By completing the flocking birds lab was interesting as it shows how realistic flocks of birds can be imitated by using three techniques which are steering behaviours, seperation, cohesion and alignment. Each bird can only percieve a local neighbourhood described by an angle and a distance. Seperation is when the agents avoid crowding local flockmates, cohesion moves towards the average position of local flockmates. Finally, alignment moves towards the average heading of local flockmates.
GOAP is a very powerful tool to use as it has actions and goals that are decoupled unlike a FSM. Multiple agents can be used and for each agent GOAP allows for action to be chosen based on preconditions and outcomes.Very realisitc behavior can be made using this technique and doing this lab helped me understand this.
By doing the steering behaviour lab I worked on the "How far to look ahead" this helped me realise how realistic movement AI can become. So instead of having a constant value you can use distance, Time to target and speed based variables to see how far you want your AI to look ahead.I have also learnt about the seek limitations as seek always move to target with max acceleration so by reducing speed when it gets closer to target, this makes it stop doing the spiral movement around target. Seek and flee, pursue and evade are all things I understand how to use now including arrival so I will look to try to learn about Wander and implement this in the steering behaviour lab.
By doing the BFS and DFS lab it really helped me understand the concept. I learn more by visually so by seeing what the code and theory is doing I can have good grasp on the topic. The implementaion has helped me understand that it really isnt hard to implement the pseudo code and there is not much difference between the two. DFS is a stack and BFS is queue.
