InteractingMultipleModel

Interacting multiple model (IMM) state transition model

Since R2025a

Description

A InteractingMultipleModel object defines the interacting-multiple-model (IMM) state transition model, which assumes that the motion of a target switches between multiple predefined models. You can use this model to define the state transition model for a CustomTarget object. The state transition model of the CustomTarget specification defines the state convention for trackers initialized with it. For an InteractingMultipleModel, the associated state convention is the same as the state convention of the first motion model with which this object is defined.

Creation

To create an InteractingMultipleModel object, use the targetStateTransitionModel function with the "interacting-multiple-model" input argument. For example:

stateTransModel = targetStateTransitionModel("interacting-multiple-model")

Properties

Models — List of Euclidean models

cell array of motion model objects

List of Euclidean models, specified as a cell array of motion model objects. You can use the targetStateTransitionModel function to generate these objects. The IMM model object supports these model objects:

• ConstantVelocityModel

• ConstantAccelerationModel

• ConstantTurnRateModel

• SingerAccelerationModel

The default Models list contains a ConstantVelocityModel object, a ConstantAccelerationModel object, and a ConstantTurnRateModel object.

Data Types: single | double

PriorModelProbabilities — Prior probabilities of each model

vector of scalars

Prior probabilities of the target in each model, specified as a vector of scalars. The size must be the same as the size of Models. By default, the probability is equal for each model.

Example: [0.3 0.4 0.3]

Data Types: single | double

TransitionProbabilities — Probability that the target switches motion models

0.9 (default) | positive real scalar | L-element vector | L-by-L matrix

Probability that the target switches motion models, specified as a positive real scalar, L-element vector, or L-by-L matrix, where L is the same as the size of Models:

• When specified as a scalar, the probability that the target stays in any model is uniform. The remaining probability (1-p) is distributed evenly across the other motion models.

• When specified as a vector, each element defines the probability that the target stays in each model. The remaining probability (1-p) is distributed evenly across the other motion models.

• When specified as a matrix, the (j,k) element defines the probability that the target switches from the jth model to the kth model. All elements must lie within the interval [0,1] and each row must sum to 1.

The transition probability defined for each model corresponds to the probability that the target switches from this model to another model in one second.

Example: 0.75

Data Types: single | double

Version History

Introduced in R2025a