public static HiddenMarkovModel CreateRobotHMM()
{
// Example adopted from Sebastian Thrun's "Probabilistic Robotics"
// Chapter 2
// A robot faces a door. The state of the door (open or closed)
// constitutes the "hidden state"
// The robot's sensor detects a "closed" or "open" state.
// (Perception may be out of synch with reality because sensors are
// probabilistic).
// The robot can either "do nothing" or "push" the door. These are the
// possible actions.
IList <string> states = new[] { HmmConstants.DoorOpen, HmmConstants.DoorClosed };
IList <string> actions = new [] { HmmConstants.DoNothing, HmmConstants.PushDoor };
IList <string> perceptions = new [] { HmmConstants.SeeDoorOpen, HmmConstants.SeeDoorClosed };
RandomVariable prior = new RandomVariable(states);
TransitionModel tm = new TransitionModel(states, actions);
// tm.SetTransitionProbability(start_state, action, end_state,
// probability)
// given a start state and an action the probability of the end state is
// probability
tm.SetTransitionProbability(HmmConstants.DoorOpen,
HmmConstants.DoNothing, HmmConstants.DoorOpen, 1.0);
tm.SetTransitionProbability(HmmConstants.DoorOpen,
HmmConstants.DoNothing, HmmConstants.DoorOpen, 1.0);
tm.SetTransitionProbability(HmmConstants.DoorOpen,
HmmConstants.DoNothing, HmmConstants.DoorClosed, 0.0);
tm.SetTransitionProbability(HmmConstants.DoorClosed,
HmmConstants.DoNothing, HmmConstants.DoorClosed, 1.0);
tm.SetTransitionProbability(HmmConstants.DoorClosed,
HmmConstants.DoNothing, HmmConstants.DoorOpen, 0.0);
tm.SetTransitionProbability(HmmConstants.DoorOpen,
HmmConstants.PushDoor, HmmConstants.DoorOpen, 1.0);
tm.SetTransitionProbability(HmmConstants.DoorOpen,
HmmConstants.PushDoor, HmmConstants.DoorClosed, 0.0);
tm.SetTransitionProbability(HmmConstants.DoorClosed,
HmmConstants.PushDoor, HmmConstants.DoorClosed, 0.2);
tm.SetTransitionProbability(HmmConstants.DoorClosed,
HmmConstants.PushDoor, HmmConstants.DoorOpen, 0.8);
SensorModel sm = new SensorModel(states, perceptions);
// sm.setSensingProbaility(state,perception,p); given a state the
// probability of a perception is p
sm.SetSensingProbability(HmmConstants.DoorOpen,
HmmConstants.SeeDoorClosed, 0.4);
sm.SetSensingProbability(HmmConstants.DoorOpen,
HmmConstants.SeeDoorOpen, 0.6);
sm.SetSensingProbability(HmmConstants.DoorClosed,
HmmConstants.SeeDoorOpen, 0.2);
sm.SetSensingProbability(HmmConstants.DoorClosed,
HmmConstants.SeeDoorClosed, 0.8);
HiddenMarkovModel hmm = new HiddenMarkovModel(prior, tm, sm);
return(hmm);
}
public HiddenMarkovModel(RandomVariable priorDistribution,
TransitionModel tm, SensorModel sm)
{
this.priorDistribution = priorDistribution;
this.TransitionModel = tm;
this.SensorModel = sm;
}
public static HiddenMarkovModel createRainmanHMM()
{
IList <string> states = new [] { HmmConstants.Raining, HmmConstants.NotRaining };
// no actions because the observer has no way of changing the hidden
// state and i spassive
IList <string> perceptions = new [] { HmmConstants.SeeUmbrella, HmmConstants.SeeNoUmbrella };
RandomVariable prior = new RandomVariable(states);
TransitionModel tm = new TransitionModel(states);
// tm.setTransitionModelValue(start_state, action, end_state,
// probability);
// given a start state and an action the probability of the end state is
// probability
tm.SetTransitionProbability(HmmConstants.Raining, HmmConstants.Raining,
0.7);
tm.SetTransitionProbability(HmmConstants.Raining,
HmmConstants.NotRaining, 0.3);
tm.SetTransitionProbability(HmmConstants.NotRaining,
HmmConstants.Raining, 0.3);
tm.SetTransitionProbability(HmmConstants.NotRaining,
HmmConstants.NotRaining, 0.7);
SensorModel sm = new SensorModel(states, perceptions);
// sm.setSensingProbaility(state,perception,p); given a state the
// probability of a perception is p
sm.SetSensingProbability(HmmConstants.Raining,
HmmConstants.SeeUmbrella, 0.9);
sm.SetSensingProbability(HmmConstants.Raining,
HmmConstants.SeeNoUmbrella, 0.1);
sm.SetSensingProbability(HmmConstants.NotRaining,
HmmConstants.SeeUmbrella, 0.2);
sm.SetSensingProbability(HmmConstants.NotRaining,
HmmConstants.SeeNoUmbrella, 0.8);
HiddenMarkovModel hmm = new HiddenMarkovModel(prior, tm, sm);
// hmm.setSensorModelValue(state,perception,p); given a state the
// probability of a perception is p
return(hmm);
}
