看了一下S7帮助，说明部分不是非常详细，但是发现帮助的例子中有一段注释如下：
Example
STL Explanation
L MD10 //Load the floating-point number into ACCU 1-L.
RND //Convert the floating-point number (32-bit, IEEE 754) into an integer (32-bit) and round off the result.
T MD20 //Transfer result (double integer number) to MD20.
上面“Load the floating-point number into ACCU 1-L”意思是将MD10装入累加器低字，这样岂不是转换成16位浮点数了吗？