要获取所需的数据，可以使用以下查询：
注：取决于 CURRDATE() 这些例子将失败。你可以用ofc。替换 CURDATE() 具有固定值，如 2018-11-27 ;
说明：
考虑到开/关紧随其后，查询将表与自身连接起来( L.id = R.id -1 )
查询正在选择任何结果，其中“on”或“off”是 today .
如果 on 是昨天，准时到了 00:00:00 今天： case when L.Time < CURDATE() then CURDATE() else L.Time end as onTime 如果最后一个条目是 on ，关闭时间“修正”为 00:00:00 明天： COALESCE(R.time, CURDATE() + Interval 1 day) （注：您可能需要 NOW() 而不是 CURDATE() + Interval 1 day ，到“现在”为止的当前秒数）
这两种方法用于计算运行秒数。
查询：

SELECT 
  L.playtime_id AS LID,
  R.playtime_id AS RID,
  case when L.Time < CURDATE() then CURDATE() else L.Time end as onTime,
  COALESCE(R.time, CURDATE() + Interval 1 day) AS offTime,
  (  
      UNIX_TIMESTAMP(COALESCE(R.time, CURDATE() + Interval 1 day)) -  
      UNIX_TIMESTAMP(case when L.Time < CURDATE() then CURDATE() else L.Time end)
  ) as RunningSeconds
FROM item0005 as L
LEFT JOIN item0005 AS R 
ON L.playtime_id = R.playtime_id -1

WHERE
  L.`value` = 1 AND
  (
    DATE(L.Time) = CURDATE() OR
    DATE (R.Time) = CURDATE()
  )
;

结果示例：

LID    RID     onTime                  offTime                 RunningSeconds
 2      3       2018-11-27T09:00:00Z    2018-11-27T11:26:24Z    8784
 4      5       2018-11-27T11:26:27Z    2018-11-27T11:28:29Z    122
 6      7       2018-11-27T11:29:39Z    2018-11-27T11:39:55Z    616
 8      (null)  2018-11-27T11:50:55Z    2018-11-28T00:00:00Z    43745

假设明天00:00:00的示例：http://sqlfiddle.com/#!公元前9/20年14/1
示例使用 NOW() 如果最后一个状态是 on : http://sqlfiddle.com/#!9/3c6227/1号
如果您只需要从中获得聚合，您可以使用另一个环境选择，并通过知道一天有86400秒来计算百分比：

SELECT
  SUM(RunningSeconds) AS RunningSeconds,
  SUM(RunningSeconds) / 86400 * 100 AS PercentageRunning
 FROM (
   ...
 ) as temp;

http://sqlfiddle.com/#!公元前9/20年14/5

----+

----+

----+
我已经准备好了：http://sqlfiddle.com/#!9/556364a号
谢谢！

----+
|值| secondsofday | percentofday|
+-------+

试试这个

SELECT VALUE, SUM(TIME_TO_SEC(TIMEDIFF(T2,T1))), 100*SUM(TIME_TO_SEC(TIMEDIFF(T2,T1)))/86400 FROM
(SELECT A.VALUE AS VALUE, COALESCE(B.TIME, TIMESTAMP(CURDATE()-2)) AS T1, A.TIME AS T2
FROM ITEM0005 A
LEFT JOIN ITEM0005 B
ON A.PLAYTIME_ID-1= B.PLAYTIME_ID
WHERE A.VALUE='0'
UNION
SELECT '1' AS VALUE, A.TIME AS T1, COALESCE(B.TIME, TIMESTAMP(CURDATE()-1)) AS T2
FROM ITEM0005 A
LEFT JOIN ITEM0005 B
ON A.PLAYTIME_ID= B.PLAYTIME_ID-1
WHERE A.VALUE='0') C
GROUP BY VALUE;

----+
| 0 | 28800 | 33.3 |
| 1 | 57600 | 66.6 |
+-------+

----+