356 nnet

build.gradle

@@ -4,11 +4,6 @@ plugins {
     id 'org.openjfx.javafxplugin' version '0.0.10'
     id 'org.beryx.jlink' version '2.24.1'
     id 'io.freefair.lombok' version '6.4.0'
-
-    // Note: the java modularity plugin is added by default with javafx, this manual addition is to override the version
-    // more info here https://discuss.gradle.org/t/unable-to-build-project-with-gradle-7-3-1-on-java-17/41601
-    // id 'org.javamodularity.moduleplugin' version '1.8.10'
-    // Removed 16 Feb - Seems to have had no effect
 }
 
 version '1.0'
@@ -39,7 +34,6 @@ tasks.withType(JavaCompile) {
 }
 
 application {
-    mainModule = 'app'
     mainClass = 'app.App'
 }
 
@@ -54,6 +48,9 @@ dependencies {
     implementation 'org.jgrapht:jgrapht-core:1.5.1'
     testImplementation("org.junit.jupiter:junit-jupiter-api:${junitVersion}")
     testRuntimeOnly("org.junit.jupiter:junit-jupiter-engine:${junitVersion}")
+
+    // https://mvnrepository.com/artifact/com.deepnetts/deepnetts-core
+    implementation group: 'com.deepnetts', name: 'deepnetts-core', version: '1.13.0'
 }
 
 test {

data_cleaning.ipynb

@@ -0,0 +1,70 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 23,
+   "id": "805f63e6",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "with open('src/main/resources/nNet/data.csv') as file:\n",
+    "    data = file.readlines()\n",
+    "    \n",
+    "clean = []"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 24,
+   "id": "2f8a53e9",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "for line in data:\n",
+    "    \n",
+    "    if not 'NaN' in line:\n",
+    "        clean.append(line)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 25,
+   "id": "120244af",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "with open('src/main/resources/nNet/data.csv', 'w') as file:\n",
+    "    file.writelines(clean)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "95f5fbfc",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.10.4"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

src/main/java/app/controller/io/FileManager.java

@@ -5,7 +5,6 @@
 import java.io.File;
 import java.nio.file.Path;
 import java.nio.file.Paths;
-import java.time.LocalDateTime;
 import java.util.Scanner;
 
 public class FileManager

src/main/java/app/controller/settings/RandomSettingsGenerator.java

@@ -35,8 +35,8 @@ public abstract class RandomSettingsGenerator
 
     public static Settings generateRandomSettings()
     {
+        clearRandomGenerator();
         initialiseTemplates();
-
         Settings randomSettings = createBaseSettings();
         createRandomObstacles(randomSettings);
 

src/main/java/app/controller/settings/Settings.java

@@ -28,8 +28,8 @@ public class Settings
     private double sprintSpeedIntruder;
     private double timeStep;
     private double scaling;
-    private AgentType guardType = AgentType.ACO;
-    private AgentType IntruderType = AgentType.WALL_FOLLOW;
+    private AgentType guardType = AgentType.ACO_MOMENTUM_SPIRAL_AVOIDANCE;
+    private AgentType IntruderType = AgentType.NEURAL_NET;
     private ArrayList<SettingsObject> furniture = new ArrayList<>();
     private ArrayList<SettingsObject> soundSources = new ArrayList<>();
     private ArrayList<SettingsObject> soundFurniture = new ArrayList<>();

src/main/java/app/model/Map.java

@@ -44,7 +44,6 @@ public class Map
 
     public Map(Settings settings)
     {
-        System.out.print("Loading settings ... ");
         this.settings = settings;
         this.width = settings.getWidth();
         this.height = settings.getHeight();
@@ -93,7 +92,6 @@ public Map(Settings settings)
             activateHuman();
 
         this.coverage = new Coverage(this);
-        System.out.print(" done");
     }
 
     public Map(Agent agent, ArrayList<Furniture> obstacles)
@@ -157,6 +155,18 @@ public MemoryGraph createFullGraph()
         return fullGraph;
     }
 
+    public void addAgent(Agent agent)
+    {
+        if(agent.getType() == Type.INTRUDER)
+                agent.setPosition(randPosition(intruderSpawn));
+        else if(agent.getType() == Type.GUARD)
+            agent.setPosition(randPosition(guardSpawn));
+
+        agent.setDirection(randDirection());
+        agents.add(agent);
+    }
+
+
     public void updateAllSeen(Agent agent)
     {
         if(agent.getType() == Type.GUARD)

src/main/java/app/model/agents/Agent.java

@@ -20,6 +20,8 @@ public interface Agent extends Boundary
 
     Vector getPosition();
 
+    void setPosition(Vector position);
+
     Vector getDirection();
 
     void setDirection(Vector direction);

src/main/java/app/model/agents/AgentImp.java

@@ -15,14 +15,12 @@
 import app.model.agents.Capture.DijkstraCaptureAgent;
 import app.view.agentView.AgentView;
 import app.view.simulation.Info;
+import java.util.ArrayList;
 import javafx.scene.canvas.GraphicsContext;
 import javafx.scene.paint.Color;
 import lombok.Getter;
 import lombok.Setter;
 
-import java.util.ArrayList;
-
-
 public class AgentImp implements Agent
 {
     @Getter @Setter protected double maxWalk = 10;
@@ -33,15 +31,15 @@ public class AgentImp implements Agent
     @Getter @Setter protected Vector typePosition;
     @Getter @Setter protected Vector tgtDirection;
     @Getter @Setter protected ArrayList<Ray> view;
+    @Getter @Setter protected World world;
+    @Getter @Setter protected Vector position;
     @Getter protected Type type;
-    @Getter protected Vector position;
     @Getter protected double radius;
     @Getter protected ArrayList<SoundVector> heard;
     @Getter protected VectorSet seen;
     @Getter protected AgentView agentViewWindow;
     protected final boolean DRAW_HEARD = false;
 
-    @Getter @Setter protected World world;
 
 
     public AgentImp(Vector position, Vector direction, double radius, Type type)
@@ -276,4 +274,22 @@ public Agent nextState()
     {
         return StateTable.lookupState(this);
     }
+
+    protected float[] normalise(float[] data)
+    {
+        float max = 0f;
+        for(int i = 0; i < data.length; i++)
+        {
+            if(data[i] > max)
+                max = data[i];
+        }
+
+        for(int i = 0; i < data.length; i++)
+        {
+            if(data[i] != -1)
+                data[i] = data[i] / 1000;
+        }
+
+        return data;
+    }
 }

src/main/java/app/model/agents/AgentType.java

@@ -3,11 +3,14 @@
 import app.controller.linAlg.Vector;
 import app.model.Type;
 import app.model.agents.ACO.*;
+import app.model.agents.Capture.BaselineCaptureAgent;
+import app.model.agents.Capture.CaptureAgent;
+import app.model.agents.Capture.DijkstraCaptureAgent;
 import app.model.agents.Evasion.*;
-import app.model.agents.Capture.*;
 import app.model.agents.WallFollow.WFMedDirHeuristic;
 import app.model.agents.WallFollow.WFHighDirHeuristic;
 import app.model.agents.WallFollow.WallFollowAgent;
+import app.model.agents.nNet.NeuralNet;
 
 public enum AgentType
 {
@@ -18,18 +21,20 @@ public enum AgentType
     ACO_RANKING,
     ACO_COLONY,
     WALL_FOLLOW,
-    WALL_FOLLOW_MED_DIR_HEURISTIC,
-    WALL_FOLLOW_HIGH_DIR_HEURISTIC,
     EVASION_RANDOM,
     EVASION_DIRECTED,
+    EVASION_RANDOMDIRECTED,
+    WALL_FOLLOW_MED_DIR_HEURISTIC,
+    WALL_FOLLOW_HIGH_DIR_HEURISTIC,
     EVASION_INTELLIGENT,
     EVASION_DISTANCE_MAX,
     EVASION_HIDEY,
     CAPTURE_BASELINE,
     EVASION_RUNAWAY,
     CAPTURE_DIJKSTRA,
     CAPTURE,
-    RANDOM;
+    RANDOM,
+    NEURAL_NET;
 
     public static Agent agentOf(AgentType agentType, Vector position, Vector direction, double radius, Type type)
     {
@@ -92,6 +97,10 @@ public static Agent agentOf(AgentType agentType, Vector position, Vector directi
             case RANDOM -> {
                 return new AgentImp(position, direction, radius, type);
             }
+            case NEURAL_NET ->
+            {
+                return new NeuralNet(position, direction, radius, type);
+            }
             default -> throw new RuntimeException("Agent type not recognized");
         }
     }

src/main/java/app/model/agents/WallFollow/WallFollowAgent.java

@@ -257,6 +257,7 @@ else if (!noWallDetected(getAngleOfLeftRay()) && leftCell.getObstacle())
         Move nextMove = new Move(newDirection, deltaPos);
         direction = nextMove.getEndDir();
         noMovesDone = false;
+
         return nextMove;
     }