College 6 Flashcards
How do optimisation models work?
They simulate an optimal condition (objective) following a set of prior conditions, criteria (constraints) and decisions variables, normally relying on mathematical techniques
Name an example of an optimalisation model
Shortest path
Name two optimalisation approaches
- Mathematical programming 2. Heurisic method
How does mathematical programming work?
- linear, non-linear and mixed integer programming - relatively strict in formulation - aims to find limited number of global optima
How doe heuristic optimalisation work?
- genetic algorithms, neural networks, evolutionary programming - flexible, fast - sets of near optimal solutions
What are the main 3 conclusions of optimalisation?
- Prescriptive not descriptive - Presented examples do not offer a single unique solution - Other optimsation methods do offfer unique solutions
What does a rule-based simulation consist of? and what are some synonyms?
- Premise 2. Conclusion know as expert or what if models
Give an exaple of a rule based simulation
Universal Soil Loss Equation - breaks up erosion process into 6 factors
What is the universal soil loss equation?
A= RKLSC*P
What are the charesteristic of Multi-Agent models?
Agents are autonomous entities that: share a common environment and take decisions tha link their behaviour to the environment
Name some Multi Agent interaction environments
- landscape - political institutions - social interaction - markets
What are the interaction types of MA models?
Agent/agent or agent/environment
What are some strenghts of Multi agent models?
- Represent decision making at level of agent 2. Promising approach
What are some weaknesses of Multi agent models?
- Parameterisation difficult 2. Data nees enormous 3. Cross-scale dynamics comlicated 4. Communication of method and results challegning
What do Mirco simulations do?
- Simulation at the level of individuals
What was the objective of the Notting HIll Carnival simulation?
Helping redesign the unsafe parade route to avoid crowding with related crime and other hazards generated by concentration in a small area
What are the benefits of Micro simulation?
Extremely dynamic models providing path depenency and emergent properties
Wha are the two components that lead to a finer resolution in the new landuse scanner?
- Descrete cells 2. new allocation algorithm
What are the two constraints of the new allocation alogrithm?
- Each cell is filled completely with one type 2. the regional demand for each landuse type
What are new allpications of the model?
regional and superanational
What are the model improvements of the landuse scanner?
Calibaration/validation, indicator development and new density to landuse types are added
What does the validation give insight to?
To the model preformance
