diff --git a/science_article_wos/starter/crawl_article_latest.py b/science_article_wos/starter/crawl_article_latest.py
index 930b2c9..960d91d 100644
--- a/science_article_wos/starter/crawl_article_latest.py
+++ b/science_article_wos/starter/crawl_article_latest.py
@@ -3,7 +3,7 @@
 # @Author  : zhaoxiangpeng
 # @File    : crawl_article_latest.py
 
-import math
+import time
 from typing import List
 import pymysql
 from pymysql import cursors
@@ -11,6 +11,7 @@ from twisted.internet import defer
 from scrapy.crawler import CrawlerProcess
 from scrapy.utils.project import get_project_settings
 from science_article_wos.spiders.wos_latest_increment import WosLatestIncrementSpider
+from science_article_wos.utils import tools
 
 CREATE_RECORD_SQL = '''insert into task_batch_record (batch_date, query_id, task_condition) VALUES ("%(batch_date)s", %(query_id)s, %(task_condition)s)'''
 SELECT_RECORD_SQL = """
@@ -31,30 +32,98 @@ WHERE
 """
 
 
+def get_connect() -> pymysql.Connection:
+    conn: pymysql.Connection = pymysql.connect(host='43.140.203.187', port=3306,
+                                                 database='science_data_dept', user='science-data-dept',
+                                                 passwd='datadept1509', )
+    return conn
+
+
+def get_strategy_id_by_name(name: str, source_type: int = 1) -> int:
+    conn = get_connect()
+    cursor = conn.cursor(cursors.DictCursor)
+    cursor.execute('''SELECT
+	r.org_id,
+	q.id AS query_id,
+	q.content,
+	q.param,
+	q.interval_unit,
+	q.disable_flag,
+	q.state
+FROM
+	relation_org_query AS r
+	JOIN task_search_strategy AS q ON r.query_id = q.id
+WHERE
+	r.org_name="%(org_name)s"
+	AND source_type = %(source_type)s
+	AND disable_flag = 0''' % {'org_name': name, 'source_type': source_type})
+    r = cursor.fetchone()
+    return r.get('query_id')
+
+
 def starter_latest_all():
     @defer.inlineCallbacks
-    def f():
-        client: pymysql.Connection = pymysql.connect(host='43.140.203.187', port=3306,
-                                                     database='science_data_dept', user='science-data-dept',
-                                                     passwd='datadept1509', )
-        cursor = client.cursor(cursors.DictCursor)
-        cursor.execute(SELECT_RECORD_SQL % {'limit': 1})
-        result = cursor.fetchone()
-        query_id = result['query_id']
-        cursor.execute('select org_id, org_name from relation_org_query where query_id=%s', (query_id,))
-        org_results: List[dict] = cursor.fetchall()
-        result['org_id'] = [org_result['org_id'] for org_result in org_results]
-        result['org_name'] = [org_result['org_name'] for org_result in org_results]
-
-        init_params = result
-        yield process.crawl(WosLatestIncrementSpider, task_obj=init_params)
+    def f(running: bool = True):
+        while running:
+            client: pymysql.Connection = pymysql.connect(host='43.140.203.187', port=3306,
+                                                         database='science_data_dept', user='science-data-dept',
+                                                         passwd='datadept1509', )
+            cursor = client.cursor(cursors.DictCursor)
+
+            # 从检索策略表插需要做的任务
+            cursor.execute('select id from task_search_strategy where disable_flag=0 and source_type=1 and state=0 limit 1')
+            search_strategy_r = cursor.fetchone()
+            if not search_strategy_r:
+                running = False
+                cursor.close()
+                client.close()
+            # 把任务加到批次表
+            strategy_id = search_strategy_r.get('id')
+            insert_sql = CREATE_RECORD_SQL % {
+                'batch_date': tools.get_today_date(),
+                'query_id': strategy_id,
+                'task_condition': '"AND PY=(2025-2026)"'
+            }
+            cursor.execute(insert_sql)
+            # 添加后标记为1
+            cursor.execute('update task_search_strategy set state=1, updated_time=CURRENT_TIMESTAMP where id=%s', (strategy_id,))
+            cursor.connection.commit()
+
+            # 从记录表取任务执行
+            cursor.execute(SELECT_RECORD_SQL % {'limit': 1})
+            result = cursor.fetchone()
+            if result:
+                query_id = result['query_id']
+                cursor.execute('select org_id, org_name from relation_org_query where query_id=%s', (query_id,))
+                org_results: List[dict] = cursor.fetchall()
+                result['org_id'] = [org_result['org_id'] for org_result in org_results]
+                result['org_name'] = [org_result['org_name'] for org_result in org_results]
+
+                init_params = result
+                cursor.close()
+                client.close()
+
+                yield process.crawl(WosLatestIncrementSpider, task_obj=init_params)
+                time.sleep(60)
+
+            else:
+                running = False
+                cursor.close()
+                client.close()
 
     process = CrawlerProcess(get_project_settings())
-    f()
+    f(True)
     process.start()
     process.stop()
 
 
+def starter_latest_one(strategy_id: int = None, org_name: str = None):
+    if strategy_id is None:
+        if org_name is None:
+            raise Exception('指定 strategy_id 或者 org_name')
+        strategy_id = get_strategy_id_by_name(org_name)
+
+
 def starter():
     process = CrawlerProcess(get_project_settings())
     process.crawl(WosLatestIncrementSpider)