Abnormal data analysis method for dynamic water level monitoring

The invention discloses an abnormal data analysis method for dynamic water level monitoring. The abnormal data analysis method includes: firstly, recording a long period of time in a certain water body and water pressure data at short time intervals and then converting the water pressure data into relative water depth data; rejecting apparent abnormal data form the water depth pressure data, thencalculating increment of the relative water depth data at two adjacent moments to form an increment sequence, using a stochastic statistical model (such as a normal distribution curve) to estimate theconfidence interval of the incremental sequence at a given confidence level, screening out abnormal increments beyond the confidence interval, and finally, based on the rainfall and water temperatureat the time of the abnormal increments, analyzing and judging weather the abnormal increments can be accepted or not to complete the analysis and identification of abnormal data. The abnormal data analysis method is simple, sc