plannerAStarGrid

A* path planner for grid map

Since R2020b

Description

The plannerAStarGrid object creates an A* path planner. The planner performs an A* search on an occupancy map and finds shortest obstacle-free path between the specified start and goal grid locations as determined by heuristic cost.

Creation

Syntax

planner = plannerAStarGrid

planner = plannerAStarGrid(map)

planner = plannerAStarGrid(___,Name,Value)

Description

planner = plannerAStarGrid creates a plannerAStarGrid object with a binaryOccupancyMap object using a width and height of 10 meters and grid resolution of 1 cell per meter.

planner = plannerAStarGrid(map) creates a plannerAStarGrid object using the specified map object map. Specify map as either a binaryOccupancyMap or occupancyMap object. The map input sets the value of the Map property.

planner = plannerAStarGrid(___,Name,Value) sets Properties using one or more name-value pairs. Unspecified properties have default values. Enclose each property name in quotes.

For example, plannerAStarGrid(map,'GCost','Manhattan') creates an A* path planner object using the Manhattan cost function.

Properties

Map — Map representation

binaryOccupancyMap object (default) | occupancyMap object

Map representation, specified as either a binaryOccupancyMap or occupancyMap object. This object represents the environment of the robot as an occupancy grid. The value of each grid cell indicates the occupancy of the associated location in the map.

Example: planner.Map = binaryOccupancyMap(zeros(50,50));

GCost — General cost of moving between any two points in grid

'Euclidean' (default) | 'Chebyshev' | 'EuclideanSquared' | 'Manhattan'

The general cost of moving between any two points in a grid, specified as one of the following predefined cost functions 'Chebyshev', 'Euclidean', 'EuclideanSquared', or 'Manhattan'.

The cost of moving between two points with Cartesian coordinates (x₁,y₁) and (x₂,y₂) are calculated as following:

• Chebyshev

  $$d=\max \left(\left|{{\displaystyle x}}_2-{{\displaystyle x}}_1\right|,\left|{{\displaystyle y}}_2-{{\displaystyle y}}_1\right|\right)$$

• Euclidean

  $$d=\sqrt{{{\displaystyle \left({{\displaystyle x}}_2-{{\displaystyle x}}_1\right)}}^2+{{\displaystyle \left({{\displaystyle y}}_2-{{\displaystyle y}}_1\right)}}^2}$$

• Euclidean Squared

  $$d={{\displaystyle \left({{\displaystyle x}}_2-{{\displaystyle x}}_1\right)}}^2+{{\displaystyle \left({{\displaystyle y}}_2-{{\displaystyle y}}_1\right)}}^2$$

• Manhattan

  $$d=\left|{{\displaystyle x}}_2-{{\displaystyle x}}_1\right|+\left|{{\displaystyle y}}_2-{{\displaystyle y}}_1\right|$$

Note

You can either use the predefined cost functions or a custom cost function. To use a custom cost function, see GCostFcn property.

Example: planner = plannerAStarGrid(map,'GCost','Manhattan');

Example: planner.GCost = 'Chebyshev';

Data Types: string | char

GCostFcn — Custom GCost function

function handle

Custom GCost function, specified as a function handle. The function handle must accept two pose inputs as [row column] vectors and return a scalar of type double.

Note

You can either use the predefined cost functions or a custom cost function. To use the predefined cost functions, see GCost property.

Example: planner = plannerAStarGrid(map,'GCostFcn',@(pose1,pose2)sum(abs(pose1-pose2),2));

Example: planner.GCostFcn = @(pose1,pose2)sum(abs(pose1-pose2),2);

Data Types: function_handle

HCost — Heuristic cost between point and goal in grid

'Euclidean' (default) | 'Chebyshev' | 'EuclideanSquared' | 'Manhattan'

The heuristic cost between a point and the goal in a grid, specified as one of the following predefined cost functions 'Chebyshev', 'Euclidean', 'EuclideanSquared', or 'Manhattan'.

The cost of moving between two points with Cartesian coordinates (x₁,y₁) and (x₂,y₂) are calculated as following:

• Chebyshev

  $$d=\max \left(\left|{{\displaystyle x}}_2-{{\displaystyle x}}_1\right|,\left|{{\displaystyle y}}_2-{{\displaystyle y}}_1\right|\right)$$

• Euclidean

  $$d=\sqrt{{{\displaystyle \left({{\displaystyle x}}_2-{{\displaystyle x}}_1\right)}}^2+{{\displaystyle \left({{\displaystyle y}}_2-{{\displaystyle y}}_1\right)}}^2}$$

• Euclidean Squared

  $$d={{\displaystyle \left({{\displaystyle x}}_2-{{\displaystyle x}}_1\right)}}^2+{{\displaystyle \left({{\displaystyle y}}_2-{{\displaystyle y}}_1\right)}}^2$$

• Manhattan

  $$d=\left|{{\displaystyle x}}_2-{{\displaystyle x}}_1\right|+\left|{{\displaystyle y}}_2-{{\displaystyle y}}_1\right|$$

Note

You can either use the predefined cost functions or a custom cost function. To use a custom cost function, see HCostFcn property.

Example: planner = plannerAStarGrid(map,'HCost','Manhattan');

Example: planner.HCost = 'Chebyshev';

Data Types: string | char

HCostFcn — Custom HCost function

function handle

Custom HCost function, specified as a function handle. The function handle must accept two pose inputs as [row column] vectors and return a scalar of type double.

Note

You can either use the predefined cost functions or a custom cost function. To use the predefined cost functions, see HCost property.

Example: planner = plannerAStarGrid(map,'HCostFcn',@(pose1,pose2)sum(abs(pose1-pose2),2));

Example: planner.HCostFcn = @(pose1,pose2)sum(abs(pose1-pose2),2);

Data Types: function_handle

TieBreaker — Toggle tiebreaker mode

'off' (default) | 'on'

Toggle tiebreaker mode, specified as either 'on' or 'off'.

When you enable the TieBreaker property, the A* path planner chooses between multiple paths of the same length by adjusting the heuristic cost value.

Example: planner = plannerAStarGrid(map,'TieBreaker','on');

Example: planner.TieBreaker = 'off';

Data Types: string | char

DiagonalSearch — Toggle diagonal search mode

'on' (default) | 'off'

Toggle diagonal search mode, specified as either 'on' or 'off'.

When you set this property to 'on', the A* path planner searches in diagonal direction along with the other four directions of the grid. When you set this property to 'off', the A* path planner searches only in the four directions of the grid.

Data Types: char | string

Object Functions

plan	Find shortest obstacle-free path between two points
show	Plot and visualize A* explored nodes and planned path

Examples

Plan Obstacle-Free Path in Grid Map Using A-Star Path Planner

Plan the shortest collision-free path through an obstacle grid map using the A* path planning algorithm.

Generate a binaryOccupancyMap object with randomly scattered obstacles using the mapClutter function.

rng('default');
map = mapClutter;

Use the map to create a plannerAStarGrid object.

planner = plannerAStarGrid(map);

Define the start and goal points.

start = [2 3];
goal = [248 248];

Plan a path from the start point to the goal point.

plan(planner,start,goal);

Visualize the path and the explored nodes using the show object function.

show(planner)

Extended Capabilities

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

Version History

Introduced in R2020b