用强化学习算法实现的猫和老鼠游戏代码资源-CSDN文库

共41个文件

class：25个

java：10个

gif：5个

强化学习

5星 · 超过95%的资源需积分: 50 148 浏览量 2009-12-03 19:07:09 上传评论 4 收藏 56KB ZIP 举报

资源推荐

资源详情

资源评论

收起资源包目录

catandmouse.zip （41个子文件）

SwingApplet$5.class 497B

SwingApplet$2.class 655B

SwingApplet$4.class 496B

SwingApplet$9.class 566B

SwingApplet$10.class 567B

boardContainer.class 338B

CatAndMouseWorld.java 7KB

SwingApplet$8.class 648B

RLPolicy.class 2KB

cheese.gif 304B

RLController.class 2KB

SwingApplet$12.class 755B

RLWorld.class 281B

RLearner.class 5KB

RLPolicy.java 5KB

SwingApplet.java 24KB

boardObject.java 1KB

CatAndMouseGame.class 3KB

SwingApplet$11.class 570B

sampleWorlds.java 3KB

chartPanel.class 2KB

SwingApplet$1.class 924B

SwingApplet$6.class 673B

mouse.gif 1KB

sampleWorlds.class 2KB

boardObject.class 1KB

boardPanel.java 4KB

catmouse.html 2KB

RLController.java 2KB

SwingApplet$7.class 712B

SwingApplet.class 15KB

boardPanel.class 3KB

RLearner.java 10KB

RLWorld.java 1017B

wall.gif 2KB

CatAndMouseGame.java 2KB

CatAndMouseWorld.class 5KB

wallSample.class 312B

cat.gif 559B

floor.gif 2KB

SwingApplet$3.class 805B

//<APPLET CODE = "SwingApplet.class" WIDTH = 700 HEIGHT = 400 ></applet> // load for early releases //import com.sun.java.swing.*; import javax.swing.*; import javax.swing.event.*; import java.awt.*; import java.awt.event.*; import java.util.*; public class SwingApplet extends JApplet implements ActionListener,Runnable{ static final int BW=300, BH=300, BX=8, BY=8, NUM_WALLS=20, SAMP_W = 100, SAMP_H = 100; static final int DEF_EPOCHS = 50000; static final long DELAY=500; static int MAXX=400, MAXY=400; CatAndMouseGame game; CatAndMouseWorld trainWorld, playWorld; // seperate world from playing world RLController rlc; RLearner rl; JTabbedPane tabbedPane; Container instructions, playPanel, trainPanel, worldPanel; // world setting components JTextField rows, cols, obst; sampleWorlds samples; boolean[][] selectedWalls; ButtonGroup worldSelGroup; boolean sampleWorld=true, designWorld=false; // instructions components JLabel instructLabel, usageLabel; final String INSTRUCT_MESSAGE = "<html>This applet demonstrates how reinforcement learning can be used to train an agent to play a simple game. In this case the game is Cat and Mouse- the mouse tries to get to the cheese and back to it's hole, the cat tries to catch the mouse.", USAGE_MESSAGE = "<html>You can train the agent by selecting the Train tab. At any time you can select the Play tab to see how well the agent is performing! Of course, the more training, the better the chance the mouse has of surviving :)"; // train panel components public static final String START="S", CONT_CHECK="C"; final String SETTINGS_TEXT = "These settings adjust some of the internal workings of the reinforcement learning algorithm.", SETTINGS_TEXT2 = "Please see the web pages for more details on what the parameters do."; JTextField alpha, gamma, epsilon, epochs, penalty, reward; JButton startTraining, stopTraining; JRadioButton softmax, greedy, sarsa, qlearn; JProgressBar progress; JLabel learnEpochsDone; // play panel components JButton startbutt, stopbutt, pausebutt; boardPanel bp; public int mousescore=0, catscore =0; JLabel catscorelabel, mousescorelabel; final String MS_TEXT = "Mouse Score:", CS_TEXT = "Cat Score:"; JSlider speed, smoothSlider; Image catImg, mouseImg; chartPanel graphPanel; JLabel winPerc; boardObject cat, mouse, cheese, back, hole, wall; public SwingApplet() { getRootPane().putClientProperty("defeatSystemEventQueueCheck",Boolean.TRUE); } public void init() { // load images catImg = getImage(getCodeBase(), "cat.gif"); mouseImg = getImage(getCodeBase(), "mouse.gif"); Image wallImg = getImage(getCodeBase(), "wall.gif"); Image cheeseImg = getImage(getCodeBase(), "cheese.gif"); Image floorImg = getImage(getCodeBase(), "floor.gif"); /* Image catImg = getImage(ClassLoader.getSystemResource("cat.gif")); Image mouseImg = getImage(ClassLoader.getSystemResource("mouse.gif")); Image wallImg = getImage(ClassLoader.getSystemResource("wall.gif")); Image cheeseImg = getImage(ClassLoader.getSystemResource("cheese.gif"));*/ // set up board objects cat = new boardObject(catImg); mouse = new boardObject(mouseImg); cheese = new boardObject(cheeseImg); back = new boardObject(floorImg); hole = new boardObject(Color.orange); wall = new boardObject(wallImg); // setup content panes tabbedPane = new JTabbedPane(); //instructions = makeInstructions(); worldPanel = makeWorldPanel(); playPanel = makePlayPanel(); trainPanel = makeTrainPanel(); tabbedPane.addTab("World", worldPanel); tabbedPane.addTab("Play", playPanel); //tabbedPane.addTab("Instructions", instructions); tabbedPane.addTab("Train", trainPanel); tabbedPane.setSelectedIndex(0); // disable panes until world created tabbedPane.setEnabledAt(1,false); tabbedPane.setEnabledAt(2,false); // set up controls //setContentPane(new JPanel()); //getContentPane().add(tabbedPane); getContentPane().add(tabbedPane); } public void worldInit(int xdim, int ydim, int numwalls) { trainWorld = new CatAndMouseWorld(xdim, ydim,numwalls); gameInit(xdim,ydim); } public void worldInit(boolean[][] givenWalls) { int xdim = givenWalls.length, ydim = givenWalls[0].length; trainWorld = new CatAndMouseWorld(xdim, ydim,givenWalls); gameInit(xdim,ydim); } private void gameInit(int xdim, int ydim) { // disable this pane tabbedPane.setEnabledAt(0,false); playWorld = new CatAndMouseWorld(xdim, ydim,trainWorld.walls); bp.setDimensions(xdim, ydim); rlc = new RLController(this, trainWorld, DELAY); rl = rlc.learner; rlc.start(); game = new CatAndMouseGame(this, DELAY, playWorld, rl.getPolicy()); game.start(); // set text fields on panels penalty.setText(Integer.toString(trainWorld.deathPenalty)); reward.setText(Integer.toString(trainWorld.cheeseReward)); alpha.setText(Double.toString(rl.getAlpha())); gamma.setText(Double.toString(rl.getGamma())); epsilon.setText(Double.toString(rl.getEpsilon())); // enable other panes tabbedPane.setEnabledAt(1,true); tabbedPane.setEnabledAt(2,true); // switch active pane tabbedPane.setSelectedIndex(1); // set first position on board updateBoard(); } // this method is triggered by SwingUtilities.invokeLater in other threads public void run() { updateBoard(); } /************ general functions ****************/ public void updateBoard() { // update score panels mousescorelabel.setText(MS_TEXT+" "+Integer.toString(mousescore)); catscorelabel.setText(CS_TEXT+" "+Integer.toString(catscore)); if (game.newInfo) { updateScore(); game.newInfo = false; } // update progress info progress.setValue(rlc.epochsdone); learnEpochsDone.setText(Integer.toString(rlc.totaldone)); if (rlc.newInfo) endTraining(); // update game board bp.clearBoard(); // draw walls boolean[][] w = game.getWalls(); for (int i=0; i<w.length; i++) { for (int j=0; j<w[0].length; j++) { if (w[i][j]) bp.setSquare(wall, i, j); } } // draw objects (cat over mouse over cheese) bp.setSquare(cheese, game.getCheese()); bp.setSquare(mouse, game.getMouse()); bp.setSquare(cat, game.getCat()); //bp.setSquare(hole, game.getHole()); // display text representation //System.out.println(bp); bp.repaint(); } void doTraining() { // begin training int episodes = Integer.parseInt(epochs.getText()); double aval = Double.parseDouble(alpha.getText()); double gval = Double.parseDouble(gamma.getText()); double eval = Double.parseDouble(epsilon.getText()); int cval = Integer.parseInt(reward.getText()); int dval = Integer.parseInt(penalty.getText()); rl.setAlpha(aval); rl.setGamma(gval); rl.setEpsilon(eval); // disable controls startTraining.setEnabled(false); epochs.setEnabled(false); reward.setEnabled(false); penalty.setEnabled(false); alpha.setEnabled(false); gamma.setEnabled(false); epsilon.setEnabled(false); softmax.setEnabled(false); greedy.setEnabled(false); sarsa.setEnabled(false); qlearn.setEnabled(false); // fix progress bar progress.setMinimum(0); progress.setMaximum(episodes); progress.setValue(0); // enable stop button stopTraining.setEnabled(true); // start training trainWorld.cheeseReward = cval; trainWorld.deathPenalty = dval; rlc.setEpisodes(episodes); } void endTraining() { // stop training rlc.stopLearner(); // enable buttons startTraining.setEnabled(true); epochs.setEnabled(true); reward.setEnabled(true); penalty.setEnabled(true); alpha.setEnabled(true); gamma.setEnabled(true); epsilon.setEnabled(true); softmax.se

评论收藏

内容反馈