train_model.py<\/h3>\n\n\n\n\"\"\"\n\u7a7a\u6c14\u8d28\u91cf\u5916\u51fa\u5efa\u8bae\u6a21\u578b\u8bad\u7ec3\u811a\u672c\n\u8bad\u7ec3\u5b8c\u6210\u540e\u4f1a\u751f\u6210 models\/air_quality_model.pkl\n\"\"\"\n\nimport pandas as pd\nimport numpy as np\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.ensemble import RandomForestClassifier\nfrom sklearn.preprocessing import LabelEncoder, StandardScaler\nfrom sklearn.metrics import accuracy_score, classification_report, confusion_matrix\nimport joblib\nimport seaborn as sns\nimport matplotlib.pyplot as plt\nimport os\nimport json\nimport warnings\nwarnings.filterwarnings('ignore')\n\n# \u8bbe\u7f6e\u4e2d\u6587\u5b57\u4f53\nplt.rcParams['font.sans-serif'] = ['SimHei', 'Microsoft YaHei', 'Noto Sans CJK JP']\nplt.rcParams['axes.unicode_minus'] = False\n\nclass AirQualityModelTrainer:\n def __init__(self, data_path='data\/cata_6716.csv'):\n \"\"\"\u521d\u59cb\u5316\u6a21\u578b\u8bad\u7ec3\u5668\"\"\"\n self.data_path = data_path\n self.df = None\n self.X = None\n self.y = None\n self.label_encoder = None\n self.scaler = None\n self.model = None\n\n def load_and_preprocess(self):\n \"\"\"\u6b65\u9aa41\uff1a\u52a0\u8f7d\u548c\u9884\u5904\u7406\u6570\u636e\"\"\"\n print(\"=== \u6b65\u9aa41: \u52a0\u8f7d\u548c\u9884\u5904\u7406\u6570\u636e ===\")\n\n # 1.1 \u8bfb\u53d6\u6570\u636e\uff08\u4f7f\u7528GBK\u7f16\u7801\u5904\u7406\u4e2d\u6587\uff09\n try:\n self.df = pd.read_csv(self.data_path, encoding='gbk')\n except:\n # \u5c1d\u8bd5\u5176\u4ed6\u7f16\u7801\n self.df = pd.read_csv(self.data_path, encoding='utf-8')\n\n print(f\"\u2713 \u6570\u636e\u52a0\u8f7d\u6210\u529f: {self.df.shape[0]} \u884c, {self.df.shape[1]} \u5217\")\n print(f\"\u2713 \u6570\u636e\u65f6\u95f4\u8303\u56f4: {self.df['\u5e74\u4efd'].min()}\u5e74 \u81f3 {self.df['\u5e74\u4efd'].max()}\u5e74\")\n print(f\"\u2713 \u57ce\u5e02: {self.df['\u57ce\u5e02\u540d'].iloc[0] if '\u57ce\u5e02\u540d' in self.df.columns else '\u672a\u6307\u5b9a'}\")\n\n # 1.2 \u68c0\u67e5\u5e76\u9009\u62e9\u7279\u5f81\u5217\n feature_columns = [\n '\u7ec6\u9897\u7c92\u7269', # PM2.5\n '\u53ef\u5438\u5165\u9897\u7c92\u7269', # PM10\n '\u4e8c\u6c27\u5316\u786b', # SO2\n '\u4e8c\u6c27\u5316\u6c2e', # NO2\n '\u4e00\u6c27\u5316\u78b3', # CO\n '\u81ed\u6c271\u5c0f\u65f6\u5e73\u5747' # O3\n ]\n\n # \u68c0\u67e5\u7279\u5f81\u662f\u5426\u5b58\u5728\n available_features = [col for col in feature_columns if col in self.df.columns]\n if len(available_features) < len(feature_columns):\n print(f\"\u26a0 \u8b66\u544a: \u90e8\u5206\u7279\u5f81\u4e0d\u5b58\u5728\uff0c\u4f7f\u7528\u53ef\u7528\u7279\u5f81: {available_features}\")\n\n # 1.3 \u521b\u5efa\u7279\u5f81\u77e9\u9635\n self.X = self.df[available_features].copy()\n print(f\"\u2713 \u4f7f\u7528\u7279\u5f81: {available_features}\")\n\n # 1.4 \u521b\u5efa\u76ee\u6807\u53d8\u91cf\uff08\u4eba\u7fa4\u7c7b\u522b\uff09\n print(\"\\n=== \u521b\u5efa\u76ee\u6807\u53d8\u91cf ===\")\n self.df['\u4eba\u7fa4\u7c7b\u522b'] = self.df.apply(self._create_crowd_label, axis=1)\n\n # 1.5 \u7f16\u7801\u76ee\u6807\u53d8\u91cf\n self.label_encoder = LabelEncoder()\n self.y = self.label_encoder.fit_transform(self.df['\u4eba\u7fa4\u7c7b\u522b'])\n\n # \u663e\u793a\u7c7b\u522b\u5206\u5e03\n label_counts = pd.Series(self.y).value_counts().sort_index()\n label_dist = pd.DataFrame({\n '\u7f16\u7801': range(len(self.label_encoder.classes_)),\n '\u4eba\u7fa4\u7c7b\u522b': self.label_encoder.classes_,\n '\u6837\u672c\u6570\u91cf': label_counts.values\n })\n print(\"\u76ee\u6807\u53d8\u91cf\u5206\u5e03:\")\n print(label_dist.to_string(index=False))\n\n # 1.6 \u7279\u5f81\u6807\u51c6\u5316\n self.scaler = StandardScaler()\n self.X_scaled = self.scaler.fit_transform(self.X)\n\n return self.X_scaled, self.y, available_features\n\n def _create_crowd_label(self, row):\n \"\"\"\u6839\u636eAQI\u548c\u6c61\u67d3\u7269\u6d53\u5ea6\u521b\u5efa\u4eba\u7fa4\u7c7b\u522b\u6807\u7b7e\"\"\"\n aqi = row['\u7a7a\u6c14\u8d28\u91cf\u6307\u6570'] if '\u7a7a\u6c14\u8d28\u91cf\u6307\u6570' in row else 50\n\n # \u57fa\u4e8e\u4e2d\u56fd\u7a7a\u6c14\u8d28\u91cf\u6807\u51c6\u548c\u5065\u5eb7\u5efa\u8bae\n if aqi <= 50:\n return '\u5168\u90e8\u4eba\u7fa4\u9002\u5b9c\u5916\u51fa'\n elif aqi <= 100:\n # \u5728\u826f\u7684\u8303\u56f4\u5185\uff0c\u6839\u636e\u4e3b\u8981\u6c61\u67d3\u7269\u8fdb\u4e00\u6b65\u7ec6\u5206\n pm25 = row['\u7ec6\u9897\u7c92\u7269'] if '\u7ec6\u9897\u7c92\u7269' in row else 35\n o3 = row['\u81ed\u6c271\u5c0f\u65f6\u5e73\u5747'] if '\u81ed\u6c271\u5c0f\u65f6\u5e73\u5747' in row else 100\n\n if pm25 > 35 or o3 > 100:\n return '\u654f\u611f\u4eba\u7fa4\u51cf\u5c11\u5916\u51fa'\n else:\n return '\u5168\u90e8\u4eba\u7fa4\u9002\u5b9c\u5916\u51fa'\n elif aqi <= 150:\n return '\u654f\u611f\u4eba\u7fa4\u907f\u514d\u5916\u51fa'\n elif aqi <= 200:\n return '\u6240\u6709\u4eba\u7fa4\u51cf\u5c11\u5916\u51fa'\n elif aqi <= 300:\n return '\u513f\u7ae5\u8001\u4eba\u907f\u514d\u5916\u51fa'\n else:\n return '\u5ba4\u5185\u6d3b\u52a8\u4e3a\u4f73'\n\n def train_model(self, test_size=0.2, random_state=42):\n \"\"\"\u6b65\u9aa42\uff1a\u8bad\u7ec3\u6a21\u578b\"\"\"\n print(\"\\n=== \u6b65\u9aa42: \u8bad\u7ec3\u6a21\u578b ===\")\n\n # 2.1 \u5212\u5206\u8bad\u7ec3\u96c6\u548c\u6d4b\u8bd5\u96c6\n X_train, X_test, y_train, y_test = train_test_split(\n self.X_scaled, self.y,\n test_size=test_size,\n random_state=random_state,\n stratify=self.y\n )\n\n print(f\"\u2713 \u8bad\u7ec3\u96c6: {X_train.shape[0]} \u6837\u672c\")\n print(f\"\u2713 \u6d4b\u8bd5\u96c6: {X_test.shape[0]} \u6837\u672c\")\n\n # 2.2 \u8bad\u7ec3\u968f\u673a\u68ee\u6797\u6a21\u578b\n print(\"\u8bad\u7ec3\u968f\u673a\u68ee\u6797\u6a21\u578b\u4e2d...\")\n self.model = RandomForestClassifier(\n n_estimators=200,\n max_depth=15,\n min_samples_split=5,\n min_samples_leaf=2,\n random_state=random_state,\n class_weight='balanced',\n n_jobs=-1,\n verbose=0\n )\n\n self.model.fit(X_train, y_train)\n print(\"\u2713 \u6a21\u578b\u8bad\u7ec3\u5b8c\u6210\")\n\n # 2.3 \u6a21\u578b\u8bc4\u4f30\n y_pred = self.model.predict(X_test)\n accuracy = accuracy_score(y_test, y_pred)\n\n print(f\"\\n=== \u6a21\u578b\u8bc4\u4f30\u7ed3\u679c ===\")\n print(f\"\u51c6\u786e\u7387: {accuracy:.4f}\")\n print(\"\\n\u8be6\u7ec6\u5206\u7c7b\u62a5\u544a:\")\n print(classification_report(y_test, y_pred,\n target_names=self.label_encoder.classes_))\n\n # 2.4 \u7279\u5f81\u91cd\u8981\u6027\u5206\u6790\n feature_importance = pd.DataFrame({\n '\u7279\u5f81': self.X.columns,\n '\u91cd\u8981\u6027': self.model.feature_importances_\n }).sort_values('\u91cd\u8981\u6027', ascending=False)\n\n print(\"\\n\u7279\u5f81\u91cd\u8981\u6027\u6392\u5e8f:\")\n print(feature_importance.to_string(index=False))\n\n # 2.5 \u53ef\u89c6\u5316\u7279\u5f81\u91cd\u8981\u6027\n self._plot_feature_importance(feature_importance)\n\n # 2.6 \u751f\u6210\u6df7\u6dc6\u77e9\u9635\n self._plot_confusion_matrix(y_test, y_pred)\n\n return accuracy, X_test, y_test, y_pred\n\n def _plot_feature_importance(self, feature_importance):\n \"\"\"\u53ef\u89c6\u5316\u7279\u5f81\u91cd\u8981\u6027\"\"\"\n plt.figure(figsize=(10, 6))\n bars = plt.barh(feature_importance['\u7279\u5f81'], feature_importance['\u91cd\u8981\u6027'],\n color=plt.cm.viridis(np.linspace(0.3, 0.9, len(feature_importance))))\n plt.xlabel('\u7279\u5f81\u91cd\u8981\u6027', fontsize=12)\n plt.title('\u6c61\u67d3\u7269\u7279\u5f81\u91cd\u8981\u6027\u6392\u5e8f', fontsize=14, fontweight='bold')\n plt.gca().invert_yaxis() # \u6700\u91cd\u8981\u7684\u663e\u793a\u5728\u6700\u4e0a\u9762\n\n # \u6dfb\u52a0\u6570\u503c\u6807\u7b7e\n for bar in bars:\n width = bar.get_width()\n plt.text(width + 0.001, bar.get_y() + bar.get_height()\/2,\n f'{width:.3f}', ha='left', va='center', fontsize=10)\n\n plt.tight_layout()\n plt.savefig('feature_importance.png', dpi=300, bbox_inches='tight')\n print(\"\u2713 \u7279\u5f81\u91cd\u8981\u6027\u56fe\u5df2\u4fdd\u5b58: feature_importance.png\")\n plt.close()\n\n def _plot_confusion_matrix(self, y_test, y_pred):\n \"\"\"\u7ed8\u5236\u6df7\u6dc6\u77e9\u9635\"\"\"\n cm = confusion_matrix(y_test, y_pred)\n\n plt.figure(figsize=(10, 8))\n sns.heatmap(cm, annot=True, fmt='d', cmap='Blues',\n xticklabels=self.label_encoder.classes_,\n yticklabels=self.label_encoder.classes_)\n plt.title('\u6df7\u6dc6\u77e9\u9635', fontsize=14, fontweight='bold')\n plt.xlabel('\u9884\u6d4b\u6807\u7b7e', fontsize=12)\n plt.ylabel('\u771f\u5b9e\u6807\u7b7e', fontsize=12)\n plt.xticks(rotation=45, ha='right')\n plt.tight_layout()\n plt.savefig('confusion_matrix.png', dpi=300, bbox_inches='tight')\n print(\"\u2713 \u6df7\u6dc6\u77e9\u9635\u5df2\u4fdd\u5b58: confusion_matrix.png\")\n plt.close()\n\n def save_model(self, model_path='models\/air_quality_model.pkl'):\n \"\"\"\u6b65\u9aa43\uff1a\u4fdd\u5b58\u6a21\u578b\u548c\u76f8\u5173\u6570\u636e\"\"\"\n print(\"\\n=== \u6b65\u9aa43: \u4fdd\u5b58\u6a21\u578b ===\")\n\n # 3.1 \u521b\u5efa\u4fdd\u5b58\u76ee\u5f55\n os.makedirs('models', exist_ok=True)\n\n # 3.2 \u51c6\u5907\u4fdd\u5b58\u7684\u6570\u636e\n model_data = {\n 'model': self.model,\n 'label_encoder': self.label_encoder,\n 'scaler': self.scaler,\n 'feature_columns': list(self.X.columns),\n 'feature_importance': pd.DataFrame({\n '\u7279\u5f81': self.X.columns,\n '\u91cd\u8981\u6027': self.model.feature_importances_\n }).sort_values('\u91cd\u8981\u6027', ascending=False).to_dict('records'),\n 'class_descriptions': {\n 0: '\u5168\u90e8\u4eba\u7fa4\u9002\u5b9c\u5916\u51fa - \u7a7a\u6c14\u8d28\u91cf\u4f18\uff0c\u9002\u5408\u6240\u6709\u6237\u5916\u6d3b\u52a8\uff0c\u5305\u62ec\u8dd1\u6b65\u3001\u9a91\u884c\u7b49\u5267\u70c8\u8fd0\u52a8',\n 1: '\u654f\u611f\u4eba\u7fa4\u51cf\u5c11\u5916\u51fa - \u7a7a\u6c14\u8d28\u91cf\u826f\uff0c\u654f\u611f\u4eba\u7fa4\u5e94\u51cf\u5c11\u957f\u65f6\u95f4\u6237\u5916\u6d3b\u52a8',\n 2: '\u654f\u611f\u4eba\u7fa4\u907f\u514d\u5916\u51fa - \u8f7b\u5ea6\u6c61\u67d3\uff0c\u513f\u7ae5\u3001\u8001\u4eba\u3001\u547c\u5438\u9053\u75be\u75c5\u60a3\u8005\u907f\u514d\u6237\u5916\u6d3b\u52a8',\n 3: '\u6240\u6709\u4eba\u7fa4\u51cf\u5c11\u5916\u51fa - \u4e2d\u5ea6\u6c61\u67d3\uff0c\u5efa\u8bae\u51cf\u5c11\u6237\u5916\u6d3b\u52a8\uff0c\u5916\u51fa\u65f6\u4f69\u6234\u53e3\u7f69',\n 4: '\u513f\u7ae5\u8001\u4eba\u907f\u514d\u5916\u51fa - \u91cd\u5ea6\u6c61\u67d3\uff0c\u654f\u611f\u4eba\u7fa4\u5e94\u7559\u5728\u5ba4\u5185\uff0c\u5173\u95ed\u95e8\u7a97',\n 5: '\u5ba4\u5185\u6d3b\u52a8\u4e3a\u4f73 - \u4e25\u91cd\u6c61\u67d3\uff0c\u6240\u6709\u4eba\u7fa4\u907f\u514d\u6237\u5916\u6d3b\u52a8\uff0c\u4f7f\u7528\u7a7a\u6c14\u51c0\u5316\u5668'\n },\n 'training_info': {\n '\u6837\u672c\u6570\u91cf': len(self.df),\n '\u7279\u5f81\u6570\u91cf': len(self.X.columns),\n '\u8bad\u7ec3\u65f6\u95f4': pd.Timestamp.now().strftime('%Y-%m-%d %H:%M:%S'),\n '\u6a21\u578b\u7c7b\u578b': 'RandomForestClassifier',\n '\u6570\u636e\u6765\u6e90': self.data_path\n }\n }\n\n # 3.3 \u4fdd\u5b58\u6a21\u578b\n joblib.dump(model_data, model_path)\n print(f\"\u2713 \u6a21\u578b\u5df2\u4fdd\u5b58\u5230: {model_path}\")\n\n # 3.4 \u4fdd\u5b58\u7279\u5f81\u8303\u56f4\uff08\u7528\u4e8eGUI\u6ed1\u5757\u8bbe\u7f6e\uff09\n feature_ranges = {}\n for col in self.X.columns:\n feature_ranges[col] = {\n 'min': float(self.df[col].min()),\n 'max': float(self.df[col].max()),\n 'mean': float(self.df[col].mean()),\n 'std': float(self.df[col].std()),\n 'q1': float(self.df[col].quantile(0.25)),\n 'q3': float(self.df[col].quantile(0.75))\n }\n\n ranges_path = 'models\/feature_ranges.json'\n with open(ranges_path, 'w', encoding='utf-8') as f:\n json.dump(feature_ranges, f, ensure_ascii=False, indent=2)\n print(f\"\u2713 \u7279\u5f81\u8303\u56f4\u5df2\u4fdd\u5b58\u5230: {ranges_path}\")\n\n # 3.5 \u4fdd\u5b58\u6a21\u578b\u5143\u6570\u636e\n metadata = {\n 'model_info': {\n 'name': 'AirQualityCrowdAdviceModel',\n 'version': '1.0.0',\n 'description': '\u57fa\u4e8e\u7a7a\u6c14\u8d28\u91cf\u6570\u636e\u7684\u4eba\u7fa4\u5916\u51fa\u5efa\u8bae\u6a21\u578b',\n 'author': 'Air Quality System',\n 'created_date': pd.Timestamp.now().strftime('%Y-%m-%d')\n },\n 'features': list(self.X.columns),\n 'classes': list(self.label_encoder.classes_),\n 'feature_ranges': feature_ranges\n }\n\n metadata_path = 'models\/model_metadata.json'\n with open(metadata_path, 'w', encoding='utf-8') as f:\n json.dump(metadata, f, ensure_ascii=False, indent=2)\n print(f\"\u2713 \u6a21\u578b\u5143\u6570\u636e\u5df2\u4fdd\u5b58\u5230: {metadata_path}\")\n\n def run_training_pipeline(self):\n \"\"\"\u8fd0\u884c\u5b8c\u6574\u7684\u8bad\u7ec3\u6d41\u7a0b\"\"\"\n print(\"=\"*60)\n print(\"\u7a7a\u6c14\u8d28\u91cf\u5916\u51fa\u5efa\u8bae\u6a21\u578b\u8bad\u7ec3\u7cfb\u7edf\")\n print(\"=\"*60)\n\n try:\n # 1. \u52a0\u8f7d\u548c\u9884\u5904\u7406\u6570\u636e\n X, y, features = self.load_and_preprocess()\n\n # 2. \u8bad\u7ec3\u6a21\u578b\n accuracy, X_test, y_test, y_pred = self.train_model()\n\n # 3. \u4fdd\u5b58\u6a21\u578b\n self.save_model()\n\n print(\"\\n\" + \"=\"*60)\n print(\"\ud83c\udf89 \u6a21\u578b\u8bad\u7ec3\u5b8c\u6210\uff01\")\n print(f\"\ud83d\udcca \u6700\u7ec8\u51c6\u786e\u7387: {accuracy:.4f}\")\n print(f\"\ud83d\udd22 \u7279\u5f81\u6570\u91cf: {len(features)}\")\n print(f\"\ud83c\udff7\ufe0f \u7c7b\u522b\u6570\u91cf: {len(self.label_encoder.classes_)}\")\n print(\"=\"*60)\n print(\"\\n\ud83d\udcc1 \u751f\u6210\u7684\u6587\u4ef6:\")\n print(\" - models\/air_quality_model.pkl (\u4e3b\u6a21\u578b\u6587\u4ef6)\")\n print(\" - models\/feature_ranges.json (\u7279\u5f81\u8303\u56f4)\")\n print(\" - models\/model_metadata.json (\u6a21\u578b\u5143\u6570\u636e)\")\n print(\" - feature_importance.png (\u7279\u5f81\u91cd\u8981\u6027\u56fe)\")\n print(\" - confusion_matrix.png (\u6df7\u6dc6\u77e9\u9635)\")\n\n return True\n\n except Exception as e:\n print(f\"\\n\u274c \u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u51fa\u73b0\u9519\u8bef: {str(e)}\")\n import traceback\n traceback.print_exc()\n return False\n\nif __name__ == \"__main__\":\n # \u8bad\u7ec3\u6a21\u578b\n trainer = AirQualityModelTrainer(data_path='data\/cata_6716.csv')\n success = trainer.run_training_pipeline()\n\n if success:\n print(\"\\n\u2705 \u4e0b\u4e00\u6b65: \u8fd0\u884c 'python predict_app.py' \u542f\u52a8GUI\u5e94\u7528\")\n else:\n print(\"\\n\u274c \u8bad\u7ec3\u5931\u8d25\uff0c\u8bf7\u68c0\u67e5\u6570\u636e\u548c\u4ee3\u7801\")\n<\/code><\/pre>\n\n\n\npredict_app.py<\/h3>\n\n\n\n\"\"\"\n\u7a7a\u6c14\u8d28\u91cf\u5916\u51fa\u5efa\u8baeGUI\u5e94\u7528\n\u4f7f\u7528\u8bad\u7ec3\u597d\u7684\u6a21\u578b\u63d0\u4f9b\u5b9e\u65f6\u5efa\u8bae\n\"\"\"\n\nimport tkinter as tk\nfrom tkinter import ttk, messagebox, font\nimport joblib\nimport numpy as np\nimport json\nimport os\nfrom PIL import Image, ImageTk\nimport pandas as pd\n\nclass AirQualityApp:\n def __init__(self, root):\n \"\"\"\u521d\u59cb\u5316GUI\u5e94\u7528\"\"\"\n self.root = root\n self.root.title(\"\u7a7a\u6c14\u8d28\u91cf\u5916\u51fa\u5efa\u8bae\u7cfb\u7edf\")\n self.root.geometry(\"1000x750\")\n self.root.configure(bg='#f0f8ff')\n\n # \u8bbe\u7f6e\u56fe\u6807\n try:\n self.root.iconbitmap('icon.ico')\n except:\n pass\n\n # \u6a21\u578b\u6570\u636e\n self.model_data = None\n self.model = None\n self.label_encoder = None\n self.scaler = None\n self.feature_columns = None\n self.feature_ranges = None\n\n # \u52a0\u8f7d\u6a21\u578b\n if not self.load_model():\n messagebox.showerror(\"\u9519\u8bef\", \"\u65e0\u6cd5\u52a0\u8f7d\u6a21\u578b\uff0c\u8bf7\u5148\u8fd0\u884ctrain_model.py\u8bad\u7ec3\u6a21\u578b\")\n return\n\n # \u8bbe\u7f6e\u6837\u5f0f\n self.setup_styles()\n\n # \u521b\u5efa\u754c\u9762\n self.create_widgets()\n\n def setup_styles(self):\n \"\"\"\u8bbe\u7f6e\u754c\u9762\u6837\u5f0f\"\"\"\n style = ttk.Style()\n style.theme_use('clam')\n\n # \u81ea\u5b9a\u4e49\u6837\u5f0f\n style.configure('Title.TLabel',\n font=('\u5fae\u8f6f\u96c5\u9ed1', 22, 'bold'),\n background='#f0f8ff',\n foreground='#1e3d59')\n\n style.configure('Subtitle.TLabel',\n font=('\u5fae\u8f6f\u96c5\u9ed1', 12),\n background='#f0f8ff',\n foreground='#2c3e50')\n\n style.configure('Result.TLabel',\n font=('\u5fae\u8f6f\u96c5\u9ed1', 16, 'bold'),\n background='#f0f8ff',\n foreground='#2c3e50')\n\n style.configure('Advice.TLabel',\n font=('\u5fae\u8f6f\u96c5\u9ed1', 13),\n background='#f0f8ff',\n foreground='#34495e',\n wraplength=400)\n\n style.configure('Slider.TLabel',\n font=('\u5fae\u8f6f\u96c5\u9ed1', 10, 'bold'),\n background='#f0f8ff',\n foreground='#2c3e50')\n\n style.configure('Value.TLabel',\n font=('\u5fae\u8f6f\u96c5\u9ed1', 10),\n background='#f0f8ff',\n foreground='#1e3d59')\n\n style.configure('Accent.TButton',\n font=('\u5fae\u8f6f\u96c5\u9ed1', 12, 'bold'),\n background='#4a90e2',\n foreground='white')\n\n def load_model(self):\n \"\"\"\u52a0\u8f7d\u8bad\u7ec3\u597d\u7684\u6a21\u578b\"\"\"\n model_path = 'models\/air_quality_model.pkl'\n\n if not os.path.exists(model_path):\n return False\n\n try:\n self.model_data = joblib.load(model_path)\n self.model = self.model_data['model']\n self.label_encoder = self.model_data['label_encoder']\n self.scaler = self.model_data['scaler']\n self.feature_columns = self.model_data['feature_columns']\n\n # \u52a0\u8f7d\u7279\u5f81\u8303\u56f4\n ranges_path = 'models\/feature_ranges.json'\n if os.path.exists(ranges_path):\n with open(ranges_path, 'r', encoding='utf-8') as f:\n self.feature_ranges = json.load(f)\n\n print(f\"\u2713 \u6a21\u578b\u52a0\u8f7d\u6210\u529f!\")\n print(f\"\u2713 \u7279\u5f81: {self.feature_columns}\")\n print(f\"\u2713 \u7c7b\u522b: {list(self.label_encoder.classes_)}\")\n\n return True\n except Exception as e:\n print(f\"\u2717 \u52a0\u8f7d\u6a21\u578b\u5931\u8d25: {str(e)}\")\n return False\n\n def create_widgets(self):\n \"\"\"\u521b\u5efa\u754c\u9762\u63a7\u4ef6\"\"\"\n # \u4e3b\u5bb9\u5668\n main_container = ttk.Frame(self.root)\n main_container.pack(fill=tk.BOTH, expand=True, padx=20, pady=20)\n\n # \u6807\u9898\u533a\u57df\n title_frame = ttk.Frame(main_container, style='Title.TFrame')\n title_frame.pack(pady=(0, 20))\n\n title_label = ttk.Label(title_frame,\n text=\"\ud83c\udf24\ufe0f \u7a7a\u6c14\u8d28\u91cf\u5916\u51fa\u5efa\u8bae\u7cfb\u7edf\",\n style='Title.TLabel')\n title_label.pack()\n\n subtitle_label = ttk.Label(title_frame,\n text=\"\u8c03\u6574\u6c61\u67d3\u7269\u53c2\u6570\uff0c\u83b7\u53d6\u9488\u5bf9\u4e0d\u540c\u4eba\u7fa4\u7684\u5065\u5eb7\u5916\u51fa\u5efa\u8bae\",\n style='Subtitle.TLabel')\n subtitle_label.pack(pady=5)\n\n # \u4e3b\u5185\u5bb9\u533a\u57df\n content_frame = ttk.Frame(main_container)\n content_frame.pack(fill=tk.BOTH, expand=True)\n\n # \u5de6\u4fa7\u53c2\u6570\u63a7\u5236\u533a\u57df\n control_frame = ttk.LabelFrame(content_frame, text=\"\u7a7a\u6c14\u8d28\u91cf\u53c2\u6570\u8bbe\u7f6e\", padding=20)\n control_frame.pack(side=tk.LEFT, fill=tk.BOTH, expand=True, padx=(0, 10))\n\n # \u5b9a\u4e49\u53c2\u6570\u914d\u7f6e\n self.param_configs = {\n '\u7ec6\u9897\u7c92\u7269': {\n 'unit': '\u03bcg\/m\u00b3',\n 'desc': 'PM2.5 - \u76f4\u5f84\u22642.5\u03bcm\u7684\u9897\u7c92\u7269',\n 'standard': '\u56fd\u5bb6\u6807\u51c6: \u226435\u03bcg\/m\u00b3',\n 'color': '#ff6b6b'\n },\n '\u53ef\u5438\u5165\u9897\u7c92\u7269': {\n 'unit': '\u03bcg\/m\u00b3',\n 'desc': 'PM10 - \u76f4\u5f84\u226410\u03bcm\u7684\u9897\u7c92\u7269',\n 'standard': '\u56fd\u5bb6\u6807\u51c6: \u226450\u03bcg\/m\u00b3',\n 'color': '#4ecdc4'\n },\n '\u4e8c\u6c27\u5316\u786b': {\n 'unit': '\u03bcg\/m\u00b3',\n 'desc': 'SO\u2082 - \u4e3b\u8981\u6765\u81ea\u71c3\u7164',\n 'standard': '\u56fd\u5bb6\u6807\u51c6: \u226460\u03bcg\/m\u00b3',\n 'color': '#45b7d1'\n },\n '\u4e8c\u6c27\u5316\u6c2e': {\n 'unit': '\u03bcg\/m\u00b3',\n 'desc': 'NO\u2082 - \u4e3b\u8981\u6765\u81ea\u673a\u52a8\u8f66\u5c3e\u6c14',\n 'standard': '\u56fd\u5bb6\u6807\u51c6: \u226440\u03bcg\/m\u00b3',\n 'color': '#96ceb4'\n },\n '\u4e00\u6c27\u5316\u78b3': {\n 'unit': 'mg\/m\u00b3',\n 'desc': 'CO - \u65e0\u8272\u65e0\u5473\u6709\u6bd2\u6c14\u4f53',\n 'standard': '\u56fd\u5bb6\u6807\u51c6: \u22644mg\/m\u00b3',\n 'color': '#feca57'\n },\n '\u81ed\u6c271\u5c0f\u65f6\u5e73\u5747': {\n 'unit': '\u03bcg\/m\u00b3',\n 'desc': 'O\u2083 - \u5149\u5316\u5b66\u70df\u96fe\u4e3b\u8981\u6210\u5206',\n 'standard': '\u56fd\u5bb6\u6807\u51c6: \u2264160\u03bcg\/m\u00b3',\n 'color': '#ff9ff3'\n }\n }\n\n # \u5b58\u50a8\u6ed1\u5757\u548c\u53d8\u91cf\n self.sliders = {}\n self.slider_vars = {}\n\n # \u521b\u5efa\u6ed1\u5757\n for i, param in enumerate(self.feature_columns):\n if param not in self.param_configs:\n continue\n\n config = self.param_configs[param]\n\n # \u83b7\u53d6\u8303\u56f4\n if self.feature_ranges and param in self.feature_ranges:\n min_val = int(self.feature_ranges[param]['min'])\n max_val = int(self.feature_ranges[param]['max'] * 1.2) # \u7559\u6709\u4f59\u91cf\n default_val = int(self.feature_ranges[param]['mean'])\n else:\n # \u9ed8\u8ba4\u8303\u56f4\n ranges = {\n '\u7ec6\u9897\u7c92\u7269': (0, 500, 50),\n '\u53ef\u5438\u5165\u9897\u7c92\u7269': (0, 600, 70),\n '\u4e8c\u6c27\u5316\u786b': (0, 200, 20),\n '\u4e8c\u6c27\u5316\u6c2e': (0, 300, 30),\n '\u4e00\u6c27\u5316\u78b3': (0, 10, 1),\n '\u81ed\u6c271\u5c0f\u65f6\u5e73\u5747': (0, 500, 50)\n }\n min_val, max_val, default_val = ranges.get(param, (0, 100, 50))\n\n # \u53c2\u6570\u6846\u67b6\n param_frame = ttk.Frame(control_frame)\n param_frame.pack(fill=tk.X, pady=(0, 15))\n\n # \u53c2\u6570\u6807\u7b7e\n label_frame = ttk.Frame(param_frame)\n label_frame.pack(fill=tk.X, pady=(0, 5))\n\n ttk.Label(label_frame,\n text=f\"{param} ({config['unit']})\",\n style='Slider.TLabel').pack(side=tk.LEFT)\n\n ttk.Label(label_frame,\n text=config['desc'],\n font=('\u5fae\u8f6f\u96c5\u9ed1', 8),\n foreground='#7f8c8d').pack(side=tk.LEFT, padx=(10, 0))\n\n # \u6ed1\u5757\u548c\u6570\u503c\u663e\u793a\u6846\u67b6\n slider_frame = ttk.Frame(param_frame)\n slider_frame.pack(fill=tk.X)\n\n # \u6570\u503c\u663e\u793a\n var = tk.DoubleVar(value=default_val)\n value_label = ttk.Label(slider_frame,\n textvariable=var,\n style='Value.TLabel',\n width=8)\n value_label.pack(side=tk.RIGHT, padx=(10, 0))\n\n # \u5355\u4f4d\u6807\u7b7e\n ttk.Label(slider_frame,\n text=config['unit'],\n font=('\u5fae\u8f6f\u96c5\u9ed1', 9),\n foreground='gray').pack(side=tk.RIGHT)\n\n # \u6ed1\u5757\n slider = ttk.Scale(slider_frame,\n from_=min_val,\n to=max_val,\n orient=tk.HORIZONTAL,\n length=350,\n variable=var,\n command=lambda v, p=param: self.on_slider_change(p, v))\n slider.pack(side=tk.LEFT, fill=tk.X, expand=True)\n\n # \u56fd\u5bb6\u6807\u51c6\u663e\u793a\n std_label = ttk.Label(param_frame,\n text=f\"\ud83d\udcca {config['standard']}\",\n font=('\u5fae\u8f6f\u96c5\u9ed1', 8),\n foreground='#27ae60')\n std_label.pack(anchor='w', pady=(2, 0))\n\n # \u5b58\u50a8\u5f15\u7528\n self.sliders[param] = slider\n self.slider_vars[param] = var\n\n # \u63a7\u5236\u6309\u94ae\u533a\u57df\n button_frame = ttk.Frame(control_frame)\n button_frame.pack(fill=tk.X, pady=(20, 0))\n\n # \u9884\u6d4b\u6309\u94ae\n predict_btn = ttk.Button(button_frame,\n text=\"\ud83d\ude80 \u83b7\u53d6\u5916\u51fa\u5efa\u8bae\",\n style='Accent.TButton',\n command=self.predict)\n predict_btn.pack(side=tk.LEFT, expand=True, fill=tk.X, padx=(0, 10))\n\n # \u91cd\u7f6e\u6309\u94ae\n reset_btn = ttk.Button(button_frame,\n text=\"\ud83d\udd04 \u91cd\u7f6e\u53c2\u6570\",\n command=self.reset_sliders)\n reset_btn.pack(side=tk.LEFT, expand=True, fill=tk.X)\n\n # \u53f3\u4fa7\u7ed3\u679c\u663e\u793a\u533a\u57df\n result_frame = ttk.LabelFrame(content_frame, text=\"\u5916\u51fa\u5efa\u8bae\", padding=20)\n result_frame.pack(side=tk.RIGHT, fill=tk.BOTH, expand=True, padx=(10, 0))\n\n # \u7ed3\u679c\u5c55\u793a\u533a\u57df\n self.result_text = tk.Text(result_frame,\n height=12,\n width=40,\n font=('\u5fae\u8f6f\u96c5\u9ed1', 12),\n wrap=tk.WORD,\n bg='#f8f9fa',\n fg='#2c3e50',\n relief=tk.FLAT,\n borderwidth=2)\n self.result_text.pack(fill=tk.BOTH, expand=True, pady=(0, 15))\n\n # \u8bbe\u7f6e\u521d\u59cb\u63d0\u793a\u6587\u672c\n initial_text = \"\"\"\ud83d\udd0d \u8bf7\u8c03\u6574\u5de6\u4fa7\u7684\u6c61\u67d3\u7269\u53c2\u6570\u6ed1\u5757\uff0c\u7136\u540e\u70b9\u51fb\"\u83b7\u53d6\u5916\u51fa\u5efa\u8bae\"\u6309\u94ae\u3002\n\n\u7cfb\u7edf\u5c06\u6839\u636e\u5f53\u524d\u7684\u7a7a\u6c14\u8d28\u91cf\u53c2\u6570\uff0c\u4e3a\u60a8\u63d0\u4f9b\u9488\u5bf9\u4e0d\u540c\u4eba\u7fa4\u7684\u5065\u5eb7\u5916\u51fa\u5efa\u8bae\u3002\n\n\ud83d\udcca \u53c2\u6570\u8bf4\u660e\uff1a\n\u2022 \u7ec6\u9897\u7c92\u7269(PM2.5): \u53ef\u5165\u80ba\u9897\u7c92\u7269\uff0c\u5065\u5eb7\u5371\u5bb3\u5927\n\u2022 \u53ef\u5438\u5165\u9897\u7c92\u7269(PM10): \u53ef\u8fdb\u5165\u547c\u5438\u9053\u7684\u9897\u7c92\u7269\n\u2022 \u4e8c\u6c27\u5316\u786b(SO\u2082): \u523a\u6fc0\u6027\u6c14\u4f53\uff0c\u4e3b\u8981\u6765\u81ea\u71c3\u7164\n\u2022 \u4e8c\u6c27\u5316\u6c2e(NO\u2082): \u523a\u6fc0\u6027\u6c14\u4f53\uff0c\u4e3b\u8981\u6765\u81ea\u673a\u52a8\u8f66\n\u2022 \u4e00\u6c27\u5316\u78b3(CO): \u65e0\u8272\u65e0\u5473\u6709\u6bd2\u6c14\u4f53\n\u2022 \u81ed\u6c27(O\u2083): \u5149\u5316\u5b66\u70df\u96fe\u4e3b\u8981\u6210\u5206\n\n\ud83d\udca1 \u63d0\u793a\uff1a\u7eff\u8272\u6807\u6ce8\u4e3a\u56fd\u5bb6\u6807\u51c6\u9650\u503c\u3002\"\"\"\n\n self.result_text.insert(1.0, initial_text)\n self.result_text.config(state=tk.DISABLED)\n\n # \u8be6\u7ec6\u5efa\u8bae\u6846\u67b6\n detail_frame = ttk.LabelFrame(result_frame, text=\"\u8be6\u7ec6\u5065\u5eb7\u5efa\u8bae\", padding=15)\n detail_frame.pack(fill=tk.BOTH, expand=True)\n\n self.detail_text = tk.Text(detail_frame,\n height=8,\n font=('\u5fae\u8f6f\u96c5\u9ed1', 11),\n wrap=tk.WORD,\n bg='#f8f9fa',\n fg='#34495e',\n relief=tk.FLAT)\n self.detail_text.pack(fill=tk.BOTH, expand=True)\n\n # \u8bbe\u7f6e\u521d\u59cb\u8be6\u7ec6\u5efa\u8bae\n detail_initial = \"\"\"\ud83d\udc65 \u9002\u7528\u4eba\u7fa4\u5206\u7c7b\uff1a\n\u2022 \u5168\u90e8\u4eba\u7fa4\uff1a\u5065\u5eb7\u6210\u5e74\u4eba\n\u2022 \u654f\u611f\u4eba\u7fa4\uff1a\u513f\u7ae5\u3001\u8001\u4eba\u3001\u5b55\u5987\u3001\u547c\u5438\u9053\u75be\u75c5\u60a3\u8005\n\u2022 \u7279\u6b8a\u4eba\u7fa4\uff1a\u5fc3\u810f\u75c5\u3001\u54ee\u5598\u7b49\u6162\u6027\u75c5\u60a3\u8005\n\n\ud83d\udee1\ufe0f \u9632\u62a4\u5efa\u8bae\u7b49\u7ea7\uff1a\n\u2022 \u2705 \u9002\u5b9c\u5916\u51fa\uff1a\u65e0\u9700\u7279\u6b8a\u9632\u62a4\n\u2022 \u26a0\ufe0f \u51cf\u5c11\u5916\u51fa\uff1a\u51cf\u5c11\u6237\u5916\u65f6\u95f4\n\u2022 \ud83d\udeab \u907f\u514d\u5916\u51fa\uff1a\u5c3d\u91cf\u7559\u5728\u5ba4\u5185\n\u2022 \ud83c\udfe0 \u5ba4\u5185\u6d3b\u52a8\uff1a\u5173\u95ed\u95e8\u7a97\uff0c\u4f7f\u7528\u7a7a\u6c14\u51c0\u5316\u5668\"\"\"\n\n self.detail_text.insert(1.0, detail_initial)\n self.detail_text.config(state=tk.DISABLED)\n\n # \u72b6\u6001\u680f\n status_frame = ttk.Frame(main_container, relief=tk.SUNKEN, borderwidth=1)\n status_frame.pack(fill=tk.X, pady=(10, 0))\n\n self.status_label = ttk.Label(status_frame,\n text=\"\u5c31\u7eea | \u6a21\u578b\u5df2\u52a0\u8f7d | \u7b49\u5f85\u8f93\u5165\u53c2\u6570\",\n font=('\u5fae\u8f6f\u96c5\u9ed1', 9),\n foreground='#7f8c8d')\n self.status_label.pack(side=tk.LEFT, padx=5)\n\n def on_slider_change(self, param, value):\n \"\"\"\u6ed1\u5757\u53d8\u5316\u56de\u8c03\"\"\"\n # \u66f4\u65b0\u72b6\u6001\u680f\n self.status_label.config(text=f\"\u8c03\u6574: {param} = {float(value):.1f}\")\n\n def reset_sliders(self):\n \"\"\"\u91cd\u7f6e\u6240\u6709\u6ed1\u5757\u5230\u9ed8\u8ba4\u503c\"\"\"\n for param, slider in self.sliders.items():\n if self.feature_ranges and param in self.feature_ranges:\n default_val = self.feature_ranges[param]['mean']\n else:\n default_val = 50\n self.slider_vars[param].set(default_val)\n\n # \u6e05\u7a7a\u7ed3\u679c\u663e\u793a\n self.result_text.config(state=tk.NORMAL)\n self.result_text.delete(1.0, tk.END)\n self.result_text.insert(1.0, \"\u53c2\u6570\u5df2\u91cd\u7f6e\uff0c\u8bf7\u91cd\u65b0\u83b7\u53d6\u5efa\u8bae\")\n self.result_text.config(state=tk.DISABLED)\n\n self.detail_text.config(state=tk.NORMAL)\n self.detail_text.delete(1.0, tk.END)\n self.detail_text.insert(1.0, \"\u7b49\u5f85\u65b0\u7684\u9884\u6d4b\u7ed3\u679c...\")\n self.detail_text.config(state=tk.DISABLED)\n\n self.status_label.config(text=\"\u53c2\u6570\u5df2\u91cd\u7f6e | \u7b49\u5f85\u8f93\u5165\")\n\n def predict(self):\n \"\"\"\u8fdb\u884c\u9884\u6d4b\u5e76\u663e\u793a\u7ed3\u679c\"\"\"\n if self.model is None:\n messagebox.showerror(\"\u9519\u8bef\", \"\u6a21\u578b\u672a\u52a0\u8f7d\uff0c\u65e0\u6cd5\u8fdb\u884c\u9884\u6d4b\")\n return\n\n try:\n # 1. \u83b7\u53d6\u5f53\u524d\u53c2\u6570\u503c\n features = []\n for param in self.feature_columns:\n value = self.slider_vars[param].get()\n features.append(value)\n\n # 2. \u8f6c\u6362\u4e3anumpy\u6570\u7ec4\u5e76\u6807\u51c6\u5316\n X = np.array([features], dtype=np.float32)\n X_scaled = self.scaler.transform(X)\n\n # 3. \u8fdb\u884c\u9884\u6d4b\n prediction = self.model.predict(X_scaled)[0]\n probability = self.model.predict_proba(X_scaled)[0]\n\n # 4. \u89e3\u7801\u9884\u6d4b\u7ed3\u679c\n label = self.label_encoder.inverse_transform([prediction])[0]\n\n # 5. \u83b7\u53d6\u9884\u6d4b\u6982\u7387\n prob_dict = {}\n for i, cls in enumerate(self.label_encoder.classes_):\n prob_dict[cls] = probability[i] * 100\n\n # 6. \u663e\u793a\u7ed3\u679c\n self.display_result(label, prob_dict, features)\n\n # 7. \u66f4\u65b0\u72b6\u6001\n self.status_label.config(text=f\"\u9884\u6d4b\u5b8c\u6210: {label} | \u7f6e\u4fe1\u5ea6: {probability[prediction]*100:.1f}%\")\n\n except Exception as e:\n messagebox.showerror(\"\u9884\u6d4b\u9519\u8bef\", f\"\u9884\u6d4b\u8fc7\u7a0b\u4e2d\u51fa\u73b0\u9519\u8bef:\\n{str(e)}\")\n self.status_label.config(text=\"\u9884\u6d4b\u5931\u8d25\")\n\n def display_result(self, label, probabilities, features):\n \"\"\"\u663e\u793a\u9884\u6d4b\u7ed3\u679c\"\"\"\n # \u7ed3\u679c\u989c\u8272\u6620\u5c04\n color_map = {\n '\u5168\u90e8\u4eba\u7fa4\u9002\u5b9c\u5916\u51fa': '#27ae60', # \u7eff\u8272\n '\u654f\u611f\u4eba\u7fa4\u51cf\u5c11\u5916\u51fa': '#f39c12', # \u6a59\u8272\n '\u654f\u611f\u4eba\u7fa4\u907f\u514d\u5916\u51fa': '#e74c3c', # \u7ea2\u8272\n '\u6240\u6709\u4eba\u7fa4\u51cf\u5c11\u5916\u51fa': '#c0392b', # \u6df1\u7ea2\n '\u513f\u7ae5\u8001\u4eba\u907f\u514d\u5916\u51fa': '#8e44ad', # \u7d2b\u8272\n '\u5ba4\u5185\u6d3b\u52a8\u4e3a\u4f73': '#2c3e50' # \u6df1\u7070\n }\n\n # \u56fe\u6807\u6620\u5c04\n icon_map = {\n '\u5168\u90e8\u4eba\u7fa4\u9002\u5b9c\u5916\u51fa': '\u2705',\n '\u654f\u611f\u4eba\u7fa4\u51cf\u5c11\u5916\u51fa': '\u26a0\ufe0f',\n '\u654f\u611f\u4eba\u7fa4\u907f\u514d\u5916\u51fa': '\ud83d\udeab',\n '\u6240\u6709\u4eba\u7fa4\u51cf\u5c11\u5916\u51fa': '\ud83d\udd34',\n '\u513f\u7ae5\u8001\u4eba\u907f\u514d\u5916\u51fa': '\ud83d\udfe3',\n '\u5ba4\u5185\u6d3b\u52a8\u4e3a\u4f73': '\ud83c\udfe0'\n }\n\n # \u8be6\u7ec6\u5efa\u8bae\u6620\u5c04\n advice_map = {\n '\u5168\u90e8\u4eba\u7fa4\u9002\u5b9c\u5916\u51fa': \"\"\"\u2705 \u5065\u5eb7\u5efa\u8bae\uff1a\n\u2022 \u6240\u6709\u4eba\u7fa4\u53ef\u6b63\u5e38\u8fdb\u884c\u6237\u5916\u6d3b\u52a8\n\u2022 \u9002\u5408\u8dd1\u6b65\u3001\u9a91\u884c\u3001\u7403\u7c7b\u7b49\u5404\u7c7b\u8fd0\u52a8\n\u2022 \u65e0\u9700\u7279\u6b8a\u9632\u62a4\u63aa\u65bd\n\u2022 \u5efa\u8bae\u5f00\u7a97\u901a\u98ce\uff0c\u4fdd\u6301\u5ba4\u5185\u7a7a\u6c14\u6d41\u901a\n\n\ud83c\udfaf \u9002\u7528\u4eba\u7fa4\uff1a\n\u2022 \u5065\u5eb7\u6210\u5e74\u4eba\n\u2022 \u513f\u7ae5\u548c\u8001\u4eba\n\u2022 \u5b55\u5987\n\u2022 \u547c\u5438\u9053\u75be\u75c5\u60a3\u8005\"\"\",\n\n '\u654f\u611f\u4eba\u7fa4\u51cf\u5c11\u5916\u51fa': \"\"\"\u26a0\ufe0f \u5065\u5eb7\u5efa\u8bae\uff1a\n\u2022 \u654f\u611f\u4eba\u7fa4\u5e94\u51cf\u5c11\u957f\u65f6\u95f4\u3001\u9ad8\u5f3a\u5ea6\u7684\u6237\u5916\u6d3b\u52a8\n\u2022 \u5065\u5eb7\u4eba\u7fa4\u53ef\u6b63\u5e38\u6d3b\u52a8\uff0c\u4f46\u907f\u514d\u5267\u70c8\u8fd0\u52a8\n\u2022 \u5916\u51fa\u65f6\u53ef\u8003\u8651\u4f69\u6234\u53e3\u7f69\n\u2022 \u5efa\u8bae\u51cf\u5c11\u5f00\u7a97\u65f6\u95f4\n\n\ud83c\udfaf \u654f\u611f\u4eba\u7fa4\u5305\u62ec\uff1a\n\u2022 \u513f\u7ae5\u3001\u8001\u4eba\u3001\u5b55\u5987\n\u2022 \u54ee\u5598\u3001\u652f\u6c14\u7ba1\u708e\u60a3\u8005\n\u2022 \u5fc3\u810f\u75c5\u3001\u9ad8\u8840\u538b\u60a3\u8005\n\u2022 \u8fc7\u654f\u4f53\u8d28\u8005\"\"\",\n\n '\u654f\u611f\u4eba\u7fa4\u907f\u514d\u5916\u51fa': \"\"\"\ud83d\udeab \u5065\u5eb7\u5efa\u8bae\uff1a\n\u2022 \u654f\u611f\u4eba\u7fa4\u5e94\u5c3d\u91cf\u907f\u514d\u6237\u5916\u6d3b\u52a8\n\u2022 \u5065\u5eb7\u4eba\u7fa4\u51cf\u5c11\u6237\u5916\u6d3b\u52a8\u65f6\u95f4\n\u2022 \u5fc5\u987b\u5916\u51fa\u65f6\u8bf7\u4f69\u6234\u9632\u62a4\u53e3\u7f69\n\u2022 \u5173\u95ed\u95e8\u7a97\uff0c\u4f7f\u7528\u7a7a\u6c14\u51c0\u5316\u5668\n\u2022 \u907f\u514d\u6668\u7ec3\u7b49\u6237\u5916\u8fd0\u52a8\n\n\ud83c\udfe5 \u91cd\u70b9\u5173\u6ce8\uff1a\n\u2022 \u513f\u7ae5\u548c\u8001\u4eba\u5c3d\u91cf\u7559\u5728\u5ba4\u5185\n\u2022 \u547c\u5438\u9053\u75be\u75c5\u60a3\u8005\u907f\u514d\u5916\u51fa\n\u2022 \u5b66\u6821\u51cf\u5c11\u6237\u5916\u4f53\u80b2\u8bfe\"\"\",\n\n '\u6240\u6709\u4eba\u7fa4\u51cf\u5c11\u5916\u51fa': \"\"\"\ud83d\udd34 \u5065\u5eb7\u5efa\u8bae\uff1a\n\u2022 \u6240\u6709\u4eba\u7fa4\u5747\u5e94\u51cf\u5c11\u6237\u5916\u6d3b\u52a8\n\u2022 \u907f\u514d\u957f\u65f6\u95f4\u505c\u7559\u5728\u5ba4\u5916\n\u2022 \u5916\u51fa\u5fc5\u987b\u4f69\u6234\u6709\u6548\u9632\u62a4\u53e3\u7f69\n\u2022 \u5173\u95ed\u95e8\u7a97\uff0c\u5f00\u542f\u7a7a\u6c14\u51c0\u5316\n\u2022 \u53d6\u6d88\u6237\u5916\u6d3b\u52a8\u548c\u96c6\u4f1a\n\n\ud83d\udee1\ufe0f \u9632\u62a4\u63aa\u65bd\uff1a\n\u2022 N95\/KN95\u7ea7\u522b\u53e3\u7f69\n\u2022 \u5c3d\u91cf\u51cf\u5c11\u6237\u5916\u505c\u7559\u65f6\u95f4\n\u2022 \u56de\u5bb6\u540e\u53ca\u65f6\u6e05\u6d17\u9762\u90e8\u548c\u9f3b\u8154\"\"\",\n\n '\u513f\u7ae5\u8001\u4eba\u907f\u514d\u5916\u51fa': \"\"\"\ud83d\udfe3 \u5065\u5eb7\u5efa\u8bae\uff1a\n\u2022 \u513f\u7ae5\u3001\u8001\u4eba\u3001\u5b55\u5987\u52a1\u5fc5\u7559\u5728\u5ba4\u5185\n\u2022 \u5065\u5eb7\u6210\u4eba\u5c3d\u91cf\u51cf\u5c11\u4e0d\u5fc5\u8981\u5916\u51fa\n\u2022 \u5fc5\u987b\u5916\u51fa\u65f6\u505a\u597d\u5168\u9762\u9632\u62a4\n\u2022 \u5b66\u6821\u5e94\u505c\u6b62\u6240\u6709\u6237\u5916\u6d3b\u52a8\n\u2022 \u517b\u8001\u9662\u3001\u5e7c\u513f\u56ed\u52a0\u5f3a\u9632\u62a4\n\n\ud83d\udc76 \u7279\u6b8a\u4fdd\u62a4\uff1a\n\u2022 \u513f\u7ae5\u514d\u75ab\u529b\u8f83\u4f4e\uff0c\u6613\u53d7\u5f71\u54cd\n\u2022 \u8001\u4eba\u547c\u5438\u7cfb\u7edf\u8106\u5f31\uff0c\u98ce\u9669\u9ad8\n\u2022 \u6162\u6027\u75c5\u60a3\u8005\u9700\u683c\u5916\u6ce8\u610f\"\"\",\n\n '\u5ba4\u5185\u6d3b\u52a8\u4e3a\u4f73': \"\"\"\ud83c\udfe0 \u5065\u5eb7\u5efa\u8bae\uff1a\n\u2022 \u6240\u6709\u4eba\u7fa4\u5e94\u907f\u514d\u6237\u5916\u6d3b\u52a8\n\u2022 \u5c3d\u91cf\u7559\u5728\u5ba4\u5185\uff0c\u5173\u95ed\u95e8\u7a97\n\u2022 \u5fc5\u987b\u5916\u51fa\u65f6\u505a\u597d\u6700\u9ad8\u7ea7\u522b\u9632\u62a4\n\u2022 \u4f7f\u7528\u9ad8\u6548\u7a7a\u6c14\u51c0\u5316\u5668\n\u2022 \u907f\u514d\u6240\u6709\u6237\u5916\u8fd0\u52a8\u548c\u6d3b\u52a8\n\n\ud83d\udea8 \u7d27\u6025\u63aa\u65bd\uff1a\n\u2022 \u5c3d\u91cf\u51cf\u5c11\u5916\u51fa\uff0c\u8fdc\u7a0b\u529e\u516c\/\u5b66\u4e60\n\u2022 \u5916\u51fa\u4f69\u6234\u4e13\u4e1a\u9632\u62a4\u8bbe\u5907\n\u2022 \u5bc6\u5207\u5173\u6ce8\u5065\u5eb7\u72b6\u51b5\n\u2022 \u5982\u6709\u4e0d\u9002\u53ca\u65f6\u5c31\u533b\"\"\"\n }\n\n # \u83b7\u53d6\u989c\u8272\u548c\u56fe\u6807\n color = color_map.get(label, '#2c3e50')\n icon = icon_map.get(label, '\ud83d\udcca')\n\n # \u6784\u5efa\u7ed3\u679c\u663e\u793a\u6587\u672c\n result_content = f\"\"\"{icon} \u3010\u5916\u51fa\u5efa\u8bae\u3011{icon}\n{label}\n\n\ud83d\udcc8 \u9884\u6d4b\u7f6e\u4fe1\u5ea6: {probabilities[label]:.1f}%\n\n\ud83d\udcca \u6240\u6709\u53ef\u80fd\u6027:\n\"\"\"\n\n # \u6309\u6982\u7387\u6392\u5e8f\n sorted_probs = sorted(probabilities.items(), key=lambda x: x[1], reverse=True)\n for cls, prob in sorted_probs:\n if prob > 1: # \u53ea\u663e\u793a\u6982\u7387\u5927\u4e8e1%\u7684\u7c7b\u522b\n result_content += f\" \u2022 {cls}: {prob:.1f}%\\n\"\n\n result_content += f\"\\n\ud83d\udd22 \u5f53\u524d\u53c2\u6570\u503c:\\n\"\n for param, value in zip(self.feature_columns, features):\n result_content += f\" \u2022 {param}: {value:.1f}\\n\"\n\n # \u66f4\u65b0\u7ed3\u679c\u6587\u672c\u6846\n self.result_text.config(state=tk.NORMAL)\n self.result_text.delete(1.0, tk.END)\n self.result_text.insert(1.0, result_content)\n\n # \u8bbe\u7f6e\u6807\u9898\u989c\u8272\n self.result_text.tag_add(\"title\", \"1.0\", \"3.0\")\n self.result_text.tag_config(\"title\", foreground=color, font=('\u5fae\u8f6f\u96c5\u9ed1', 14, 'bold'))\n\n self.result_text.config(state=tk.DISABLED)\n\n # \u66f4\u65b0\u8be6\u7ec6\u5efa\u8bae\n self.detail_text.config(state=tk.NORMAL)\n self.detail_text.delete(1.0, tk.END)\n self.detail_text.insert(1.0, advice_map.get(label, \"\u6682\u65e0\u8be6\u7ec6\u5efa\u8bae\"))\n self.detail_text.config(state=tk.DISABLED)\n\ndef main():\n \"\"\"\u4e3b\u51fd\u6570\"\"\"\n root = tk.Tk()\n app = AirQualityApp(root)\n\n # \u8bbe\u7f6e\u7a97\u53e3\u5c45\u4e2d\n root.update_idletasks()\n width = root.winfo_width()\n height = root.winfo_height()\n x = (root.winfo_screenwidth() \/\/ 2) - (width \/\/ 2)\n y = (root.winfo_screenheight() \/\/ 2) - (height \/\/ 2)\n root.geometry(f'{width}x{height}+{x}+{y}')\n\n # \u8bbe\u7f6e\u6700\u5c0f\u7a97\u53e3\u5927\u5c0f\n root.minsize(900, 600)\n\n root.mainloop()\n\nif __name__ == \"__main__\":\n main()\n<\/code><\/pre>\n\n\n\n\u6b64\u5904\u4e3aReadme\u5c55\u793a\uff08\u7531AI\u751f\u6210\uff09\uff1a<\/p>\n\n\n\n
\u7a7a\u6c14\u8d28\u91cf\u5916\u51fa\u5efa\u8bae\u7cfb\u7edf\uff08readme.md\u59cb\uff09<\/h1>\n\n\n\n
\u57fa\u4e8e\u673a\u5668\u5b66\u4e60\u7684\u7a7a\u6c14\u8d28\u91cf\u5206\u6790\u4e0e\u4eba\u7fa4\u5065\u5eb7\u5efa\u8bae\u7cfb\u7edf\uff0c\u53ef\u6839\u636e\u6c61\u67d3\u7269\u6d53\u5ea6\u9884\u6d4b\u9002\u5b9c\u5916\u51fa\u4eba\u7fa4\u3002<\/p>\n\n\n\n
\u529f\u80fd\u7279\u70b9<\/h2>\n\n\n\n\n- \u6570\u636e\u5206\u6790<\/strong>\uff1a\u5bf9\u5386\u53f2\u7a7a\u6c14\u8d28\u91cf\u6570\u636e\u8fdb\u884c\u6df1\u5165\u5206\u6790<\/li>\n\n\n\n
- \u6a21\u578b\u8bad\u7ec3<\/strong>\uff1a\u4f7f\u7528\u968f\u673a\u68ee\u6797\u7b97\u6cd5\u8bad\u7ec3\u9884\u6d4b\u6a21\u578b<\/li>\n\n\n\n
- \u5b9e\u65f6\u9884\u6d4b<\/strong>\uff1a\u901a\u8fc7GUI\u754c\u9762\u5b9e\u65f6\u8c03\u6574\u53c2\u6570\u83b7\u53d6\u5efa\u8bae<\/li>\n\n\n\n
- \u5065\u5eb7\u5efa\u8bae<\/strong>\uff1a\u9488\u5bf9\u4e0d\u540c\u4eba\u7fa4\u63d0\u4f9b\u8be6\u7ec6\u5916\u51fa\u5efa\u8bae<\/li>\n\n\n\n
- \u53ef\u89c6\u5316<\/strong>\uff1a\u751f\u6210\u7279\u5f81\u91cd\u8981\u6027\u3001\u6df7\u6dc6\u77e9\u9635\u7b49\u5206\u6790\u56fe\u8868<\/li>\n<\/ol>\n\n\n\n
\u5feb\u901f\u5f00\u59cb<\/h2>\n\n\n\n1. \u73af\u5883\u914d\u7f6e<\/h3>\n\n\n\n
bash
\u514b\u9686\u6216\u4e0b\u8f7d\u9879\u76ee
git clone
cd air_quality_system
\u521b\u5efa\u865a\u62df\u73af\u5883\uff08\u63a8\u8350\uff09
python -m venv venv
\u6fc0\u6d3b\u865a\u62df\u73af\u5883
Windows:
venv\\Scripts\\activate
Linux\/Mac:
source venv\/bin\/activate
\u5b89\u88c5\u4f9d\u8d56
pip install -r requirements.txt<\/p>\n\n\n\n
2. \u51c6\u5907\u6570\u636e<\/h3>\n\n\n\n\n- \u5c06\u60a8\u7684\u7a7a\u6c14\u8d28\u91cf\u6570\u636eCSV\u6587\u4ef6\uff08\u5982
cata_6716.csv<\/code>\uff09\u653e\u5165data\/<\/code>\u76ee\u5f55<\/li>\n\n\n\n- \u786e\u4fdd\u6570\u636e\u5305\u542b\u4ee5\u4e0b\u5b57\u6bb5\uff08\u6216\u7c7b\u4f3c\u5b57\u6bb5\uff09\uff1a<\/li>\n<\/ol>\n\n\n\n
![\"\"](\"https:\/\/cn-sy1.rains3.com\/ntblog\/2026\/03\/image-1-1024x810.png\")
<\/div><\/figure>\n\n\n\n![\"\"](\"https:\/\/cn-sy1.rains3.com\/ntblog\/2026\/03\/image-2-1024x810.png\")
<\/div><\/figure>\n\n\n\n![\"\"](\"https:\/\/cn-sy1.rains3.com\/ntblog\/2026\/03\/image-3-1024x810.png\")
<\/div><\/figure>\n\n\n\n