24小时热门版块排行榜

返回列表

当前只显示满足指定条件的回帖，点击这里查看本话题的所有回帖

zhongnanliuhui

木虫 (著名写手)

小木虫

应助: 13 (小学生)
金币: 2379.3
散金: 509
红花: 5
帖子: 1343
在线: 808.4小时
虫号: 490715
注册: 2008-01-03
性别: GG
专业: 控制理论与方法

[交流] 投了Journal of Wind Engineering & Industrial Aerodynamics,两年被拒已有4人参与

郁闷啊等了2年中间主编发过一次信让我不要着急
现在被拒不过好像还可以修改再投
因为提出的问题虽然多但是好像好回答
有投过该刊的牛人说说

评语：
Reviewers' comments:

Reviewer #1: The author's are to be commended for tackling a difficult and timely subject. However, I am afraid that I have a number of issues with the work and the way it has been presented.  For example, the introduction section is too brief and although it mentions a number of other similar studies it does not discuss the relative benefits/disadvantages of these.  More importantly it does not address the fundamental question of how the current approach is better than that which already exists.  In addition, the first seven references are written in Chinese and as such will not be acceptable to the majority of readers.  While I have no problem with the authors quoting their previous work I feel that seven examples are somewhat excessive.  In addition and perhaps more importantly a number of key references have also been missed, e.g. Delaunay et al. (2006) and Ding et al. (2008).

翻译：对作者所研究的这个困难而及时的课题表示肯定。但是我提出如下问题：摘要过于简单，没有比较提到文献的优缺点，提出方法的优势在哪里？此外，前面7篇参考文献均是中文，许多读者对这会介意的。我虽然不介意作者引用之前的工作基础，但是我觉得还是有点多，这样很容易遗失其他更加重要的参考文献：如e.g. Delaunay et al. (2006) and Ding et al. (2008).

Section 2 outlines the warning system.  In my view this section is also too brief and would benefit from a more in depth discussion of local features which may or may not influence the wind speed.  For example, what is the surrounding terrain near the anemometers?  What height are the anemometers positioned at and what is their sampling resolution?  Section 2.2 outlines the requirements of the system but no background information has been provided.  For example, why is a 1 minute forecast acceptable and how does this relate to train speed/anemometer separation etc?

翻译：关于预警系统，我个人认为仍然描述的过于简单，最好能够进一步深入讨论可能影响风速的当地地貌情况。例如，风速计周围的地形？风速计安装的位置？以及对应的采用方案？第2.2节虽然提出了预测要求指标，但是没有进一步说明其提出的背景信息。例如，为什么有超前1分钟预测？他们跟车速及行车安全有什么关系？

Section 3 considers the actual wind speed forecast.  Why are the authors only concerned with wind speed and not the stream wise and lateral velocity components?  Why is the data from the 51st - 220th wind speeds used to build the forecast?  The discussion relating to AR vs. ARMA vs. ARIMA does satisfactory address the issue of why a ARIMA (7,1,0) model was chosen.  Importantly no consideration appears to have been given as to whether the predicted wind speeds have the appropriate spectral form or whether the predicted wind speeds at different locations are appropriately correlated.
The conclusions could be expanded significantly if more work has been undertaken. I would advise the authors to resubmit the paper only after the proposed future work has been taken into account.

翻译：关于实际风速预测。为什么作者仅仅是关心风速，而不关心风速的流向和侧向分量呢？为什么建模数据取51st-220th这段样本？文章涉及AR/MA/ARMA/ARIMA模型，但是没有令人满意的解释为什么选择ARIMA(7,1,0)? 最重要的是，作者没有考虑预测风速的适用区间或是说预测风速与不同测风点或是监控点之间的关系。此外，作者的结论可以根据历史研究结果进一步扩展。我个人建议作者完成上述提出的问题后重新投稿。

References quoted:
Ding, Y, Sterling, M and Baker, C. J. (2008).An alternative approach to modelling train stability in high cross winds. Proceedings of the Institute of mechanical Engineers Part F: Journal of Rail and Rapid Transport. Volume 222, Number 1, 85-97.
Delaunay, D., Baker, C. J. Cheli, F, Morvan, H, Beger, L., Casazza, M, Gomez, C, Cleac'j C.Le., Saffel, R, Gregoire, R and Vinuales, A (2006) Development of wind alarm systems for road and rail vehicles: Presentation of the WEATHER project. SIRWEC 13th International Road Weather Conference, Toriono (Italy).

翻译：专家提出的相关参考文献。

Reviewer #2: Comments to Authors:
1. The paper deals with anemological data collected along the Qinghai-Tibet railway line. Anemological data are used to derive a forecast model, but in the paper there is no explanation concerning the instrumentation nor the sampling rate of data. Which is the sampling rate of raw data? Raw data has been averaged? Collected wind speed have to be intended as mean value? In this case, on which time span? A clear explanation is necessary.
翻译：文章是研究如何对青藏铁路沿线实测气象数据建立预测模型以保障行车的问题。但是文章没有介绍所选用的测试系统（仪器）以及对于的采样频率等设置。原如始数据的采样频率是多少？原始数据是否经过平均化处理？实测数据是否不得不经过平均化处理？处理的时间长度？这些问题都必须清楚地交代。

2. The data is analysed by means of a time-series approach based on Box-Jenkins model, but the explanations about the derivation of the method are not detailed. The forecast of wind speed is carried out 1, 3 and 5 minutes ahead, but there's no indication concerning the number of steps corresponding to such time intervals.
翻译：数据建模方法是使用时间序列分析理论中的詹金斯建模方案，但是缺少对这种建模方法的详细描述。预测是针对超前1、3、5分钟进行的，但是没有进一步指出采用数据的处理间隔以及超前预测的步数。

3. ARIMA(7,1,0) model is initially chosen to forecast wind speeds 1, 3 and 5 minutes ahead, and for all the meteorological station. It is stated that the model is calibrated on the data of the "27th meteorological station" (page 2-3), and that the model is requested to have an accuracy such that the mean absolute percentage error is less than given thresholds. The estimation of the error with the ARIMA(7,1,0) model (Table 1) is based to a whole time series of a single anemometer? It refers to a single anemological station (station 27?) or is the averaged value from different stations? This is not clear. The same considerations can be applied to Table 2. Moreover, the mean percentage error is not a clear estimator of the goodness of the forecast, since underestimations and overestimations of the actual values tend to cancel each other. The RMSE is a better estimator, but since it is dimensional it should be referred to velocity ranges. Percentage error (mean value and
confidence bounds, as standard deviation) is preferable.
翻译：作者提到所见的模型ARIMA(7,1,0)将运用到全部测风站预测，并且预测精度必须满足门槛要求。那么有如下疑问：模型的精度评价是仅仅使用27号册封站的全部样本数据，还是全部测风站样本数据的平均处理后的样本？这个问题没有交代清楚。此外，百分比误差指标不是一个清楚的考核精度指标，因为欠预测和过预测精度容易相互抵消。而RMSE指标更加理想，但是其涉及速度范围，不好计算。因此，平均值加上置信区间的误差指标更加受欢迎。

4. Since strong winds are responsible of safety issues, a clear indication of the error in the estimation of strong events could be a clear indicator of the performance of the model. Moreover, a comparison of the performance of adopted models with a persistent model could be helpful to understand the effective performance: the errors of the proposed model are smaller or larger than those of a persistent one?
翻译：强风将影响列车行车安全，因此模型的预测性能是强风评价考核的重要因素。建议将采用模型和现有模型的预测性能进行进一步比较分析，看到底采用模型的误差比现有模型大还是小?

5. Equation 4 does not correspond to a percentage value.
6. The so called "modified method" seems to be based on the ARIMA(7,1,0) that updates the estimation of the model parameter based on forecasted value. In this case, it should not be intended as an effective "improvement" of the method, but a recursive estimation algorithm.
翻译：作者将提出的方法命名为改进算法，实际上是基于模型参数估计的实时刷新。我个人认为这应该称为迭代优化算法更加合适。

7. Figures 6-10 show real time wind speeds and forecasted values. The predicted time series' seem to be similar to the actual values, but shifted in time. The time shifting increases as the forecast horizon increases, and in this case the similarity of actual and forecasted time series decreases. Do the authors have any explanation for this phenomenon?
翻译：文章图6-10指出，预测风速与实测风速较为接近，但是存在明显的预测延时现象。延时主要出现在风速阶跃点，这种现象将降低风速预测精度，作者能对这个现象进行解释吗？

8. References 11 and 12 are not cited within the text. The style used for references is not uniform and need to be revised. The inclusion of references to articles and book related to Box-Jenkins model and its applications to wind forecast is strongly suggested.
9. There are many grammar and spelling errors: the text need to be revised carefully.
10. The text style (e.g. indentation) is not uniform.
翻译：参考文献11和12没有在文中标注。参考文祥的格式不统一，需要修改。此外，建议参考文献补充关于詹金斯预测的书和文章。文章许多拼写和语法错误，请作者认真修改。

不知道还有没有胜算？还是换一个刊物？

回复此楼