameters and the proposed method has more accurate results than the default parameters LSSVM method.

Improved RBF network applied to short-term load forecasting

From the practical application of short-term load forecasting, this article introduced the radial basis function network and use nearest neighbor clustering algorithm to determine the width of radial basis function, select the cluster centers and weights. The predicted results show that the method is faster and has higher precision.

Notice of Retraction Research on Chinese cities comprehensive competitiveness based on principal component analysis and factor analysis in SPSS

Factor analysis and principal component analysis are commonly used in multivariate statistical analysis. This paper has shown similarities and differences of principal component analysis and factor analysis in mathematical model and solution procedure. And we select some important indexes of Chinese cities comprehensive competitiveness, use economics software SPSS to carry on the principal component analysis and factor analysis, and get comprehensive score. At last, we analyze the influential factors of cities competitiveness to provide support and protection for the development of cities.

Analysis of Electricity Demand Forecasting in Inner Mongolia Based on Gray Markov Model

Accurate electricity demand forecasting is the foundation of power system operation and planning, the basic of placing development plans, business strategy and tactics of the power companies. Electricity consumption is a gray system which is impacted by economic development, industrial structure, income levels and national policies. The paper counted Inner Mongolia electricity data from four factors, used the gray model GM (1, n) to predict, established Markov model from the percentage of relative deviation δ of the fitted value and actual value, and revised forecast values obtained from gray model. At last, the paper predicted the demand for electric power industry in Inner Mongolia by gray Markov model.

Wavelet neural network embedded expert system used in short-term load forecasting

The wavelet neural network has been widely used as a new load forecasting method. This paper has analyzed wavelet neural network which consists with wavelet and artificial neural network. Summarize historical load data to fine the long-term variation, gain the empirical knowledge from professionals, and form a series of rules to simulate a human expert reasoning and judging process, which is the expert system. Finally, use expert system to correct data in wavelet neural network, and predicted results can be obtained. The case study, about the load forecasting of a city grid, is very well embedded in the wavelet neural network expert system approach, and the prediction accuracy has get improved as well.

Empirical analysis on Shanxi electric energy efficiency by DEA model

With the rapid development of Chinese economy, Chinas power shortage problem has been gradually emerging, and widespread power cuts often occur occasionally, which has brought serious constraints to Chinas economic development and affected peoples normal life. How to improve electric energy efficiency and fully utilize power have become an urgent task and also the requirement to build a saving society determined by the ‘Eleventh Five-Year Plan’. This paper establishes a indicator system of total factor energy efficiency, putting total power consumption, fixed capital investment and labor forces as input indicators and GDP as a output indicator; then adopts the DEA model to make an empirical analysis of electric energy efficiency of Shanxi Province.

Study on Chaos Characteristics of Electricity Price Based on Power-Law Distribution

Electricity market is a complex system involving multi-factors and multi-links. As an important economic lever, the electricity price shows the evolution of complex irregular movement. It’s significant to reveal the internal law of this seemingly stochastic evolution. the study firstly determined the sequence method as the research tool after comparative analysis of four different power indexes. And then, the study compared the single time sequence and the time-sharing time sequence in PJM market. The former sequence better reflected the complex factors. So exponential of the single time sequence was studied by sequential method of least squares fit. We found that the time-sharing time sequence has a distinct power-law distribution. The exponential is in the range of 2.262-3.375. At last, the study summarized the main factors affecting price volatility. It’s a new research idea for the complex nature of electricity price and the complex system of electricity market.