N57xx：LAN/VISA API上的数据接收超时

您好，我们在一个应用中使用多个N5768A电源，通过查询电压输出电平，电流输出电平，输出状态和保护标志来监控仪器状态。
这些参数由相关的SCPI请求单独请求，通过LAN / VISA API发送到仪器。
100ms后读取对每个请求的响应。
这种通信方式似乎对我们的大多数电源都很好，因为我们能够在数小时/数天内监控它们而不会出现任何故障。
其中一个电源反应不同：几个小时后（4个小时似乎是平均值），我们的应用程序无法读取对SCPI请求之一的响应。
关闭/重新打开连接后，我们可以再次与仪器通信，但这种解决方法对于生产代码是不可接受的。
SCPI序列：viPrintf：MEAS：VOLT？
睡眠：100ms viScanf：MEAS：CURR？
睡眠：100ms viScanf：OUTP：STAT？
睡眠：100ms viScanf：STAT：QUES？
睡眠：100毫秒我已经通过文档查看了一些线索，现在还没找到。
你做 ？
谢谢，弗雷德里克



     以下为原文

  Hello,

We are using several N5768A power supplies in one application which monitors the instrument state by polling the voltage output level, current output level, output state and protect flags. These parameters are requested individually by the related SCPI request, sent to the instrument through the LAN/VISA API. The response to each request is read 100ms later.

This way to communicate seems fine for most of our power supplies, as we are able to monitor them during hours/days without having any glitch. One of the power supply reacts differently however : after a few hours (4 hours seems the average), one response to one of the SCPI request is not read by our application. We are able to communicate again with the instrument after closing/reopening the connection but this workaround is not acceptable for production code.

SCPI sequence :  
viPrintf : MEAS:VOLT?
Sleep : 100ms
viScanf : MEAS:CURR?
Sleep : 100ms
viScanf : OUTP:STAT?
Sleep : 100ms
viScanf : STAT:QUES?
Sleep : 100ms

I have looked through the documentation for some clue, and haven't find any for now. Do you ?
Thanks,

Frederic