"Fixed window plus even interval" is the key: it narrows the field from "someone deliberately knocking" to something period-driven, such as a device or a pipe.
There is a dull, rhythmic knocking from upstairs every night between about 23:00 and 01:00 — what is it most likely to be?
依据:Fixed window and even interval fit a program cycle; a low thud is consistent with vibration transmitted through the floor slab
依据:Heating or hot-water pipes tick as they expand and cool, usually when heating stops — but less evenly spaced than this
依据:Wood moves as temperature and humidity shift, typically around a temperature turning point
依据:Possible, but "once every 30 seconds for two hours" does not match natural activity
依据:Needs extra assumptions; ranked last by Occam's razor
If it is a night-time wash/dry cycle, the sound should change in stages — fill, tumble, spin — and stop entirely after an hour or two. If it is pipe expansion, the noise clusters around when water is heated or stops being heated, and lines up with building-wide water use. The two differ clearly in whether the sound evolves in stages.
The value here is not guessing correctly but using "fixed window + even interval" to narrow to something periodic, then spending a few nights distinguishing a device from a pipe. It costs nothing, needs no confrontation, and can be settled within a week.
本页是一次推演示例:结论为概率性推断,不是事实认定,也不构成任何专业建议。你的线索与结果不会被保存。想自己推一次,见推演台;方法见方法论。