TW200419046A - Wide range optimum factory control device - Google Patents
Wide range optimum factory control device Download PDFInfo
- Publication number
- TW200419046A TW200419046A TW093103935A TW93103935A TW200419046A TW 200419046 A TW200419046 A TW 200419046A TW 093103935 A TW093103935 A TW 093103935A TW 93103935 A TW93103935 A TW 93103935A TW 200419046 A TW200419046 A TW 200419046A
- Authority
- TW
- Taiwan
- Prior art keywords
- plan
- water
- control device
- factory
- time
- Prior art date
Links
Classifications
-
- A—HUMAN NECESSITIES
- A62—LIFE-SAVING; FIRE-FIGHTING
- A62B—DEVICES, APPARATUS OR METHODS FOR LIFE-SAVING
- A62B13/00—Special devices for ventilating gasproof shelters
-
- A—HUMAN NECESSITIES
- A62—LIFE-SAVING; FIRE-FIGHTING
- A62B—DEVICES, APPARATUS OR METHODS FOR LIFE-SAVING
- A62B5/00—Other devices for rescuing from fire
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P80/00—Climate change mitigation technologies for sector-wide applications
- Y02P80/10—Efficient use of energy, e.g. using compressed air or pressurized fluid as energy carrier
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Landscapes
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Feedback Control In General (AREA)
- Flow Control (AREA)
Abstract
Description
(1) 200419046 玖、發明說明 【發明所屬之技術領域】(1) 200419046 发明. Description of the invention [Technical field to which the invention belongs]
本發明是關於以複數個設施爲流體供應對象之廣範圍 工廠，例如對於廣範圍自來水廠的運用’使原水的取水量 或淨工廠的總濾過量、從淨工廠送到配水池的送水量之時 間上變動盡量降低，並且幫浦的耗電量也盡量降低，而實 現高速最適化且穩定又有效率的水運用控制之廣範圍工廠 的最適當運用控制裝置。 【先前技術】 用計算機來支援自來水廠運用的計劃時，預測當作對 象之配水區域的需求，根據該預測値來運算最適當的運用 計劃。以複數個設施爲對象之廣範圍自來水廠，由於控管 設施數大增，其運用方法也變複雜，將運用計劃最適化之 運算時間增長，而會有實用上造成障礙的可能性。The present invention relates to a wide range of plants where a plurality of facilities are used as a fluid supply target. For example, for the use of a wide range of water plants, 'the amount of raw water intake or the total filtration of a net plant is excessive, and the amount of water sent from the net plant to a distribution tank is The time variation is minimized, and the power consumption of the pump is also reduced as much as possible, and it is the most suitable operation control device for a wide range of plants that realizes high-speed optimization, stable and efficient water operation control. [Prior art] When a computer is used to support a water plant operation plan, the demand for the water distribution area as an object is predicted, and the most appropriate operation plan is calculated based on the forecast. For a wide range of water plants that target multiple facilities, the number of control facilities has increased dramatically, and their use has become more complex. The operation time that has been optimized for the use plan has increased, which may cause practical obstacles.
關於自來水廠的自動控制，當根據計算機所作成之運 用計劃來運用控制工廠時，淨水能穩定生產且不會斷水， 順利供應到需求戶爲主要考量。同時，工廠的運用上，從 運用成本或設備檢修費用等的層面，也要求有效運用控制 工廠。因此，重點爲是否能作成滿足這個觀點的要求之最 適當或近乎最適當之工廠運用計劃。 近年，隨著工廠的複雜化或大規模化，對於在實質時 間內迅速運算如同上述的最適當或近乎最適當之工廠運用 計劃會有困難。特別是廣範圍工廠，自來水事業體的淨工 (2) 200419046 廠或配工廠的設施散置在各處，統籌這些設施所監控之廣 範圍監視的水流正逐漸安定中。在這種情況，迅速運算每 曰的工廠運用計劃越加困難。 因此，爲了解決這種問題，必須實現即使是複雜或大 規模的工廠，仍能運算高速最適化或近乎最適當之運用計 劃，實現有效率且穩定的工廠運用之廣範圍工廠的最適當 運用控制裝置。Regarding the automatic control of the water plant, when the control plant is applied according to the application plan made by the computer, the purified water can be produced stably without being cut off, and the smooth supply to the customers is the main consideration. At the same time, the use of factories requires effective control of factories from the aspects of operating costs and equipment maintenance costs. Therefore, the focus is on whether the most appropriate or nearly the most appropriate plant operation plan to meet the requirements of this viewpoint can be made. In recent years, with the complexity or scale of factories, it has been difficult to quickly calculate the most appropriate or nearly the most appropriate factory operation plan as described above in real time. Especially for wide-area factories, the net work of the water supply business body (2) 200419046 The facilities of the plant or distribution plant are scattered all over the place, and the wide-range monitored water flow monitored by these facilities is gradually stabilized. In this case, it becomes more difficult to quickly calculate each plant operation plan. Therefore, in order to solve such a problem, it is necessary to realize a high-speed optimized or near-optimal operation plan for a complex or large-scale factory, and to implement an optimal operation control of a wide range of factories for efficient and stable factory operation. Device.
用計算機來支援自來水廠運用的計劃時，根據當作對 象之配水區域的需求預測來運算最適當的運用計劃（例如 ，參照日本專利特開平 8 - 1 8 6 0 7 8號公報及特開 2 0 0 1 -5 5 7 6 3號公報）。以複數個設施爲對象之廣範圍自來水、 污工廠，由於管控的設施數增多，其運用方法也變複雜， 因而用於運用計劃之運算時間大增，而會有實用上造成障 礙的可能性。When a computer is used to support a water plant operation plan, the most appropriate operation plan is calculated based on the demand forecast of the target water distribution area (for example, refer to Japanese Patent Laid-Open No. 8-1 8 6 0 7 8 and JP 2). 0 0 1 -5 5 7 6 3). For a wide range of tap water and sewage plants that target multiple facilities, the number of facilities to be controlled and the methods of using them will become more complicated. As a result, the calculation time for the use plan will increase significantly, which may cause practical obstacles.
【發明內容】 隨著管理•運用設施的統籌化•管範圍監視化，必須 實現能實現有效率又穩定的自來水廠運用，且不失實用性 作成高速最適當的工廠運用計劃之工廠運用控制裝置。 本發明鑒於這點，其目的是提供對於以複數個設施爲 對象之廣範圍工廠的運用，使原水的取水量或淨廠的總濾 過量、從淨工廠送到配水池的送水量之時間上變動盡量降 低，並且幫浦的耗電量也盡量降低，而實現高速最適化且 穩定又有效率的水運用控制之廣範圍工廠的最適當運用控 -6 - (3) (3)200419046 制裝置。 爲了達成上述目的，本發明提供第1〜1 4項的專利申 請項目。 本發明的申請專利範圍第1項，是針對將以複數個設 1 施爲對象之廣範圍工廠作最適當的運用控制之廣範圍工廠 的最適當運用控制裝置，其特徵爲具備有：輸入必要的設 定値或條件之資料輸入部、及儲存處理程序資料的計測値 或種種的參數設定値等的資料之實績資料庫部、及參照藉 | 由資料輸入部所輸入的天候資或儲存在實績資料庫部之過 去的實績需要値來預測運轉該日以後每單位時間的需要量 之需要預測部、及根據需要預測部所取得每單位時間的預 測需要量和處理程序的計測値，把該日每單位時間的運用 流量及其流量的時刻所組成之變數組當作基因，只有運用 流量及其流量的時刻被更改時依照存有該變數組之當作可 變長度的基因列使用之基因運算來運算以複數個設施爲對 象之廣範圍工廠的最適當或近乎最適當的運用計劃之運用 f 計劃部、輸出運用計劃部所取得的廣範圍工廠運用運算結 果和必要的其他資料之資料輸出部。 本發明的申請專利範圍第2項，如申請專利範圍第1 項之廣範圍工廠的最適當運用控制裝置，其中具備：運用 計劃部生成初期個體之際，判定執行能滿足限制條件的解 之初期個體是否在一定時間內生成，若無法生成則中止初 期個體的生成之初期個體生成計時器部。 本發明的申請專利範圍3項，如申請專利範圍第2項 -7- (4) (4)200419046 之廣範圍工廠的最適當運用控制裝置，其中具備：把儲存 在實績資料庫部之過去的工廠運用當作初期個體來生成後 提供給運用計劃部之實績初期個體生成部。 本發明的申請專利範4項，如申請專利範圍第2項之 廣範圍工廠的最適當運用控制裝置，其中具備：把人爲用 資料輸入手段所設定的工廠運用案當作初期個體來生成後 提供給運用計劃部之探試初期個體生成部。 本發明的申請專利範5項，如申請專利範圍第2項之 廣範圍工廠的最適當運用控制裝置，其中具備··以其他的 最適當手法，對前述運用計劃部所取得之工廠運用計劃再 度進行最適化的運算之合成最適化部。 本發明的申請專利範6項，如申請專利範圍第2項之 廣範圍工廠的最適當運用控制裝置，其中具備：運用計劃 部沒有取得最適當的工廠運用計劃時，對運用計劃部緩和 限制條件後再進行最適化的運算之限制緩和運用計劃部。 本發明的申請專利範7項，如申請專利範圍第2項之 廣範圍工廠的最適當運用控制裝置，其中具備：依時間觀 看運用計劃部所取得之運用計劃而局部成爲凹狀時，促使 最適化的評比値與限制條件一面作比較，一面將該運用計 劃平滑化之凹狀平滑化部。 本發明的申請專利範8項，如申請專利範圍第2項之 廣範圍工廠的最適當運用控制裝置，其中具備：依時間觀 看運用計劃部所取得之運用計劃而局部成爲凸狀時，促使 最適化的評比値與限制條件一面作比較，一面將該運用計 (5) 200419046 劃平滑化之凸狀平滑化部。 本發明的申請專利範9項，如申請專利範圍第2 廣範圍工廠的最適當運用控制裝置，其中具備··依時 看運用計劃部所取得之運用計劃而成爲向上階梯狀時 使最適化的評比値與限制條件一面作比較’一面將該 計劃平滑化之向上階梯狀平滑化部。 本發明的申請專利範1 0項，如申請專利範圍第 之廣範圍工廠的最適當運用控制裝置，其中具備：依 觀看運用計劃部已取得之運用計劃而成爲向下階梯狀 促使最適化的評比値與限制條件一面作比較，一面將 用計劃平滑化之向下階梯狀平滑化部。 本發明的申請專利範1 1項，如申請專利範圍第 之廣範圍工廠的最適當運用控制裝置，其中具備：工 有局部控制的控制器，含有工廠運用控制裝置無法控 機器時，模擬局部控制，將其結果輸入運用計劃部後 工廠的運用計劃之局部控制模擬計劃部。 本發明的申請專利範1 2項，如申請專利範圍第 之廣範圍工廠的最適當運用控制裝置，其中具備：從 到會有時間延遲的地方工廠供應所必要的流量時，衡 間的延遲來修正工廠觀看運用計劃部所取得之運用計 必要量之時間延遲修正部。 本發明的申請專利軺1 3項，如申請專利範圍第 之廣範圍工廠的最適當運用控制裝置，其中具備：廣 工廠爲廣範圍自來水廠，定期監視各下游設施的水位[Summary of the Invention] As the management, operation facilities are coordinated, and the scope of management is monitored, it is necessary to realize an efficient and stable water plant operation, and a plant operation control device that is the most appropriate high-speed factory operation plan without losing practicality. . In view of this point, the present invention aims to provide a wide range of plants targeted at a plurality of facilities, so that the amount of raw water intake or the total filtration of the net plant is excessive, and the time of the amount of water sent from the net plant to the distribution tank Minimize changes and minimize power consumption of pumps, and realize the most appropriate operation control of a wide range of plants for high-speed optimal, stable, and efficient water operation control. -6-(3) (3) 200419046 . To achieve the above object, the present invention provides items 1 to 14 of patent applications. The first item of the scope of patent application of the present invention is the most suitable operation control device for a wide-range factory that is the most suitable for controlling a wide-range factory targeted at a plurality of facilities, and is characterized by having: input necessary Data input section for setting, or condition, and measurement database for storing process program data, or various parameter settings, etc., and a reference database for borrowing | weather data entered by the data input section or stored in actual results The past performance of the database department needs to measure the demand forecasting unit per unit time after the day of operation, and measure the forecasted demand and processing procedures per unit time obtained by the forecasting unit based on the demand. The variable flow composed of the applied flow rate and its flow time per unit time is regarded as a gene, and only when the applied flow rate and its flow time are changed is the genetic operation used as a variable-length gene array in which the variable array is stored To calculate the most appropriate or nearly the most appropriate use plan for a wide range of plants targeted at multiple facilities. F Planning Department, The use of a wide range of factory planning department made use of the information output unit calculates the results and other information as necessary. The most suitable operation control device for a wide range of factories in the scope of patent application No. 2 of the present invention, such as the scope of patent application No. 1, includes: the use of the planning department to generate the initial individuals, and to determine the initial stage of execution of the solution that can meet the restrictions. Whether the individual is generated within a certain period of time, and if the individual cannot be generated, the initial individual generation timer section that suspends the initial individual generation is suspended. The present invention has three patent application scopes, such as the second patent application scope -7- (4) (4) 200419046 The most appropriate operation control device for a wide range of factories, which includes: the past stored in the performance database department The results are provided to the operation planning department after the factory operations are generated as initial individuals. There are 4 items of the patent application scope of the present invention. For example, the most suitable operation control device for a wide range of factories in the scope of the patent application No. 2 includes: after the factory operation plan set by artificial data input means is taken as an initial entity and generated. It is provided to the initial stage individual generation department of the operation planning department. There are 5 items of patent application scope of the present invention. For example, the most suitable operation control device for a wide range of factories in the second item of the patent application scope, which includes ... the factory operation plan obtained by the aforementioned operation planning department again by other most appropriate methods. A synthesis optimization unit that performs an optimization operation. There are 6 items of patent application scope of the present invention. For example, the most suitable operation control device for a wide range of factories in the second item of the patent application scope includes the following: when the operation planning department does not obtain the most appropriate factory operation plan, the operation planning department relaxes the restrictions. Restriction mitigation operation planning department that performs optimization calculations later. The most suitable operation control device for a wide range of factories in the patent application scope of the present invention, such as item 7 of the patent application, includes: when the operation plan obtained by the operation planning department is viewed in time and partially becomes concave, the optimum The concave evaluation section 化 is compared with the restrictive conditions while smoothing the operation plan. The eighth patent application scope of the present invention, such as the most appropriate operation control device for a wide-range factory applying for the second scope of the patent application, includes: when the operation plan obtained by the operation planning department is watched over time and partially becomes convex, the optimum operation is promoted. The comparison of the evaluation rate and the restriction conditions are compared, and the convex smoothing part of the application plan (5) 200419046 is smoothed. There are 9 patent applications of the present invention. For example, the most suitable operation control device of the second wide-range factory for which the patent application scope is applied, which includes ... The comparison 値 is compared with the restriction conditions, and an upward step-like smoothing section that smoothes the plan. There are 10 items of the patent application scope of the present invention. For example, the most suitable operation control device of the wide-range factory with the scope of patent application, which includes: a downward step-like evaluation according to the operation plan obtained by the operation planning department to promote optimization.値 Compared with the restrictions, the plan will be smoothed down with a stepped down side. The item 11 of the patent application of the present invention, for example, the most appropriate operation control device for a wide range of factories in the scope of patent application, includes: a controller with local control, and simulation of local control when the factory operation control device cannot control the machine. And input the result to the local control simulation planning department of the factory's operation planning after the operation planning department. There are 12 items in the patent application scope of the present invention. For example, the most suitable control device for a wide range of factories in the scope of patent application, which includes: when the necessary flow is supplied from a local factory with a time delay, the delay between scales Correction of the time delay correction section necessary for the factory to view the necessary amount of operation meter obtained by the operation planning department. There are 13 patent applications for this invention, such as the most appropriate control device for a wide range of plants in the scope of patent application, which includes: the wide plant is a wide-range water plant, and the water levels of various downstream facilities are regularly monitored.
項之 間觀 ，促 運用 2項 時間 時， 該運 2項 廠存 制的 運算 2項 工廠 量時 劃的 2項 範圓 計畫1J -9- (6) 200419046 是否在預定的範圍內，在範圍內時繼續監視，不在範圍內 時’首先只有該設施重新計劃，其結果若是滿足其他設施 的限制條件則根據重新計劃結果來運用廣範圍工廠；若是 不滿足其他設施的限制條件則也包括位於該設施上游的設 施促使逐一反覆重新計劃之重新計劃判定部。Viewing between items, when two items of time are promoted, the two items of the plant's inventory system to calculate the two items of time scale of the two factory circle plans 1J -9- (6) 200419046 are within the predetermined range, within Continue monitoring when in range, if not in range, only the facility will be rescheduled first. If the results meet the constraints of other facilities, use the wide-range factory according to the results of the replanning. The facilities upstream of this facility have prompted the replanning judgement department, which is replanned one after another.
本發明的申請專利範1 4項，如申請專利範圍第1項 之廣範圍工廠的最適當運用控制裝置，其中具備：進行模 擬來促使能夠確認所期望的運用設定値是否滿足工廠運用 上的限制條件之模擬部。 [實施方式】 以下，用圖面來說明本發明的實施例。 第1圖表示本發明的最適當運用控制裝置的一個實施 形態之方塊圖。第2圖表示使用第1圖的最適當運用控制 裝置之廣範圍自來水廠之系統圖。There are 14 items of patent application scope of the present invention. For example, the most suitable operation control device for a wide range of factories in the scope of patent application No. 1 includes simulations to enable confirmation of desired operation settings and whether or not factory operation restrictions are met. Conditions simulation department. [Embodiment] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. Fig. 1 is a block diagram showing an embodiment of the most suitable control device of the present invention. Fig. 2 shows a system diagram of a wide range of waterworks using the most suitable control device of Fig. 1.
<實施形態之構成> 第2圖所示之廣範圍自來水廠，廣範圍自來水設施以 複數個（圖中爲8處）配水池4 1〜4 8爲對象，從共同的淨 水廠4 0配水給這些配水池4 1〜4 8。淨水廠4 0用集水幫浦 從河川等取得原水，經過集水井在混合池注入藥品，藉由 凝集•沉澱•過濾之淨水手段來進行淨水處理。所得到的 淨水用氯來進行殺菌處理過後，經淨水池，藉由送水幫浦 ，經由各配水池配水給各別的配水區。當然也有可能直接 -10- (7) (7)200419046 從淨水廠4 0配水給一般家庭等的配水區域。第2圖所示 的廣範圍自來水設施’大體上2分爲第1群的配水 '池 4 1〜4 4及第2群的配水池4 5〜4 8。圖中管路彼此連結的節 點用圖號5 1表示；另外若有必要在各管路或配水池中間 串接水閥5 2 ;若有必要在各處設置測定各管路流過的水 流量之流量計（例如’電磁流量計）5 3。 第1圖所示的最適當運用控制裝置爲最適當運用控制 第2圖所示之廣範圍自來水廠之裝置，具備有：資料輸入 部2、資料輸出部4、實績D B部（實績資料庫部）6、需 要預測部8、運用計劃部1 0、、實績初期個體生成部1 4 、探試初期個體生成部1 6、合成最適化部1 8、限制緩和 運用計劃部20、凹狀平滑化部22、凸狀平滑化部24、向 上階梯平滑化部2 6、向下階梯平滑化部2 8、局部控制模 擬計劃部3 0、時間延遲修正部3 2、重新計劃判定部3 4及 模擬部3 6。 資料輸入部2爲在廣範圍工廠的最適當運用上輸入必 要的設定値或條件之輸入處理手段。資料輸出部4爲輸出 最適當運用控制裝置之廣範圍工廠的最適當運用運算結果 及其他的必要資料之輸出處理手段。實績DB部6爲保存 處理程序的計測値或種種的參數設定値等的資料之資料庫 。需要預測部8係以天候資訊或過去的實績需要値等來預 測運轉該日以後每單位時間的需要量。運用計劃部1 〇係 把需要預測部8所取得每單位時間的預測需要量及根據處 理程序的計測値，把該日每單位時間的運用水量及其流量 -11 - (8) (8)200419046 之時間的變數組當作基因，只有運用流量及其流量的時刻 被更改時，依照存有該變數組之當作可變長度的基因列使 用之基因運算來運算以複數個設施爲對象之廣範圍工廠的 最適當或近乎最適當之運用計劃。 初期個體生成計時器部1 2係在生成初期個體之際， 判定滿足限制條件下可執行的解之初期個體是否能再一定 時間內生成，若無法生成時，用來中止初期個體的生成之 計時器。實績初期個體生成部1 4係把儲存在實績DB部6 之過去的工廠運用當作初期個體使用之手段。探試初期個 體生成部1 6係把人爲從輸入手段所設定的工廠運用案當 作初期個體使用之手段。合成最適化部1 8係將已取得工 廠運用計劃再度用除此以外的最適化手段進行最適化運算 。限制緩和運用計劃部2 0係在得不到最適當工廠運用計 劃時，緩和限制條件才進行最適化運算。 凹狀平滑化部22係在已用運用計劃部1 0運算過的運 用計劃成爲凹狀時，將促使最適化之評比値與限制條件一 面作比較，一面將其運用計劃平滑化。同樣的，凸狀平滑 化部24係運用計劃成爲凸狀時，將促使最適化之評比値 與限制條件一面作比較，一面將其運用計劃平滑化。向上 階梯平滑化部26係在運用計劃成爲向上階梯狀時，將促 使最適化之評比値與限制條件一面作比較，一面將其運用 計劃平滑化。進而，向下階梯平滑化部2 8係運用計劃成 爲向下階梯狀時，將促使最適化之評比値與限制條件一面 作比較，一面將其運用計劃平滑化。 -12- (9) 200419046 局部控制模擬計劃部3 0係在工廠 器，且含有工廠運用控制裝置無法控制 部控制，將其結果輸入運用計劃部來運 。時間延遲修正部3 2係從工廠到會有 工廠供應所必要的量時衡量時間的延 爲的必要量。 重新計劃判定部3 4係依照預定的 施（配工廠4 1〜4 8 )是否滿足限制條件 視，若有未滿足的設施時則首先只有該 結果：若滿足其他設施的限制條件則根 運用廣範圍工廠。若不滿足其他設施的 位於該設施的上游之設施逐一反覆重新 模擬部3 6係進行模擬來促使能夠 2所輸入之所期望的運用設定値是否滿 制條件。 本竇施型態係利用由以上的構成部 運用控制裝置，對於以複數個設施爲對 工廠的運用，使原水的取水量或淨水廠 工廠水送到配水池4 1〜4 8的送水量之時 ’並且幫浦的耗電量也盡量降低’而實 疋且有效率的水運用控制之廣軔Η工廠 裝置。 <實施型態的作用〉 有局部控制的控制 的機器時，模擬局 算工廠的運用計劃 時間延遲的地方， 遲來修正工廠所視 順序定期監視各設 ，若滿足時繼續監 設施重新計劃。其 據重新計劃結果來 限制條件則也包括 計劃。 確認從資料輸入部 足工廠運用上的限 分所組成之最適當 象之廣範圍自來水 的總濾過量、從淨 間上變動盡量降低 現高速最適化又穩 的最適當運用控制 -13- (10) (10)200419046 首先，說明第i圖所不之最適當運用控制裝置的基本 功會g 。 需要預測部8之配水的需要預測方法能想到的有統計 的手法或最小次方法、GMDH ( Groupmg Method Data H a n d 1 i n g )等的各種同等手法、神經元網路的方法等，不 過本發明並不拘限於特定的手法，任何的手法皆可。需要 預測部8及運用計劃部1 0 —日一次定時啓動。首先，直 到一定時刻之前，藉由資料輸入部2以手動或自動輸入需 要預測所必要的資料。此處，資料例如是指需要預測日的 天氣預報、或最高氣溫或者最低氣溫的天氣預報、或至此 所得到天候資訊的實績値、需要量實績値等。需要預測部 8之需要預測的結果每一定早位時間至少輸出一日量的時 間。當然若有必要進行二日以上的需要預測亦可。 運用計劃部1 〇根據需要預測結果及工廠的現在預測 値（若是淨水的送水計劃則爲配水池的水位或送水幫浦的 流量、運轉台數、配水量等；若是淨工廠內總濾過流量的 計劃則爲淨水池水位、或各濾過持的洗淨時間點、濾過池 洗淨水量、排水池水位等）；及送水幫浦流量特性、配水 池4 1〜4 8或淨水廠4 0、排水池的容量（運用水位上下限 値）等的參數，將脫離淨水廠4 0或配水池4 1〜4 8的運用 水位上下限、或不足配水量預測値且引起總濾過量或者送 水量的激烈變化之水應用計劃作最適當運算。此時，若有 必要也考量在耗電尖峰時間的時間帶限制幫浦的可用台數 -14- (11) (11)200419046 經此方式，把該日每單位時間的運用水量及其流量的 時間之變數組當作基因，只有運用水量及其流量的時間被 更改時，依照存有該變數組之作爲可變長度的基因列使用 之基因運算的運用計劃部1 0，每單位時間至少輸出一日 量的時間。當然若有必要進行二日以上的運用計劃亦可。 現在推測第2圖的廣範圍自來水廠當中藉由送水幫浦 從淨水池4 0經由配水池4 1〜4 8將淨水配送到畫配水區之 程序加以說明。當然也有以自然流下，時而從淨水廠40 送水到配水池 4 1〜4 8，時而從配水池配水到配水區之情形 ’不過此處，被認爲即使以送水幫浦送水，以配水幫浦配 水’若是設定進水池側的流量調節閥或淨水池出水側的流 量調節閥的其中一調節閥的張開度設定値，則這些水閥張 開度設定値的離散値相當於幫浦1台的吐出流量，因而不 至於喪失一般性。 另外，送水處理程序的思考方法係依配水需要來判定 從淨水廠4 0送到各配水池4 1〜4 8的送水量，並且濾過池 的洗淨水或廠內所使用的其他水量也含在配水需要，依配 水池來判定淨水池，依送水幫浦來判定濾過池，且洗淨時 間點之濾過池停歇視同送水幫浦的限制條件，因而能幾乎 原樣因應於總濾過流量計劃，所以也可以同樣的進行淨水 廠內之總濾過流量計劃的最適化。當然統籌將從送水到總 濾過流量計劃爲止最適化也能用同樣的思考方法。 在一定時刻k從淨水池送到配水池的送水量Q p ( k ) 藉由所啓動的送水幫浦或安裝在配管之水閥來控制。其目 -15- (12) (12)200419046 標的流量以離散且以幾個階段來設定。這個流量稱爲流量 階段。例如，不管是水閥或幫浦，總之送水量的物理最大 値假設爲100m3/h，若此値分成5區份，則流纛爲〇、20 ' 40、60、100 m3/h，這些値代表送水計劃所取得的送水 量 ° 現在只用轉速（迴轉速度）固定的定速幫浦n台來送 水，定速幫浦的台數與送水計劃所取得的送水量之離散値 對應於1對1。爲了使說明簡單，以下推定這個情況來作 說明。 把每單位時間的運用水量及其水量之時間的變數組當 作基因，只有運用流量及其流量的時刻被更改時’存有該 變數組之當作可變長度的基因列使用之基因運算則是如第 3圖所示來表現基因。 第3圖表現第2圖所示之各設施（配水池4 1〜4 8 )在 各時刻的進水量。現在配水池爲N個（圖示的例子爲8 個）時，配水池從1依序加上圖號’此圖號稱爲設施編號 。設施編號i的進水量計劃値以（時刻、流量階段）的組 來表現。把這個組當作基因。因此，時間爲取用從計劃開 始時刻至結束時刻爲止·之整數値；流量階段係依預定流1量: 離散的目標値來取用實數値。此基因的組係依照本發明， 並不是針對應該計劃的時間帶才存在，只關於進水量變更 的時刻才持有。因而，進水量爲一定時基因的組一組即可 ，可以節約記憶體的容量，結果是也連帶縮小探索最適當 解的對象領域範圍。如此’本發明中基因組（列）的長度 -16- (13) (13)< Structure of the embodiment > The wide-area waterworks shown in Fig. 2 has a wide-area water-supply facility with a plurality of (eight places in the figure) distribution tanks 4 1 to 4 8 and a common water purification plant 4 0 distribute water to these distribution pools 4 1 ~ 48. The water purification plant 40 uses water collection pumps to obtain raw water from rivers, etc., and inject chemicals into the mixing tank through the water collection wells, and performs water purification treatment by means of water purification by aggregation, sedimentation, and filtration. After the obtained purified water is sterilized with chlorine, the water is distributed to the respective water distribution areas through the water purification pumps and water distribution pumps through the water purification pumps. Of course, it is also possible to directly distribute water from the water purification plant 40 to the water distribution area of ordinary households, such as -10- (7) (7) 200419046. The wide-range tap water facility 'shown in Fig. 2 is roughly divided into water distribution' pools 4 1 to 4 4 of the first group and water distribution pools 4 5 to 4 8 of the second group. The nodes connecting the pipelines in the figure are represented by the figure No. 51. In addition, if it is necessary to connect a water valve 5 2 in series between each pipeline or the distribution tank; if necessary, measure the flow of water flowing through each pipeline. Flowmeter (eg 'electromagnetic flowmeter') 5 3. The most suitable operation control device shown in FIG. 1 is the most suitable device for controlling a wide range of waterworks shown in FIG. 2. The device includes: a data input section 2, a data output section 4, and a performance DB section (performance database section). ) 6. Needs forecasting section 8, operation planning section 10, initial stage individual generation section 14, initial stage trial individual generation section 16, synthesis optimization section 18, restricted mitigation operation planning section 20, concave smoothing Section 22, convex smoothing section 24, upward step smoothing section 2 6, downward step smoothing section 2 8, local control simulation planning section 30, time delay correction section 3 2, replanning determination section 34, and simulation Department 3 6. The data input unit 2 is an input processing means for inputting necessary settings or conditions for the most suitable operation of a wide range of factories. The data output unit 4 is an output processing means for outputting the most suitable operation calculation results and other necessary data from a wide range of factories where the control device is most suitably used. The actual performance DB section 6 is a database that stores data such as the measurement data of the processing program and various parameter settings. The demand forecasting unit 8 uses weather information or past actual performance requirements to estimate the demand per unit time after the day of operation. The operation planning department 10 is based on the forecasted demand per unit time obtained by the demand forecasting unit 8 and the measurement based on the processing program. The operation water quantity and flow rate per unit time on that day are -11-(8) (8) The variable array at the time of 200419046 is regarded as a gene. Only when the flow rate and the time of the flow rate are changed, the calculation is performed on a plurality of facilities according to the genetic algorithm used as a variable-length gene array in which the variable array is stored. The most appropriate or nearly the most appropriate use plan for a wide range of plants. The initial entity generation timer unit 12 determines whether the initial entity that can execute the solution that meets the restrictions can be generated within a certain time when the initial entity is generated. If it cannot be generated, it is used to stop the timing of the initial entity generation. Device. The initial individual production unit 14 uses the past factory operations stored in the actual performance DB unit 6 as a means of initial individual use. In the early stage of the trial, the individual production unit 16 used the factory operation plan set by artificial means as an initial means for individual use. The synthesis and optimization department 18 series will again use the optimization method other than the optimization method to obtain the plant operation plan. The Restriction and Mitigation Planning Department 20 performs optimization calculations only when the optimum plant utilization plan is not obtained. The concave smoothing section 22 smoothes the operation plan while comparing the optimized evaluation with the constraints when the operation plan that has been calculated by the operation planning section 10 becomes concave. Similarly, when the convex-shaped smoothing unit 24 is convex, the evaluation plan that promotes the optimization will be compared with the restrictions, and the operation plan will be smoothed. The upward step smoothing unit 26 smoothes the operation plan while comparing the optimization evaluation with the constraints when the operation plan becomes stepped up. Furthermore, when the down-step smoothing section 28 series operation plan becomes a down-step, it compares the optimized evaluation with the constraints and smoothes its use plan. -12- (9) 200419046 The local control simulation planning department 30 is located in the factory and contains factory operation control devices that cannot be controlled by the control department. The results are input to the operation planning department for operation. The time delay correction unit 3 2 measures the necessary amount of time delay from the time when the factory supplies the necessary amount to the factory. The re-planning judgment section 34 is based on whether the predetermined facilities (distribution plants 4 1 to 4 8) meet the restrictions. If there are unsatisfied facilities, the only result is the first one: if the restrictions of other facilities are met, it will be widely used. Scope factory. If it does not meet the other facilities, the facilities located upstream of the facility are repeated one by one. The simulation department 36 performs simulations to enable the desired operation settings entered by the 2 to be satisfied. This sinus application type uses the operation control device of the above components to make the raw water intake or the water from the water purification plant to the distribution tanks 4 1 to 4 8 for the operation of the factory with a plurality of facilities. At that time, and the power consumption of the pump was also minimized, the Hiroshima factory installed a practical and efficient water use control. < The role of the implementation type> When there is a locally controlled machine, the place where the operation plan of the plant is simulated is delayed, and the plant's view of the plant is delayed to periodically monitor each facility. If it is satisfied, the facility is re-planned. Restrictions based on replanning results also include planning. Confirm the widest range of tap water in the wide range of tap water from the data input department and the limit of the plant's operation. Excessive total filtration, and change from the clean room. Minimize the existing high-speed optimal and stable operation. -13- (10 (10) 200419046 First, the basic skill g of the most appropriate use of the control device which is not shown in Fig. I will be described. The method for predicting the water distribution required by the prediction unit 8 includes statistical methods, minimum order methods, various equivalent methods such as GMDH (Groupmg Method Data H and 1 ing), and methods of neuron networks. It is not limited to a specific technique, and any technique is acceptable. The forecasting department 8 and the operation planning department need to be activated at regular intervals once a day. First, until a certain time, data necessary for prediction is manually or automatically inputted by the data input section 2. Here, the data refers to, for example, a weather forecast for a forecast day, a weather forecast for a maximum temperature or a minimum temperature, or a performance record of weather information obtained so far, a demand performance report, and the like. The required prediction section 8 needs to output at least one day of time at a certain early bit time. Of course, if it is necessary to forecast the demand for more than two days. Operation planning department 10: Forecast results according to needs and the current forecast of the plant (if the water supply plan of the purified water is the water level of the distribution pool or the flow of the water pump, the number of operating units, the amount of water distribution, etc .; The plan is the water level of the clean pool, or the washing time of each filter, the amount of water in the filter pool, the water level of the drainage pool, etc.); and the flow characteristics of the water pump, the distribution pool 4 1 ~ 4 8 or the water purification plant 4 0, the capacity of the drainage pool (the upper and lower limits of the water level) and other parameters will deviate from the upper and lower limits of the water level of the water purification plant 40 or the distribution tank 4 1 to 48, or the prediction of insufficient water distribution will cause excessive total filtration or The water application plan for drastic changes in water supply volume is the most appropriate calculation. At this time, if necessary, also consider the limit of the number of available pumps in the time zone of the peak power consumption time. -14- (11) (11) 200419046 In this way, the amount of water used per unit time and its flow rate on that day The time-varying array is regarded as a gene. Only when the amount of water and its flow time is changed, according to the operation planning section of the genetic calculation using the genetic array used as a variable-length gene array that stores the variable array, 10 per unit time. Output at least one day of time. Of course, if it is necessary to carry out an operation plan for more than two days. Now, it is estimated that the procedure of distributing purified water to the water distribution area from the clean water tank 40 through the water distribution tanks 4 to 4 8 in the wide-range water supply plant in FIG. 2 will be explained. Of course, there may be natural flow, sometimes from the water purification plant 40 to the water distribution tank 4 1 ~ 4 8, and sometimes from the water distribution tank to the water distribution area '. However, it is considered that even if the water is sent by the water pump, If the opening degree setting of one of the flow regulating valve on the inlet side or the flow regulating valve on the outlet side of the clean water tank is set, then the opening degree setting of these water valve openings is discrete, which is equivalent to one pump. The output flow rate is not lost. In addition, the method of thinking of the water treatment program is to determine the amount of water to be sent from the water purification plant 40 to each of the distribution tanks 41 to 48 according to the needs of the water distribution. The amount of washing water in the filtration tank or other water used in the plant is also determined. Contained in the need for water distribution, the clean pool is determined according to the distribution pool, the filtration pool is determined according to the water pump, and the stopping of the filtration pool at the time of cleaning is the same as the restriction of the water pump, so it can almost correspond to the total filtration flow. Planning, so you can also optimize the total filtration flow plan in the water purification plant. Of course, the same way of thinking can be used to coordinate the optimization from the water delivery to the total filtration flow plan. The amount of water Q p (k) sent from the clean water tank to the distribution tank at a certain time k is controlled by the activated water supply pump or a water valve installed in the piping. Its purpose -15- (12) (12) 200419046 The target flow is set in discrete and in several stages. This flow is called the flow phase. For example, whether it is a water valve or a pump, the physical maximum of the water supply is assumed to be 100m3 / h. If this area is divided into 5 areas, the flow rate is 0, 20 '40, 60, 100 m3 / h. Represents the amount of water delivered by the water supply plan. Currently, only fixed-speed pumps with a fixed speed (swing speed) are used to send water. The number of fixed-speed pumps and the water supply obtained by the water supply plan are discrete. 1. In order to make the explanation simple, this case is assumed to be explained below. Take the amount of water used per unit time and its variable array of time as genes, and only use the flow rate and the time of the flow rate to be changed. 'Genetic calculations using the variable array as a variable-length gene array are stored. The genes are expressed as shown in Figure 3. Fig. 3 shows the water inflow of each facility (distribution tanks 41 to 48) shown in Fig. 2 at each time. Now when there are N distribution pools (8 in the example shown in the figure), the distribution pools are sequentially added with a drawing number 'from 1'. This drawing number is called the facility number. The water intake plan for facility number i is expressed in groups of (time, flow rate). Think of this group as genes. Therefore, time is taken as an integer 从 from the planned start time to the end time; the flow phase is based on the predetermined flow 1 amount: discrete target 値 to take the real number 値. The genome of this gene according to the present invention does not exist only for the time period that should be planned, but only holds at the time when the water intake volume is changed. Therefore, a set of genes can be used when the amount of water is constant, which can save the memory capacity. As a result, the scope of the most suitable target area for exploration can also be reduced. Thus, the length of the genome (column) in the present invention -16- (13) (13)
200419046 爲可變，此基因長度稱爲可變長度基因列。可 列若爲如同第4圖的.進水計劃假設爲設施編號 如同第3圖的設施編號1來表現。 現在，各配水池4 1〜4 8之進水計劃量總和 4 0的送水計劃量，所以將想像中送水計劃的 如下述加以公式化。公式化的方法係依欲將那 劃最適化而變化所以並不是隨意，不過即使那 化，若是以下的組合稱爲最適化問題之公式化 每單位時間的運用水量及其流量之時刻的變數 ，只有運用流量及其流量的時刻被更改時’依 數組之當作可變長度的基因列使用之基因運算 化或近乎最適化。 <數式1 > 目的函數： minimize f =W1xf1+W2xf2-^W3x /., 此處 Λ =3 ^ \k) x -/\k) + vv2. (k) x (k) + >v3/ (k) x h. (k) 八=i ('⑷x ^丨㈨)、 凡㈨> 』之時 ί^ί k^b 凡·⑷sC//G J之時 /3 =名|2 也㈨)-β(#-1))1 f:目的函數 X, ( k):以設施編號i的時刻k的離散 水量階段 (若只有定速幫浦，與幫浦運轉台數同値： 變長度基因 1，則可以 爲從淨水廠 最適化問題 一種送水計 一種的公式 ，則把該日 組當作基因 照存有該變 ，就能最適 …⑴ …⑵ …（3) …⑷ …⑸ 性決定之送 -17- (14) 200419046200419046 is variable, and the length of this gene is called a variable-length gene sequence. Can be listed as shown in Figure 4. The water intake plan is assumed to be represented by the facility number as shown in Figure 3. At present, the sum of the water supply plan amounts of each of the distribution tanks 41 to 48 is 40, so the imaginary water supply plan is formulated as follows. The method of formulating is to change the optimization of the plan as desired, so it is not random, but even if it is a combination, the following combination is called the optimization problem. The formula for the amount of water used per unit of time and the variable of the flow rate is only When the flow rate and the moment of the flow rate are changed, the genetic algorithm used as a variable-length gene array according to the array is calculated or nearly optimized. < Equation 1 > Purpose function: minimize f = W1xf1 + W2xf2- ^ W3x /., where Λ = 3 ^ \ k) x-/ \ k) + vv2. (k) x (k) + > v3 / (k) x h. (k) eight = i ('⑷x ^ 丨 ㈨), where ㈨㈨> ^ 时 ί ^ ί k ^ b ⑷ · SC /// GJ Hour / 3 = name | 2 also ㈨ ) -β (#-1)) 1 f: objective function X, (k): discrete water quantity phase at time k at facility number i (if there is only fixed-speed pump, the same as the number of pumps running: variable-length gene 1, it can be a formula for one type of water meter from the optimization problem of the water purification plant. If the day group is stored as a genetic photo with this change, it can be optimal ... ⑴ ⑴ ⑵ （(3) ⑷ ⑸ ⑸ ⑸ Giveaway-17- (14) 200419046
Yi ( k ) = | Xi ( k ) - Xi ( k - 1 ) I :以設施編號 i 的時刻k的離散性決定之送水量變更階段數（若只有定速 幫浦，與幫浦運轉變更台數同値） h, ( k ):設施編號i的時刻k之配水池水位 hA, ( k):設施編號i的時刻k之目標水位 CHG — P, ··設施編號i每單位時間的離散性決定之送水 量變更最大階段數。（若只有定速幫浦，與幫浦最大變更 台數同値） T :最終時刻。Yi (k) = | Xi (k)-Xi (k-1) I: Number of water supply volume change stages determined by the discreteness of time k at facility number i (if there is only a fixed speed pump, and the pump operation change station Numbers are the same) h, (k): water level of the distribution pool at time k at facility number i, (k): target water level at time k at facility number i CHG — P, ·· The discreteness of unit number i per unit time is determined The maximum number of stages of water supply change. (If there is only a fixed-speed pump, it is the same as the maximum number of pumps to be changed.) T: The final moment.
Wji ( k ):設施編號i的時刻k之最適化加權値，j = 1〜4Wji (k): the optimal weighting at time k for facility number i, j = 1 ~ 4
Wm :促使廣範圍最適化的最適化加權値，m = 1〜3 Q i ( k ):設施編號i的離散性決定之送水量階段的 送水量（m3 / h ) (若只有定速幫浦，與幫浦k台的送水量（m3/h )同 <數式2 > …⑹ …（7)Wm: The optimization weighting that promotes wide-range optimization, m = 1 ~ 3 Q i (k): The water supply volume (m3 / h) at the water supply stage determined by the discreteness of the facility number i (if only constant speed pumps) ， It is the same as the water supply volume (m3 / h) of the pump station K < Equation 2 >… ⑹… (7)
A hi ( k):設施編號i的時刻k之配水池水位下限値 (m ) h i_ ( k ):設施編號i的時刻k之配水池水位上限 値（m ) -18 - (15) (15)200419046 Q« ( k ):設施編號i的離散性決定之送水量階段的 送水量（m3 / h ) (若只有定速幫浦，與幫浦k台的送水量（m3/h )同 値） qdi ( k):設施編號i的時刻k之需要預測値（m3/h )A hi (k): lower limit of water level of distribution pool at time k at facility number i (m) h i_ (k): upper limit of water level of distribution pool at time k at facility number i (m) -18-(15) (15 ) 200419046 Q «(k): The water supply volume (m3 / h) of the water supply volume stage determined by the discreteness of the facility number i (if there is only a fixed speed pump, the same as the water supply volume of the pump k station (m3 / h)) qdi (k): the forecast of the time k of the facility number i (m3 / h)
Ai :設施編號i的配水池底面積（m2 ) 表現設施編號i的進水量計劃之基因（時刻、流量階 段）與上述的目的函數和限制條件之變數的對應爲： (時刻、流量階段）=(k、X i ( k )) 這種問題一般稱爲組合最適化問題，如第3圖所示， 把該日每單位時間的運用水量及其流量之時刻的變數組當 作基因，只有運用水量及其流量的時刻被變更時，依照存 有該變數組之當作可變長度的基因列使用之基因運算，就 能高速地取得最適當或接近最適當的運用計劃。其運算程 序如下所示。 基因運算 < step-Ι >初期個體群的生成 從分別預先已定義之個體的個數中產生隨機分配基因 生成的個體。不滿足限制條件時重新隨機分配基因來生成 < step-2〉各個體的評比 -19- (16) (16)200419046 計算各個體的的適應度f及平均適應度。 < s t e p - 3 >淘汰處理 存有未滿足限制條件的個體或預先已定義的個體數以 上的個體時，直到適應度較差（適應度小）的個體成爲該 已定義的個數爲止才淘汰（剔除）° < step-4〉增殖處理 個體數比預先已定義的個體數還少時，增殖適應度最 好的個體（複製）。 < step-5>交叉處理 隨機進行配對。配對只對與全個體數成比例（交叉率 )的份量才進行，每對都隨機選擇基因座（基因的地方） 使其成交叉（從所選擇基因的地方交替變換基因的設定） < step-6>突然變遷處理 只對與全個體數成比例（突然變遷率）的份量隨機選 擇個體，使任意（隨機決定）基因座的基因變更。 < step-7 >結束判定處理 反覆 <step-2> 〜<step-6>。只不過在於 <step-2> ，該世代的平均適應度與前述世帶的平均適應度作比較， -20- (17) (17)200419046 若爲一定値ε (任意的設定値）以上或一定世代數（反覆 次數的上限値）以上則結束運算。 另外，若是<step-l>〜<step-7>在預先已定義的時 間內仍未結束則強制結束運算。 然而，個體係指第3圖所示的1個基因列。 此處，說明< step-5 >所述的「交叉」。如第5圖所 示，隨機選擇2個設施編號相同（圖示例子爲「2」）的 基因列。進而，以亂數來選擇該基因組的地方，即是以亂 數來選擇有（時刻、流量階段）一定個數的地方並作更換 。第5圖的例子則是在第3個與第4個之間作更換。此時 ，更換過的結果：會有生成含有相同時刻以後的基因之基 因列的情形（第5圖中相對於位於下方的子之1 1時以後 的流量階段爲5時，則存有1 0時以後的流量階段6之基 0 )。此情況，之後所附加的基因優先。也就是之前出現 白勺相同時刻之基因爲之後的基因優先而被忽略（第5圖中 從1 〇日寸至1 1時爲止’流重階段爲6，1 1時以後’流量階 段爲5)。此點是對全體的個體數依照已被決定的比率來 操作。此點稱爲交叉。 進而，在< step-6 >所述過之突然變遷，如第6圖所 禾，一定個體（父）被隨機選擇時，例如在基因組的最後 ，新的基因（1 5、4 )隨機被突然追加、變換則稱爲突然 變遷。突然變遷係依照全體的基因當中被決定的比例（突 然變遷率）來操作。針對交叉及突然變遷所增加的基因組 ，在時刻重疊時，之後附加的基因組優先。 -21 - (18) (18)200419046 針對基因運算的< step-1 >，生成初期個體之際，無 法在初期個體生成計時器部1 2已設定的時間內生成能滿 足數式（6 )的限制條件之個體時，停止生成亂數的初期 個體，在於探試初期個體生成部1 6依照以下的順序’用 試探（根據經驗所發現的）來生成初期個體。 若在作成計劃之對象時間帶的先頭時刻，例如在該曰 〇時到翌日〇時止之2 4小時（1日）的計劃，則可能會有 把該日〇 : 〇 〇時間點的進水量當作初期値，每1日都以該 f 原本的初期値來進水的情形。如果配水池的水位在於全部 對象時間內滿足數式（6 )則把該配水池的進水流量當作 初期個體。若已知在一定時刻仍未滿足數式（6 )則變更 進水流量來滿足數式（6 )，之後的時間帶設成變更後的 流量計劃。然而，此操作在全時間內直到滿足數式（6 ) 爲止持續變更。然而，若滿足數式（6 )的進水流量作成 一個來作爲初期個體，則即使生成一定個數η的個數亦可Ai: The floor area (m2) of the distribution pool at facility number i. The correspondence between the gene (time, flow stage) representing the water intake plan of facility number i and the variables of the above-mentioned objective function and constraints: (time, flow stage) = (k, X i (k)) This type of problem is generally called a combination optimization problem. As shown in Figure 3, the variable amount of water used per unit time and its flow rate at that day is regarded as a gene. Only When the time of use of the amount of water and its flow rate is changed, the most suitable or close to the most suitable operation plan can be obtained at high speed in accordance with the genetic calculation used as a variable-length gene array in which the variable array is stored. The operation procedure is shown below. Gene calculation < step-I > Initial generation of individual populations Individuals generated from randomly assigned genes are generated from the number of individuals that have been defined in advance. When the constraints are not met, the genes are redistributed randomly to generate < step-2> evaluation of each individual -19- (16) (16) 200419046 calculate the fitness f and average fitness of each individual. < step-3 > When eliminating individuals with unsatisfied restrictions or having more than a predefined number of individuals, they will not be eliminated until the poor fitness (small fitness) individuals become the defined number. (Cut Out) ° < step-4> When the number of proliferation-treated individuals is smaller than the number of individuals that have been defined in advance, the individuals with the best proliferation fitness (duplication). < step-5 > Cross processing Pairing is performed randomly. Pairing is performed only for weights that are proportional to the number of individuals (crossover rate), and each pair randomly selects the loci (where the genes are) to make them cross (alternatively changes the setting of the genes from the place of the selected genes) < step -6 > Sudden transition processing only selects individuals at random in proportion to the total number of individuals (sudden transition rate), so that genes at arbitrary (randomly determined) loci are changed. < step-7 > End determination processing Repeat < step-2 > to < step-6 >. It is only in < step-2 > that the average fitness of this generation is compared with the average fitness of the previous generation, -20- (17) (17) 200419046 if it is a certain 値 ε (arbitrarily set) or more The calculation ends after a certain number of generations (the upper limit of the number of iterations 値). In addition, if it is < step-l > to < step-7 > that has not ended within a predetermined time, the operation is forcibly ended. However, each system refers to one gene column shown in FIG. 3. Here, "cross" described in < step-5 > will be described. As shown in Figure 5, two gene sequences with the same facility number (the example shown in the figure is "2") are randomly selected. Furthermore, the random number is used to select the place of the genome, that is, the random number is used to select a certain number of places (time, flow stage) and replaced. The example in Figure 5 is a replacement between the third and fourth. At this time, the result of the replacement: there may be cases where a gene sequence containing genes after the same time is generated. The base of the flow stage 6 after the hour is 0). In this case, the gene added later takes precedence. That is, genes that appeared at the same moment before are ignored for later genes (from the 10th inch to 11 o'clock in Figure 5, the 'flow weight stage is 6, and after 11 o'clock, the flow stage is 5) . This point is based on the ratio of the total number of individuals. This point is called crossing. Furthermore, in the sudden change described in < step-6 >, as shown in Fig. 6, when a certain individual (parent) is randomly selected, for example, at the end of the genome, a new gene (1, 5, 4) is randomly selected Sudden additions and changes are called sudden changes. Sudden transitions operate according to a determined ratio (the abrupt transition rate) among all genes. For cross-over and sudden changes in the genome, when the time overlaps, the additional genome is given priority. -21-(18) (18) 200419046 For gene operation < step-1 >, when initial individuals are generated, it is not possible to generate a formula that satisfies the formula (6) within the time set by the initial individual generation timer unit 12 In the case of individuals with restrictions of), the initial individuals who stop generating random numbers are in the initial stage of the trial. The initial unit 16 generates the initial individuals by trial (found through experience) in the following order. If the first time of the target time zone of the plan is made, for example, a plan of 24 hours (1 day) from 00:00 to the next day, there may be water inflow at the time of 0: 00 It is regarded as an initial stage, and water is fed into the initial stage of f every day. If the water level of the distribution pool satisfies Equation (6) in all the target time, the inflow of the distribution pool is regarded as the initial individual. If it is known that the formula (6) is not satisfied at a certain time, the inlet water flow is changed to satisfy the formula (6), and the subsequent time zone is set to the changed flow plan. However, this operation is continuously changed until the expression (6) is satisfied throughout the time. However, if one of the inflow flows satisfying the formula (6) is made as an initial individual, then even a certain number η can be generated.
如< step-Ι >所示，不只是隨機生成初期個體，也可 以由過去的實績値轉換到基因列，只有預先已定義的個數 載入來作爲初期個體。 對於廣範圍工廠的最適當運用控制裝置，能夠以人爲 由資料輸入部2輸入。上述的例子是由實績値轉換到基因 列來載入作爲初期個體，不過此處藉由操作員等的使用者 以人爲輸入，能夠讓有效地以當作可變長度的基因列使用 之基因運算所算出的運用計劃接近最適當的解。 -22- (19) 200419046 以當作可變長度的基因列使用之基因運算 用計劃並不是最適化，而是近乎最適當解的近 時’將該解當作其他最適化的初期値來輸入就 適當的解。最適當的手法列舉有分歧限定法或 ’不過並沒有特別的限定。 運用計劃部1 〇運算過的運用計劃在預定 有完成時、或未滿足預定的工廠狀態的上下限 由限制緩和運用計劃部2 0，將預定的工廠狀 範圍從限制條件中除去，改而將物理上的上下 在新的限制條件中，因而緩和限制條件；進而 最適化運算中的目的函數，包括運用計劃的最 的目的函數中對脫離工廠狀態的上下限範圍之 補償，就能夠運算運用計劃。 解決此最適化問題之運算在預定的時間內 分鐘內沒有完成時、或是未滿足預定的工廠狀 範圍（例如，配水池的運用水位上下限値）時 廠狀態的上下限範圍之限制條件緩和，運用計 運算下的目的函數中對脫離工廠狀態的上下限 量加以補償，依照以下的方式再度運算最適當 <數式3 > 限制條件：除去改而以新的 所得到的運 似最適當解 可以算出最 動態計劃法 的時間內沒 範圍時，藉 態的上下限 限範圍設定 運用計劃的 適化運算下 脫離量加以 ，例如在 1 態的上下限 ，預定的工 劃的最適化 範圍之脫離 的運用計劃 -23 - (20) (20)200419046 只不過' MM分別表示物理上的上下限範圍，設 flpi (^) ^ - hpi (k) ^ h{(k) 反一 ° 另外，在數式（2 )的目的函數 f!中加入數式（9 ) ，變爲數式（8 )。 <數式4 > 八》名％ X (名(vvh·⑹X么·(小％ (Λ) X L (小νν3ί⑷X卜办Μ 只不過W」i ( k ):設施編號I的時刻k之最適化權値 j，j = 5、6。 關於在於基因運算的< S t e p - 1 >所得到的最適當或近 乎最適當之運用計劃的一定時間帶，成爲如同第7 ( a ) 圖左側所示的凹狀計劃時（在一定時間帶變化爲向下凸的 圖形時）、成爲如同第7 ( b )圖左側所示的凸狀計劃時 (在一定時間帶變化爲向上凸的圖形時）、成爲如同第7 (c )圖左側所示的向上階梯狀計劃時、成爲如同第7 ( d )圖左側所示的向下階梯狀計劃時，若是促使那時的最適 化之評比値及限制條件（目的函數或適應度）沒有惡化’ 則能夠用平滑化部來對各計劃施予平滑化處理，使第7 ( a )圖〜第7 ( d )圖分別從左圖變爲右圖的狀態。 因而，平滑化部分別爲凹狀平滑化部22 (參照第7 ( a )圖）、凸狀平滑化部24 (參照第7 ( b )圖）、向上階 梯平滑化部2 6 (參照第7 ( c )圖）、及向下階梯平滑化 部2 8 (參照第7 ( d )圖）。這些平滑化部爲了讓把該曰 -24- (21) (21)200419046 每單位時間的運用水量及其流量的時刻之變數組當作基因 ，只有運用水量及其流量的時刻被更改時，存省該變數組 之當作可變長度的基因列使用之基因運算更有效作用，而 爲已導入的探試性的操作；主要在於數式（2 )中含有y， (k )的項發揮改善效果。特別是重要點爲只有促使那時 的最適化之評比値及限制條件沒有惡化時，且即使平滑化 仍滿足限制條件時，進行平滑化，不是這種情況則不進行 平滑化。 工廠中存有局部控制的控制器，且含有工廠運用控制 裝置無法控制的機器時，能具備在工廠運用控制裝置模擬 局部控制，將其結果輸入工廠運用計劃部來運算工廠的運 用計劃之局部控制模擬計劃部3 0。 例如，如第8圖所示，在淨水池5 0至配水區之5 3之 間，配置第1送水幫浦群6 1、第1配水池5 1、第2送水 幫浦群6 2、第2配水池5 2、及配水幫浦群6 3 ;進而將從 淨水池5 0送到第1配水池5 1的送水計劃最適化時，對於 相當於時刻k的需要量支配水區5 3的配水需要預測値qd; (k )則是現在從第1配水池5 1送到第2配水池5 2之送 水幫浦的啓動停止之控制與送水計劃無關，受到局部控制 〇 此情況，對於第2配水池5 2的需要預測値若進行需 次預測即可取得’不過相當於對第1配水池5 1的需要預 測的量（從第1配水池5 1送到第2配水池5 2的送水量） 則不知道。因此，第2送水幫浦群62的控制時序有種種 -25- (22) 200419046 的想法’不過對於極單純地以適當定義第2配水池5 2 水位h 2 ( k )之關於第2配水池5 2的水位h 2 ( k )的上 臨界値H m a X及下限臨界値H m i η，若h 2 ( k ) > H m a X 讓第2送水幫浦群6 2全部停止，若h 2 ( k ) < H m i n則 第2送水幫浦群62全部啓動。K有時是1分鐘週期有 是5分鐘週期，設定適當的計算週期。然而，計算第2 水池5 2的水位h2 ( k )使用數式（7 )即可。 經由此方式就會知道第2送水幫浦群6 2該在那個 間點啓動或停止’所以調整模擬運用計劃部1 〇所計算 最適當運用計劃的之時間上的計劃單位（一小時爲單位 2 4小時份量、3 0分鐘爲單位的2 4小時份量等的〇〇單 )而得到的結果，輸入到運用計劃部1 〇，則可以得到第 送水幫浦群6 1的最適當或近乎最適當的送水計劃。 例如，在時刻1 : 34以l〇〇rn3/h第2送水幫浦群 運轉至3 : 0 0爲止時，輸入給運用計劃部1 〇則〗點鐘 圍內爲 lOOx ( 60— 34) /60=43.3m3/h，2 點鐘範圍內 1 0 0m3/h。這種運算在必要的時間部分進行，則會知道 第1配水池5 1依照需要預測値送至第2配水池5 2之送 量，就能夠運算最適當或近乎最適當的送水計劃。 如此，即使存有局部控制的控制器，且含有工廠運 控制裝置無法控制的機器時，在裝置內模擬局部控制， 其結果輸入運用計劃部1 〇，就能夠運算最適當或近乎 適當的送水計劃。 比如配水池與設置送水幫浦的地方隔離數km遠時 的 限 則 讓 時 配 時 的 的 位 6 2 範 爲 從 水 用 將 最 -26- (23) 200419046 就會發生控制上的問題。例如，淨水從送水幫浦的設置地 方送到配水池要1 〇分鐘時，送水量在時刻1 1 ·· 〇 〇時沒有 從100 m3/h增加到150 m3/h則配水池的水位低於下限時 ’必須在1 0分鐘之前推測該量來增加送水量。 因此，如第9 ( a )圖所示，以1 0分鐘爲單位進行需 胃預測時，若是淨水從送水幫浦的設置地方送到配水池要 1 〇分鐘，則如如第9 ( b )圖所示，單純地以1 0分鐘即是 以1階段份量先送水的量作爲需要預測値。第9 ( a )圖 f 表示當初的需要預測値；第9 ( b )圖表示1階段份量先 送水的狀態，即是表示圖上向左側偏移1階段份量的狀態 °如果淨水從送水幫浦的設置地方送到配水池要2 0分鐘As shown in < step-I >, not only the initial individuals are randomly generated, but also the past performance can be converted to the gene sequence, and only a predefined number is loaded as the initial individuals. The most suitable operation control device for a wide-range factory can be input by the data input unit 2 by human. The above example is converted from the actual performance to the gene sequence to be loaded as an initial individual. However, by using human input by a user such as an operator, the gene can be effectively used as a variable-length gene sequence. The operation plan calculated by the calculation is close to the most appropriate solution. -22- (19) 200419046 The genetic algorithm for use as a variable-length gene sequence is not an optimization, but a near-term solution of the most appropriate solution. The solution is input as the other initial phase of the optimization. Just the right solution. The most suitable method is a divergence limitation method or ', but it is not particularly limited. When the planned operation plan calculated by the operation planning department 10 is completed, or the upper and lower limits of the factory state are not met, the operation planning department 20 will be relaxed by the restriction, and the planned plant-like range will be removed from the restrictions and replaced by The physical up and down are in the new restrictions, so the restrictions are eased; furthermore, the objective function in the optimization operation, including the compensation of the upper and lower limits of the factory state in the most objective function of the application plan, can calculate the operation plan. . When the operation to solve this optimization problem is not completed within a predetermined time and minutes, or the predetermined plant-like range is not met (for example, the upper and lower limits of the operating water level of the distribution tank), the restrictions on the upper and lower limits of the plant state are relaxed. In order to compensate for the upper and lower limits of leaving the factory state in the objective function under calculation, the calculation is most appropriate in the following manner < Equation 3 > Restrictions: It is most appropriate to remove and use the new obtained operation. When the solution can calculate the most dynamic planning method within the time limit, the upper and lower limits of the borrowed state are set using the plan's adaptive calculation to deduct the amount, for example, at the upper and lower limits of the 1 state, the optimized range of the predetermined schedule The use of the plan -23-(20) (20) 200419046 except that 'MM represents the upper and lower limits of the physical, respectively, let flpi (^) ^-hpi (k) ^ h {(k) inverse ° ° In addition, in The objective function f! Of the formula (2) is added into the formula (9), and becomes the formula (8). < Equation 4 > Eight '' name% X (name (vvh · ⑹X 么 · (小 ％ (Λ) XL (小 νν3ί⑷X 卜 办 Μ) but W ″ i (k): the best time for facility number I Turning right jj, j = 5, 6. Regarding a certain time period of the most appropriate or nearly the most appropriate use plan obtained by < S tep-1 > in the genetic operation, it becomes like the left side of Fig. 7 (a) The concave plan shown (when changing to a downwardly convex figure at a certain time zone), and the convex plan shown at the left of Figure 7 (b) (when changing to a convexly upward figure at a certain time zone) ). When it becomes the step-up plan as shown on the left side of Fig. 7 (c), and when it becomes the step-down plan as shown on the left side of Fig. 7 (d), if the evaluation of the optimization at that time is promoted and Constraints (objective function or fitness) have not deteriorated ', the smoothing section can be used to smooth each plan, and the graphs 7 (a) to 7 (d) can be changed from left to right, respectively. Therefore, the smoothing sections are a concave smoothing section 22 (see FIG. 7 (a)) and a convex smoothing section 24 (see Figure 7 (b)), step-up smoothing section 2 6 (see Figure 7 (c)), and step-down smoothing section 2 8 (see Figure 7 (d)). These smoothing sections The day -24- (21) (21) 200419046 per unit of time is used as a gene for the change in the amount of water and its flow rate, and only when the time of the use of water and its flow is changed, save the variable array The genetic algorithm used as a variable-length gene sequence is more effective, and it is an exploratory operation that has been introduced; it is mainly because the terms containing y and (k) in equation (2) exert an improvement effect. It is particularly important The point is that smoothing is performed only when the evaluation that promoted the optimization at that time and the limiting conditions have not deteriorated, and when the smoothing conditions still meet the limiting conditions. Otherwise, smoothing is not performed. In the factory, there are local controls. If the controller contains a device that cannot be controlled by the plant operation control device, it can be provided with a local control simulation planning section 30 that simulates local control at the plant operation control device and inputs the results to the plant operation planning section to calculate the plant operation plan. For example, as shown in FIG. 8, the first water supply group 6 1, the first water supply group 5 1, and the second water supply group 6 2 are arranged between the clean water pool 50 and the water distribution area 5 3. The second distribution pool 5 2 and the distribution pump group 6 3; when the water distribution plan from the clean pool 50 to the first distribution pool 51 1 is optimized, the water area 5 is dominated by the demand equivalent to the time k (K) is now the control of the start and stop of the water delivery pump from the first water distribution pool 51 to the second water distribution pool 52, which is not related to the water delivery plan, and is subject to local control. In this case, The demand forecast for the second distribution pool 5 2 値 If you need to make a prediction, you can get 'but it is equivalent to the forecast of the first distribution pool 51 1 (from the first distribution pool 51 to the second distribution pool 5 2) The amount of water delivered) is unknown. Therefore, the control timing of the second water supply pump group 62 has various ideas of -25- (22) 200419046 'But for the second distribution pool, the definition of the second distribution pool 5 2 water level h 2 (k) is very simple. Upper limit 値 H ma X and lower limit 値 H mi η of the water level h 2 (k) of 5 2 if h 2 (k) > H ma X stop all the second water pump group 6 2 if h 2 (k) < H min, the second water pump group 62 is all started. K may be a 1-minute cycle or a 5-minute cycle, and an appropriate calculation cycle is set. However, to calculate the water level h2 (k) of the second pond 5 2, the equation (7) may be used. In this way, we will know that the second water pump group 6 2 should start or stop at that point. So adjust the planning unit (the hour is 2) for the most appropriate use plan calculated by the simulation operation planning department 10. 4 hours, 30 minutes, 24 hours, etc. (00 orders), and the results obtained by inputting to the operation planning department 10, you can get the most appropriate or nearly the most appropriate water pump group 61 Water delivery plan. For example, when the second water supply pump group is operated at 100: 3rn / h at time 1:34 until 3:00, it is inputted to the operation planning department, and the number within the circle is 100x (60—34) / 60 = 43.3m3 / h, 100m3 / h within 2 o'clock. This calculation is performed in the necessary time part, and then the first distribution pool 51 can predict the amount of water sent to the second distribution pool 52 according to needs, and can calculate the most appropriate or nearly the most appropriate water distribution plan. In this way, even if there is a local control controller and there are machines that cannot be controlled by the plant transportation control device, the local control is simulated in the device, and the result is input to the operation planning unit 10 to calculate the most appropriate or nearly appropriate water delivery plan. . For example, a time limit of several kilometers when the distribution pool is separated from the place where the water pump is installed. The time limit of the time distribution is 62. The range is -26- (23) 200419046, which will cause control problems. For example, when it takes 10 minutes for the purified water to be sent to the distribution pool from the place where the water pump is installed, and the water supply volume does not increase from 100 m3 / h to 150 m3 / h at time 1 1 ·· 00, the water level of the distribution pool is low At the lower limit, the amount must be estimated before 10 minutes to increase the water supply. Therefore, as shown in Fig. 9 (a), when the gastric demand prediction is performed in units of 10 minutes, if it takes 10 minutes for the purified water to be sent from the installation place of the water pump to the distribution pool, as in Fig. 9 (b As shown in the figure, simply taking 10 minutes is the amount of water to be sent first in one step as the demand prediction. Figure 9 (a) f indicates the original need forecast; Figure 9 (b) shows the state of the first stage of the water supply, that is, the state of the stage is shifted to the left by the first stage of the ° ° If the purified water from the water supply help It takes 20 minutes for the setting place in Pu to be sent to the distribution pool.
’貝U 20分鐘前先送水即是2階段份量先送水即可。然後 ’配合運用計劃部1 0所計算的最適當運用計劃的之時間 上的計劃單位（一小時爲單位的24小時份量、3 0分鐘爲 單位的24小時份量等的〇〇單位），將需要預測値加工 使用。例如，以1 0 : 0 0〜1 1 : 0 0之間的需要預測値P d i ( 1 1 )爲1小時單位，如上述過衡量時間延遲來求取時’首 先因原本假定爲以1 0分鐘單位來預測需要，所以如下表 所示，未時間延遲時以1 〇分鐘單位的需要預測値得到1 曰份量。 -27- (24) 200419046 表1 時間 10:00 〜 10:10 〜 10:20〜 10:30 〜 10:40〜 10:50〜 11:00 〜 延遲 10:10 10:20 10:30 10:40 10:50 11:00 11:10 沒有 110 120 130 120 130 130 120 有 120 130 120 130 130 120 應用此表求出1 〇點鐘範圍內、1小時份量的需要預 測値爲：110+120+130+120+130+130=610 現在1 〇分鐘的時間延遲，所以1 〇分鐘即是1階段先 送水，則 1 0點鐘範圍內、1小時份量的需要預測値爲： 120+130+120+130+130+120=620 操作1日份量的此値’將結果輸入運用計劃部。 經由此方法，從工廠到會有時間延遲的地方’即使從 工廠輸送必要的水量時’仍能作成已考量時間延遲在內之 運用計劃。 關於最適當運用控制裝置所作成的運用計瘦1 ’定期監 視各設施的水位計劃是否在一定的範圍’在範圍內則繼續 監視，不在範圍內則首先只有該設施重新計劃。其結果： 若滿足其他設施的限制條件就根據重新計劃結果來運用 廣範圍工廠。若不滿足其他設施的限制條件則包括位於該 設施上游的設施逐一反覆重新計劃。因而能防止因脫離數 個設施的計劃及實績所造成的頻繁修正計劃。 另外，在於最適當運用控制裝置，具備操作員依照所 -28- (25) 200419046 期望的運用設定値來檢討工 確認是否滿足如同數式（6 件（例如，池中的運用上下 想。此模擬部爲用資料輸入 量，以配水量的需要預測値 模擬之後的水位變化。因此 用上下限水位範圍，而提供 何處理之有效資料。然而， 算出的最適當運用計劃來修 劃部1 0的運算結果即可。 <其他的實施形態> 上述過的最適當運用控 明過，不過本發明不僅能用 區的幫浦廠到廢水處理廠之 留設施在運轉雨水排水幫浦 當然能使用，也同樣能用於 溫水或蒸氣之熱源機器、或 適當運轉，依據由複數個發 劣化或電力需要之最適當工 來水的淨工廠、配水池、水 廠、幫浦廠、水閘（或水閥 水閥對分別應於區間冷暧氣 ’自來水的流量，包括蒸氣 廠運用時，進行水收支模擬來 )所示之工廠運用上的限制條 限水位）之模擬部3 6較爲理 部2任意設定各配水池的進水 及配水池的現在水位爲依據， ，爲了確認是否在所設定的運 操作員用來檢討之後的運用如 操作員根據運用計劃部1 〇所 正之際的檢討也是設定運用計 制以自來水廠爲控制對象作說 於自來水廠，從散佈在廣大地 家庭排水等的流入量或雨水貯 所形成的雨水排水量之平滑化 區間冷暖氣工廠的用來供應冷 考慮到熱設施的能量效率之最 電所等所組成之發電廠的機器 廠自動控制。該情況，例如自 閥分別對應於廢水的廢水處理 );另外，自來水的淨工廠及 工廠之熱源工廠及水閥。進而 量等，一般廢水對應於廢水量 -29- (26) (26)200419046 ，區間冷暖氣對應於蒸氣量、溫水量、冷水量’電力對應 於電力量。 〔發明之效果〕 如上述過依據本發明，提供以複數個設施爲對象之廣 範圍工廠運用，使原水的取水量或淨工廠的總濾過量、從 淨工廠至配水池的送水量等之時間上的變動盡量降低’且 幫浦的耗電量也盡量降低，而實現高速最適化，又安定且 有效率的流體運用控制之廣範圍工廠的最適當運用控制裝 置。 【圖式簡單說明】 第1圖爲表示本發明的最適當運用控制裝置的一個實 施形態之方塊圖。 第2圖爲表示使用第1圖的最適當運用控制裝置之廣 範圍自來水廠的一種構成例之系統圖。 第3圖爲說明可變長度基因列的例子之圖。 第4圖爲說明表現可變長度基因列之進水流量的形像 之圖。 第5圖爲使用可變長度的基因列之基因運算的交叉方 法之說明圖。 第6圖爲使用可變長度的基因列之基因運算的突然變 遷方法之說明圖。 第7圖爲說明凹狀、凸狀、向上階梯狀、及向下階梯 -30- (27) (27)200419046 狀的計劃之圖。 弟8圖爲g兌明包含局部控制的處理程序例之圖。 第9圖爲說明時間延遲修正部之圖。 元件對照表 2 :資料輸入部 4 :資料輸出部 6 :實績DB部 8 :需要預測部 1 〇 :運用計劃部 1 2 :初期個體生成計時器部 1 4 :實績初期個體生成部 1 6 :探試初期個體生成部 1 8 :合成最適化部 20 :限制緩和運用計劃部 22 :凹狀平滑化部 24 :凸狀平滑化部 26 :向上階梯平滑化部 28 :向下階梯平滑化部 3 0 :局部控制模擬計劃部 3 2 :時間延遲修正部 3 4 :重新計劃判定部 3 6 :模擬部 4 0 :淨水廠 -31 - (28) (28)200419046 4 1〜48 :配水池 5 1 :節點 5 2 :水閥 5 3·· 流量計′ 贝 U Water is delivered first 20 minutes ago. Then, in conjunction with the planning unit (the unit of 24 hours for one hour, the unit of 24 hours for 30 minutes, etc.) calculated by the use planning department 10 for the most appropriate use of the plan, you will need Prediction 値 processing use. For example, if the demand prediction between 1 0: 0 0 to 1 1: 0 0 is used, P di (1 1) is a unit of 1 hour. When the time delay is calculated as described above, it is assumed to be 1 0 Minutes are used to predict the demand, so as shown in the table below, the demand is forecasted in units of 10 minutes without a time delay. -27- (24) 200419046 Table 1 Time 10:00 ～ 10:10 ～ 10: 20 ～ 10:30 ～ 10: 40 ～ 10: 50 ～ 11:00 ～ Delay 10:10 10:20 10:30 10: 40 10:50 11:00 11:10 No 110 120 130 120 130 130 120 Yes 120 130 120 130 130 120 Use this table to find the forecast of the amount of 1 hour in the range of 10 o'clock: 110 + 120 + 130 + 120 + 130 + 130 = 610 Now the time delay is 10 minutes, so 10 minutes is the first stage of water delivery, so the forecast of the 1-hour portion in the range of 10 o'clock is: 120 + 130 + 120 + 130 + 130 + 120 = 620 This result is input to the operation planning department for the 1-day serving. In this way, from the factory to the place where there is a time delay, even when the necessary amount of water is delivered from the factory, it is possible to create an operation plan that takes into account the time delay. About the operation plan made by using the most appropriate control device 1 ′ Regularly monitor whether the water level plan of each facility is within a certain range ”and continue to monitor it. If it is not within range, only the facility is re-planned first. Result: If the constraints of other facilities are met, a wide range of plants will be used based on the results of the replanning. Failure to meet the constraints of other facilities includes re-planning of facilities located upstream of the facility one by one. As a result, frequent revisions to the plan due to plans and performance results from leaving several facilities can be prevented. In addition, it is the most appropriate use of the control device, and the operator has to check whether the operator satisfies the formula (6 pieces (for example, the application in the pool) according to the desired operation setting of -28- (25) 200419046. This simulation The department uses data input to predict the water level change after the simulation based on the need for water distribution. Therefore, it uses the upper and lower limits of the water level range to provide effective data on how to handle it. However, the calculated most appropriate application plan is used to modify the department's 10 The calculation result is sufficient. ≪ Other embodiments > The above-mentioned most suitable operation control has been used, but the present invention can be used not only in the district's pump plant to the wastewater treatment plant's retention facility for operating the rainwater drainage pump, of course. , Can also be used for warm water or steam heat source machines, or properly operated, according to a number of clean factories, distribution pools, water plants, pump plants, sluices (or water plants) The water valve and the water valve pair should be separately cooled in the interval of the tap water's tap water flow, including the water revenue and expenditure simulation when the steam plant is used. Water level limit) The simulation department 3 6 compares with the management department 2 to arbitrarily set the inlet water of each distribution pool and the current water level of the distribution pool as the basis. In order to confirm whether the set operation operator is used to review the operation after the review, such as the operator Based on the review at the time of the operation planning department 10, the use of a water meter is also set to control the water plant as a control object. The water plant is smoothed from the inflow of household drainage, etc., or rainwater storage, which is distributed in a large area. The district heating and cooling plant is used to supply the cooling power of the power plant, which takes into account the energy efficiency of the heating facility. The plant is automatically controlled by the power plant. In this case, for example, the valve corresponds to the wastewater treatment of wastewater); in addition, tap water Net factory and factory heat source factory and water valve. In addition, the amount of waste water generally corresponds to the amount of waste water -29- (26) (26) 200419046, and the interval cooling and heating correspond to the amount of steam, warm water, and cold water. The power corresponds to the amount of power. [Effects of the Invention] As described above, according to the present invention, a wide range of factories targeted at a plurality of facilities are provided, and the time for the raw water intake or the total filtration of the net factory to be excessive, and the time from the net factory to the water distribution tank, etc. Minimize changes and minimize the power consumption of the pump, and realize the most appropriate operation control device for a wide range of plants that optimizes high-speed, stable and efficient fluid operation control. [Brief Description of the Drawings] Fig. 1 is a block diagram showing an embodiment of the most suitable control device of the present invention. Fig. 2 is a system diagram showing a configuration example of a wide-range waterworks using the most suitable control device of Fig. 1. Fig. 3 is a diagram illustrating an example of a variable-length gene sequence. Fig. 4 is a diagram illustrating an image showing the inflow of a variable-length gene sequence. Fig. 5 is an explanatory diagram of a crossover method for gene arithmetic using a variable-length gene sequence. Fig. 6 is an explanatory diagram of a method of abrupt change in gene arithmetic using a variable-length gene sequence. Figure 7 is a diagram illustrating a plan of a concave shape, a convex shape, a step-up shape, and a step-down shape -30- (27) (27) 200419046. Figure 8 is a diagram of an example of a processing program including local control in ghuiming. Fig. 9 is a diagram illustrating a time delay correction unit. Component comparison table 2: Data input section 4: Data output section 6: Performance DB section 8: Need forecasting section 1 〇: Operation planning section 12: Initial individual generation timer section 14: Initial individual generation generating section 16: Probe Initial test individual generation unit 1 8: Synthesis optimization unit 20: Restriction mitigation operation planning unit 22: Concave smoothing unit 24: Convex smoothing unit 26: Up-step smoothing unit 28: Down-step smoothing unit 30 : Local control simulation planning unit 3 2: Time delay correction unit 3 4: Replan determination unit 3 6: Simulation unit 4 0: Water purification plant -31-(28) (28) 200419046 4 1 ~ 48: Distribution tank 5 1 : Node 5 2: Water valve 5 3 ·· Flowmeter
-32--32-
Claims (1)
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2003041423A JP4365598B2 (en) | 2003-02-19 | 2003-02-19 | Optimal operation control system for wide area plant |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| TW200419046A true TW200419046A (en) | 2004-10-01 |
| TWI273156B TWI273156B (en) | 2007-02-11 |
Family
ID=33025007
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| TW093103935A TWI273156B (en) | 2003-02-19 | 2004-02-18 | Wide range optimum factory control device |
Country Status (4)
| Country | Link |
|---|---|
| JP (1) | JP4365598B2 (en) |
| KR (1) | KR20040074626A (en) |
| CN (1) | CN1317676C (en) |
| TW (1) | TWI273156B (en) |
Families Citing this family (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4511435B2 (en) * | 2005-09-05 | 2010-07-28 | 三菱電機株式会社 | Water supply operation planning device |
| JP2007257190A (en) * | 2006-03-22 | 2007-10-04 | Toshiba Corp | Total monitoring diagnosis device |
| JP5019197B2 (en) * | 2006-03-31 | 2012-09-05 | 株式会社東芝 | Water distribution information management device |
| JP2009103028A (en) * | 2007-10-23 | 2009-05-14 | Toshiba Corp | Control device and control method of rain water pump |
| JP5416729B2 (en) * | 2011-03-22 | 2014-02-12 | 株式会社日立製作所 | Water central monitoring and control device, water monitoring control system and water monitoring control program |
| JP6327498B2 (en) * | 2013-02-20 | 2018-05-23 | 三菱重工業株式会社 | Solution search device, solution search method and program, schedule generation device, schedule generation method and program, and charging control system |
| JP5701919B2 (en) * | 2013-03-14 | 2015-04-15 | 株式会社東芝 | Reservoir water level setting device, distribution reservoir water level setting method and distribution reservoir water level setting system |
| KR101638553B1 (en) * | 2014-01-28 | 2016-07-20 | 한국농어촌공사 | a operation management technology for controlling pump and press using ICT |
| KR101656433B1 (en) * | 2014-01-28 | 2016-09-22 | 한국농어촌공사 | A water-loop system for allocationing and supplying multy-water source using ICT |
| CN104374060B (en) * | 2014-11-04 | 2017-02-15 | 清华大学 | Water pump system control method based on centerless network |
| CN114035432B (en) * | 2021-11-05 | 2023-09-08 | 扬州大学 | Pump station control method considering start-stop loss based on linear integer-dynamic programming combination algorithm |
| CN116757424B (en) * | 2023-06-21 | 2024-03-12 | 成都同飞科技有限责任公司 | Peak regulation control method and system for water supply network pool |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3352153B2 (en) * | 1993-06-17 | 2002-12-03 | 株式会社東芝 | Water distribution flow prediction device |
| JP3397475B2 (en) * | 1994-10-31 | 2003-04-14 | 株式会社東芝 | Water operation planning device |
| WO2000007113A1 (en) * | 1998-07-31 | 2000-02-10 | Cet Technologies Pte Ltd. | Automatic freeway incident detection system using artificial neural networks and genetic algorithms |
| JP2000342257A (en) * | 1999-06-08 | 2000-12-12 | Japan Science & Technology Corp | Method and apparatus for estimating bond site structure of control element of gene |
| JP3756705B2 (en) * | 1999-08-18 | 2006-03-15 | 株式会社東芝 | Wide-area optimum water operation equipment for waterworks plants |
| FI112856B (en) * | 2000-03-03 | 2004-01-30 | Kone Corp | Method and apparatus for passenger allocation by genetic algorithm |
-
2003
- 2003-02-19 JP JP2003041423A patent/JP4365598B2/en not_active Expired - Fee Related
-
2004
- 2004-02-18 TW TW093103935A patent/TWI273156B/en not_active IP Right Cessation
- 2004-02-18 KR KR1020040010758A patent/KR20040074626A/en not_active Application Discontinuation
- 2004-02-19 CN CNB200410006834XA patent/CN1317676C/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN1540460A (en) | 2004-10-27 |
| JP2004250961A (en) | 2004-09-09 |
| JP4365598B2 (en) | 2009-11-18 |
| CN1317676C (en) | 2007-05-23 |
| TWI273156B (en) | 2007-02-11 |
| KR20040074626A (en) | 2004-08-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Jiménez-Bello et al. | Methodology for grouping intakes of pressurised irrigation networks into sectors to minimise energy consumption | |
| TW200419046A (en) | Wide range optimum factory control device | |
| Darwazeh et al. | Review of peak load management strategies in commercial buildings | |
| Mirzaei et al. | Robust network-constrained energy management of a multiple energy distribution company in the presence of multi-energy conversion and storage technologies | |
| Shao et al. | The chiller’s electricity consumption simulation by considering the demand response program in power system | |
| Arunkumar et al. | Water demand analysis of municipal water supply using epanet software | |
| CN102331759A (en) | Optimization system using an iterative expert engine | |
| CN108805329B (en) | Method and system for realizing real-time scheduling of cascade reservoir | |
| CN112909933B (en) | Intraday rolling optimization scheduling method containing pumped storage unit under spot market environment | |
| Golmohamadi et al. | Hierarchical flexibility potentials of residential buildings with responsive heat pumps: A case study of Denmark | |
| Zhang et al. | An optimal regulation method for parallel water-intake pump group of drinking water treatment process | |
| CN115062904A (en) | Digital water supply pipe network scheduling method and system | |
| Kefer et al. | Simulation-based optimization of residential energy flows using white box modeling by genetic programming | |
| CN111626603B (en) | Oil field ordered electricity utilization system with genetic algorithm and ordered electricity utilization distribution method thereof | |
| AbdelMeguid | Pressure, leakage and energy management in water distribution systems | |
| Le Quiniou et al. | Optimization of drinking water and sewer hydraulic management: Coupling of a genetic algorithm and two network hydraulic tools | |
| CN116307094A (en) | Urban water supply optimal scheduling method based on multi-target particle swarm algorithm | |
| Fracasso et al. | Optimized control for water utilities | |
| CN111639796B (en) | Ordered oil field electricity utilization system with topological structure and ordered electricity utilization distribution method thereof | |
| Kazantzis et al. | A new methodology for optimizing the daily operations of a pumping plant | |
| JP6603156B2 (en) | Sewerage facility operation support apparatus, system and method thereof | |
| Jia et al. | Optimization design of industrial water supply pump station considering the influence of atmospheric temperature on operation cost | |
| Al-Ani et al. | Optimal operation of water pumping stations | |
| CN113205250A (en) | Wind-solar-water complementary power generation plan compilation method considering reservoir scheduling risk | |
| JP4099334B2 (en) | Plant operation control device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| MM4A | Annulment or lapse of patent due to non-payment of fees |