higorpo / puzzle8-resolver

This repository uses some search algorithms like uniform cost, A* and heuristics to find the solution to problems involving the Puzzle8 game.

Algorithm for solving Puzzle 8 problems

This project consists in the implementation of an algorithm for solving puzzles in the Puzzle 8 game, using the Dart language. The algorithm can use A* with basic or advanced heuristics, or just use Uniform Cost to determine the best solution for a given problem.

Requirements

To run the algorithm, you will need to have the Dart SDK installed on your machine. To install Dart, you can follow the instructions available in the official documentation.

How to use

To use the algorithm, it is necessary to execute the file main.dart through the terminal, passing as a parameter the initial state of the board and the algorithm to be used. The initial state must be passed as a 3x3 matrix of integers, where the number 0 represents the empty position of the board. The algorithm can be chosen from AStar, SimpleAStar and UniformCost.

Example of execution with the initial state [1, 2, 3], [4, 5, 6], [7, 8, 0] and the algorithm A*:

dart run main.dart [[1,2,3], [4,5,6], [7,8,0]] AStar

Algorithm operation

The algorithm follows the graph search model, where each state of the board is represented by a node in the graph. For each node, the algorithm searches for its successor states, that is, the possible moves that can be made from the current state to reach a new state.

A* uses heuristics to determine the order of expansion of the nodes, always aiming to expand the nodes that have the lowest cost. Simple A* is a simplified version of A*, which uses a simpler and faster heuristic to calculate. The Uniform Cost, as its name suggests, expands the nodes with less cost, but does not use heuristics.

Optimizations:

• The algorithm was created optimized to use binary search when inserting elements in the list of open nodes in an ordered way, which results in a significant reduction in execution time.
• For this, a Dart list implementation called SortedList was created using the collections package, which already has binary search and ordered insertion methods implemented.
• Furthermore, the heuristic used in the algorithm was chosen in order to be efficient in terms of execution time and ability to guide the algorithm towards the goal state.
• The code was written in a clear and organized manner, following good programming practices and clean code standards, which facilitates algorithm maintenance and optimization.