Adaptive Lookup Table (nD Stair-Fit)

Adaptive lookup table of arbitrary dimension

Library

Simulink Design Optimization

• 

Description

The Adaptive Lookup Table (nD Stair-Fit) block creates an adaptive lookup table of arbitrary dimension by dynamically updating the underlying lookup table. The block uses the outputs of your system to do the adaptations.

Each indexing parameter may take a value within a set of adapting data points, which are called breakpoints. Breakpoints in each dimension define a cell. The set of all breakpoints in one of the dimensions defines a grid. In the n-dimensional case, each cell has two n breakpoints and is an (n-1) hypersurface.

You can use the Adaptive Lookup Table (nD Stair-Fit) block to model time-varying systems with 2 or more inputs.

Data Type Support

Doubles only

Parameters

Number of table dimensions

The number of dimensions for the adaptive lookup table.

Table breakpoints (cell array)

A set of one-dimensional vectors that contains possible block input values for the input variables. Each input row must be monotonically increasing, but the rows do not have to be the same length. For example, if the Number of table dimensions is 3, you can set the table breakpoints as follows:

{[1 2 3], [5 7], [1 3 5 7]}

Make initial table an input

Selecting this check box forces the Adaptive Lookup Table (nD Stair-Fit) block to ignore the Table data (initial) parameter, and creates a new input port Tin. Use this port to input the table data.

Table data (initial)

The initial table output values. This (n-D) array must be of size (n-1)-by-(n-1) ... -by- (n-1), (D times), where D is the number of dimensions and n is the number of input breakpoints.

Table numbering data

Number values assigned to cells. This vector must be the same size as the table data array, and each value must be unique.

Adaptation method

Choose Sample mean or Sample mean with forgetting. Sample mean averages all the values received within a cell. Sample mean with forgetting gives more weight to the new data. How much weight is determined by the Adaptation gain parameter. For more information, see Selecting an Adaptation Method.

Adaptation gain (0 to 1)

A number from 0 to 1 that regulates the weight given to new data during the adaptation. A 0 means short memory (last data becomes the table value), and 1 means long memory (average all data received in a cell).

Make adapted table an output

Selecting this check box creates an additional output port Tout for the adapted table.

Note

The Adaptive Lookup Table (n-D Stair Fit) block cannot output a table of 3 or more dimensions.

Add adaptation enable/disable/reset port

Selecting this check box creates an additional input port Enable that enables, disables, or resets the adaptive lookup table. A signal value of 0 applied to the port disables the adaptation, and signal value of 1 enables the adaptation. Setting the signal value to 2 resets the table values to the initial table data.

Add cell lock enable/disable port

Selecting this check box creates an additional input port Lock that provides the means for updating only specified cells during a simulation run. A signal value of 0 unlocks the specified cells and signal value of 1 locks the specified cells.

Action for out-of-range input

Ignore or Adapt by extrapolating beyond the extreme breakpoints.

Extended Capabilities

C/C++ Code Generation
Generate C and C++ code using Simulink® Coder™.

Version History

Introduced in R2009a