""" 群体管理器 负责创建和管理群体成员 """ from panda3d.core import Vec3, Point3, LVector3f import random import math import time class SwarmManager: """ 群体管理器 负责创建和管理群体成员，实现完整的群体智能行为 """ def __init__(self, world): self.world = world self.members = [] self.obstacles = [] self.predators = [] self.paused = False self.task = None self.swarm_type = "鸟类" self.target = None self.path = [] self.bounds = { 'min_x': -50, 'max_x': 50, 'min_y': -50, 'max_y': 50, 'min_z': 0, 'max_z': 30 } # 多群体支持 self.swarms = [] # 存储多个群体 self.swarm_id_counter = 0 # 性能统计 self.frame_count = 0 self.last_update_time = time.time() self.fps = 0 # 群体参数 self.cohesion_weight = 1.0 self.separation_weight = 1.5 self.alignment_weight = 1.0 self.max_speed = 5.0 self.perception_radius = 10.0 self.member_count = 20 # 初始化高级Boids算法 from .advanced_boids import AdvancedBoidsAlgorithm from .config import Config from .interaction_manager import InteractionManager from .learning_manager import LearningManager from .environment_manager import EnvironmentManager self.config = Config() self.boids_algorithm = AdvancedBoidsAlgorithm(self.config) self.interaction_manager = InteractionManager(self.config) self.learning_manager = LearningManager(self.config) self.environment_manager = EnvironmentManager(self.config) # 启动更新任务 self.task = self.world.taskMgr.add(self._update_task, "swarm_update_task") def create_example_swarm(self, swarm_type=None, member_count=None, position_offset=Vec3(0, 0, 0)): """ 创建示例群体 根据群体类型创建不同类型的群体成员 """ # 如果是创建新的独立群体，不清理现有群体 if swarm_type is None and member_count is None: # 清理现有群体 self.cleanup() # 获取群体类型和数量 swarm_type = swarm_type or self.swarm_type count = member_count or self.config.get("member_count", self.member_count) # 创建指定数量的群体成员 new_members = [] for i in range(count): # 根据群体类型设置随机位置范围，并应用位置偏移 if swarm_type == "鸟类": pos = Vec3( random.uniform(-20, 20), random.uniform(-20, 20), random.uniform(5, 25) ) + position_offset elif swarm_type == "鱼类": pos = Vec3( random.uniform(-15, 15), random.uniform(-15, 15), random.uniform(2, 15) ) + position_offset elif swarm_type == "昆虫": pos = Vec3( random.uniform(-10, 10), random.uniform(-10, 10), random.uniform(1, 10) ) + position_offset else: # 自定义 pos = Vec3( random.uniform(self.bounds['min_x'], self.bounds['max_x']), random.uniform(self.bounds['min_y'], self.bounds['min_y']), random.uniform(self.bounds['min_z'], self.bounds['max_z']) ) + position_offset # 根据群体类型加载不同的模型 if swarm_type == "鸟类": member_model = "models/misc/sphere" # 实际应用中应使用鸟类模型 scale = 0.5 elif swarm_type == "鱼类": member_model = "models/misc/sphere" # 实际应用中应使用鱼类模型 scale = 0.3 elif swarm_type == "昆虫": member_model = "models/misc/sphere" # 实际应用中应使用昆虫模型 scale = 0.2 else: member_model = "models/misc/sphere" scale = 0.5 # 加载模型 try: member_node = self.world.loader.loadModel(member_model) except: # 如果模型加载失败，使用默认球体 member_node = self.world.loader.loadModel("models/misc/sphere") member_node.reparentTo(self.world.render) member_node.setPos(pos) member_node.setScale(scale) # 随机初始速度 velocity = Vec3( random.uniform(-1, 1), random.uniform(-1, 1), random.uniform(-1, 1) ) # 添加到群体成员列表 member = { 'node': member_node, 'position': pos, 'velocity': velocity, 'index': i, # 成员索引，用于队形计算 'type': swarm_type, # 成员类型 'swarm_id': self.swarm_id_counter # 群体ID } new_members.append(member) # 为成员添加环境传感器 if hasattr(self, 'environment_manager'): self.environment_manager.add_sensor(id(member)) # 如果是创建新的独立群体 if swarm_type is not None or member_count is not None: # 创建新的群体对象 swarm = { 'id': self.swarm_id_counter, 'type': swarm_type, 'members': new_members, 'position_offset': position_offset } self.swarms.append(swarm) self.swarm_id_counter += 1 # 将新成员添加到主成员列表 self.members.extend(new_members) print(f"已创建包含{len(new_members)}个成员的{swarm_type}群体 (ID: {swarm['id']})") return swarm else: # 更新主成员列表 self.members = new_members print(f"已创建包含{len(self.members)}个成员的{swarm_type}群体") def create_multiple_swarms(self, swarm_configs: List[Dict]): """ 创建多个群体 :param swarm_configs: 群体配置列表，每个配置包含type, count, position_offset """ self.cleanup() # 清理现有群体 self.swarms.clear() # 清理群体列表 self.swarm_id_counter = 0 # 重置ID计数器 for config in swarm_configs: swarm_type = config.get('type', '鸟类') member_count = config.get('count', 20) position_offset = config.get('position_offset', Vec3(0, 0, 0)) self.create_example_swarm(swarm_type, member_count, position_offset) print(f"已创建{len(self.swarms)}个群体") def get_swarm_by_id(self, swarm_id: int): """ 根据ID获取群体 """ for swarm in self.swarms: if swarm['id'] == swarm_id: return swarm return None def get_swarm_by_type(self, swarm_type: str): """ 根据类型获取群体 """ return [swarm for swarm in self.swarms if swarm['type'] == swarm_type] def create_obstacle(self, position, radius): """ 创建障碍物 """ obstacle = { 'position': position, 'radius': radius } self.obstacles.append(obstacle) return obstacle def create_predator(self, position): """ 创建捕食者 """ # 创建捕食者模型 try: predator_node = self.world.loader.loadModel("models/misc/sphere") # 实际应用中应使用捕食者模型 except: predator_node = self.world.loader.loadModel("models/misc/sphere") predator_node.reparentTo(self.world.render) predator_node.setPos(position) predator_node.setScale(1.0) predator_node.setColor(1, 0, 0, 1) # 红色表示捕食者 # 创建捕食者对象 predator = { 'node': predator_node, 'position': position, 'velocity': Vec3(0, 0, 0), 'target_index': 0 # 目标群体成员索引 } self.predators.append(predator) return predator def set_target(self, target): """ 设置群体目标位置 """ self.target = target def set_path(self, path): """ 设置群体移动路径 """ self.path = path def _update_task(self, task): """ 更新任务 """ if not self.paused: self._update_swarm() # 更新性能统计 self.frame_count += 1 current_time = time.time() if current_time - self.last_update_time >= 1.0: self.fps = self.frame_count self.frame_count = 0 self.last_update_time = current_time return task.cont def _update_swarm(self): """ 更新群体行为 """ # 更新捕食者行为（如果存在） self._update_predators() # 更新群体间交互 interaction_forces = {} if hasattr(self, 'interaction_manager'): # 为每个成员计算交互力 for member in self.members: interaction_forces[id(member)] = {} # 应用合作力 if 'cooperative_force' in member: interaction_forces[id(member)]['cooperative_force'] = member['cooperative_force'] del member['cooperative_force'] # 应用竞争力 if 'competition_force' in member: interaction_forces[id(member)]['competition_force'] = member['competition_force'] del member['competition_force'] # 应用捕食力 if 'predation_force' in member: interaction_forces[id(member)]['predation_force'] = member['predation_force'] del member['predation_force'] # 应用逃避力 if 'escape_force' in member: interaction_forces[id(member)]['escape_force'] = member['escape_force'] del member['escape_force'] # 更新环境状态 if hasattr(self, 'environment_manager'): self.environment_manager.update_environment_state( self.obstacles, self.swarms, self.predators, getattr(self, 'resources', []), getattr(self, 'hazards', []) ) # 更新学习和适应 if hasattr(self, 'learning_manager') and self.config.get("learning_enabled", False): # 计算环境状态 environment_state = self._calculate_environment_state() # 更新环境模型 self.learning_manager.update_environment_model(environment_state) # 适应环境 self.learning_manager.adapt_to_environment(environment_state) # 获取适应后的参数 adapted_params = self.learning_manager.get_adapted_parameters() # 更新Boids算法参数 self._update_boids_parameters(adapted_params) # 检查是否启用并行计算 if self.config.get("parallel_computation", False) and len(self.members) > 50: # 使用并行计算更新群体成员 self._update_swarm_parallel() else: # 使用串行计算更新群体成员 for i, member in enumerate(self.members): # 计算邻居 neighbors = self._find_neighbors(member, i) # 获取扩展参数 obstacles = self.obstacles if self.config.get("obstacle_avoidance_enabled", True) else [] bounds = self.bounds if self.config.get("boundary_enabled", True) else None predators = self.predators if self.config.get("predator_avoid_enabled", False) else [] target = self.target if self.config.get("seek_enabled", False) else None path = self.path if self.config.get("path_following_enabled", False) else None interactions = interaction_forces.get(id(member), {}) if interaction_forces else {} # 计算奖励（如果启用学习） if hasattr(self, 'learning_manager') and self.config.get("learning_enabled", False): reward = self.learning_manager.calculate_reward(member, neighbors, obstacles, environment_state) # 创建状态表示 state = { 'neighbor_count': len(neighbors), 'obstacle_distance': self._calculate_min_obstacle_distance(member), 'boundary_distance': self._calculate_boundary_distance(member), 'velocity': member['velocity'].length(), 'position_x': member['position'].x, 'position_y': member['position'].y, 'position_z': member['position'].z } # 添加经验（简化） # 在实际应用中，这里需要更复杂的动作表示 action = "maintain" # 简化动作 next_state = state.copy() # 简化下一状态 self.learning_manager.add_experience(state, action, reward, next_state) # 获取环境响应力 environment_force = Vec3(0, 0, 0) if (hasattr(self, 'environment_manager') and self.config.get("environment_awareness_enabled", True)): environment_force = self.environment_manager.get_environmental_response( member, member['position'], member['velocity'] ) # 使用Boids算法更新成员状态，添加环境响应力 self.boids_algorithm.update_member( member, neighbors, obstacles, target, bounds, path, predators, interactions ) # 应用环境响应力 if self.config.get("environment_awareness_enabled", True): member['velocity'] += environment_force * 0.1 # 环境力权重 # 更新成员节点位置 member['node'].setPos(member['position']) # 更新成员朝向（面向移动方向） if member['velocity'].length() > 0: # 创建一个向前的向量 forward = member['velocity'].normalized() # 设置节点朝向 member['node'].lookAt(member['position'] + forward) # 更新行为策略（定期更新） if hasattr(self, 'learning_manager') and self.config.get("learning_enabled", False): if self.frame_count % 60 == 0: # 每秒更新一次（假设60FPS） self.learning_manager.update_behavior_policies() def _update_swarm_parallel(self): """ 使用并行计算更新群体行为 """ # 获取扩展参数 obstacles = self.obstacles if self.config.get("obstacle_avoidance_enabled", True) else [] bounds = self.bounds if self.config.get("boundary_enabled", True) else None predators = self.predators if self.config.get("predator_avoid_enabled", False) else [] target = self.target if self.config.get("seek_enabled", False) else None path = self.path if self.config.get("path_following_enabled", False) else None # 准备交互力 interaction_forces = {} if hasattr(self, 'interaction_manager'): # 为每个成员计算交互力 for member in self.members: interaction_forces[id(member)] = {} # 应用合作力 if 'cooperative_force' in member: interaction_forces[id(member)]['cooperative_force'] = member['cooperative_force'] del member['cooperative_force'] # 应用竞争力 if 'competition_force' in member: interaction_forces[id(member)]['competition_force'] = member['competition_force'] del member['competition_force'] # 应用捕食力 if 'predation_force' in member: interaction_forces[id(member)]['predation_force'] = member['predation_force'] del member['predation_force'] # 应用逃避力 if 'escape_force' in member: interaction_forces[id(member)]['escape_force'] = member['escape_force'] del member['escape_force'] # 准备环境力参数 environment_forces = {} if (hasattr(self, 'environment_manager') and self.config.get("environment_awareness_enabled", True)): for member in self.members: environment_forces[id(member)] = self.environment_manager.get_environmental_response( member, member['position'], member['velocity'] ) else: # 如果环境感知被禁用，设置为零力 for member in self.members: environment_forces[id(member)] = Vec3(0, 0, 0) # 使用并行算法更新群体成员 self.boids_algorithm.update_swarm_parallel( self.members, self.members, obstacles, target, bounds, path, predators, interaction_forces, environment_forces ) # 更新所有成员的节点位置和朝向 for member in self.members: member['node'].setPos(member['position']) # 更新成员朝向（面向移动方向） if member['velocity'].length() > 0: # 创建一个向前的向量 forward = member['velocity'].normalized() # 设置节点朝向 member['node'].lookAt(member['position'] + forward) def _update_predators(self): """ 更新捕食者行为 """ for predator in self.predators: if self.members: # 简单的捕食者行为：寻找最近的群体成员 nearest_member = None nearest_distance = float('inf') for member in self.members: distance = (predator['position'] - member['position']).length() if distance < nearest_distance: nearest_distance = distance nearest_member = member if nearest_member: # 向最近的成员移动 direction = (nearest_member['position'] - predator['position']).normalized() predator['velocity'] = direction * 3.0 # 捕食者速度 predator['position'] += predator['velocity'] predator['node'].setPos(predator['position']) def _calculate_environment_state(self): """ 计算环境状态 :return: 环境状态字典 """ if not self.members: return {} # 计算群体密度 member_count = len(self.members) total_area = (self.bounds['max_x'] - self.bounds['min_x']) * (self.bounds['max_y'] - self.bounds['min_y']) density = member_count / total_area if total_area > 0 else 0 # 计算平均速度 total_speed = sum(member['velocity'].length() for member in self.members) avg_speed = total_speed / member_count if member_count > 0 else 0 # 计算与障碍物的平均距离 total_obstacle_distance = 0 obstacle_count = 0 for member in self.members: min_distance = float('inf') for obstacle in self.obstacles: distance = (member['position'] - obstacle['position']).length() - obstacle.get('radius', 1.0) min_distance = min(min_distance, distance) if min_distance != float('inf'): total_obstacle_distance += min_distance obstacle_count += 1 avg_obstacle_distance = total_obstacle_distance / obstacle_count if obstacle_count > 0 else 10.0 # 威胁水平（基于与障碍物的距离） threat_level = max(0.0, min(1.0, 1.0 - avg_obstacle_distance / 20.0)) # 资源丰富度（简化为群体密度的反比） resource_abundance = max(0.0, min(1.0, 1.0 - density * 10.0)) return { 'density': density, 'avg_speed': avg_speed, 'avg_obstacle_distance': avg_obstacle_distance, 'threat_level': threat_level, 'resource_abundance': resource_abundance, 'member_count': member_count } def _calculate_min_obstacle_distance(self, member): """ 计算成员与最近障碍物的距离 :param member: 群体成员 :return: 最小距离 """ if not self.obstacles: return float('inf') min_distance = float('inf') for obstacle in self.obstacles: distance = (member['position'] - obstacle['position']).length() - obstacle.get('radius', 1.0) min_distance = min(min_distance, distance) return min_distance def _calculate_boundary_distance(self, member): """ 计算成员与边界的距离 :param member: 群体成员 :return: 最小边界距离 """ min_distance = float('inf') # 计算到各边界的距离 distances = [ member['position'].x - self.bounds['min_x'], self.bounds['max_x'] - member['position'].x, member['position'].y - self.bounds['min_y'], self.bounds['max_y'] - member['position'].y, member['position'].z - self.bounds['min_z'], self.bounds['max_z'] - member['position'].z ] return min(distances) if distances else float('inf') def _update_boids_parameters(self, adapted_params): """ 更新Boids算法参数 :param adapted_params: 适应后的参数 """ # 更新速度 current_max_speed = self.config.get("max_speed", 5.0) self.config.set("max_speed", current_max_speed * adapted_params['speed']) # 更新行为权重 current_cohesion = self.config.get("cohesion_weight", 1.0) self.config.set("cohesion_weight", current_cohesion * adapted_params['cohesion']) current_separation = self.config.get("separation_weight", 1.5) self.config.set("separation_weight", current_separation * adapted_params['separation']) current_alignment = self.config.get("alignment_weight", 1.0) self.config.set("alignment_weight", current_alignment * adapted_params['alignment']) def _find_neighbors(self, member, index): """ 查找邻居 使用空间分区优化邻居查找（简化实现） """ neighbors = [] perception_radius = self.config.get("perception_radius", self.perception_radius) for i, other in enumerate(self.members): if i != index: # 排除自己 distance = (member['position'] - other['position']).length() if distance < perception_radius: neighbors.append(other) return neighbors def add_member(self): """ 添加一个新的群体成员 """ # 随机位置（靠近现有群体） if self.members: # 在现有群体附近创建新成员 existing_member = random.choice(self.members) pos = Vec3( existing_member['position'].x + random.uniform(-5, 5), existing_member['position'].y + random.uniform(-5, 5), existing_member['position'].z + random.uniform(-2, 2) ) else: # 如果没有现有成员，在随机位置创建 pos = Vec3( random.uniform(-20, 20), random.uniform(-20, 20), random.uniform(5, 15) ) # 加载模型 try: member_node = self.world.loader.loadModel("models/misc/sphere") except: member_node = self.world.loader.loadModel("models/misc/sphere") member_node.reparentTo(self.world.render) member_node.setPos(pos) member_node.setScale(0.5) # 随机初始速度 velocity = Vec3( random.uniform(-1, 1), random.uniform(-1, 1), random.uniform(-1, 1) ) # 添加到群体成员列表 member = { 'node': member_node, 'position': pos, 'velocity': velocity, 'index': len(self.members), # 成员索引 'type': self.swarm_type } self.members.append(member) self.config.set("member_count", len(self.members)) return member def remove_member(self): """ 移除一个群体成员 """ if self.members: # 移除随机成员 member = self.members.pop() member['node'].removeNode() self.config.set("member_count", len(self.members)) return member return None def update_swarm_type(self, swarm_type): """ 更新群体类型 """ self.swarm_type = swarm_type # 重新创建群体以应用新的类型 self.cleanup() self.create_example_swarm() def get_stats(self): """ 获取群体统计信息 """ return { 'member_count': len(self.members), 'paused': self.paused, 'fps': self.fps, 'swarm_type': self.swarm_type } def toggle_pause(self): """ 切换暂停状态 """ self.paused = not self.paused print(f"群体模拟{'已暂停' if self.paused else '已恢复'}") def cleanup(self): """ 清理资源 """ # 移除任务 if self.task: self.world.taskMgr.remove(self.task) # 移除所有群体成员 for member in self.members: member['node'].removeNode() # 移除所有捕食者 for predator in self.predators: predator['node'].removeNode() # 清空列表 self.members.clear() self.obstacles.clear() self.predators.clear() def reset(self): """ 重置群体 """ self.cleanup() self.create_example_swarm()