package com.smartlagfix.ai; import com.smartlagfix.SmartLagFix; import org.bukkit.Bukkit; import org.bukkit.scheduler.BukkitTask; import java.util.ArrayDeque; import java.util.Deque; public class AIMode { private final SmartLagFix plugin; private BukkitTask aiTask; private final Deque tpsHistory; private final Deque entityHistory; private int sampleCount; private boolean learningComplete; private double adaptiveCriticalTPS; private double adaptiveWarningTPS; public AIMode(SmartLagFix plugin) { this.plugin = plugin; this.tpsHistory = new ArrayDeque<>(); this.entityHistory = new ArrayDeque<>(); this.sampleCount = 0; this.learningComplete = false; this.adaptiveCriticalTPS = plugin.getConfigManager().getCriticalTPS(); this.adaptiveWarningTPS = plugin.getConfigManager().getWarningTPS(); } public void start() { this.aiTask = Bukkit.getScheduler().runTaskTimerAsynchronously(plugin, this::analyze, 100L, 100L); plugin.getLogger().info("AI Mode started - learning phase active"); } public void stop() { if (aiTask != null && !aiTask.isCancelled()) { aiTask.cancel(); } } private void analyze() { collectData(); if (!learningComplete) { if (sampleCount >= plugin.getConfigManager().getLearningPeriod()) { completeLearning(); } return; } if (plugin.getConfigManager().isAdaptiveThresholds()) { adjustThresholds(); } makeOptimizationDecision(); } private void collectData() { double currentTPS = plugin.getTpsMonitor().getCurrentTPS(); int entityCount = plugin.getEntityMonitor().getTotalEntities(); tpsHistory.addLast(currentTPS); entityHistory.addLast(entityCount); if (tpsHistory.size() > 100) { tpsHistory.removeFirst(); entityHistory.removeFirst(); } sampleCount++; } private void completeLearning() { learningComplete = true; double avgTPS = tpsHistory.stream().mapToDouble(Double::doubleValue).average().orElse(20.0); int avgEntities = entityHistory.stream().mapToInt(Integer::intValue).sum() / entityHistory.size(); plugin.getLogger().info("AI learning complete - Baseline TPS: " + String.format("%.2f", avgTPS) + ", Avg Entities: " + avgEntities); } private void adjustThresholds() { double recentAvgTPS = tpsHistory.stream() .skip(Math.max(0, tpsHistory.size() - 20)) .mapToDouble(Double::doubleValue) .average() .orElse(20.0); int aggressiveness = plugin.getConfigManager().getAggressiveness(); double adjustment = (aggressiveness - 5) * 0.2; this.adaptiveCriticalTPS = Math.max(12.0, plugin.getConfigManager().getCriticalTPS() + adjustment); this.adaptiveWarningTPS = Math.max(15.0, plugin.getConfigManager().getWarningTPS() + adjustment); if (plugin.getConfigManager().isDebug()) { plugin.getLogger().info("AI adjusted thresholds - Critical: " + adaptiveCriticalTPS + ", Warning: " + adaptiveWarningTPS); } } private void makeOptimizationDecision() { double currentTPS = plugin.getTpsMonitor().getCurrentTPS(); int aggressiveness = plugin.getConfigManager().getAggressiveness(); if (currentTPS < adaptiveCriticalTPS) { if (aggressiveness >= 7) { Bukkit.getScheduler().runTask(plugin, () -> { plugin.getAutoOptimizer().optimize(); }); } } else if (currentTPS < adaptiveWarningTPS) { if (aggressiveness >= 5 && shouldOptimizePreventively()) { Bukkit.getScheduler().runTask(plugin, () -> { plugin.getAutoOptimizer().optimize(); }); } } } private boolean shouldOptimizePreventively() { if (tpsHistory.size() < 10) return false; double trend = calculateTrend(); return trend < -0.5; } private double calculateTrend() { if (tpsHistory.size() < 10) return 0; double[] recent = tpsHistory.stream() .skip(Math.max(0, tpsHistory.size() - 10)) .mapToDouble(Double::doubleValue) .toArray(); double firstHalf = 0; double secondHalf = 0; int halfSize = recent.length / 2; for (int i = 0; i < halfSize; i++) { firstHalf += recent[i]; } for (int i = halfSize; i < recent.length; i++) { secondHalf += recent[i]; } return (secondHalf / (recent.length - halfSize)) - (firstHalf / halfSize); } }