"""
ATC 警觉力衰退分类 → 汽车驾驶对应场景
"""
from dataclasses import dataclass
from typing import List

@dataclass
class VigilanceDecrement:
    """警觉力衰退类型"""
    name: str
    atc_scenario: str
    driving_scenario: str
    detection_signal: str
    severity: str

DECREMENTS = [
    VigilanceDecrement(
        name="低负荷脱节",
        atc_scenario="低流量时段，空管员脱节",
        driving_scenario="高速公路长时间单调驾驶",
        detection_signal="θ功率上升，α反弹",
        severity="high"
    ),
    VigilanceDecrement(
        name="过载失效",
        atc_scenario="高密度扇区，空管员处理能力下降",
        driving_scenario="复杂城市交通+多源信息干扰",
        detection_signal="瞳孔收缩，HRV降低",
        severity="critical"
    ),
    VigilanceDecrement(
        name="自动化消极",
        atc_scenario="自动化处理后，空管员监控注意力下降",
        driving_scenario="ADAS开启后驾驶员注意力分散",
        detection_signal="眨眼频率增加，反应时延长",
        severity="high"
    ),
    VigilanceDecrement(
        name="时间压力协调错误",
        atc_scenario="紧急情况多任务时间压力",
        driving_scenario="紧急变道+刹车场景",
        detection_signal="HRV极低，EEG beta 降低",
        severity="critical"
    ),
]

for d in DECREMENTS:
    print(f"{'⚠️' if d.severity=='high' else '🔴'} {d.name}")
    print(f"  ATC: {d.atc_scenario}")
    print(f"  汽车: {d.driving_scenario}")
    print(f"  信号: {d.detection_signal}")
    print()

"""
ATC 警觉力监测多模态融合框架
迁移到汽车 DMS
"""
import numpy as np

class MultimodalVigilanceFusion:
    """多模态警觉力融合检测"""
    
    def __init__(self):
        # 各传感器权重（来自ATC研究经验）
        self.weights = {
            "eeg_theta_alpha_ratio": 0.30,   # 最强预测力但侵入性高
            "eye_blink_rate": 0.25,          # DMS摄像头可实现
            "hrv_lf_hf_ratio": 0.20,         # 方向盘/座椅传感器
            "pupil_diameter": 0.15,          # DMS摄像头可实现
            "steering_entropy": 0.10,        # 车辆行为数据
        }
        
    def compute_vigilance_score(self, features: dict) -> float:
        """
        计算综合警觉力评分
        
        Args:
            features: 各传感器提取的特征
        
        Returns:
            vigilance_score: 0(完全衰退) - 1(最佳状态)
        """
        score = 1.0
        
        # EEG θ/α比升高 → 警觉力下降
        if "eeg_theta_alpha_ratio" in features:
            ratio = features["eeg_theta_alpha_ratio"]
            # ATC研究: 正常<2.0, 衰退>3.0
            eeg_impact = max(0, (ratio - 2.0) / 2.0) * self.weights["eeg_theta_alpha_ratio"]
            score -= eeg_impact
        
        # 眨眼率升高 → 警觉力下降
        if "eye_blink_rate" in features:
            rate = features["eye_blink_rate"]  # 次/分钟
            # 正常10-20, 衰退>25
            blink_impact = max(0, (rate - 20) / 15) * self.weights["eye_blink_rate"]
            score -= blink_impact
        
        # HRV降低 → 警觉力下降
        if "hrv_lf_hf_ratio" in features:
            ratio = features["hrv_lf_hf_ratio"]
            # 正常2-4, 衰退<1.5
            hrv_impact = max(0, (2.0 - ratio) / 2.0) * self.weights["hrv_lf_hf_ratio"]
            score -= hrv_impact
        
        # 瞳孔直径变化 → 认知负荷/疲劳
        if "pupil_diameter_cv" in features:
            cv = features["pupil_diameter_cv"]  # 变异系数
            pupil_impact = min(0.15, cv * 0.5) * self.weights["pupil_diameter"]
            score -= pupil_impact
        
        # 方向盘熵升高 → 警觉力下降
        if "steering_entropy" in features:
            entropy = features["steering_entropy"]
            steer_impact = max(0, (entropy - 0.5) / 0.5) * self.weights["steering_entropy"]
            score -= steer_impact
        
        return max(0, min(1, score))
    
    def get_alert_level(self, score: float) -> dict:
        """根据评分获取告警级别"""
        if score > 0.7:
            return {"level": "normal", "action": "none", "color": "green"}
        elif score > 0.5:
            return {"level": "caution", "action": "audio_alert", "color": "yellow"}
        elif score > 0.3:
            return {"level": "warning", "action": "hud_alert+seat_vibration", "color": "orange"}
        else:
            return {"level": "critical", "action": "pull_over_safely", "color": "red"}

# 测试
fusion = MultimodalVigilanceFusion()

# 模拟正常状态
normal_features = {
    "eye_blink_rate": 15,
    "hrv_lf_hf_ratio": 3.0,
    "pupil_diameter_cv": 0.05,
    "steering_entropy": 0.3,
}
score1 = fusion.compute_vigilance_score(normal_features)
print(f"正常驾驶 - 警觉力评分: {score1:.2f}")
print(f"  → {fusion.get_alert_level(score1)}")

# 模拟疲劳状态
fatigued_features = {
    "eye_blink_rate": 28,
    "hrv_lf_hf_ratio": 1.2,
    "pupil_diameter_cv": 0.12,
    "steering_entropy": 0.7,
}
score2 = fusion.compute_vigilance_score(fatigued_features)
print(f"疲劳驾驶 - 警觉力评分: {score2:.2f}")
print(f"  → {fusion.get_alert_level(score2)}")