Passing Strategy Optimization Based on Kernel Density Negative Example Learning Within RoboCup 2D Simulation

RoboCup2D simulation soccer is a multi-intelligence collaborative simulation experiment platform. In RoboCup2D simulation soccer platform, the passing module is one of the of the team, and an effective passing strategy will improve the overall team strength. Since the parameters of the passing module are too complicated and the data distribution is too heterogeneous. In order to be able to classify the passing data effectively, we propose a method based on kernel density negative example learning, which extracts the action chain feature data conforming to our preset by parsing the game log files based on kernel density analysis, and weakens the weights of common features of passing data through negative example learning to improve the classification of different kinds of passing. The classification effect of feature data is improved by weakening the weight of common features of passing data through negative example learning. By analyzing the similarities and differences of the passing action chain features when facing different teams, we extract the key areas of attack to optimize the team's passing strategy and improve the average number of goals scored in the test with several teams.