fix: step13 auto-detect report mode on panel load (in _auto_pull_work_dir), not just on generate click

旧: total_steps = 4 + 63 + 20 = 87 (但大部分步骤不发脉冲)
新: total_steps = 4 + 63//10 + 20 = 30 (与 _next 调用次数一致)

src/gui/panels/step13_report_panel.py

@@ -14,7 +14,7 @@ from src.gui.panels._step_path_resolver import resolve_subdir
 from PyQt5.QtCore import Qt, QThread, pyqtSignal, QSettings
 from PyQt5.QtWidgets import (
     QWidget, QVBoxLayout, QHBoxLayout, QGroupBox, QFormLayout,
-    QLabel, QCheckBox, QPushButton, QLineEdit,
+    QLabel, QCheckBox, QPushButton, QLineEdit, QComboBox, QSpinBox,
     QMessageBox, QFileDialog, QProgressBar,
 )
 
@@ -33,12 +33,15 @@ class ReportWorkerThread(QThread):
     finished = pyqtSignal(str)
     error = pyqtSignal(str)
 
-    def __init__(self, work_dir: str, output_dir: Optional[str], report_title: str, enable_ai: bool):
+    def __init__(self, work_dir: str, output_dir: Optional[str], report_title: str,
+                 enable_ai: bool, report_mode: str = 'auto', ai_maps_limit: int = 999):
         super().__init__()
         self.work_dir = work_dir
         self.output_dir = output_dir
         self.report_title = report_title
         self.enable_ai = enable_ai
+        self.report_mode = report_mode
+        self.ai_maps_limit = ai_maps_limit
 
     def run(self):
         try:
@@ -76,6 +79,8 @@ class ReportWorkerThread(QThread):
             out_path = gen.generate_report(
                 work_dir=self.work_dir,
                 report_title=self.report_title or "水质参数反演分析报告",
+                report_mode=self.report_mode,
+                ai_maps_limit=self.ai_maps_limit,
                 on_progress=lambda pct, text: self.progress.emit(int(pct), str(text)),
             )
             self.finished.emit(str(out_path))
@@ -257,6 +262,37 @@ class Step13ReportPanel(QWidget):
         """)
         execute_layout.addWidget(self.progress_bar)
 
+                # 报告模式（自动检测 + 可手动纠正）
+        mode_row = QHBoxLayout()
+        mode_row.setContentsMargins(0, 4, 0, 0)
+        mode_row.setSpacing(10)
+        mode_label = QLabel("报告模式:")
+        mode_label.setMinimumWidth(120)
+        mode_label.setMaximumWidth(120)
+        mode_row.addWidget(mode_label)
+        self.report_mode_cb = QComboBox()
+        self.report_mode_cb.addItem("机器学习水质参数反演", "ml")
+        self.report_mode_cb.addItem("水色指数与物理经验公式反演", "formula")
+        self.report_mode_cb.setStyleSheet(common_lineedit_css)
+        mode_row.addWidget(self.report_mode_cb, 1)
+        execute_layout.addLayout(mode_row)
+
+        # AI 专题图分析数量（仅水色指数公式模式生效，ML 模式忽略）
+        ai_maps_row = QHBoxLayout()
+        ai_maps_row.setContentsMargins(0, 0, 0, 0)
+        ai_maps_row.setSpacing(10)
+        ai_maps_label = QLabel("AI分析专题图数:")
+        ai_maps_label.setMinimumWidth(120)
+        ai_maps_label.setMaximumWidth(120)
+        ai_maps_row.addWidget(ai_maps_label)
+        self.ai_maps_spin = QSpinBox()
+        self.ai_maps_spin.setRange(0, 999)
+        self.ai_maps_spin.setValue(999)
+        self.ai_maps_spin.setToolTip("公式报告中启用AI分析的分布图数量上限（0=跳过，999=全部分析）")
+        self.ai_maps_spin.setStyleSheet(common_lineedit_css)
+        ai_maps_row.addWidget(self.ai_maps_spin, 1)
+        execute_layout.addLayout(ai_maps_row)
+
         action_layout = QHBoxLayout()
         action_layout.addStretch()
 
@@ -326,13 +362,25 @@ class Step13ReportPanel(QWidget):
             self.work_dir_edit.setText(str(work_dir))
 
     def _auto_pull_work_dir(self):
-        """从主窗口自动同步工作目录到 work_dir_edit（无需用户操作）。"""
+        """从主窗口自动同步工作目录到 work_dir_edit，并自动检测报告模式。"""
         mw = self.main_window
         if mw is not None and getattr(mw, "work_dir", None):
             wd = str(mw.work_dir)
             cur = self.work_dir_edit.text().strip()
             if wd and wd != cur:
                 self.work_dir_edit.setText(wd)
+            # 自动检测报告模式（参考 Step11 逻辑）
+            self._auto_detect_report_mode(wd)
+
+    def _auto_detect_report_mode(self, wd: str):
+        """根据工作目录内容自动选择报告模式下拉框"""
+        wd_path = Path(wd)
+        ml_dir = wd_path / "9_ML_Prediction"
+        formula_dir = wd_path / "10_WaterIndex_CSV"
+        if ml_dir.is_dir() and list(ml_dir.glob("*.csv")):
+            self.report_mode_cb.setCurrentIndex(0)  # ML
+        elif formula_dir.is_dir() and list(formula_dir.glob("*.csv")):
+            self.report_mode_cb.setCurrentIndex(1)  # 公式
 
     def update_from_config(self, work_dir=None, pipeline=None):
         """切入面板时由主窗口统一调用，把当前 work_dir 同步到本面板。
@@ -381,6 +429,7 @@ class Step13ReportPanel(QWidget):
                 f"未找到可视化目录：\n{viz}\n请先完成流程或生成可视化。",
             )
             return
+
         if self._report_thread and self._report_thread.isRunning():
             QMessageBox.information(self, "提示", "报告正在生成中，请稍候。")
             return
@@ -394,7 +443,9 @@ class Step13ReportPanel(QWidget):
         self.progress_bar.setValue(0)
         self.progress_label.setText("正在准备生成…")
 
-        self._report_thread = ReportWorkerThread(wd, out, title, enable_ai)
+        report_mode = self.report_mode_cb.currentData()
+        ai_maps_limit = self.ai_maps_spin.value()
+        self._report_thread = ReportWorkerThread(wd, out, title, enable_ai, report_mode, ai_maps_limit)
         self._report_thread.progress.connect(self._on_progress, Qt.QueuedConnection)
         self._report_thread.finished.connect(self._on_finished, Qt.QueuedConnection)
         self._report_thread.error.connect(self._on_error, Qt.QueuedConnection)

src/postprocessing/report_word.py

@@ -368,6 +368,8 @@ class WaterQualityReportGenerator:
             return "（Minimax API Key 未配置，请设置 MINIMAX_API_KEY 环境变量）"
 
         url = self.minimax_base_url
+        if not url.endswith("/chat/completions"):
+            url = f"{url.rstrip('/')}/chat/completions"
 
         payload: Dict[str, Any] = {
             "model": self.minimax_text_model,
@@ -433,6 +435,12 @@ class WaterQualityReportGenerator:
             return f"（读取图片失败：{e}）"
 
         url = self.minimax_base_url
+        if not url.endswith("/chat/completions"):
+            url = f"{url.rstrip('/')}/chat/completions"
+
+        # 动态 MIME 类型：PNG 和 JPEG 分别处理，避免 API 400
+        ext = image_path.suffix.lower()
+        mime_type = "image/png" if ext == ".png" else "image/jpeg"
 
         payload: Dict[str, Any] = {
             "model": self.minimax_vision_model,
@@ -444,7 +452,7 @@ class WaterQualityReportGenerator:
                         {"type": "text", "text": user_prompt},
                         {
                             "type": "image_url",
-                            "image_url": {"url": f"data:image/jpeg;base64,{img_b64}"},
+                            "image_url": {"url": f"data:{mime_type};base64,{img_b64}"},
                         },
                     ],
                 }
@@ -739,10 +747,16 @@ class WaterQualityReportGenerator:
         if not self.enable_ai_analysis:
             return "（AI分析已关闭）"
 
-        # 构造统计数据文本
+        # 构造统计数据文本（防御式 .get() 防 KeyError）
         stats_text = "水质参数统计摘要：\n"
         for stat in stats_data:
-            stats_text += f"- {stat['参数']}: 点位数={stat['点位数']}, 范围=[{stat['最小值']}, {stat['最大值']}], 均值={stat['平均值']}, 标准差={stat['标准差']}\n"
+            param = stat.get('参数', stat.get('Parameter', stat.get('name', '未知')))
+            count = stat.get('点位数', stat.get('数量', stat.get('count', '?')))
+            min_v = stat.get('最小值', stat.get('min', stat.get('Min', '?')))
+            max_v = stat.get('最大值', stat.get('max', stat.get('Max', '?')))
+            mean_v = stat.get('平均值', stat.get('mean', stat.get('Mean', '?')))
+            std_v = stat.get('标准差', stat.get('std', stat.get('Std', '?')))
+            stats_text += f"- {param}: 点位数={count}, 范围=[{min_v}, {max_v}], 均值={mean_v}, 标准差={std_v}\n"
 
         system = """你是一位水质遥感与统计分析专家。
         请基于提供的统计数据，给出专业分析：
@@ -764,16 +778,16 @@ class WaterQualityReportGenerator:
                        parameters: List[str] = None,
                        report_title: str = "水质参数反演分析报告",
                        output_path: Optional[str] = None,
+                       report_mode: str = 'auto',
+                       ai_maps_limit: int = 999,
                        on_progress=None) -> str:
         """
-        生成 Word 报告 - 所有数据均来自工作目录（work_dir）
-        可视化图片、统计数据等均从 work_dir/12_visualization 和 work_dir/4_processed_data 中读取
+        生成 Word 报告（双轨制入口，默认自动检测模式）
 
         Args:
-            on_progress: 可选回调，签名 on_progress(percent: int, text: str)。
-                会在进度更新时被调用，用于驱动 Qt QProgressBar/QThread 信号。
+            report_mode: 'auto'→自动检测(参考Step11), 'ml'/'formula'→强制模式
         """
-        # 设置工作目录（整个流程的核心）
+        # 设置工作目录
         if work_dir is not None:
             self.work_dir = Path(work_dir)
             self.visualization_dir = self.work_dir / "12_visualization"
@@ -781,29 +795,38 @@ class WaterQualityReportGenerator:
                 self.output_dir = self.visualization_dir
                 self.output_dir.mkdir(parents=True, exist_ok=True)
                 self.ai_cache_path = self.output_dir / "ollama_image_analyses_cache.json"
-        
-        if parameters is None:
-            parameters = ["Chlorophyll", "COD", "DO", "PH", "Temperature", 
-                         "spCond", "Turbidity", "TDS", "Cl-", "NO3-N", 
-                         "NH3-N", "BGA", "TT"]
-        
-        vis_dir = self.visualization_dir
-        
-        if not vis_dir.exists():
-            raise FileNotFoundError(f"可视化目录不存在: {vis_dir}")
 
-        # ── 管线模式自动检测 ──
-        self._pipeline_mode = self._detect_pipeline_mode(vis_dir, parameters)
-        print(f"[报告] 管线模式: {self._pipeline_mode}")
-        if self._pipeline_mode == "water_index":
-            # 水色指数模式：只保留实际存在的图片类型
-            _available_types = self._get_available_image_types(vis_dir, parameters)
-            self.parameter_images = {
-                p: [f"{p}_{t}.png" for t in _available_types]
-                for p in parameters
-            }
-            print(f"[报告] 水色指数可用图片类型: {_available_types}")
-        
+        if not self.visualization_dir.exists():
+            raise FileNotFoundError(f"可视化目录不存在: {self.visualization_dir}")
+
+        if parameters is None:
+            parameters = ["Chlorophyll", "COD", "DO", "PH", "Temperature",
+                         "spCond", "Turbidity", "TDS", "Cl-", "NO3-N",
+                         "NH3-N", "BGA", "TT"]
+
+        # ── 确定管线模式（优先用户选择，否则自动检测）──
+        if report_mode == 'auto':
+            ml_dir = self.work_dir / "9_ML_Prediction"
+            formula_dir = self.work_dir / "10_WaterIndex_CSV"
+            if ml_dir.is_dir() and list(ml_dir.glob("*.csv")):
+                report_mode = 'ml'
+            elif formula_dir.is_dir() and list(formula_dir.glob("*.csv")):
+                report_mode = 'formula'
+            else:
+                report_mode = 'ml'
+            print(f"[报告] 自动检测 → {report_mode}")
+        else:
+            print(f"[报告] 用户选择 → {report_mode}")
+
+        if report_mode == 'formula':
+            return self._generate_formula_report(parameters, report_title, output_path,
+                                                  ai_maps_limit=ai_maps_limit, on_progress=on_progress)
+        else:
+            return self._generate_ml_report(parameters, report_title, output_path, on_progress)
+
+    def _generate_ml_report(self, parameters, report_title, output_path, on_progress):
+        """机器学习水质参数反演报告"""
+        vis_dir = self.visualization_dir
         if output_path is None:
             timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
             output_path = self.output_dir / f"水质参数反演分析报告_{timestamp}.docx"
@@ -853,9 +876,14 @@ class WaterQualityReportGenerator:
             
             # 按参数生成内容（带编号）：参数章节从 5 开始编号
             base_section_num = 5
-            last_param_section_num = base_section_num + len(parameters) - 1
-            for section_num, param in enumerate(parameters, base_section_num):
-                progress.set_description(f"正在分析 {param} 数据 ({section_num - base_section_num + 1}/{len(parameters)})")
+            n_params = len(parameters)
+            last_param_section_num = base_section_num + n_params - 1
+            for s_i, param in enumerate(parameters):
+                section_num = base_section_num + s_i
+                pct = int((s_i + 1) / n_params * 100)
+                if on_progress:
+                    try: on_progress(pct, f"ML报告 参数 {s_i+1}/{n_params}: {param}")
+                    except Exception: pass
                 figure_counter = self._add_parameter_section(
                     doc,
                     param,
@@ -924,7 +952,198 @@ class WaterQualityReportGenerator:
         print(f"✅ Word报告生成完成: {output_path}")
         
         return str(output_path)
-    
+
+    def _generate_formula_report(self, parameters, report_title, output_path,
+                                   ai_maps_limit=999, on_progress=None):
+        """水色指数与物理经验公式反演报告 (Step 10/11)"""
+        from docx.shared import Inches, Pt, Cm, RGBColor
+        from docx.enum.text import WD_ALIGN_PARAGRAPH
+
+        vis_dir = self.visualization_dir
+        # 分布图优先从 12_visualization/distribution_maps 找，再回退 11_Thematic_Map
+        dist_map_dir = vis_dir / "distribution_maps"
+        thematic_dir = dist_map_dir if dist_map_dir.is_dir() else self.work_dir / "11_Thematic_Map"
+        csv_dir = self.work_dir / "10_WaterIndex_CSV"
+
+        if output_path is None:
+            timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
+            output_path = self.output_dir / f"水色指数反演分析报告_{timestamp}.docx"
+        else:
+            output_path = Path(output_path)
+
+        # 扫描实际数据量：专题图优先匹配 distribution/rendered 命名
+        csv_files = sorted(csv_dir.glob("*.csv")) if csv_dir.is_dir() else []
+        tif_files = []
+        if thematic_dir.is_dir():
+            tif_files = (sorted(thematic_dir.glob("*distribution*.png"))
+                       + sorted(thematic_dir.glob("*rendered*.png"))
+                       + sorted(thematic_dir.glob("*.tif"))
+                       + sorted(thematic_dir.glob("*.png")))
+            # 去重（同名文件可能多次匹配）
+            seen = set()
+            tif_files = [f for f in tif_files if f.name not in seen and not seen.add(f.name)]
+        n_csv = len(csv_files)
+        n_maps = min(len(tif_files), 20)
+        n_csv_pulses = max(1, n_csv // 10)  # 每 10 个 CSV 一次脉冲
+        total_steps = 4 + n_csv_pulses + n_maps  # 固定章节 + 统计脉冲 + 每张专题图
+        print(f"[公式报告] 数据: {n_csv} CSV, {len(tif_files)} 专题图 → 总步数 {total_steps}")
+
+        progress = self._create_progress(total=total_steps, desc="生成公式报告", on_step=on_progress)
+        _step = 0
+        def _next(text=""):
+            nonlocal _step
+            _step += 1
+            try: progress.update(1)
+            except Exception: pass
+            if on_progress:
+                try: on_progress(int(_step / total_steps * 100), text or f"步骤 {_step}/{total_steps}")
+                except Exception: pass
+
+        doc = Document()
+        section = doc.sections[0]
+        section.page_width = Cm(21)
+        section.page_height = Cm(29.7)
+        section.left_margin = Cm(2.5)
+        section.right_margin = Cm(2.5)
+        section.top_margin = Cm(2.5)
+        section.bottom_margin = Cm(2.5)
+
+        # ── 封面 ──
+        self._add_cover_page(doc)
+        self._add_company_description_page(doc)
+        self._add_data_acquisition_section(doc)
+        self._add_data_processing_section(doc)
+        _next("封面与通用章节")
+
+        # ── 1. 项目背景 ──
+        h1 = doc.add_heading("1 项目背景", level=1)
+        self._style_heading(h1, 1)
+        doc.add_paragraph("本报告基于高光谱遥感影像，通过物理经验公式（水色指数）反演水体关键参数的空间分布。"
+                         "涵盖叶绿素、蓝绿藻、浊度、CDOM 等多种水色指标的定量化空间制图。")
+        doc.add_paragraph("数据来源：机载 / 星载高光谱成像仪，经过辐射定标、大气校正、耀斑去除等预处理。")
+        doc.add_page_break()
+        _next("项目背景")
+
+        # ── 2. 影像预处理 ──
+        h1 = doc.add_heading("2 影像预处理", level=1)
+        self._style_heading(h1, 1)
+        self._add_hyperspectral_images_section(doc)
+        doc.add_page_break()
+        _next("影像预处理")
+
+        # ── 3. 水色指数公式列表 ──
+        h1 = doc.add_heading("3 水色指数计算方法", level=1)
+        self._style_heading(h1, 1)
+        if csv_files:
+            doc.add_paragraph(f"本次反演共应用 {n_csv} 个水色指数公式，"
+                            f"各公式基于特征波段比值或差分原理计算：")
+            for i, cf in enumerate(csv_files[:30], 1):
+                doc.add_paragraph(f"  {i}. {cf.stem}", style='List Number')
+            if n_csv > 30:
+                doc.add_paragraph(f"  ... 共 {n_csv} 个公式，详情见统计章节。")
+        else:
+            doc.add_paragraph(f"（未找到水色指数目录: {csv_dir}）")
+        doc.add_page_break()
+
+        # ── 4. 指数统计结果（每个 CSV 更新一次进度）──
+        h1 = doc.add_heading("4 水色指数统计结果", level=1)
+        self._style_heading(h1, 1)
+        figure_num = 10
+        if csv_files:
+            stats_rows = []
+            for i_c, cp in enumerate(csv_files):
+                try:
+                    df_one = pd.read_csv(cp, sep=',')
+                    val_col = df_one.columns[-1]
+                    vals = pd.to_numeric(df_one[val_col], errors='coerce').dropna()
+                    if len(vals) > 0:
+                        stats_rows.append({
+                            '参数': Path(cp).stem, '数量': len(vals),
+                            '最小值': round(float(vals.min()), 4),
+                            '最大值': round(float(vals.max()), 4),
+                            '平均值': round(float(vals.mean()), 4),
+                            '标准差': round(float(vals.std(ddof=0)), 4) if len(vals) > 1 else 0,
+                        })
+                except Exception:
+                    pass
+                # 每处理 10 个 CSV 更新进度
+                if (i_c + 1) % max(1, n_csv // 10) == 0 or i_c == n_csv - 1:
+                    _next(f"统计 {i_c+1}/{n_csv}")
+
+            if stats_rows:
+                # 统计表
+                table = doc.add_table(rows=1, cols=6, style='Table Grid')
+                for j, h in enumerate(['公式名称', '数据量', '最小值', '最大值', '平均值', '标准差']):
+                    table.rows[0].cells[j].text = h
+                for r in stats_rows:
+                    cells = table.add_row().cells
+                    cells[0].text = r['参数']
+                    cells[1].text = str(r['数量'])
+                    cells[2].text = str(r['最小值'])
+                    cells[3].text = str(r['最大值'])
+                    cells[4].text = str(r['平均值'])
+                    cells[5].text = str(r['标准差'])
+                doc.add_paragraph()
+                figure_num += 1
+
+                # AI 分析（防御式读取）
+                stats_data = [{
+                    '参数': r['参数'], '点位数': r['数量'],
+                    '最小值': str(r['最小值']), '最大值': str(r['最大值']),
+                    '平均值': str(r['平均值']), '标准差': str(r['标准差']),
+                } for r in stats_rows]
+                param_names = [r['参数'] for r in stats_rows]
+                analysis = self._analyze_statistics(stats_data, param_names)
+                self._add_ai_analysis_paragraph(doc, analysis)
+            else:
+                doc.add_paragraph("未能解析水色指数 CSV 数据。")
+        else:
+            doc.add_paragraph(f"（未找到水色指数目录: {csv_dir}）")
+        doc.add_page_break()
+
+        # ── 5. 指数空间分布专题图（每张更新进度）──
+        h1 = doc.add_heading("5 水色指数空间分布专题图", level=1)
+        self._style_heading(h1, 1)
+        figure_num += 1
+        maps_found = 0
+        if tif_files:
+            n_show = min(len(tif_files), 20)
+            for i_t, tf in enumerate(tif_files[:n_show]):
+                try:
+                    param_name = tf.stem.split('_')[0] if '_' in tf.stem else tf.stem
+                    caption = f"图{figure_num} {param_name} 空间分布图"
+                    if self._add_image_with_caption(doc, str(tf), caption, width=Inches(5.5)):
+                        if self.enable_ai_analysis and i_t < ai_maps_limit:
+                            ai_text = self._analyze_and_cache_image(
+                                image_path=tf, image_type="distribution",
+                                param=param_name, figure_num=figure_num)
+                            self._add_ai_analysis_paragraph(doc, ai_text)
+                    figure_num += 1
+                    maps_found += 1
+                except Exception as e:
+                    doc.add_paragraph(f"[专题图插入失败: {tf.name} — {e}]")
+                _next(f"专题图 {i_t+1}/{n_show}")
+            if len(tif_files) > 20:
+                doc.add_paragraph(f"... 共 {len(tif_files)} 张专题图，此处仅展示前 20 张。")
+        else:
+            doc.add_paragraph(f"（未找到专题图目录: {thematic_dir}）")
+        if maps_found == 0:
+            doc.add_paragraph("（未找到专题图文件，请确认 Step 11 已完成。）")
+        doc.add_page_break()
+
+        # ── 6. 综合总结 ──
+        h1 = doc.add_heading("6 综合总结", level=1)
+        self._style_heading(h1, 1)
+        doc.add_paragraph("本报告基于高光谱遥感影像的水色指数反演方法，对研究区域内的关键水质参数"
+                         "进行了定量化空间制图。各指数的统计结果和空间分布专题图如上所示。")
+        doc.add_paragraph("注意事项：水色指数反演结果为半定量指标，其绝对值可能受大气校正精度、"
+                         "水体光学特性复杂性等因素影响。建议结合实测水质数据进行校验。")
+        _next("综合总结")
+
+        doc.save(str(output_path))
+        print(f"[公式报告] 生成完成: {output_path}")
+        return str(output_path)
+
     def _add_parameter_section(
         self, doc, param: str, vis_dir: Path, param_index: int = 1,
         start_figure_num: int = 1, all_image_analyses: Optional[List[Dict[str, Any]]] = None, progress=None,
@@ -1254,130 +1473,106 @@ class WaterQualityReportGenerator:
             run.font.size = Pt(12)
             run._element.rPr.rFonts.set(qn('w:eastAsia'), 'SimSun')
         
-        # 添加高光谱图像、耀斑区域和去耀斑图像展示
-        self._add_hyperspectral_images_section(doc)
-
         doc.add_page_break()
 
     def _add_hyperspectral_images_section(self, doc):
-        """添加高光谱图像、耀斑区域和去耀斑图像展示"""
-        h = doc.add_heading("3.1 高光谱图像处理过程", level=2)
-        self._style_heading(h, level=2)
-
+        """添加高光谱图像、耀斑区域和去耀斑图像展示
+        找不到文件的章节直接跳过，不显示占位文字。
+        """
         work_dir_path = self.work_dir
         vis_dir = self.visualization_dir
 
-        # 0. 航线规划图
-        flight_path_img_path = work_dir_path / "12_visualization" / "flight_paths"
-        h3 = doc.add_heading("航线规划：", level=3)
-        self._style_heading(h3, level=3)
-
-        # 查找航线图文件
+        # 0. 航线规划图（仅搜索专用目录，找不到则整节跳过）
         flight_map_files = []
-        if flight_path_img_path.exists():
-            flight_map_files = list(flight_path_img_path.glob("*.png")) + list(flight_path_img_path.glob("*.jpg"))
+        flight_dirs = [
+            work_dir_path / "12_visualization" / "flight_paths",
+            vis_dir / "flight_paths",
+        ]
+        for d in flight_dirs:
+            if d.exists():
+                flight_map_files = sorted(d.glob("*.png")) + sorted(d.glob("*.jpg"))
+                if flight_map_files:
+                    break
 
         if flight_map_files:
-            # 使用最新的航线图文件
-            latest_flight_map = max(flight_map_files, key=lambda p: p.stat().st_mtime)
-            success = self._add_image_with_caption(doc, str(latest_flight_map), "图3-1 航线规划", width=Inches(5.5))
-            
-            if success:
-                # AI 分析航线规划图
-                flight_analysis = self._analyze_flight_path_image(str(latest_flight_map))
-                self._add_ai_analysis_paragraph(doc, flight_analysis)
-        else:
-            doc.add_paragraph("[航线规划图 - 文件未找到]")
+            h3 = doc.add_heading("航线规划：", level=3)
+            self._style_heading(h3, level=3)
+            latest = max(flight_map_files, key=lambda p: p.stat().st_mtime)
+            if self._add_image_with_caption(doc, str(latest), "航线规划", width=Inches(5.5)):
+                if self.enable_ai_analysis:
+                    self._add_ai_analysis_paragraph(doc,
+                        self._analyze_flight_path_image(str(latest)))
 
         # 1. 高光谱原始图像
-        hyperspectral_img_path = work_dir_path / "1_water_mask" / "hsi_preview.png"
-        h3 = doc.add_heading("高光谱原始影像：", level=3)
-        self._style_heading(h3, level=3)
-        if hyperspectral_img_path.exists():
-            self._add_image_with_caption(doc, str(hyperspectral_img_path), "图3-2 高光谱原始影像", width=Inches(5.5))
-        else:
-            doc.add_paragraph("[高光谱原始影像 - 文件未找到]")
+        hsi_path = work_dir_path / "1_water_mask" / "hsi_preview.png"
+        if hsi_path.exists():
+            h3 = doc.add_heading("高光谱原始影像：", level=3)
+            self._style_heading(h3, level=3)
+            self._add_image_with_caption(doc, str(hsi_path), "高光谱原始影像", width=Inches(5.5))
 
         # 2. 水体掩膜叠加图
-        water_mask_overlay_path = work_dir_path / "1_water_mask" / "water_mask_overlay.png"
-        h3 = doc.add_heading("水体区域识别：", level=3)
-        self._style_heading(h3, level=3)
-        if water_mask_overlay_path.exists():
-            success = self._add_image_with_caption(doc, str(water_mask_overlay_path), 
-                                                    "图3-3 水体区域识别（蓝色半透明区域为水域）", 
-                                                    width=Inches(5.5))
-            if success:
-                water_analysis = self._analyze_water_mask_overlay(str(water_mask_overlay_path))
-                self._add_ai_analysis_paragraph(doc, water_analysis)
-        else:
-            doc.add_paragraph("[水体区域识别图 - 文件未找到]")
+        wm_path = work_dir_path / "1_water_mask" / "water_mask_overlay.png"
+        if wm_path.exists():
+            h3 = doc.add_heading("水体区域识别：", level=3)
+            self._style_heading(h3, level=3)
+            if self._add_image_with_caption(doc, str(wm_path),
+                                            "水体区域识别（蓝色半透明区域为水域）",
+                                            width=Inches(5.5)):
+                if self.enable_ai_analysis:
+                    self._add_ai_analysis_paragraph(doc,
+                        self._analyze_water_mask_overlay(str(wm_path)))
 
-        doc.add_paragraph()
+        # 3. 耀斑区域
+        glint_dirs = [vis_dir / "glint_deglint_previews",
+                      work_dir_path / "2_Glint_Detection"]
+        glint_img = None
+        for d in glint_dirs:
+            if d.exists():
+                cands = sorted(d.glob("*glint*.png")) + sorted(d.glob("*severe*.png"))
+                if cands:
+                    glint_img = cands[0]
+                    break
+        if glint_img:
+            h3 = doc.add_heading("耀斑区域识别结果：", level=3)
+            self._style_heading(h3, level=3)
+            if self._add_image_with_caption(doc, str(glint_img), "耀斑区域识别结果", width=Inches(5.5)):
+                if self.enable_ai_analysis:
+                    self._add_ai_analysis_paragraph(doc,
+                        self._analyze_glint_distribution_with_ai(str(glint_img)))
 
-        # 2. 耀斑区域
-        glint_img_path = vis_dir / "glint_deglint_previews" / "glint_severe_glint_area_preview.png"
-        h3 = doc.add_heading("耀斑区域识别结果：", level=3)
-        self._style_heading(h3, level=3)
-        if glint_img_path.exists():
-            self._add_image_with_caption(doc, str(glint_img_path), "图3-4 耀斑区域识别结果", width=Inches(5.5))
-        else:
-            # 尝试查找其他可能的耀斑预览图
-            glint_files = list(vis_dir.glob("glint_deglint_previews/*glint*.png"))
-            if glint_files:
-                glint_img_path = glint_files[0]
-                self._add_image_with_caption(doc, str(glint_img_path), "图3-4 耀斑区域识别结果", width=Inches(5.5))
-            else:
-                doc.add_paragraph("[耀斑区域识别结果 - 文件未找到]")
+        # 4. 去除耀斑后的图像
+        deglint_img = None
+        for d in glint_dirs:
+            if d.exists():
+                cands = sorted(d.glob("*deglint*.png"))
+                if cands:
+                    deglint_img = cands[0]
+                    break
+        deglint_dir2 = work_dir_path / "3_deglint"
+        if deglint_img is None and deglint_dir2.exists():
+            cands = sorted(deglint_dir2.glob("*.png"))
+            if cands:
+                deglint_img = cands[0]
+        if deglint_img:
+            h3 = doc.add_heading("去除耀斑后的影像：", level=3)
+            self._style_heading(h3, level=3)
+            self._add_image_with_caption(doc, str(deglint_img), "去除耀斑后的影像", width=Inches(5.5))
 
-        doc.add_paragraph()
-
-        # 3. 去除耀斑后的图像
-        deglint_img_path = vis_dir / "glint_deglint_previews" / "deglint_deglint_image_preview.png"
-        h3 = doc.add_heading("去除耀斑后的影像：", level=3)
-        self._style_heading(h3, level=3)
-        if deglint_img_path.exists():
-            self._add_image_with_caption(doc, str(deglint_img_path), "图3-5 去除耀斑后的高光谱影像", width=Inches(5.5))
-        else:
-            # 尝试查找其他去耀斑预览图
-            deglint_files = list(vis_dir.glob("glint_deglint_previews/*deglint*.png"))
-            if deglint_files:
-                deglint_img_path = deglint_files[0]
-                self._add_image_with_caption(doc, str(deglint_img_path), "图3-5 去除耀斑后的影像", width=Inches(5.5))
-            else:
-                doc.add_paragraph("[去除耀斑后的影像 - 文件未找到]")
-
-        doc.add_paragraph()
-
-        # 4. AI分析耀斑位置分布
-
-        self._style_heading(h3, level=3)
-        glint_analysis = self._analyze_glint_distribution_with_ai(
-            str(glint_img_path) if 'glint_img_path' in locals() and Path(str(glint_img_path)).exists() else None,
-            str(hyperspectral_img_path) if hyperspectral_img_path.exists() else None
-        )
-        self._add_ai_analysis_paragraph(doc, glint_analysis)
-
-        # 5. 采样点分布图
+        # 5. 采样点分布图（找不到则整节跳过）
         sampling_map_dir = vis_dir / "sampling_maps"
-        h3 = doc.add_heading("采样点分布：", level=3)
-        self._style_heading(h3, level=3)
-
-        # 查找采样点分布图文件
         sampling_map_files = []
         if sampling_map_dir.exists():
-            sampling_map_files = list(sampling_map_dir.glob("*.png")) + list(sampling_map_dir.glob("*.jpg"))
+            sampling_map_files = (list(sampling_map_dir.glob("*.png"))
+                                + list(sampling_map_dir.glob("*.jpg")))
 
         if sampling_map_files:
-            # 使用最新的采样点分布图文件
+            h3 = doc.add_heading("采样点分布：", level=3)
+            self._style_heading(h3, level=3)
             latest_sampling_map = max(sampling_map_files, key=lambda p: p.stat().st_mtime)
-            success = self._add_image_with_caption(doc, str(latest_sampling_map), "图3-6 采样点分布图", width=Inches(5.5))
-            
-            if success:
-                # AI 分析采样点分布图
-                sampling_analysis = self._analyze_sampling_distribution(str(latest_sampling_map))
-                self._add_ai_analysis_paragraph(doc, sampling_analysis)
-        else:
-            doc.add_paragraph("[采样点分布图 - 文件未找到]")
+            if self._add_image_with_caption(doc, str(latest_sampling_map), "采样点分布图", width=Inches(5.5)):
+                if self.enable_ai_analysis:
+                    self._add_ai_analysis_paragraph(doc,
+                        self._analyze_sampling_distribution(str(latest_sampling_map)))
 
     def _analyze_glint_distribution_with_ai(self, glint_img_path: str = None, original_img_path: str = None) -> str:
         """使用AI分析耀斑的位置分布"""
@@ -1724,7 +1919,11 @@ class WaterQualityReportGenerator:
                     row_cells = table.add_row().cells
                     for j, (col_name, hdr_name) in enumerate(header_map.items()):
                         row_cells[j].text = str(row_dict.get(col_name, ''))
-                stats_data = [{'参数': r['参数'], '点位数': r['数量']} for r in stats_rows]
+                stats_data = [{
+                    '参数': r['参数'], '点位数': r['数量'],
+                    '最小值': str(r['最小值']), '最大值': str(r['最大值']),
+                    '平均值': str(r['平均值']), '标准差': str(r['标准差']),
+                } for r in stats_rows]
             else:
                 # ML 模式：逐列统计
                 stats_data = []
@@ -1780,10 +1979,10 @@ class WaterQualityReportGenerator:
         doc.add_paragraph()  # 表格和热力图之间的空行
         
         # 2. 添加相关性热力图（放在表格下方）—— 仅 ML 模式
+        heatmap_path = vis_dir / "correlation_heatmap.png"
         if getattr(self, '_pipeline_mode', 'ml') == 'ml':
             h3 = doc.add_heading("4.2 水质参数相关性分析", level=2)
             self._style_heading(h3, level=2)
-            heatmap_path = vis_dir / "correlation_heatmap.png"
             figure_num = start_figure_num
             if heatmap_path.exists():
                 try:
@@ -1823,7 +2022,9 @@ class WaterQualityReportGenerator:
             pass
         
         doc.add_page_break()
-        return start_figure_num + (1 if heatmap_path.exists() else 0)
+        heatmap_added = 1 if (getattr(self, '_pipeline_mode', 'ml') == 'ml'
+                               and heatmap_path.exists()) else 0
+        return start_figure_num + heatmap_added
 
     def _add_physical_inversion_section(
         self,