頁籤選單縮合
題名 | 自來水加氯處理的AR(p)預估模式=AR(p) Prediction Models for Chlorine Dosage Used in Water Supply Processing Systems |
---|---|
作者姓名(中文) | 溫敏杰; 張虎明; 溫清光; | 書刊名 | 中國統計學報 |
卷期 | 36:2 1998.06[民87.06] |
頁次 | 頁155-172 |
分類號 | 445.256 |
關鍵詞 | 統計診斷檢驗; AR(p)模式建立; 加氯預估; 誤差分析比較; Statistical diagnostic checking; AR(p) model building; Chlorine dosage prediction; Error analysis and survey; |
語文 | 中文(Chinese) |
中文摘要 | 本文首先利用統計診斷檢驗來確定選用自我迴歸模型擬合自來水處理工程中所用 氯劑量時間序列的合理性;繼而使用定階和參數估計等兩個步驟建立了有效的AR(p)加氯預 估模式;最後對誤差的分析和比較,確立了加氯預估模式的適用性。本文有別於文獻[2]的關 鍵在於增加了建立模式時解決定階這一不可或缺的程序。而本文優於文獻[2]的地方,則在於 撇開對水質中各種有害化學成份和有機物的考量,單純地從氯劑量時間序列本身出發,進而 找到了問題的解決之道,並增加了它在實際應用時的實用性。 |
英文摘要 | This paper first demonstrates that, by statistical diagnostic checking, autoregressive models may well fit the time series of observations from the Chlorine dosage used in water supply processing systems. Second, it builds, via two steps of order determination and parameters estimation, effective AR(p) prediction models for Chlorine dosage. Finally, through analysis and survey to corresponding prediction errors, it confirms the model's applicability to real world. The key difference between results given here and in Reference [2] lies in that this paper adds the order-determination procedure which is absolutely necessary in autoretgressive model building. The major advantage over Reference [2] is that this paper yields an operation-ready formulation for Chlorine dosage, which is merely derived from the Chlorine time series itself, without considering any other harmful chemical or organic elements. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。