AIDA-Glossary of Terms 专业术语-2

译文：
Empirical Models 经验模型 从数据推断（实验）得出模型。
Error 代表模型拟合后输出的原因不明的误差。它是预测输出与实际输出之间的误差。又称为残差。
Estimation 从数据确定经验模型的过程。
Excitation 激励 对一过程施加扰动以获得数据。
Filter 滤波 对一个信号进行数学方程处理，使得处理后信号仅保留该信号的某些特征。
F2P 通过拟合FIR阶跃响应而不针对数据得到的参数化模型。F2P模型是用作FIR和PAR 拟合的桥梁。
FIR 有限脉冲响应。
First-principles Models 第一性原理模型 通过对系统物理数学分析得到的模型。
**Fit **利用回归估计出对一组数据解释的最佳模型的过程。在辨识和估计时其也指估计的结果。在AIDAPro中它指的是在拟合进行的探索树的目的。
**Fit Group **拟合组 拟合的集合。定义输入、输出和中间变量并对所有拟合子集施加组合影响。
Fit Statistic 拟合统计 拟合优度代表信噪比的量度，即预测分量输出对非预测分量的波动。 (也成为拟合统计)。
Gray-box Models 灰箱模型 基于流程和工程知识，实验模型通过因果结构相互连接。
**Horizon **时域 指向未来的一个时间窗。
ID ID是应用于PCTP的身份验证。
Identification 辨识 使用过程数据估算经验模型的过程。
Intermediate Variable 中间变量 某一变量是一个系统的输出，又是另一个系统的输入。
Inverse Response 反向响应 输入的阶跃使输出发生响应，输出响应改变开始时与最终方向是相反的称为反方向响应。反向响应是输出的两个反向影响的结果 (也称为非最小相位系统)。
Linear Models 线性模型 模型的数学表达基于线性方程组。
MD 监测与诊断。
**MIMO **多输入多输出。
Model 对系统的描述。在这里是数学描述。
Model Order 模型阶数 参数模型的维数。它确定了描述动态过程模型所需要的参数数目。 (在传递函数形式中它是分母多项式的维数。)
MPC 模型预测控制。

---

原文：
Empirical Models Models inferred from (experimental) data.
Error Represents the unexplained variation of an output after fitting of a model. It is the difference between the actual output and the predicted output. Also known as the residual.
Estimation The process of determining empirical models from data.
Excitation Perturbation of a process to obtain data.
Filter A mathematical function applied on a signal that permits only certain characteristics of the signal to end up in the resulting signal.
F2P A parametric model whose parameters are determined by fitting the FIR step response rather than against data. The F2P model is used as a bridge between FIR and PAR fits.
FIR Finite Impulse Response.
First-principles Models Models obtained from a physical mathematical analysis of the system.
Fit The process of estimating the best model that explains a set of data using regression. In identification and estimation it could also refer to the result of an estimate. In AIDAPro it refers to the object on the tree explorer where fitting is performed.
**Fit Group ** A collection of Fits. Defines the inputs, outputs, intermediate variables and the binding cause-effect structure for all children fits.
Fit Statistic A measure of the goodness-of-fit representative of the signal-to-noise ratio, that is, a variation of the predictable component of an output to the unpredicted component. (Also referred to as Fit Stat.)
**Gray-box Models **Empirical models interconnected with each other using cause-effect structures based on process and engineering knowledge.
Horizon A window of time into the future.
ID The PCTP identification application was known as ID.
Identification The process of estimating empirical models using process data.
Intermediate Variable A variable that is an output from a system, and an input to another.
Inverse Response A response of an output to a step input change that starts off in an opposite direction to the ultimate direction of the response. An inverse response is the result of two opposing effects on the output. (Also known as non-minimum phase systems.)
Linear Models Models whose mathematical equations are based on linear equations.
**MD ** Monitoring and Diagnosis.
MIMO Multiple-Input Multiple-Output.
Model A description of a system. Descriptions are mathematical in our context.
**Model Order **The dimension of a parametric model. It determines the number of parameters required to describe the dynamics of the process model. (In transfer function form it is the dimension of the denominator polynomial.)
MPC Model Predictive Control.

---

2016/4/15