Research And Application Of Intelligent Algorithm In The Tank Robot

With widely used in various fields, artificial intelligence has become increasinglyimportant. To the theory of intelligent algorithms based on genetic algorithm, neural networks,machine learning and game-specific algorithms to explore artificial intelligence algorithmsapplied to the AI-TANK tank robot battle simulation engine robots move, attack,strategy research and application. The main specific work is as follows:(1) Research the AI-TANK simulation engine, I am extremely skilled in the platform usingC++programming language to implement my own ideas as a basis to do variousexperiments.(2) Research the BP artificial neural network. Shooting strategy of tank robot used BP neuralnetwork design, and I implemented this algorithm on the platform. I found that the robotfire various enemy not bad, and it has a relatively stable percentage of hits on differentopponents.(3) Research the genetic algorithm, and then I make some improvements to the selectoperator. Then I applied to the tank robot to the move strategy. I found that this tank havea certain degree of adaptability to different opponents, but sometimes the training processwill be some twists and turns, the overall performance is still not reached the ideal state.(4) In AI-TANK, select the object to attack is important to the fighting as a team. I narratehow to fulfill a system based on the Bayesian classifier object selection, then use thenecessary information as input, and finally by continuously updating the Bayesiannetwork, making the fighting as a team, select the target is more favorable to our side.