运营区边缘计算：区域数据缓存

ebike-operations/src/main/java/com/cdzy/operations/component/GridIndexBuilder.java

@@ -0,0 +1,339 @@
+package com.cdzy.operations.component;
+
+import com.cdzy.operations.utils.RedisUtil;
+import com.fasterxml.jackson.databind.ObjectMapper;
+import jakarta.annotation.Resource;
+import org.locationtech.jts.geom.*;
+import org.locationtech.jts.io.WKTReader;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.stereotype.Component;
+
+import java.util.*;
+import java.util.stream.Collectors;
+
+/**
+ * 网格索引构建器（Spring Bean 版本，无本地缓存）
+ * 将运营区多边形边界切分为网格，每个网格关联其覆盖的边界线段。
+ * 提供构建索引、发布到 Redis、以及实时判断车辆是否靠近边缘的功能。
+ * 每次判断均直接访问 Redis，保证数据实时性。
+ */
+@Component
+public class GridIndexBuilder {
+    private static final Logger log = LoggerFactory.getLogger(GridIndexBuilder.class);
+    private static final double EARTH_RADIUS = 6371000; // 地球半径（米）
+    private static final GeometryFactory GEOMETRY_FACTORY = new GeometryFactory();
+    private static final WKTReader WKT_READER = new WKTReader();
+    private final ObjectMapper objectMapper = new ObjectMapper();
+
+    // 配置参数（可从 application.yml 注入）
+    @Value("${fence.threshold.meters:100}")
+    private double thresholdMeters;
+
+    @Resource
+    private RedisUtil redisUtil;
+
+    // 无参构造（Spring使用）
+    public GridIndexBuilder() {}
+
+    // 带参构造（方便手动调用或单元测试）
+    public GridIndexBuilder(double thresholdMeters, RedisUtil redisUtil) {
+        this.thresholdMeters = thresholdMeters;
+        this.redisUtil = redisUtil;
+    }
+
+    // ==================== 索引构建与发布 ====================
+
+    /**
+     * 构建网格索引并发布到 Redis
+     * @param polygon  运营区多边形
+     * @param regionId 运营区编号
+     */
+    public void buildIndex(Polygon polygon, Long regionId) {
+        // 1. 获取外包矩形
+        Envelope envelope = polygon.getEnvelopeInternal();
+        double minLon = envelope.getMinX();
+        double maxLon = envelope.getMaxX();
+        double minLat = envelope.getMinY();
+        double maxLat = envelope.getMaxY();
+
+        // 2. 计算网格尺寸（米转度）
+        double centerLat = (minLat + maxLat) / 2;
+        double lonPerMeter = 1 / (111320 * Math.cos(Math.toRadians(centerLat)));
+        double latPerMeter = (double)1 / 110574;
+        double gridLonWidth = (thresholdMeters / 2) * lonPerMeter;
+        double gridLatWidth = (thresholdMeters / 2) * latPerMeter;
+
+        // 3. 提取所有边界线段
+        List<Coordinate[]> edges = extractEdges(polygon);
+
+        // 4. 构建网格索引
+        Map<String, List<Coordinate[]>> gridIndex = new HashMap<>();
+        for (Coordinate[] seg : edges) {
+            Coordinate a = seg[0];
+            Coordinate b = seg[1];
+
+            double minSegLon = Math.min(a.x, b.x) - gridLonWidth;
+            double maxSegLon = Math.max(a.x, b.x) + gridLonWidth;
+            double minSegLat = Math.min(a.y, b.y) - gridLatWidth;
+            double maxSegLat = Math.max(a.y, b.y) + gridLatWidth;
+
+            int minLonIdx = (int) Math.floor((minSegLon - minLon) / gridLonWidth);
+            int maxLonIdx = (int) Math.floor((maxSegLon - minLon) / gridLonWidth);
+            int minLatIdx = (int) Math.floor((minSegLat - minLat) / gridLatWidth);
+            int maxLatIdx = (int) Math.floor((maxSegLat - minLat) / gridLatWidth);
+
+            for (int lonIdx = minLonIdx; lonIdx <= maxLonIdx; lonIdx++) {
+                for (int latIdx = minLatIdx; latIdx <= maxLatIdx; latIdx++) {
+                    String gridId = lonIdx + "_" + latIdx;
+                    gridIndex.computeIfAbsent(gridId, k -> new ArrayList<>()).add(seg);
+                }
+            }
+        }
+
+        // 5. 创建元数据
+        GridMetadata metadata = new GridMetadata(minLon, minLat, gridLonWidth, gridLatWidth, thresholdMeters);
+
+        // 6. 发布到 Redis
+        try {
+            publishToRedis(gridIndex, metadata, polygon, regionId);
+            log.info("网格索引发布成功，regionId: {}, 网格数量: {}", regionId, gridIndex.size());
+        } catch (Exception e) {
+            throw new RuntimeException("发布网格索引到 Redis 失败", e);
+        }
+    }
+
+    /**
+     * 从多边形提取所有边界线段
+     */
+    private List<Coordinate[]> extractEdges(Polygon polygon) {
+        List<Coordinate[]> edges = new ArrayList<>();
+        LineString exterior = polygon.getExteriorRing();
+        addEdgesFromLineString(exterior, edges);
+        return edges;
+    }
+
+    private void addEdgesFromLineString(LineString lineString, List<Coordinate[]> edges) {
+        Coordinate[] coords = lineString.getCoordinates();
+        for (int i = 0; i < coords.length - 1; i++) {
+            edges.add(new Coordinate[]{coords[i], coords[i + 1]});
+        }
+    }
+
+    /**
+     * 发布网格索引、元数据和多边形到 Redis
+     */
+    private void publishToRedis(Map<String, List<Coordinate[]>> gridIndex, GridMetadata metadata,
+                                Polygon polygon, Long regionId) throws Exception {
+        String gridKey = "bike:grid:" + regionId;
+        String metaKey = "bike:grid:meta:" + regionId;
+        String wktKey = "bike:fence:wkt:" + regionId;
+
+        // 清理可能存在的类型错误键
+        checkAndCleanKey(gridKey, "hash");
+        checkAndCleanKey(metaKey, "string");
+        checkAndCleanKey(wktKey, "string");
+
+        // 写入网格索引（Hash）
+        Map<String, Object> hashMap = new HashMap<>();
+        for (Map.Entry<String, List<Coordinate[]>> entry : gridIndex.entrySet()) {
+            hashMap.put(entry.getKey(), serializeSegments(entry.getValue()));
+        }
+        redisUtil.hSetAll(gridKey, hashMap,RedisUtil.Database.DB2);
+
+        // 写入元数据（String JSON）
+        String metaJson = objectMapper.writeValueAsString(metadata);
+        redisUtil.set(RedisUtil.Database.DB2,metaKey, metaJson);
+
+        // 写入多边形 WKT（用于包含判断）
+        redisUtil.set(RedisUtil.Database.DB2,wktKey, polygon.toText());
+    }
+
+    /**
+     * 检查并清理不匹配类型的 key
+     */
+    private void checkAndCleanKey(String key, String expectedType) {
+        String type = redisUtil.type(key, RedisUtil.Database.DB2);
+        if (!"none".equals(type) && !expectedType.equals(type)) {
+            redisUtil.delete(RedisUtil.Database.DB2,key);
+            log.warn("清理类型不匹配的 key: {}, 原类型: {}, 期望类型: {}", key, type, expectedType);
+        }
+    }
+
+    // ==================== 实时判断方法 ====================
+
+    /**
+     * 判断车辆是否接近运营区边缘（从内部骑出）
+     * 每次调用均直接访问 Redis，保证数据实时性。
+     * @param regionId 运营区ID
+     * @param lon      经度
+     * @param lat      纬度
+     * @return true 表示需要提醒（距离小于阈值且车辆在多边形内），否则 false
+     */
+    public boolean isNearEdge(Long regionId, double lon, double lat) {
+        try {
+            // 1. 从 Redis 获取多边形 WKT 并判断是否在多边形内部
+            String wktKey = "bike:fence:wkt:" + regionId;
+            Object wktObj = redisUtil.get(RedisUtil.Database.DB2,wktKey);
+            if (wktObj == null) {
+                log.warn("多边形数据不存在，regionId: {}", regionId);
+                return false;
+            }
+            Polygon polygon = (Polygon) WKT_READER.read(wktObj.toString());
+            Point point = GEOMETRY_FACTORY.createPoint(new Coordinate(lon, lat));
+            if (!polygon.contains(point)) {
+                return false; // 不在内部，不处理
+            }
+
+            // 2. 从 Redis 获取网格元数据
+            String metaKey = "bike:grid:meta:" + regionId;
+            Object metaObj = redisUtil.get(RedisUtil.Database.DB2,metaKey);
+            if (metaObj == null) {
+                log.warn("网格元数据不存在，regionId: {}", regionId);
+                return false;
+            }
+            GridMetadata metadata = objectMapper.readValue(metaObj.toString(), GridMetadata.class);
+
+            // 3. 计算网格及邻域
+            int lonIdx = (int) Math.floor((lon - metadata.minLon) / metadata.gridLonWidth);
+            int latIdx = (int) Math.floor((lat - metadata.minLat) / metadata.gridLatWidth);
+            List<String> neighborIds = getNeighborGridIds(lonIdx, latIdx);
+
+            // 4. 从 Redis 获取这些网格的线段数据
+            String gridKey = "bike:grid:" + regionId;
+            List<Coordinate[]> candidateSegments = new ArrayList<>();
+            for (String gridId : neighborIds) {
+                Object jsonObj = redisUtil.hGet(gridKey, gridId);
+                if (jsonObj != null) {
+                    candidateSegments.addAll(deserializeSegments(jsonObj.toString()));
+                }
+            }
+
+            // 5. 计算距离并判断
+            return isDistanceLessThanThreshold(lon, lat, candidateSegments, metadata.threshold);
+        } catch (Exception e) {
+            log.error("判断边缘距离异常，regionId: {}, lon: {}, lat: {}", regionId, lon, lat, e);
+            return false;
+        }
+    }
+
+    // ---------- 距离计算工具 ----------
+    private boolean isDistanceLessThanThreshold(double lon, double lat, List<Coordinate[]> segments, double threshold) {
+        if (segments == null || segments.isEmpty()) return false;
+        Coordinate point = new Coordinate(lon, lat);
+        double minDist = Double.MAX_VALUE;
+        for (Coordinate[] seg : segments) {
+            double d = distanceToSegment(point, seg[0], seg[1]);
+            if (d < minDist) {
+                minDist = d;
+                if (minDist < threshold) return true;
+            }
+        }
+         return minDist < threshold;
+    }
+
+    private double distanceToSegment(Coordinate p, Coordinate a, Coordinate b) {
+        double refLat = Math.toRadians((a.y + b.y) / 2);
+        double[] pM = toMeters(p, refLat);
+        double[] aM = toMeters(a, refLat);
+        double[] bM = toMeters(b, refLat);
+
+        double dx = bM[0] - aM[0];
+        double dy = bM[1] - aM[1];
+        double lenSq = dx * dx + dy * dy;
+
+        if (lenSq == 0) {
+            return Math.hypot(pM[0] - aM[0], pM[1] - aM[1]);
+        }
+
+        double t = ((pM[0] - aM[0]) * dx + (pM[1] - aM[1]) * dy) / lenSq;
+        t = Math.max(0, Math.min(1, t));
+
+        double projX = aM[0] + t * dx;
+        double projY = aM[1] + t * dy;
+
+        return Math.hypot(pM[0] - projX, pM[1] - projY);
+    }
+
+    private double[] toMeters(Coordinate p, double refLat) {
+        double lonRad = Math.toRadians(p.x);
+        double latRad = Math.toRadians(p.y);
+        double x = lonRad * Math.cos(refLat) * EARTH_RADIUS;
+        double y = latRad * EARTH_RADIUS;
+        return new double[]{x, y};
+    }
+
+    // ==================== 静态工具方法 ====================
+
+    /**
+     * 获取网格ID
+     */
+    public static String getGridId(double lon, double lat, GridMetadata metadata) {
+        int lonIdx = (int) Math.floor((lon - metadata.minLon) / metadata.gridLonWidth);
+        int latIdx = (int) Math.floor((lat - metadata.minLat) / metadata.gridLatWidth);
+        return lonIdx + "_" + latIdx;
+    }
+
+    /**
+     * 获取邻域网格ID列表（3x3）
+     */
+    public static List<String> getNeighborGridIds(int lonIdx, int latIdx) {
+        List<String> ids = new ArrayList<>();
+        for (int dl = -1; dl <= 1; dl++) {
+            for (int db = -1; db <= 1; db++) {
+                ids.add((lonIdx + dl) + "_" + (latIdx + db));
+            }
+        }
+        return ids;
+    }
+
+    /**
+     * 反序列化线段 JSON
+     */
+    public static List<Coordinate[]> deserializeSegments(String json) throws Exception {
+        ObjectMapper mapper = new ObjectMapper();
+        List<List<double[]>> segList = mapper.readValue(json,
+                mapper.getTypeFactory().constructCollectionType(List.class,
+                        mapper.getTypeFactory().constructCollectionType(List.class, double[].class)));
+        List<Coordinate[]> result = new ArrayList<>();
+        for (List<double[]> seg : segList) {
+            Coordinate a = new Coordinate(seg.get(0)[0], seg.get(0)[1]);
+            Coordinate b = new Coordinate(seg.get(1)[0], seg.get(1)[1]);
+            result.add(new Coordinate[]{a, b});
+        }
+        return result;
+    }
+
+    /**
+     * 序列化线段列表为 JSON
+     */
+    private String serializeSegments(List<Coordinate[]> segments) throws Exception {
+        List<List<double[]>> segList = segments.stream()
+                .map(seg -> Arrays.asList(
+                        new double[]{seg[0].x, seg[0].y},
+                        new double[]{seg[1].x, seg[1].y}))
+                .collect(Collectors.toList());
+        return objectMapper.writeValueAsString(segList);
+    }
+
+    // ==================== 元数据内部类 ====================
+    public static class GridMetadata {
+        public double minLon;
+        public double minLat;
+        public double gridLonWidth;
+        public double gridLatWidth;
+        public double threshold;
+
+        // 无参构造（Jackson 反序列化需要）
+        public GridMetadata() {}
+
+        public GridMetadata(double minLon, double minLat, double gridLonWidth, double gridLatWidth, double threshold) {
+            this.minLon = minLon;
+            this.minLat = minLat;
+            this.gridLonWidth = gridLonWidth;
+            this.gridLatWidth = gridLatWidth;
+            this.threshold = threshold;
+        }
+    }
+}

ebike-operations/src/main/java/com/cdzy/operations/service/impl/EbikeRegionServiceImpl.java

@@ -2,6 +2,7 @@ package com.cdzy.operations.service.impl;
 
 import cn.dev33.satoken.stp.StpUtil;
 import com.cdzy.common.ex.EbikeException;
+import com.cdzy.operations.component.GridIndexBuilder;
 import com.cdzy.operations.enums.RegionStatus;
 import com.cdzy.operations.mapper.EbikeOperationConfigMapper;
 import com.cdzy.operations.mapper.EbikeOperationLockConfigMapper;
@@ -55,7 +56,11 @@ public class EbikeRegionServiceImpl extends ServiceImpl<EbikeRegionMapper, Ebike
     @Resource
     private EbikeOperationReturnConfigMapper returnConfigMapper;
 
+    @Resource
+    private GridIndexBuilder gridIndexBuilder;
+
     @Override
+    @Transactional
     public void save(EbikeRegionVo ebikeRegion) {
         EbikeRegion entity = EbikeRegion.builder()
                 .operatorId(ebikeRegion.getOperatorId())
@@ -66,6 +71,7 @@ public class EbikeRegionServiceImpl extends ServiceImpl<EbikeRegionMapper, Ebike
                 .createdBy(StpUtil.getLoginIdAsLong())
                 .build();
         this.mapper.insert(entity);
+        gridIndexBuilder.buildIndex(entity.getRegionPolygon(),entity.getRegionId());
     }
 
     @Override
@@ -80,6 +86,7 @@ public class EbikeRegionServiceImpl extends ServiceImpl<EbikeRegionMapper, Ebike
         region.setRegionPolygon(ebikeRegion.getRegionPolygon());
         region.setRegionName(ebikeRegion.getRegionName());
         region.setRegionSimpleName(ebikeRegion.getRegionSimpleName());
+        gridIndexBuilder.buildIndex(region.getRegionPolygon(),region.getRegionId());
         this.mapper.update(region);
     }
 

ebike-operations/src/main/java/com/cdzy/operations/utils/RedisUtil.java

@@ -1,7 +1,6 @@
 package com.cdzy.operations.utils;
 
 import com.cdzy.common.enums.EbikeContents;
-import com.cdzy.operations.model.vo.EbikeEcuSnInfoVo;
 import org.springframework.beans.factory.annotation.Autowired;
 import org.springframework.beans.factory.annotation.Qualifier;
 import org.springframework.data.redis.core.RedisTemplate;
@@ -180,8 +179,8 @@ public class RedisUtil {
      * @param key 键
      * @param map 多个Hash键值对
      */
-    public void hSetAll(String key, Map<String, Object> map) {
+    public void hSetAll(String key, Map<String, Object> map,Integer DB) {
-        redisTemplate.opsForHash().putAll(key, map);
+        getRedisTemplate(DB).opsForHash().putAll(key, map);
     }
 
     /**
@@ -310,7 +309,7 @@ public class RedisUtil {
      * @param timeout 超时时间
      * @param unit    时间单位
      */
-    public Boolean tryLock(String key, Object value, long timeout, TimeUnit unit) {
+      public Boolean tryLock(String key, Object value, long timeout, TimeUnit unit) {
         return redisTemplate.opsForValue().setIfAbsent(key, value, timeout, unit);
     }
 
@@ -332,6 +331,7 @@ public class RedisUtil {
         getRedisTemplate(database).opsForValue().set(key, value);
     }
 
+
     /**
      * 设置键值对到指定数据库并设置过期时间
      */
@@ -603,4 +603,13 @@ public class RedisUtil {
 
         return result;
     }
+
+    /**
+     * 获取 key 的类型
+     * @param key 键
+     * @return 类型字符串（如 "string", "hash", "none" 等）
+     */
+    public String type(String key,Integer DB) {
+        return getRedisTemplate(DB).type(key).code();
+    }
 }