Of course. A comprehensive glossary is an essential resource for students. Here is a detailed glossary of key terms used throughout the textbook, designed to be saved as docs/appendix/glossary.md.
Appendix: Glossary¶
This glossary provides definitions for the key terms and concepts used throughout the textbook "Real-Time Simulation Modeling for Digital Twins."
A¶
ABM (Agent-Based Modeling) : A simulation paradigm that models a system from the bottom up as a population of autonomous, decision-making agents. System-level behavior is not programmed directly but emerges from the local interactions of these agents. Best suited for modeling complex adaptive systems.
Acausal Modeling : A modeling approach, typified by the Modelica language, where the modeler defines the physical equations of components without pre-defining the direction of causality (i.e., which variables are inputs and which are outputs). The simulation engine automatically solves the entire system of equations. Contrasts with Causal Modeling.
Actuator : A physical device that receives a command from a control system and effects a change on the real-world asset. It is the "muscle" of a Digital Twin. (e.g., a valve, a motor, a relay).
AFAP (As-Fast-As-Possible) : An execution mode for a simulation where the virtual clock advances as quickly as the computer can process events, without regard to wall-clock time. Used for predictive "what-if" analysis and optimization.
Agent : The fundamental building block of an ABM. An autonomous software entity with its own internal state, behavioral rules, and actions it can perform on its environment.
API (Application Programming Interface) : A set of rules and protocols that allows different software components to communicate with each other. Microservices in a DT architecture communicate via APIs.
B¶
Balancing Feedback Loop : In System Dynamics, a feedback loop that seeks stability and counteracts change. It is goal-seeking and is also known as a negative feedback loop.
Bayesian Calibration : A statistical method for updating the belief about a model's parameters using real-world data. It starts with a prior probability distribution for a parameter and uses data to calculate a more accurate posterior distribution.
C¶
Causal Loop Diagram (CLD)
: A qualitative diagram used in System Dynamics to map the feedback structures of a system. It consists of variables connected by arrows that indicate causal influence, marked with a polarity (+ or -).
Causal Modeling : A modeling approach where the modeler must explicitly define the input-output relationships and the flow of calculation. Block-diagram tools like Simulink are primarily causal. Contrasts with Acausal Modeling.
Co-Simulation : An approach to hybrid simulation where multiple, independent simulation models, often running in different software tools, are executed in parallel. A master algorithm coordinates the exchange of data and the advancement of time between them.
Containerization : The process of packaging a software application and all its dependencies into a single, isolated, and portable unit called a container. Docker is the leading containerization technology.
D¶
Data Assimilation : The process of incorporating real-world observation data into a running simulation model to correct its state and improve its accuracy. A core component of continuous validation.
DES (Discrete-Event Simulation) : A simulation paradigm that models a system's operation as a chronological sequence of discrete events. The simulation clock jumps from one event to the next. Best suited for modeling processes, workflows, and resource-constrained systems.
Digital Shadow : A system where there is an automated, one-way flow of data from a physical asset to a digital model. The model can show the current state of the asset, but the link is not bidirectional.
Digital Twin : A living, simulation-powered model of a physical asset or system that is continuously synchronized with its real-world counterpart via an automated, two-way data link. It is used for real-time monitoring, prediction, and optimization.
Docker : The leading software platform for containerization.
E¶
Edge Computing : A distributed computing paradigm where computation is performed on-site, on or near the physical asset, rather than in a centralized cloud. Used to reduce latency and improve reliability.
Emergence : The arising of novel, coherent structures and patterns at a macro-level from the simple, local interactions of agents at a micro-level. It is a hallmark of complex adaptive systems and is captured by ABM.
Entity : The fundamental dynamic object that flows through a DES model. (e.g., a customer, a part, a message).
Event Calendar : The core data structure in a DES engine. It is a time-ordered list of all future events that are scheduled to occur.
F¶
Fidelity : The degree to which a simulation model accurately represents its real-world counterpart. Fidelity is a multi-dimensional concept (physical, visual, process, etc.).
FMI (Functional Mock-up Interface) : A free, industry-wide standard that defines a common API for packaging simulation models into a Functional Mock-up Unit (FMU). FMUs can be easily shared and used in different co-simulation environments.
H¶
Hard Real-Time : A system constraint where missing a computational deadline constitutes a total system failure. Common in safety-critical control systems.
Hybrid Simulation : The use of multiple simulation paradigms (e.g., ABM and DES) within a single, composite model to capture behaviors at different scales or from different domains.
J¶
JSON (JavaScript Object Notation) : A lightweight, human-readable data-interchange format. It is the de facto standard for structuring the payload of messages in IoT systems.
K¶
Kalman Filter : A powerful algorithm used for state estimation. It produces an optimal estimate of a system's true state by recursively fusing predictions from a model with noisy measurements from sensors.
Kubernetes : The leading open-source platform for orchestrating containerized applications. It automates the deployment, scaling, and management of microservices.
M¶
Microservices : An architectural style where a complex application is composed of small, independent services, each responsible for a specific capability. They communicate over APIs and can be deployed and scaled independently.
Modelica : A non-proprietary, object-oriented, equation-based language for acausal modeling of complex physical systems.
MQTT (Message Queuing Telemetry Transport) : A lightweight, publish-subscribe network protocol that is the de facto standard for IoT messaging.
O¶
ODE (Ordinary Differential Equation) : A mathematical equation that describes the rate of change of a system's state variables. It is the foundation of physics-based modeling for dynamical systems.
Orchestration : The automated management, coordination, and scaling of containerized applications. Kubernetes is the leading orchestration tool.
P¶
Prescriptive Analytics : The most advanced form of analytics, which goes beyond predicting the future to recommend the best course of action to achieve a goal. Simulation-based optimization and RL are prescriptive techniques.
Publish/Subscribe (Pub/Sub) : A messaging pattern where senders (publishers) do not send messages directly to receivers (subscribers). Instead, they publish messages to a central broker, which then delivers them to all interested subscribers. MQTT is a pub/sub protocol.
R¶
Reinforcing Feedback Loop : In System Dynamics, a feedback loop that amplifies change, leading to exponential growth or collapse. Also known as a positive feedback loop.
Reinforcement Learning (RL) : A branch of machine learning where an agent learns to make optimal decisions by taking actions in an environment and receiving rewards. A Digital Twin can act as a high-speed, risk-free training environment for an RL agent.
Resource : A static, capacity-constrained object in a DES model that provides a service to entities. (e.g., a machine, an operator, a server).
S¶
SD (System Dynamics) : A simulation paradigm that models the long-term behavior of complex systems from a top-down, aggregate perspective, focusing on stocks, flows, and feedback loops.
Sensor : A physical device that detects and measures a physical property of an asset and converts it into a data signal. It is the "sense organ" of a Digital Twin.
Soft Real-Time : A system constraint where missing a deadline degrades performance but does not cause a system failure. This is the most common constraint for Digital Twins.
State Vector : The set of all variables required to completely describe the state of a simulation model at a single point in time.
Stigmergy : A form of indirect communication where an agent modifies its environment, and other agents respond to that environmental change at a later time.
Stock : The fundamental building block of an SD model representing an accumulation of something. A stock can only be changed by its in-flows and out-flows.
Stochastic : A process or model that incorporates randomness. Contrasts with Deterministic.
T¶
Time-Series Database : A database specifically designed to store and query large volumes of timestamped data, such as sensor readings from a Digital Twin.
Topic (MQTT) : A hierarchical string used in MQTT to label and route messages. Publishers send messages to a topic, and subscribers receive messages by subscribing to a topic.
U¶
Uncertainty Quantification (UQ) : The process of identifying, characterizing, and quantifying the uncertainty in a simulation model's inputs, parameters, and structure, and then propagating that uncertainty to its outputs.
V¶
Validation : The process of determining if a simulation model is an accurate representation of the real-world system for its intended purpose. "Are we building the right model?"
Verification : The process of determining if a simulation model is implemented correctly and matches its conceptual design. "Are we building the model right?" ```