用戶:Wolfch/先進過程控制
 | 本用戶頁目前為「先進過程控制」的草稿。 如有任何疑問，請至討論頁發起討論。 |
在控制理論中，先進過程控制（en:Advanced process control，簡稱APC）是指許多工業過程控制系統中會用到的技術及技巧。先進過程控制一般是在「基本」的過程控制以外，選擇性布署的部份。基本的過程控制是根據過程本身所設計及建構，方便基本的操作、控制以及自動化的需求。先進過程控制一般是循序增加的，可能是在幾年以後增加，目的是希望此程序中達到性能或是經濟上的提昇。
此處的過程控制一般是指連續流製程製造業（process industries），包括化工業、石化業、石油和礦物精煉、食品加工、制藥業、電廠等。這些產業的特點是連續性的加工程序以及流體處理，和分立零件製造（如汽車業及電子業）不同。過程自動化和過程控制在本質上是類似的。
過程控制會佈置在過程控制系統內，過程控制系統可能是分散控制系統（DCS）、可程式邏輯控制器（PLC）、或（及）監督控制用的電腦。所用的DCS及PLC是針對工業應用進行強化的，且有容錯的特性。監督控制電腦多半沒有工業上的強化的，也沒有容錯，不過可以加強系統的運算能力，可以處理重要（但非關鍵性的）先進控制應用。依應用的不同，先進控制可能會在DCS內，也可能在監督控制電腦內。而基本控制會在DCS及PLC內。
先進過程控制的分類
以下是幾個先進過程控制的類別：
- Advanced regulatory control (ARC) refers to several proven advanced control techniques, such as override or adaptive gain (but in all cases, "regulating or feedback"). ARC is also a catch-all term used to refer to any customized or non-simple technique that does not fall into any other category. ARCs are typically implemented using function blocks or custom programming capabilities at the DCS level. In some cases, ARCs reside at the supervisory control computer level.
- Advanced process control (APC) refers to several proven advanced control techniques, such as feedforward, decoupling, and inferential control. APC can also include Model Predictive Control, described below. APC is typically implemented using function blocks or custom programming capabilities at the DCS level. In some cases, APC resides at the supervisory control computer level.
- Multivariable 模型預測控制 (MPC) is a popular technology, usually deployed on a supervisory control computer, that identifies important independent and dependent process variables and the dynamic relationships (models) between them, and often uses matrix-math based control and optimization algorithms to control multiple variables simultaneously. One requirement of MPC is that the models must be linear across the operating range of the controller. MPC has been a prominent part of APC ever since supervisory computers first brought the necessary computational capabilities to control systems in the 1980s.
- Nonlinear MPC: Similar to Multivariable MPC in that it incorporates dynamic models and matrix-math based control; however, it does not have the requirement for model linearity. Nonlinear MPC is capable of accommodating processes with models that have varying process gains and dynamics (i.e. dead-times and lag times).
- Inferential Measurements: The concept behind inferentials is to calculate a stream property from readily available process measurements, such as temperature and pressure, that otherwise might be too costly or time-consuming to measure directly in real time. The accuracy of the inference can be periodically cross-checked with laboratory analysis. Inferentials can be utilized in place of actual online analyzers, whether for operator information, cascaded to base-layer process controllers, or multivariable controller CVs.
- Sequential control refers to discontinuous time- and event-based automation sequences that occur within continuous processes. These may be implemented as a collection of time and logic function blocks, a custom algorithm, or using a formalized 順序功能流程圖 methodology.
- 智能控制 is a class of control techniques that use various 人工智能 computing approaches like Lua錯誤：bad argument #1 to 'gsub' (string expected, got nil)。, 貝葉斯概率, 模糊邏輯, 機器學習, 進化計算 and 遺傳算法.
相關技術
The following technologies are related to APC and in some contexts can be considered part of APC, but are generally separate technologies having their own (or in need of their own) Wiki articles.
- Lua錯誤：bad argument #1 to 'gsub' (string expected, got nil)。 (SPC), despite its name, is much more common in discrete parts manufacturing and batch process control than in continuous process control. In SPC, 「process」 refers to the work and quality control process, rather than continuous process control.
- Batch process control (see ANSI/ISA-88) is employed in non-continuous batch processes, such as many pharmaceuticals, chemicals, and foods.
- Simulation-based optimization incorporates dynamic or steady-state computer-based process simulation models to determine more optimal operating targets in real-time, i.e. on a periodic basis, ranging from hourly to daily. This is sometimes considered a part of APC, but in practice it is still an emerging technology and is more often part of MPO.
- Manufacturing planning and optimization (MPO) refers to ongoing business activity to arrive at optimal operating targets that are then implemented in the operating organization, either manually or in some cases automatically communicated to the process control system.
- Lua錯誤：bad argument #1 to 'gsub' (string expected, got nil)。 refers to a system that is independent of the process control system, both physically and administratively, whose purpose is to assure basic safety of the process.
APC Business and Professionals
Those responsible for the design, implementation and maintenance of APC applications are often referred to as APC Engineers or Control Application Engineers. Usually their education is dependent upon the field of specialization. For example, in the process industries many APC Engineers have a chemical engineering background, combining process control and chemical processing expertise.
Most large operating facilities, such as oil refineries, employ a number of control system specialists and professionals, ranging from field instrumentation, regulatory control system (DCS and PLC), advanced process control, and control system network and security. Depending on facility size and circumstances, these personnel may have responsibilities across multiple areas, or be dedicated to each area. There are also many process control service companies that can be hired for support and services in each area.
人工智能與製程控制
The use of Artificial Intelligence, Machine Learning and Deep Learning techniques in Process Control is also considered as an advanced process control approach in which intelligence is used to further optimize operational parameters.
Operations and Logics in process control systems in oil and gas and for decades are based only on physics equations that dictates parameters along with operators』 interactions based on experience and operating manuals. Artificial Intelligence and Machine Learning algorithms can look into the dynamic operational conditions, analyse them and suggest optimized parameters that can either directly tune logic parameters or give suggestion to operators. Interventions by such intelligent models leads to optimization in cost, production and safety.[1]
詞語
- APC: Advanced process control, including feedforward, decoupling, inferentials, and custom algorithms; usually implies DCS-based.
- ARC: Advanced regulatory control, including adaptive gain, override, logic, fuzzy logic, sequence control, device control, and custom algorithms; usually implies DCS-based.
- Base-Layer: Includes DCS, SIS, field devices, and other DCS subsystems, such as analyzers, equipment health systems, and PLCs.
- BPCS: Basic process control system (see "base-layer")
- DCS: Distributed control system, often synonymous with BPCS
- MPO: Manufacturing planning optimization
- MPC: 多變數模型預測控制
- SIS: Lua錯誤：bad argument #1 to 'gsub' (string expected, got nil)。
- SME: Subject matter expert
參考資料
- ^ Oil and Gas, AI, and the Promise of a Better Tomorrow. SparkCognition Inc. 2016-04-06 [2018-03-23] （美國英語）.
外部連結
- Article about Advanced Process Control.