用户:Wolfch/先进过程控制
 | 本用户页目前为“先进过程控制”的草稿。 如有任何疑问，请至讨论页发起讨论。 |
在控制理论中，先进过程控制（en:Advanced process control，简称APC）是指许多工业过程控制系统中会用到的技术及技巧。先进过程控制一般是在“基本”的过程控制以外，选择性布署的部分。基本的过程控制是根据过程本身所设计及建构，方便基本的操作、控制以及自动化的需求。先进过程控制一般是循序增加的，可能是在几年以后增加，目的是希望此程序中达到性能或是经济上的提升。
此处的过程控制一般是指连续流制程制造业（process industries），包括化工业、石化业、石油和矿物精炼、食品加工、制药业、电厂等。这些产业的特点是连续性的加工程序以及流体处理，和分立零件制造（如汽车业及电子业）不同。过程自动化和过程控制在本质上是类似的。
过程控制会布置在过程控制系统内，过程控制系统可能是分散控制系统（DCS）、可编程逻辑控制器（PLC）、或（及）监督控制用的电脑。所用的DCS及PLC是针对工业应用进行强化的，且有容错的特性。监督控制电脑多半没有工业上的强化的，也没有容错，不过可以加强系统的运算能力，可以处理重要（但非关键性的）先进控制应用。依应用的不同，先进控制可能会在DCS内，也可能在监督控制电脑内。而基本控制会在DCS及PLC内。
先进过程控制的分类[编辑]
以下是几个先进过程控制的类别：
- 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 模型预测控制 (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 顺序功能流程图 methodology.
- 智能控制 is a class of control techniques that use various 人工智能 computing approaches like Lua错误：bad argument #1 to 'gsub' (string expected, got nil)。, 贝叶斯概率, 模糊逻辑, 机器学习, 进化计算 and 遗传算法.
相关技术[编辑]
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错误：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错误：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[编辑]
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.
人工智能与制程控制[编辑]
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]
词语[编辑]
- 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: 多变数模型预测控制
- SIS: Lua错误：bad argument #1 to 'gsub' (string expected, got nil)。
- SME: Subject matter expert
参考资料[编辑]
- ^ Oil and Gas, AI, and the Promise of a Better Tomorrow. SparkCognition Inc. 2016-04-06 [2018-03-23] （美国英语）.
外部链接[编辑]
- Article about Advanced Process Control.