Considering the instability of the output power of photovoltaic(PV)generation system,to improve the power regulation ability of PV power during grid-connected operation,based on the quantitative analysis of meteorolog...Considering the instability of the output power of photovoltaic(PV)generation system,to improve the power regulation ability of PV power during grid-connected operation,based on the quantitative analysis of meteorological conditions,a short-term prediction method of PV power based on LMD-EE-ESN with iterative error correction was proposed.Firstly,through the fuzzy clustering processing of meteorological conditions,taking the power curves of PV power generation in sunny,rainy or snowy,cloudy,and changeable weather as the reference,the local mean decomposition(LMD)was carried out respectively,and their energy entropy(EE)was taken as the meteorological characteristics.Then,the historical generation power series was decomposed by LMD algorithm,and the hierarchical prediction of the power curve was realized by echo state network(ESN)prediction algorithm combined with meteorological characteristics.Finally,the iterative error theory was applied to the correction of power prediction results.The analysis of the historical data in the PV power generation system shows that this method avoids the influence of meteorological conditions in the short-term prediction of PV output power,and improves the accuracy of power prediction on the condition of hierarchical prediction and iterative error correction.展开更多
A new method of short-term forecasting for water consumption in municipal supply water networks based on wavelet transformation is introduced. By wavelet decomposing commonly used in the signal field, water consumptio...A new method of short-term forecasting for water consumption in municipal supply water networks based on wavelet transformation is introduced. By wavelet decomposing commonly used in the signal field, water consumption per hour is decomposed into many series. Trend item, cycle item and random item are separated from the original time series in this way.Then by analyzing, building a model, forecasting every series and composing the results, the forecasting value of the original consumption is received. Simulation results show that this forecasting method is faster and more accurate, of which the error is less than 2%, indicating that the wavelet analytical method is practicable.展开更多
Objective The objective of this retrospective study was to investigate the prognostic factors associated with survival among patients with extensive stage-smal cel lung cancer （ES-SCLC）. Methods Clinical data from 6...Objective The objective of this retrospective study was to investigate the prognostic factors associated with survival among patients with extensive stage-smal cel lung cancer （ES-SCLC）. Methods Clinical data from 66 patients with ES-SCLC diagnosed via histopathology or cytology between July 2005 and July 2009 at Anyang Tumor Hospital （China） were analyzed. Univariate and multivariate Kaplan-Meier, log-rank, and Cox proportional hazard regression analyses were conducted. Results The 12-, 24-, and 36-month survival rates among patients with ES-SCLC were 40.9%, 13.6%, and 6.1%, respectively. The median survival time （MST） was 10 months. Univariate analyses indicated that weight loss, eficacy of first-line chemotherapy, total number of chemotherapy cycles, treatment meth-od, and serum sodium levels significantly influenced survival among patients with ES-SCLC. Multivariate analyses suggested that the eficacy of first-line chemotherapy, total number of chemotherapy cycles, and serum sodium levels were independent prognostic factors associated with survival. Conclusion The eficacy of first-line chemotherapy, total number of chemotherapy cycles, and serum sodium levels are important prognostic factors for patients with ES-SCLC.展开更多
Flood frequency analysis procedure was performed on annual maximum discharge data of River Oshun at Iwo in Osun State, Nigeria for the period 1985 to 2002 utilizing three probability distribution models namely： Extre...Flood frequency analysis procedure was performed on annual maximum discharge data of River Oshun at Iwo in Osun State, Nigeria for the period 1985 to 2002 utilizing three probability distribution models namely： Extreme EVI （value Type-l）, LN （Log normal） and LPIII （Log Pearson Type III）. The models were used to predict and compare corresponding flood discharge estimates at 2, 5, 10, 25, 50, 100 and 200 years return periods. The results indicated that Extreme Value Type 1 distribution predicted discharge values ranging from 26.6 m3/s for two years to 431.8 m3/s for 200 years return periods; the Log Pearson Type III distribution predicted discharge values ranging from 127.2 m3/s for two years to 399.54 m3/s for 200 years return periods and the Log normal distribution predicted discharge values ranging from 116.2 m3/s for two years to 643.9 m3/s for 200 years return periods. From the results~ it was concluded that for lower return periods （T_〈 50 yrs） Extreme Value Type 1 and Log Pearson Type III could be used to estimate flood quantile values at the station while for higher return periods （T 〉 50 yrs） Log Normal probability distribution model which gives higher estimates could be utilized for safe design in view of the short length of discharge records used for the analysis.展开更多
The application of support vector machines to forecasting problems is becoming popular, lately. Several comparisons between neural networks trained with error backpropagation and support vector machines have shown adv...The application of support vector machines to forecasting problems is becoming popular, lately. Several comparisons between neural networks trained with error backpropagation and support vector machines have shown advantage for the latter in different domains of application. However, some difficulties still deteriorate the performance of the support vector machines. The main one is related to the setting of the hyperparameters involved in their training. Techniques based on meta-heuristics have been employed to determine appropriate values for those hyperparameters. However, because of the high noneonvexity of this estimation problem, which makes the search for a good solution very hard, an approach based on Bayesian inference, called relevance vector machine, has been proposed more recently. The present paper aims at investigating the suitability of this new approach to the short-term load forecasting problem.展开更多
This paper investigates a risk-averse inventory model by balancing the expected profit and conditional value-at-risk （CVaR） in a newsvendor model setting. We find out that： i） The optimal order quantity is increas...This paper investigates a risk-averse inventory model by balancing the expected profit and conditional value-at-risk （CVaR） in a newsvendor model setting. We find out that： i） The optimal order quantity is increasing in the shortage cost for both the CVaR only criterion and the tradeoff objective, ii） For the case of zero shortage cost, the optimal order quantity to the CVaR criterion or tradeoff objective is increasing in the selling price, respectively. However, it may not be monotonic in the selling price when incorporating a substantial shortage cost. Moreover, it may be larger or less than the risk-neutral solution, iii） Under the tradeoff objective function, although the optimal order quantity for the model without shortage cost is increasing in the weight put on the expected profit, this property may not be true in general for the model with a substantial shortage cost. Some numerical examples are conducted to verify our results and observations.展开更多
基金supported by National Natural Science Foundation of China(No.516667017).
文摘Considering the instability of the output power of photovoltaic(PV)generation system,to improve the power regulation ability of PV power during grid-connected operation,based on the quantitative analysis of meteorological conditions,a short-term prediction method of PV power based on LMD-EE-ESN with iterative error correction was proposed.Firstly,through the fuzzy clustering processing of meteorological conditions,taking the power curves of PV power generation in sunny,rainy or snowy,cloudy,and changeable weather as the reference,the local mean decomposition(LMD)was carried out respectively,and their energy entropy(EE)was taken as the meteorological characteristics.Then,the historical generation power series was decomposed by LMD algorithm,and the hierarchical prediction of the power curve was realized by echo state network(ESN)prediction algorithm combined with meteorological characteristics.Finally,the iterative error theory was applied to the correction of power prediction results.The analysis of the historical data in the PV power generation system shows that this method avoids the influence of meteorological conditions in the short-term prediction of PV output power,and improves the accuracy of power prediction on the condition of hierarchical prediction and iterative error correction.
文摘A new method of short-term forecasting for water consumption in municipal supply water networks based on wavelet transformation is introduced. By wavelet decomposing commonly used in the signal field, water consumption per hour is decomposed into many series. Trend item, cycle item and random item are separated from the original time series in this way.Then by analyzing, building a model, forecasting every series and composing the results, the forecasting value of the original consumption is received. Simulation results show that this forecasting method is faster and more accurate, of which the error is less than 2%, indicating that the wavelet analytical method is practicable.
文摘Objective The objective of this retrospective study was to investigate the prognostic factors associated with survival among patients with extensive stage-smal cel lung cancer （ES-SCLC）. Methods Clinical data from 66 patients with ES-SCLC diagnosed via histopathology or cytology between July 2005 and July 2009 at Anyang Tumor Hospital （China） were analyzed. Univariate and multivariate Kaplan-Meier, log-rank, and Cox proportional hazard regression analyses were conducted. Results The 12-, 24-, and 36-month survival rates among patients with ES-SCLC were 40.9%, 13.6%, and 6.1%, respectively. The median survival time （MST） was 10 months. Univariate analyses indicated that weight loss, eficacy of first-line chemotherapy, total number of chemotherapy cycles, treatment meth-od, and serum sodium levels significantly influenced survival among patients with ES-SCLC. Multivariate analyses suggested that the eficacy of first-line chemotherapy, total number of chemotherapy cycles, and serum sodium levels were independent prognostic factors associated with survival. Conclusion The eficacy of first-line chemotherapy, total number of chemotherapy cycles, and serum sodium levels are important prognostic factors for patients with ES-SCLC.
文摘Flood frequency analysis procedure was performed on annual maximum discharge data of River Oshun at Iwo in Osun State, Nigeria for the period 1985 to 2002 utilizing three probability distribution models namely： Extreme EVI （value Type-l）, LN （Log normal） and LPIII （Log Pearson Type III）. The models were used to predict and compare corresponding flood discharge estimates at 2, 5, 10, 25, 50, 100 and 200 years return periods. The results indicated that Extreme Value Type 1 distribution predicted discharge values ranging from 26.6 m3/s for two years to 431.8 m3/s for 200 years return periods; the Log Pearson Type III distribution predicted discharge values ranging from 127.2 m3/s for two years to 399.54 m3/s for 200 years return periods and the Log normal distribution predicted discharge values ranging from 116.2 m3/s for two years to 643.9 m3/s for 200 years return periods. From the results~ it was concluded that for lower return periods （T_〈 50 yrs） Extreme Value Type 1 and Log Pearson Type III could be used to estimate flood quantile values at the station while for higher return periods （T 〉 50 yrs） Log Normal probability distribution model which gives higher estimates could be utilized for safe design in view of the short length of discharge records used for the analysis.
文摘The application of support vector machines to forecasting problems is becoming popular, lately. Several comparisons between neural networks trained with error backpropagation and support vector machines have shown advantage for the latter in different domains of application. However, some difficulties still deteriorate the performance of the support vector machines. The main one is related to the setting of the hyperparameters involved in their training. Techniques based on meta-heuristics have been employed to determine appropriate values for those hyperparameters. However, because of the high noneonvexity of this estimation problem, which makes the search for a good solution very hard, an approach based on Bayesian inference, called relevance vector machine, has been proposed more recently. The present paper aims at investigating the suitability of this new approach to the short-term load forecasting problem.
基金This research was supported by the Social Science Foundation of the Ministry of Education of China under Grant No. 07JA630015, the National Natural Science Foundation of China under Grant Nos. 70901059 and 70901029, and the Fundamental Research Funds for the Central Universities under Grant No. 105-275171.
文摘This paper investigates a risk-averse inventory model by balancing the expected profit and conditional value-at-risk （CVaR） in a newsvendor model setting. We find out that： i） The optimal order quantity is increasing in the shortage cost for both the CVaR only criterion and the tradeoff objective, ii） For the case of zero shortage cost, the optimal order quantity to the CVaR criterion or tradeoff objective is increasing in the selling price, respectively. However, it may not be monotonic in the selling price when incorporating a substantial shortage cost. Moreover, it may be larger or less than the risk-neutral solution, iii） Under the tradeoff objective function, although the optimal order quantity for the model without shortage cost is increasing in the weight put on the expected profit, this property may not be true in general for the model with a substantial shortage cost. Some numerical examples are conducted to verify our results and observations.
